# APPLICATIONS OF NANOBIOTECHNOLOGY IN PHARMACOLOGY

EDITED BY : Wei Tao, Shahed Behzadi, Jianxun Ding and Chao Wang PUBLISHED IN : Frontiers in Pharmacology

#### Frontiers eBook Copyright Statement

The copyright in the text of individual articles in this eBook is the property of their respective authors or their respective institutions or funders. The copyright in graphics and images within each article may be subject to copyright of other parties. In both cases this is subject to a license granted to Frontiers. The compilation of articles constituting this eBook is the property of Frontiers.

Each article within this eBook, and the eBook itself, are published under the most recent version of the Creative Commons CC-BY licence. The version current at the date of publication of this eBook is CC-BY 4.0. If the CC-BY licence is updated, the licence granted by Frontiers is automatically updated to the new version.

When exercising any right under the CC-BY licence, Frontiers must be attributed as the original publisher of the article or eBook, as applicable.

Authors have the responsibility of ensuring that any graphics or other materials which are the property of others may be included in the CC-BY licence, but this should be checked before relying on the CC-BY licence to reproduce those materials. Any copyright notices relating to those materials must be complied with.

Copyright and source acknowledgement notices may not be removed and must be displayed in any copy, derivative work or partial copy which includes the elements in question.

All copyright, and all rights therein, are protected by national and international copyright laws. The above represents a summary only. For further information please read Frontiers' Conditions for Website Use and Copyright Statement, and the applicable CC-BY licence.

ISSN 1664-8714 ISBN 978-2-88963-375-3 DOI 10.3389/978-2-88963-375-3

#### About Frontiers

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

#### Frontiers Journal Series

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

#### Dedication to Quality

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

#### What are Frontiers Research Topics?

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# APPLICATIONS OF NANOBIOTECHNOLOGY IN PHARMACOLOGY

Topic Editors: Wei Tao, Harvard Medical School, United States Shahed Behzadi, Harvard Medical School, United States Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China Chao Wang, Soochow University, China

Citation: Tao, W., Behzadi, S., Ding, J., Wang, C., eds. (2020). Applications of Nanobiotechnology in Pharmacology. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88963-375-3

# Table of Contents


Qingqing Xiong, Mangmang Cui, Ge Yu, Jian Wang and Tianqiang Song


Na Kong, Mei Deng, Xiu-Na Sun, Yi-Ding Chen and Xin-Bing Sui


Zhenqiang Dong, Yang Kang, Qijuan Yuan, Manli Luo and Zhipeng Gu


Liang Ge, Xinru You, Jun Huang, Yuejian Chen, Li Chen, Ying Zhu, Yuan Zhang, Xiqiang Liu, Jun Wu and Qian Hai


Wenhao Nan, Li Ding, Houjie Chen, Fahim U. Khan, Lu Yu, Xinbing Sui and Xiaojun Shi


Phei Er Saw, Ao Zhang, Yan Nie, Lei Zhang, Yingjie Xu and Xiaoding Xu


Shi Liang, Junmeng Zheng, Wei Wu, Quan Li, Phei Er Saw, Jianing Chen, Xiaoding Xu, Herui Yao and Yandan Yao

*188 Enhanced Anti-tumor of Pep-1 Modified Superparamagnetic Iron Oxide/ PTX Loaded Polymer Nanoparticles*

Baoyan Wang, Weijun Wu, Hongjin Lu, Zhi Wang and Hongliang Xin

*197 Polypeptide Nanogels With Different Functional Cores Promote Chemotherapy of Lung Carcinoma*

Kai Niu, Nan Li, Yunming Yao, Chunjie Guo, Yuanyuan Ge and Jianmeng Wang

*212 Ceramide-Graphene Oxide Nanoparticles Enhance Cytotoxicity and Decrease HCC Xenograft Development: A Novel Approach for Targeted Cancer Therapy*

Shi-Bing Wang, Ying-Yu Ma, Xiao-Yi Chen, Yuan-Yuan Zhao and Xiao-Zhou Mou


Xiao Xiao, Ting Wang, Leijiao Li, Zhongli Zhu, Weina Zhang, Guihua Cui and Wenliang Li

*242 Efficient Treatment of* Sporothrix globosa *Infection Using the Antibody Elicited by Recombinant Phage Nanofibers*

Feng Chen, Rihua Jiang, Shuai Dong and Bailing Yan


Nansha Gao, Chenyang Xing, Haifei Wang, Liwen Feng, Xiaowei Zeng, Lin Mei and Zhengchun Peng

*280 Targeted Delivery of Chlorin e6* via *Redox Sensitive Diselenide-Containing Micelles for Improved Photodynamic Therapy in Cluster of Differentiation 44-Overexpressing Breast Cancer*

Chan Feng, Donglei Zhu, Lv Chen, Yonglin Lu, Jie Liu, Na Yoon Kim, Shujing Liang, Xia Zhang, Yun Lin, Yabin Ma and Chunyan Dong

*289 Synergistic Effect of Retinoic Acid Polymeric Micelles and Prodrug for the Pharmacodynamic Evaluation of Tumor Suppression*

Yan-Hua Zhu, Ning Ye, Xin-Feng Tang, Malik Ihsanullah Khan, Hong-Liang Liu, Ning Shi and Li-Feng Hang

# Editorial: Applications of Nanobiotechnology in Pharmacology

*Baowen Qi1, Chao Wang2\*, Jianxun Ding1,3\* and Wei Tao1\**

1 Center for Nanomedicine and Department of Anesthesiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, United States, 2 Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, Institute of Functional Nano & Soft Materials (FUNSOM), Soochow University, Suzhou, China, 3 Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, China

Keywords: pharmaceutical nanotechnology, drug delivery, nanomedicine, theranostics, biomaterial, regenerative medicine

**Editorial on the Research Topic**

#### **Applications of Nanobiotechnology in Pharmacology**

Nanobiotechnology, by definition, is a multi-strategic technique that combines nanotechnology and biotechnology to engineer the properties of therapeutic agents, e.g., target delivery of therapeutics by nanoparticles, in a unique way as paradigm shifts from fundamental biological study to clinical pharmacology. A well-defined nanosystem has controllable dimensions and properties being able to carry various functional biomolecules, such as small molecules, proteins, genes, and so forth. Those unusual characteristics enable them to exhibit prominent efficacies for the diagnostic and/ or treatment of numerous diseases like cancer *via* precisely tuning the size, morphology, and surface property. Moreover, strategies to achieve a better therapeutic purpose rely on "responsive" nanomaterials that release the active substances under specific stimuli, such as pH, redox potential, temperature, enzymes, or other external stimuli dependent on their unique physicochemical conditions. It is particularly noteworthy that the synergistic combination of nanoparticles with different target ligands facilitates the development of more efficient "active" drug delivery systems.

Furthermore, the sensitivity of *in vivo* real-time diagnosis can be enhanced by combing nanotechnology with a contrast agent for next-generation precision medicine. However, the safety issues of nanosystems as well as their clinical efficacies remain controversial, which concern intrinsically originated from four aspects: 1) synthetic nanoparticles are generally composed of inorganic or organic materials, which may trigger cytotoxic pathways; 2) nanoparticles alters the biodistribution of the carried agents, which may, in turn, change the toxicological behaviors of the agents as well; 3) the enhanced permeability and retention (EPR) effect-based tumortargeting nanomedicines is hard to be proved in humans, which was explained by the drastically difference between murine and human tumor tissues; 4) the complicated protein coronas on the surface of nanoparticles disturbs their metabolism behaviors. Although there is still a fierce debate over the fate of this emerging field, we still hold very optimistic attitudes toward the revolution of nanobiotechnology brought to the future pharmacological realms.

In the current topic, an overview of applications of nanobiotechnology in pharmacology is provided through 26 articles by 206 authors, which contains 2 reviews and 24 original research papers (Total views: 40,096; as of Oct 1st, 2019). One of the reviews summarized the currently available knowledge on the role of clinically approved poly(ethylene glycol)–polylactide (PEG– PLA) copolymer micelles as nanocarriers for therapy of malignancies (Wang et al.). Another review summarized the studies using carbon-based nanomaterials, including carbon nanotubes, graphene oxide (GO), and graphene quantum dots, being extensively investigated for various applications, e.g., biosensing and cancer therapy, owing to their structural diversities (Maiti et al.).

#### Edited and reviewed by:

Salvatore Salomone, University of Catania, Italy

#### \*Correspondence:

Chao Wang cwang@suda.edu.cn Jianxun Ding jxding@ciac.ac.cn Wei Tao wtao@bwh.harvard.edu

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 02 October 2019 Accepted: 12 November 2019 Published: 04 December 2019

#### Citation:

Qi B, Wang C, Ding J and Tao W (2019) Editorial: Applications of Nanobiotechnology in Pharmacology. Front. Pharmacol. 10:1451. doi: 10.3389/fphar.2019.01451

1 **6**

The special topic contains many original studies covering critical aspects of nanobiotechnology. Among them, a few selfassembled nanoparticles were meticulously designed and prepared using biocompatible polymers. To start with, PEG as a hydrophilic polymer is widely used for self-assembly with other sets of hydrophobic components to form "core–shell" nano-structures, for instance, PEG-based platinum(II) (Pt(II)) nanoformulation was developed to combat multidrug-resistant cancer (Tsai et al.). Similarly, PEG–PLA nanocarrier containing both Pt(IV) and capecitabine was developed using a co-assembly approach (Xiao et al.). Self-assembled nanoparticle prepared by the conjugation PEG to β-cyclodextrin (PEG-CD) was able to efficiently deliver doxorubicin (DOX) and sorafenib in a rational manner (Xiong et al.), and a multi-functional peptide-modified PEG-CD complex was developed for diabetes and immunotherapy (Dong et al.). PEG-poly(ε-caprolactone) (PEG-*b*-PCL) was able to deliver cytokines in order to prohibit the migration of breast cancer cells (Liang et al.) and PEG–poly(amino acid) was effectively used to deliver DOX as a benefit of the EPR effect (Niu et al.).

Pluronic F127 is a triblock copolymer composed of poly(ethylene glycol)–poly(propylene glycol)–poly(ethylene glycol) (PEG–PPG–PEG). The hydrophobic drug retinoic acid was stabilized with Pluronic F127 to form micelle for combination therapy against solid tumor (Zhu et al.). In order to enhance the efficacy of antitumor drug docetaxel (DTX), aptamer-polydopamine-functionalized nanoparticle was prepared and utilized to enhance chemo-photothermal therapy for breast cancer (Kong et al.), and hydrophobic poly(ester amide) nanoparticle was used to encapsulate DTX for the suppression of non-small-cell lung cancers (Chen et al.).

Chitosan (CS), hyaluronic acid (HA), and polygalactose are polysaccharide biomaterials, which are also discussed in this special issue. For instance, CS oligosaccharide was used to inhibit the intrinsic coagulation pathway due to its biocompatible activity (Guo et al.). A natural product quercetin was encapsulated into CS nanoparticle as a potential therapeutic agent for topical application against ultraviolet B radiation (Nan et al.). Hydrophilic HA can specifically target CD44 on the cancer cells, which was able to self-assemble with hydrophobic photosensitizer chlorin e6 (Ce6) to form into a micelle. The micellar system demonstrated a redox-responsive kinetics to controllably release of payloads for the enhancement efficacy of photodynamic therapy (PDT) (Feng et al.). Besides, glycopolymer formed nanoparticle by crosslinking with the hydrophobic drug, therefore possessing both redox-responsive and pH-sensitive characteristics for precise hepatoma therapy (Wu et al.).

In addition to polymer nanoparticles, liposomes or lipid nanoparticles also played vital roles in pharmacology. One study was to encapsulate brinzolamide into hydropropylβ-cyclodextrin inclusion complex and was then formulated

as nanoliposome by the thin-film dispersion method (Wang et al.). In the same research group, the authors synthesized mannose-cholesterol conjugate by click reaction with PEG of different molecular weights to prepare another unique nanoliposome (Wang et al.). Moreover, encapsulating peptide into lipid nanoparticle possessed both the superior nature of polymer nanoparticle and liposome. Therefore a high oral bioavailability and sustained release kinetics were achieved (Zhao et al.).

Nowadays, inorganic materials also demonstrated high potentials to be used in pharmacology. The superparamagnetic iron-oxide nanoparticle was attractive due to its unique properties to deliver paclitaxel, offering both magnetic targeting and receptor-mediated targeting outcome (Wang et al.). The ultra-small nano-ceramide-GO nanoparticle was devised for treating hepatoma (Wang et al.). Polymer-coated black phosphorus nanosheet showed long circulation kinetics and an excellent cell uptake capacity *in vivo*, providing a synergistic cancer therapeutic strategy (Gao et al.).

Overall, this research topic discussed a few proofs of concept by taking the advantages of nanobiotechnology in pharmacology. Such an emerging field provided insightful thought from fundamental nanotechnology researches to translational medicine, meanwhile presenting a future perspective to achieve the success of nanotechnology as a return.

# AUTHOR CONTRIBUTIONS

BQ, CW, JD and WT contributed to writing this Editorial.

# ACKNOWLEDGMENTS

The editors appreciate the contributions of all authors to this Research Topic, the constructive comments of all the reviewers, and the editorial support from Frontiers throughout the publication process. This work is partly supported by Collaborative Innovation Center of Suzhou Nano Science & Technology, the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the 111 Project.

**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2019 Qi, Wang, Ding and Tao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Facile Fabrication of Reduction-Responsive Supramolecular Nanoassemblies for Co-delivery of Doxorubicin and Sorafenib toward Hepatoma Cells

Qingqing Xiong<sup>1</sup> , Mangmang Cui 1,2, Ge Yu<sup>1</sup> , Jian Wang<sup>3</sup> \* and Tianqiang Song<sup>1</sup> \*

*<sup>1</sup> Department of Hepatobiliary Cancer, National Clinical Research Center for Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin's Clinical Research Center for Cancer, Tianjin Medical University Cancer Institute and Hospital, Tianjin, China, <sup>2</sup> Hebei province Cangzhou Hospital of Integrated Traditional and Western Medicine, Cangzhou, China, <sup>3</sup> Department of Immunology, National Clinical Research Center for Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin's Clinical Research Center for Cancer, Tianjin Medical University Cancer Institute and Hospital, Tianjin, China*

#### Edited by:

*Chao Wang, University of North Carolina at Chapel Hill, United States*

#### Reviewed by:

*Yanqi Ye, University of North Carolina at Chapel Hill, United States Fuping Gao, Chinese Academy of Sciences, China Wenjun Miao, Nanjing Tech University, China*

#### \*Correspondence:

*Jian Wang wangjian112358@163.com Tianqiang Song tjchi@hotmail.com*

#### Specialty section:

*This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology*

> Received: *18 December 2017* Accepted: *17 January 2018* Published: *06 February 2018*

#### Citation:

*Xiong Q, Cui M, Yu G, Wang J and Song T (2018) Facile Fabrication of Reduction-Responsive Supramolecular Nanoassemblies for Co-delivery of Doxorubicin and Sorafenib toward Hepatoma Cells. Front. Pharmacol. 9:61. doi: 10.3389/fphar.2018.00061* Combination of doxorubicin with sorafenib (SF) was reported to be a promising strategy for treating hepatocellular carcinoma (HCC). In this study, we designed a reduction-responsive supramolecular nanosystem based on poly (ethylene glycol)-β-cyclodextrin (PEG-CD) and a disulfide-containing adamantine-terminated doxorubicin prodrug (AD) for efficient co-delivery of doxorubicin and sorafenib. PEG-CD/AD supramolecular amphiphiles were formed through host-guest interaction between cyclodextrin and adamantine moieties, and then self-assembled into regular spherical nanoparticles with a uniform size of 166.4 nm. Flow cytometry analysis and confocal laser scanning microscopy images showed that PEG-CD/AD nanoparticles could be successfully taken up by HepG2 cells and then released doxorubicin into the cell nuclei. Moreover, sorafenib could be facilely encapsulated into the hydrophobic cores to form PEG-CD/AD/SF nanoparticles with a slightly larger size of 186.2 nm. PEG-CD/AD/SF nanoparticles sequentially released sorafenib and doxorubicin in a reduction-response manner. *In vitro* cytotoxicity assay showed that PEG-CD/AD/SF nanoparticles had an approximately 4.7-fold decrease in the IC<sup>50</sup> value compared to that of PEG-CD/AD and SF physical mixtures, indicating stronger inhibitory effect against HepG2 cells by co-loading these two drugs. In summary, this novel supramolecular nanosystem provided a simple strategy to co-deliver doxorubicin and sorafenib toward hepatoma cells, which showed promising potential for treatment of HCC.

Keywords: reduction-responsive, supramolecular nanoassemblies, doxorubicin, sorafenib, hepatocellular carcinoma

# INTRODUCTION

Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related death (International Agency for Research on Cancer. GLOBOCAN, 2012), characterized by its insidious onset, poor diagnosis, and intrinsic resistance to chemotherapy agents. Doxorubicin (DOX), a widely used chemotherapeutic drug, plays an undisputed key role in transcatheter arterial chemoembolization

**8**

(TACE) for HCC (Vilaseca et al., 1978). However, the therapeutic efficacy of DOX for HCC is often limited because of the emergence of drug resistance (Deng et al., 2015; Galun et al., 2017) and its irreversible cardiotoxicity (Singal and Iliskovic, 1998). Sorafenib (SF), an oral multiple kinase inhibitor, is successful to prolong the survival of advanced HCC patients and has been approved by FDA for the standard treatment of patients with unresectable HCC (Llovet et al., 2015). However, during the clinical application, it was found that SF was only beneficial to about 33% HCC patients while showed low tumor response to other majority of patients (Cheng et al., 2009). It has been reported that combining two or more drugs of different molecular mechanisms might be a promising alternative option for long-term application of a single drug and would exert better therapeutic effects on HCC (Lin et al., 2016; Sun et al., 2017; Zhao et al., 2017). Moreover, clinical trial studies have demonstrated that combination of DOX with SF exhibited remarkable improvement in the overall survival of HCC patients (Hutchinson, 2011; Pazo Cid et al., 2011).

However, due to the hydrophobic properties and varied pharmacokinetic profiles of these two drugs, the co-delivery of both drugs toward HCC is still a major challenge. With the development of nanotechnology in biomedical fields, nanoparticles provide a valid platform to achieve co-loading of multiple drugs and deliver them toward tumor cells simultaneously (Hu et al., 2016). Among these nanoparticles, supramolecular nanosystems, especially formed by host-guest interaction, have attracted researchers' great attention in drug delivery (Feng et al., 2017; Ping et al., 2017). Cyclodextrins (CDs), a series of natural cyclic oligosaccharides composed of D-glucose units, are the most commonly used macrocyclic hosts to construct supramolecular nanosystems in virtue of their low toxicity and low immunogenicity (Zhang et al., 2013; Feng et al., 2017; Xiong et al., 2017). The most prominent structural feature of CDs is their hydrophobic cavities, which can accommodate a variety of guest molecules such as adamantine (Luo et al., 2012) and trans-azobenzene (Li et al., 2013) via hostgust interaction. During the past few decades, CDs and their derivatives have been extensively utilized in the construction of drug delivery system. Zhang and coworkers established a supramolecular photosensitizer delivery system through the self-assembly of supramolecular amphiphiles constructed by the host–guest interaction between poly (ethylene glycol)-βcyclodextrin and adamantane-terminated porphyrin derivatives (Liu et al., 2015). In our previous study, a series of well-defined cyclodextrin-based polymers were synthesized by atom transfer radical polymerization and then further used for drug delivery (Xiong et al., 2014; Zhang M. et al., 2014).

Recently, stimuli-responsive nanoparticles have drawn a lot of attentions due to their responsiveness to tumor or intracellular micro-environmental stimuli, such as low pH-values (Chen et al., 2016), high concentration of certain enzymes (Zhang et al., 2017), and more reductive environment (Wang et al., 2013), and thus releasing their payloads in a controlled manner. According to the previous studies (Wang et al., 2012; Yin et al., 2013), the cytosolic glutathione (GSH) concentration is around 2–10 mM, which is substantially higher than that in the extracellular fluids and blood (2–10µM). Disulfide bond, which is stable in blood circulation but can be cleaved under reducing intracellular microenvironment, has been widely used as a linker in the design of reduction-responsive drug delivery system. In Sun's work, they reported a prodrug-based nanoplatform self-assembled by the disulfide bond linked conjugates of paclitaxel (PTX) and oleic acid, achieving the rapid and differential release of PTX in tumor cells (Luo et al., 2016). Moreover, the prodrug strategy provides several advantages such as enhanced in vivo stability, the prolonged half-life in blood, the improved water solubility, etc. (Yang et al., 2015; Li D. et al., 2016; Li W. et al., 2016).

In this paper, we prepared a novel reduction-responsive supramolecular nanosystem by a facile method for co-delivery of DOX and SF toward hepatoma cells (**Figure 1**). A reductionresponsive prodrug of doxorubicin (Ada-ss-DOX) was first synthesized by conjugation of doxorubicin with amantadine through disulfide bond, and then complexed with poly (ethylene glycol)-β-cyclodextrin (PEG-CD) via host-guest interaction between CD cavity and Ada moiety to form PEG-CD/AD supramolecular amphiphiles (I). Using a modified nanoprecipitation method, PEG-CD/AD amphiphiles could self-assemble into spherical nanoparticles in aqueous medium (II). The reduction-responsive disassembly of PEG-CD/AD nanoparticles triggered by dithiothreitol (DTT) was then investigated (III). Moreover, SF could also be incorporated into the hydrophobic cores of the nanoparticles by hydrophobic interaction and π-π stacking interaction with DOX (IV), and thus achieved the co-loading of DOX and SF. The size and morphology of these nanoparticles were respectively characterized by dynamic light scattering (DLS) and transmission electron microscopy (TEM). The cellular uptake and intracellular locations of PEG-CD/AD nanoparticles were investigated by flow cytometry analysis and confocal laser scanning microscopy (CLSM). In addition, the cytotoxicity of PEG-CD/AD/SF nanoparticles against hepatoma HepG2 cells were evaluated by CCK-8 assay in detail.

#### MATERIALS AND METHODS

# Materials

Amantadine hydrochloride (Ada), and dithiodiglycolic acid (DTGA, ≥98%) 4′ ,6-diamidino-2-phenylindole (DAPI) were obtained from Sigma-Aldrich (St. Louis, MO, USA) and used without further purification. Doxorubicin hydrochloride (DOX·HCl) and sorafenib (SF) base was purchased from Dalian Meilun Biotech Co., Ltd. (Dalian, China). 1-(3- Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC·HCl), 1-hydroxybenzotriazole hydrate (HOBt, 99%), O-(7-azabenzotriazol-1-yl)-N,N,N′ ,N′ -tetramethyluronium hexafluorophosphate (HATU, 99%), and N,Ndiisopropylethylamine (DIPEA, 99%) was purchased from J&K Scientific (Beijing, China). Methoxypoly(ethylene glycol) succinimidyl succinate (PEG-NHS, M<sup>n</sup> = 2000) was purchased from JenKem Technology Co., Ltd (Beijing, China). Other chemical reagents in the study were analytical grade and obtained from commercial sources. PEG-CD was synthesized according to the previous study (Peng et al., 2014) and the synthetic route was described in the supplementary data (Figure S1A). The <sup>1</sup>H NMR spectrum of PEG-CD was displayed in Figure S1B.

# Synthesis of Amantadine-Terminated Doxorubicin Prodrug (Ada-ss-DOX)

Ada-ss-DOX was synthesized by the following two steps. DTGA (5 mmol, 0.9110 g) was dissolved 15 mL of dichloromethane. Subsequently, HOBt (4.5 mmol, 0.6070 g), EDC·HCl (4.5 mmol, 0.8612 g) and triethylamine (15 mmol, 1.5178 g) were added. After the mixture was stirred at 20◦C for 1 h, Ada (3 mmol, 0.4549 g) was added. The resulting mixture was stirred at 20◦C for another 3 h and filtered to give the filtrate. The filtrate was concentrated to give the crude product and then purified by prep-HPLC system (Instrument: GILSON 281, Column: ASB C18 150<sup>∗</sup> 25 mm. Mobile phase: A: deionized water containing 0.1% HCl; B: acetonitrile, Gradient: B from 40 to 70% in 10 min. Flow rate: 25 mL/min) to obtain Ada-ss-COOH as a brown solid (Yield 6.8%).

Subsequently, the synthesized Ada-ss-COOH (3.17 mmol, 1.00 g) and DOX·HCl (2.85 mmol, 1.65 g) were co-dissolved in 20 mL dimethyl formamide (DMF) and then HATU (4.76 mmol, 1.81 g) and DIPEA (9.51 mmol, 1.23 g) were added. The mixture was stirred at 20◦C for 2 h and then diluted by 20 mL of water. The resulting mixture was extracted with ethyl acetate. The organic phases were collected, washed with brine and dried with anhydrous Na2SO4, filtered and finally concentrated under vacuum to give Ada-ss-DOX (hereafter "AD") as a red solid powder (Yield 45.5%).

# Preparation of PEG-CD/AD and PEG-CD/AD/SF Nanoparticles

PEG-CD/AD nanoparticles were prepared by a modified nanoprecipitation and dialysis method. PEG-CD and AD were dissolved in DMF respectively to obtain solutions with a concentration of 1 mg/mL. Then, PEG-CD solution and Adass-DOX solution with molar ratio of 1:1 was mixed and stirred overnight to obtain PEG-CD/AD supramolecular amphiphiles. Afterwards, an equal volume of water was slowly dropped into the above mixed solution under stirring at 400 rpm. The solution was stirred for another 4 h, followed by dialyzed against deionized water for 24 h to remove DMF (MWCO 3500). The dialysate was further concentrated by ultra-filtration membrane (MWCO 5000).

PEG-CD/AD/SF nanoparticles were prepared by a similar method. A determined volume of SF solution in DMF was stirred with PEG-CD/AD mixed solution for 4 h and then water was added. The solution was stirred, dialyzed and concentrated as described above to obtain PEG-CD/AD/SF nanoparticles.

#### Characterization

<sup>1</sup>H NMR analysis was carried out on a AVANCE III HD nuclear magnetic resonance spectrometer at 400 MHz (Bruker, Germany). The molecular weight of AD was analyzed by mass spectrometry (Agilent, 6510 Q-TOF LC/MS, USA). The size, size distribution, and zeta potential of the nanoparticles in aqueous solution were determined by a Zetasizer Nano ZS instrument (Malvern, England) at 25◦C. The nanoparticle morphology was visualized using Tecnai G2 20STwin transmission electron microscope (FEI, USA).

#### Reduction-Triggered Disassembly of PEG-CD/AD Nanoparticles

PEG-CD/AD nanoparticles were added into DTT solutions with different concentrations and then treated for determined periods. The fluorescence intensities of the mixed solutions were measured on a F-4500 fluorescence spectrometer (Hitachi, Japan) at an excitation wavelength of 485 nm. Meanwhile, the size and morphology changes of PEG-CD/AD nanoparticles were monitored by DLS and TEM.

# Cellular Uptake and Intracellular Distributions of PEG-CD/AD Nanoparticles

Flow cytometry was employed to determine the cellular uptake of PEG-CD/AD nanoparticles in HepG2 cells. Briefly, the cells were seeded on a 12-well plate and cultured for 24 h. Then the cultural media were replaced with fresh media containing free DOX or PEG-CD/AD nanoparticles at a DOX concentration of 2µg/mL. After incubation for determined periods, the cells were collected and analyzed on a BD FACSVerseTM flow cytometer (BD Biosciences, USA).

For confocal laser scanning microscope (CLSM) study, the cells were placed onto 12-well glass plates. After incubation for 24 h, the cells were treated as the same as flow cytometry analysis. Then the cells were washed with PBS, fixed with 4% formaldehyde and the cell nuclei were stained by DAPI. Thereafter, the cells were observed under FV 1000 CLSM (Olympus, Japan).

#### In vitro Drug Releases of DOX and SF from PEG-CD/AD/SF Nanoparticles

The in vitro releases of DOX and SF from PEG-CD/AD/SF nanoparticles were evaluated using dynamic dialysis method in four kinds of release media (I: pH 7.4 PBS; II: pH 7.4 PBS with 10 mM DTT; III: pH 5.0 PBS; IV: pH 5.0 PBS with 10 mM DTT). Briefly, PEG-CD/AD/SF nanoparticle solutions (1 mL) were placed into dialysis bags (MWCO 7000) and dialyzed against above media at 37◦C in the air bath shaking with 100 rpm. At designated time points, 0.5 mL of release media was withdrawn and replaced with an equal volume of fresh release media. The amounts of DOX and SF in the release media were determined by UV/Vis spectrophotometer (Beckman DU-640, USA) at a wavelength of 490 and 267 nm, respectively (Malarvizhi et al., 2014). The release tests were repeated for three times and the accumulative releases of drug were calculated.

#### In vitro Cytotoxicity Analysis

The cytotoxicities of PEG-CD/AD and PEG-CD/AD/SF nanoparticles compared with free DOX, free SF, and PEG-CD/AD nanoparticles and SF physical mixtures were evaluated in HepG2 cells using CCK-8 regents. Briefly, the cells were seeded onto 96-well plate and cultured for 24 h, and then the media were replaced with fresh media containing free DOX, free SF, PEG-CD/AD nanoparticles, PEG-CD/AD and free SF physical mixtures, and PEG-CD/AD/SF nanoparticles followed by further incubation for 48, 72, and 96 h, respectively. Next, the media were removed and 10 µL of CCK-8 regents and 90 µL of fresh media were added into each well. After incubated for 2 h, the plates were gently shaken for 2 min and the absorbance was measured at 450 nm using an Epoch 2 microplate spectrophotometer (BioTek, Winooski, USA). The cell viabilities were calculated as the ratio of the absorbance values of treated cells to those of untreated cells. The inhibitory effect of PEG-CD on HepG2 cell growth for 48 h was also assessed as the similar method.

### Statistical Analysis

Each experiment was repeated for three times. All data were presented as mean ± standard deviation and compared using one-way ANOVA. The differences were significant when p < 0.05.

# RESULTS AND DISCUSSION

# Synthesis and Characterization of Ada-ss-DOX

DTGA, with two active carboxyl groups, was chosen as the disulfide reagent to produce Ada-ss-DOX prodrug. The synthesis route is displayed in **Figure 2A**. Firstly, Ada-ss-COOH was synthesized via the condensation reaction between Ada and DTGA. To ensure that the product with only one of the carboxyl groups of DTGA conjugated with Ada was collected, excess DTGA was added during the reaction process and the product was further purified by prep-HPLC. From the <sup>1</sup>H NMR spectrum of Ada-ss-COOH in **Figure 2B**, the integration ratio of the peaks at 3.66 and 3.38 ppm, corresponding to the methylene protons of DTGA (signal a and b), to the peaks at 1.75 ppm, originating from the methylene protons of Ada (signal d), was calculated to be 1:3, which indicated that the molar ratio of Ada and DTPA is 1:1. Moreover, the molecular weight of Ada-ss-COOH (C14H22NO3S + 2 ) determined by ESI-MS in **Figure 2C** was found to be 315.9 g/mol (calcd 316.5 g/mol), further indicating the success synthesis of Ada-ss-COOH.

Subsequently, Ada-ss-DOX was obtained by the reaction between the end carboxyl group of Ada-ss-COOH and the reactive amine group of DOX. The characteristic peaks of hydroxyl protons of DOX (signal a' and b') located at 13.93 and 13.20 ppm were obviously observed in the <sup>1</sup>H NMR spectrum of Ada-ss-DOX (**Figure 2B**). Due to the sterichindrance effect after conjugation of DOX, the signals attributed to the methylene protons of DTGA (signal a and b) significantly decreased in the1H NMR spectrum of Ads-ss-DOX. The ESI-MS result showed that the molecular weight of Ada-ss-DOX (C41H49N2O13S + 2 ) was 841.3 g/mol, which was consistence with calculated 841.9 g/mol (**Figure 2C**). Altogether, the above results demonstrated that Ada-ss-DOX was successfully prepared.

#### Preparation and Characterization of PEG-CD/AD Nanoparticles

To our knowledge, the formation of nanoparticles was often due to the self-assembly process of amphiphilic block copolymers driven by hydrophobic interaction among the hydrophobic segments (Zhao et al., 2012). In this study, PEG-CD/AD nanoparticles were self-assembled from PEG-CD/AD supramolecular amphiphiles with hydrophilic segment PEG-CD and hydrophobic segment DOX (**Figure 1**). The DMF solution of PEG-CD and AD was stirred overnight to obtain the supramolecular amphiphiles via host-guest interaction between CD cavities and Ada moieties, and then the deionized water was slowly added to induce the selfassembly of PEG-CD/AD nanoparticles. Finally, the solution

was dialyzed against deionized water to remove DMF. The DOX content was calculated to be 13.5% as the molar ratio of PEG-CD and AD was 1:1. The hydrodynamic diameter and morphology of PEG-CD/AD nanoparticles was determined by DLS and TEM. As shown in **Figure 3B**, the average size of PEG-CD/AD nanoparticles was 166.4 nm and polydispersity (PDI) of 0.089. TEM image in **Figure 3Ca** showed that PEG-CD/AD nanoparticles had spherical shape and the average particle size was about 100 nm, which was significantly smaller that measured by DLS. It was perhaps because that the PEG chains in the outer surface of PEG-CD/AD nanoparticles were extended in aqueous solution while collapsed in dry samples.

#### Reduction-Responsive Disassembly of PEG-CD/AD Nanoparticles

DTT was used to mimic the reductive environment as it was a prevailing GSH stimulant (Dai et al., 2011; Zhang B. et al., 2014). As shown in **Figure 3A**, the fluorescence intensity of the PEG-CD/AD solution containing 10 mM DTT gradually increased as time prolonged. By contrast, almost no change in the emission spectra of the PEG-CD/AD solution was observed for 48 h in the absence of DTT (Figure S2). Moreover, the emission spectra of the nanoparticles were also detected after incubation with DTT of different concentrations for 2 h. It was obviously found that the fluorescence intensity increased with the DTT concentration increasing from 1 to 50 mM (Figure S3). The above results suggested that the disulfide bond could be cleaved under the stimulation of DTT, resulting in the disassembly of PEG-CD/AD nanoparticles and release of DOX.

The hydrodynamic diameters of PEG-CD/AD nanoparticles were also measured after incubation with DTT for 4 h and 24 h (**Figure 3B**). After 4 h, the average size of the nanoparticles increased to 683.6 nm with a corresponding increase of PDI to 0.29. The nanoparticles changed to be larger than 1µm with PDI of 0.89 after 24 h incubation. TEM images in **Figures 3Cb**,**c** showed that the morphology of the nanoparticles was irregular and large aggregated particles could be observed, which further confirmed that the disassembly due to reductive cleavage of the disulfide bonds. To be noticed, the hydrodynamic diameters of PEG-CD/AD nanoparticles after incubation with DTT seemed larger than the sizes of nanoparticles displayed in TEM images. This might be attributed that the average size measured by DLS was determined by intensity, and in this case, the size would be larger than the actual size when very large aggregated particles existed.

#### Cellular Uptake and Intracellular Locations of PEG-CD/AD Nanoparticles

The cellular uptake of PEG-CD/AD nanoparticles was investigated in HepG2 cells by flow cytometry and CLSM. HepG2 cells were incubated with PEG-CD/AD nanoparticles or free DOX for 2 h at DOX concentration of 2.0µg/mL and then collected for flow cytometry analysis. As shown in **Figure 4A**, PEG-CD/AD nanoparticles exhibited much stronger fluorescence intensity than the control, which indicated that the nanoparticles could be effectively internalized into HepG2 cells. However, PEG-CD/AD nanoparticles showed slightly lower intracellular fluorescence intensity than that of free DOX. It was probably because DOX could not be completely released from the nanoparticles in 2 h.

Next, CLSM was employed to track the intracellular locations of PEG-CD/AD nanoparticles after internalization. HepG2 cells were treated with PEG-CD/AD nanoparticles for predetermined time and the nuclei were stained by DAPI (blue). As shown in **Figure 4B**, the fluorescence signals of DOX (red) in the cells exhibited time-dependent manner. Weak fluorescence signals were observed in the cytoplasm after 0.5 h of incubation and the signals tended to be stronger after 2 h of incubation. Minor red fluorescence signals could be observed in the nuclei after 8 h of incubation, suggesting that DOX was released from the nanoparticles and began to accumulate in the nuclei. After 24 h, the red fluorescence signals were almost overlapped with the shrunken nuclei, which indicated that most of DOX entered the nuclei and exerted an evident cytotoxic effect. Besides, the dynamic intracellular process of free DOX was also investigated and the results are shown in Figure S4. Unlike PEG-CD/AD nanoparticles, free DOX could rapidly enter the nuclei since the red fluorescence signals were obviously observed in the nuclei after 0.5 h. The drug release from free DOX was much faster than that from DOX-loaded nanoparticles (Figure S5), thus, this difference might be attributed to the fact that free DOX can quickly transport into the cells via passive diffusion, while PEG-CD/AD nanoparticles entered the cells through endocytosis and then released DOX into the nuclei, thus showing a lagged effect. According to the above results, we might draw the conclusion that, after PEG-CD/AD nanoparticles were taken up, the disulfide bond could be cleaved by the high GSH environment in the cytoplasm and PEG-CD/AD nanoparticles disassembled to release free DOX (**Figure 4C**).

#### In vitro Cytotoxicity of PEG-CD/AD Nanoparticles against HepG2 Cells

The in vitro cytotoxicity of PEG-CD/AD nanoparticles against HepG2 cells was evaluated by CCK-8 agents. Firstly, we determined the inhibitory effect of PEG-CD on the growth of the cells and no significant cytotoxicity was detected since the cell viabilities were larger than 95% at the concentration range of 5– 50µM (**Figure 5A**). By comparison, PEG-CD/AD nanoparticles exhibited obvious inhibition on cell growth in time- and dosedependent manners (**Figure 5B**). The half maximal inhibitory concentrations (IC<sup>50</sup> values), denoted as the concentration of DOX causing 50% of cell growth, were calculated to be 19.47, 9.31, and 6.64µM for 48, 72, and 96 h, respectively. In addition, PEG-CD/AD nanoparticles showed weaker cytotoxicity in HepG2 cells than that of free DOX (Figure S6). According to other disulfide linked DOX-prodrug reported previously (Song et al., 2016), we deduced that it was because PEG-CD/AD showed a sustained release of DOX and thus exhibited delayed therapeutic effects, while free DOX entered the cell nuclei directly and exerted their therapeutic effects instantly.

### Physicochemical Characterization of PEG-CD/AD/SF Nanoparticles

Sorafenib (SF) was further integrated into the nanosystem via hydrophobic interaction and π-π stacking interaction (Zhao et al., 2015) between DOX and SF. The core of the nanosystem was composed of DOX and SF, while the shell was a layer of PEG-CD. According to a previous study (Zhang et al., 2016), the CI values (combination index determined by IC50) in DOX-based combinations (DOX/SF molar ratio: 1/0.1, 1/0.2, 1/0.5) were much lower than that of SF-based combinations (DOX/SF molar ratio: 0.1/1, 0.2/1), indicated that DOX-based combinations present better cytotoxic effect. In our work, the hydrodynamic

diameters of PEG-CD/AD/SF nanoparticles at different molar ratios were evaluated and the results are displayed in **Table 1**. The average size of PEG-CD/AD/SF nanoparticles increased as the molar ratio of PEG-CD: AD: SF changed from 1/1/0.1 to 1/1/2. Meanwhile, the PDI of PEG-CD/AD/SF nanoparticles tended to be large when more SF was added, indicating that excess SF had a negative effect on the stability of PEG-CD/AD/SF nanoparticles. Taken a comprehensive consideration of CI values and sizes, 1/1/0.2 was believed as the optimal ratio to prepare PEG-CD/AD/SF nanoparticles. As shown in **Figure 6A**, PEG-CD/AD/SF nanoparticles prepared at this ratio had an average size of 186.2 nm with a relatively narrow size distribution (PDI = 0.114). TEM images showed that PEG-CD/AD/SF nanoparticles displayed uniform spherical structure (**Figure 6B**).

The in vitro releases of DOX and SF from PEG-CD/AD/SF nanoparticles were investigated in different media and the release profiles are displayed in **Figures 6C,D**. Obviously, DOX releases


from PEG-CD/AD/SF nanoparticles in the media containing DTT were significantly faster than that in the media without DTT. DOX release amounts reached up to 20.2 and 33.0% in 84 h with the presence of DTT at pH 7.4 and pH 5.0, respectively, whereas were less than 10% in the media without DTT at both pH. We believed that the higher DOX release rates in pH 5.0 PBS with DTT was attributed to the reduction sensitive properties of disulfide bond as well as better DOX solubility in slightly acid media. The same trend was also observed in the releases profiles of SF from PEG-CD/AD/SF nanoparticles in different media. However, the overall release rates of SF were much higher than that of DOX at the same condition. For example, SF release amounts were 76.3 and 92.8% in 84 h with the presence of DTT at pH 7.4 and pH 5.0, respectively. This difference could be attributed to their distinct drug-loading modes. SF was incorporated into the nanoparticles by physical π-π stacking interaction and hydrophobic interaction, while DOX was loaded by chemical bond and thus exhibited sustained release property.

# In vitro Cytotoxicity of PEG-CD/AD/SF Nanoparticles against HepG2 Cells

The cytotoxicity of free SF, PEG-CD/AD+SF physical mixtures and PEG-CD/AD/SF nanoparticles against HepG2 cells were investigated for different time (**Figure 7**). Compared to PEG-CD/AD nanoparticles (**Figure 5B**) and free SF, PEG-CD/AD+SF

FIGURE 6 | Characterization of PEG-CD/AD/SF nanoparticles. Hydrodynamic diameter distribution (A) and TEM image (B) of PEG-CD/AD/SF nanoparticles; *in vitro* releases of DOX (C), and SF (D) from PEG-CD/AD/SF nanoparticles in different release media. Data are means ± standard deviation for three separate experiments.

physical mixtures showed an additive cytotoxicity toward HepG2 cells. PEG-CD/AD/SF nanoparticles exhibited higher cytotoxicity than PEG-CD/AD+SF physical mixtures (p < 0.05) and the IC<sup>50</sup> values of PEG-CD/AD/SF nanoparticles were approximately 1.9-, 2.0-, and 4.7-fold decrease compared to that of PEG-CD/AD+ SF physical mixtures after treating HepG2 cells for 48, 72, and 96 h, respectively. The results suggested that PEG-CD/AD/SF nanoparticles exert stronger inhibitory effect against HepG2 cells by co-loading DOX and SF into the same nanosystem. This might perhaps because the solubility of sorafenib greatly improved after loading into the PEG-CD/AD/SF nanoparticles. Moreover, by comparing the IC<sup>50</sup> values of PEG-CD/AD/SF nanoparticles at different time, we could easily found that the cytotoxicity of PEG-CD/AD/SF nanoparticles increased remarkably as the time increased. This sustained cytotoxicity would be beneficial for treatment of HCC in vivo.

# CONCLUSION

In this study, we successfully synthesized a disulfide linked prodrug of DOX, which could complex with PEG-CD to form supramolecular amphiphiles. PEG-CD/AD amphiphiles could then self-assemble into nanostructures in aqueous solution with a uniform size of 166.4 nm and meanwhile load SF to form PEG-CD/AD/SF nanoparticles with a size of 186.2 nm. PEG-CD/AD nanoparticles possessed reduction-responsive property and could be effectively uptaken by HepG2 cells. The in vitro release of DOX and SF from PEG-CD/AD/SF nanoparticles exhibited reduction-responsive manners. Finally, the in vitro cytotoxicity assay illustrated that PEG/AD/SF nanoparticles showed two- to five-fold increased cytotoxicity toward HepG2 cells than PEG-CD/AD+SF physical mixtures. Altogether, this reduction-responsive nanoparticle system showed great potential for HCC treatment.

# AUTHOR CONTRIBUTIONS

QX and TS: designed the experiments and wrote the manuscript; MC, JW, and GY: carried out the research and analysis of data and contributed to manuscript writing.

#### ACKNOWLEDGMENTS

This work was financially supported by National Natural Science Foundation of China (Grant No. 81501575 and 81602569), Key research Project of Tianjin health industry (Grant No. 14KG142) and The Science & Technology Development Fund of Tianjin Education Commission for Higher Education (Grant No. 2017KJ197).

#### REFERENCES


#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2018.00061/full#supplementary-material

prodrug as potent chemotherapeutic nanomedicine. Small 12, 6353–6362. doi: 10.1002/smll.201601597


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer YY and handling Editor declared their shared affiliation.

Copyright © 2018 Xiong, Cui, Yu, Wang and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-09-00091 February 9, 2018 Time: 18:26 # 1

# Nanoliposome-Encapsulated Brinzolamide-hydropropylβ-cyclodextrin Inclusion Complex: A Potential Therapeutic Ocular Drug-Delivery System

Fazhan Wang<sup>1</sup>† , Xingting Bao<sup>1</sup>† , Aiping Fang1,2† , Huili Li<sup>1</sup> , Yang Zhou<sup>1</sup> , Yongmei Liu<sup>1</sup> , Chunling Jiang<sup>1</sup> , Jinhui Wu<sup>1</sup> \* and Xiangrong Song<sup>1</sup> \*

<sup>1</sup> State Key Laboratory of Biotherapy, Geriatrics and Cancer Center, West China Hospital and Collaborative Innovation Center for Biotherapy, Sichuan University, Chengdu, China, <sup>2</sup> West China School of Public Health, Sichuan University,

#### Edited by:

Chengdu, China

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Yanqi Ye, The University of North Carolina at Chapel Hill, United States Tianjiao Ji, Boston Children's Hospital, Harvard University, United States

#### \*Correspondence:

Xiangrong Song songxr@scu.edu.cn Jinhui Wu wujinhui@scu.edu.cn

†These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 30 December 2017 Accepted: 26 January 2018 Published: 13 February 2018

#### Citation:

Wang F, Bao X, Fang A, Li H, Zhou Y, Liu Y, Jiang C, Wu J and Song X (2018) Nanoliposome-Encapsulated Brinzolamide-hydropropylβ-cyclodextrin Inclusion Complex: A Potential Therapeutic Ocular Drug-Delivery System. Front. Pharmacol. 9:91. doi: 10.3389/fphar.2018.00091 Novel ocular drug delivery systems (NODDSs) remain to be explored to overcome the anatomical and physiological barriers of the eyes. This study was to encapsulate brinzolamide (BRZ)-hydropropyl-β-cyclodextrin (HP-β-CD) inclusion complex (HPβ-CD/BRZ) into nanoliposomes and investigate its potential as one of NODDS to improve BRZ local glaucomatous therapeutic effect. HP-β-CD/BRZ was firstly prepared to enhance the solubility of poorly water-soluble BRZ. The HP-β-CD/BRZ loaded nanoliposomes (BCL) were subsequently constructed by thin-film dispersion method. After the optimization of the ratio of BRZ to HP-β-CD, the optimal BCL showed an average size of 82.29 ± 6.20 nm, ζ potential of −3.57 ± 0.46 mV and entrapment efficiency (EE) of 92.50 ± 2.10% with nearly spherical in shape. The X-ray diffraction (XRD) confirmed the formation of HP-β-CD/BRZ and BCL. The in vitro release study of BCL was evaluated using the dialysis technique, and BCL showed moderate sustained release. BCL (1 mg/mL BRZ) showed a 9.36-fold increase in the apparent permeability coefficient and had a sustained and enhanced intraocular pressure reduction efficacy when compared with the commercially available formulation (BRZ-Sus) (10 mg/mL BRZ). In conclusion, BCL might have a promising future as a NODDS for glaucoma treatment.

Keywords: brinzolamide, cyclodextrin, inclusion complex, nanoliposomes, intraocular pressure, ocular drug delivery

#### INTRODUCTION

Despite easy accessibility for topical drug administration, overcoming the anatomical and physiological barriers of the eye remains one of the greatest challenges for ocular drug delivery (Bucolo et al., 2012; Zhang et al., 2015). Drug retention is impeded by tear reflex, blinking, and nasolacrimal drainage (Morrison et al., 2013). Cornea barrier protects the eye from the passage of any foreign molecules including drugs into the eye, and thus only a small fraction of the topically applied drug penetrates the cornea and reaches intraocular tissues (Mun et al., 2014). As a result, repeated dosing is needed, which increases the patient discomfort and other adverse effects (Luo et al., 2011). A system which behaves like a solution and at the same time can lead to retention of drug in the eye and increase corneal permeability is of great urgency.

fphar-09-00091 February 9, 2018 Time: 18:26 # 2

Nanoliposomes, as biodegradable, biocompatible and nonimmunogenic drug delivery system, have been widely applied (van Rooijen and van Nieuwmegen, 1980; Mugabe et al., 2006; Arias, 2013; Sercombe et al., 2015; Ji et al., 2017). They can encapsulate hydrophilic drugs into the inner aqueous phase, whereas entrap hydrophobic drugs into the hydrophobic lipid phase (Allen and Cullis, 2004; Chen et al., 2009). However, it is difficult to obtain high drug-loading capacity in the lipid phase because of the limited space offered by the lipid phase, and a large amount of hydrophobic drugs can destabilize the structural integrity of the liposomal layers (Song et al., 2012). Inclusion complexes are one of the most useful strategies in the pharmaceutical field, which can solubilize a wide range of hydrophobic drugs (Morrison et al., 2013; Zhang et al., 2013; Alomrani et al., 2014; Ji et al., 2016; Maria et al., 2017; Soliman et al., 2017). The binding force between the guest drug molecule and the cyclodextrin host molecule is usually weak, possibly leading to rapid dissociation of the inclusion complexes due to dilution by plasma or extracellular fluids (Chen et al., 2014). Encapsulation of hydrophobic drugs in the form of inclusion complexes into nanoliposomes has been investigated as a new strategy to combine the advantages of inclusion complexes and nanoliposomes, namely the drug-incyclodextrin-in-nanoliposome (DCL) system (Piel et al., 2006; Hatzi et al., 2007; Zhang et al., 2015). The presence of cyclodextrin in the aqueous compartment of nanoliposomes would not affect the characteristics of conventional liposomes but prolong drug release compared to conventional liposomes (Maestrelli et al., 2010; Zhang et al., 2015). Although a number of studies on DCL systems have been reported (Dhule et al., 2012; Rahman et al., 2012; Lapenda et al., 2013; Alomrani et al., 2014), to date, the therapeutic application of DCL systems as an ocular delivery system remains to be explored.

Nanoliposome systems are applicable in ocular delivery, which have the potential of increasing corneal permeability and improving retention time (de Sa et al., 2015; Li et al., 2016). Moreover, some functional liposomes have been constructed (Eriksen et al., 2017; Huang et al., 2017; Tan et al., 2017). Cyclodextrin, with a hydrophilic outer surface and a cavity at its center and low renal toxicity, has the ability to form hydrophilic inclusion complexes by a molecular complexation with a lot of hydrophobic drugs (Ji et al., 2016; Maria et al., 2016, 2017; Abd El-Gawad et al., 2017). Hydropropyl-β-cyclodextrin (HP-β-CD), a type of cyclodextrin, is commonly used for ocular drug delivery and has been approved by the FDA (Bozkir et al., 2012). Many in vitro and in vivo studies also showed that the inclusion complex of poor water-soluble drugs by HP-β-CD could increase drug solubility, corneal permeability, and ophthalmic bioavailability (Ito et al., 2013; Morrison et al., 2013). Encouraged by these previous results, we postulated that DCL might have the potential to enhance ocular permeation of the hydrophobic model drug brinzolamide (BRZ) and enhance its therapeutic efficacy for glaucoma.

Previously, we have used inclusion complex and nanoliposomes to encapsulate the low water-soluble drug BRZ, respectively. Both formulations improved the corneal permeation and therapeutic efficacy of BRZ to some extent (Zhang et al., 2013; Li et al., 2016). Here, we further combined HP-β-CD and nanoliposomes to prepare BRZ-HP-β-CD inclusion complex (HP-β-CD/BRZ)-loaded nanoliposomes (BCL) as shown in **Figure 1**. The pharmaceutical properties including UV/vis spectra, particle size, surface charge, entrapment efficiency (EE) of BRZ, X-ray diffraction (XRD), and in vitro release profile were then characterized systemically. The ex vivo corneal permeability, in vivo intraocular pressure (IOP) reduction efficiency and preliminary safety of BCL were finally evaluated.

# MATERIALS AND METHODS

#### Materials

BRZ (purity > 99%) was obtained from Dalian Meilun Biology Technology Company (Dalian, China). HP-β-CD (purity > 99%, DS = 7) was supplied by Xi'an Deli Chemicals Corporation (Xi'an, China). Soybean phosphatidylcholine (SPC) was provided by Shanghai A.V.T. Pharmaceutical, Co., Ltd. (Shanghai, China). Cholesterol was purchased from Shanghai Yuanju Biology Technology Company (Shanghai, China). The commercial formulation AZOPT <sup>R</sup> (BRZ-Sus) was supplied by Alcon Laboratories (Puurs, Belgium). All the other reagents were of analytical grade unless otherwise stated.

#### Animals

White New Zealand rabbits (male, 2.0–2.5 kg) were used for each formulation (n = 6). Animals were housed individually and allowed free access to food and water with a 12:12 h cyclic lighting schedule. After a week of adaptation, animals were admitted to experiments. All animal experiments were approved and supervised by the State Key Laboratory of Biotherapy Animal Care and Use Committee (Sichuan University, Chengdu, Sichuan, China).

# Preparation of H-β-CD/BRZ Inclusion Complex

According to previously described methods by Zhang et al. (2013), the inclusion complex was prepared with some

modifications. Briefly, BRZ and HP-β-CD at different molar ratio (2:1, 1:1, 1:2, and 1:3) were dissolved by ethanol in a roundbottomed flask. After 1 h of ultrasound at 37◦C, the ethanol was removed using a rotary evaporator in vacuum. To dissolve the inclusion complex, PBS was added to the flask followed by filtering through a 0.22 µm membrane. The content of BRZ in the inclusion complex was analyzed by high-performance liquid chromatography (HPLC, Waters, Milford, MA, United States) at 252 nm. The mobile phase, at 1 mL/min flow rate, was composed of acetonitrile and water at a volume ratio of 40:60.

#### Preparation of BCL

fphar-09-00091 February 9, 2018 Time: 18:26 # 3

BCL was prepared using a modified thin-film dispersion method (Lapenda et al., 2013; Li et al., 2016). Briefly, SPC and cholesterol were dissolved in a mixture solvent of chloroform/methanol (4:1, v/v). The organic solvents were evaporated using a rotary evaporator at 37◦C, and then the formed thin film was further dried under high vacuum for 1 h. The lipid film was hydrated with HP-β-CD/BRZ inclusion complexes at 60◦C for 1 h to obtain a suspension. To obtain BCL, the above suspension was sonicated for 3 min at 100 W in an ice bath and filtered through a 0.22-µm membrane filter.

The BRZ-loaded nanoliposomes (LP/BRZ) were obtained using the similar procedure with the addition of BRZ into the chloroform/methanol (4:1 v/v) solvent mixture and hydrating lipid film with sterile PBS. The blank liposomes (Blank LP) were also prepared using the similar preparation process of LP/BRZ without adding BRZ.

# Characterization of BCL

#### Ultraviolet-Visible (UV/vis) Spectroscopy

The UV/vis absorption spectra of BRZ, HP-β-CD, Blank LP, inclusion complex, LP/BRZ, and BCL were recorded using a SHIMADZU UV-2600 UV/vis spectrophotometer in the range of 200–600 nm.

#### Particle Size and ζ Potential

The average particle size, size distribution [polydispersity index (PDI)], and ζ potential of diluted BCL were recorded by Zetasizer Nano ZS90 (Malvern Instruments, Malvern, United Kingdom). All measurements were carried out at 25◦C after 3 min of equilibration and were conducted in triplicate. All the data were presented as mean ± standard deviation (SD).

#### Morphology

The morphological examination of BCL was performed by transmission electron microscopy (TEM, H-600, Hitachi, Japan). Briefly, a drop of liposomal suspension was placed onto copper electron microscopy grids, and then they were negatively stained with 2% phosphotungstic acid for observation at an acceleration voltage of 100 kV.

#### Drug Entrapment Efficiency

According to previously described methods (Rahman et al., 2012), the drug content of BCL was determined by separating BCL from the unentrapped drug using cooling centrifugation at 50,000 rpm for 30 min at a temperature of 4◦C. Following the removal of the supernatant, 1 ml of acetonitrile was added into the BCL sediment and sonicated for 20 min to extract BRZ from BCL. After centrifugation at 13,000 rpm for 10 min, the content of BRZ in the supernatant diluted with acetonitrile was quantified by HPLC as described previously. Analysis was performed in triplicate and the values were expressed as mean ± SD. EE was calculated by the following calculation equation:

EE = (amount of BRZ in BCL/initial BRZ amount) × 100%.

#### Stability of BCL

The optimized formulation was stored at 4◦C for 2 weeks in order to investigate the preliminary stability. In the test, the physicochemical characteristics such as particle size and EE were recorded and changes over time were evaluated.

#### XRD Analysis

X-ray diffraction was performed to investigate the crystal structure according to the former study (Tao et al., 2017). Different samples (pure BRZ, HP-β-CD, HP-β-CD/BRZ, and BCL) were examined using an X-ray diffractometer (X'Pert Pro Philips, Netherlands) at a voltage of 40 kV and a current of 40 mA. The scans were carried out at a scanning rate of 10◦C/min in the 2θ range from 0 to 50◦C.

#### In Vitro BRZ Release from BCL

In vitro release studies were carried out using the dynamic dialysis technique in sink conditions. Briefly, 2 mL of different formulations containing the drug were placed into different dialysis bags (molecular weight cut-off = 3,500 Da) and immersed in 40 mL of simulated tear fluid (STF) and continuously agitated in an orbital shaker maintained at 37◦C. At given time intervals, 1 mL of receptor phase was removed and replaced with an equal volume of STF. The drug concentration was quantified by the HPLC method as described above. All data were presented as mean ± SD of three independent measurements.

#### Corneal Permeability

The transcorneal permeability of BCL was evaluated using a Franz diffusion chamber which consisted of a donor and a receiver compartment (with a volume of 1.0 and 2.0 mL, respectively). The diffusion chamber was maintained at a constant temperature (37 ± 0.2◦C) by a thermostatic water bath, under mixing conditions using a magnetic stirrer. The cornea together with a 2-mm ring of sclera was excised immediately after the rabbits (n = 3) were sacrificed and stored in glutathione bicarbonate ringer (GBR) buffer after rinsing with cold saline before use. One mL of the GBR solution was added to the endothelial (receptor) side of the cornea, while 0.5 mL of different formulations was added to the epithelial (donor) side of the cornea. Different samples from the receptor chamber were withdrawn at 0, 0.5, 1, 1.5, 2, 3, 4, 5 and 6 h and replaced with fresh GBR buffer-receptor solution. The content of drug permeation across the cornea was determined by the HPLC method as described above. The experiment was done in triplicate.

The apparent permeation coefficient (Papp, cm/s) of each formulation was calculated using the following equation fphar-09-00091 February 9, 2018 Time: 18:26 # 4

$$\text{P}\_{\text{app}} = (1/\text{AC}\_0)^\* \text{ (dM/dt)}$$

where A is the available corneal surface area for diffusion (cm<sup>2</sup> ), C<sup>0</sup> is the initial drug concentration (µg/mL) in the donor compartment, and dM/dt is the flux across the cornea (µg/cm2h).

#### Corneal Hydration Level

According to the literature (Vega et al., 2008), the corneal hydration level (HL) was evaluated by measuring total water content by the gravimetric method. At the end of the corneal permeability study (n = 3), the wet weight of each cornea (W1) and dry corneal weight (W2) after a desiccation at 100◦C for 6 h were obtained, respectively. The corneal HL was calculated using the following equation: HL% = (W1-W2)/W<sup>1</sup> × 100%.

#### In Vivo IOP Measurement

IOP was recorded in mmHg with a Tono-pen XL <sup>R</sup> tonometer (Reichert, NY, United States) calibrated according to the manufacturer's instructions. IOP was recorded until the animals (n = 6) were accustomed to the experimental procedure. Each formulation (50 µL) was instilled topically into the right eye, while the left eye received no treatment to minimize the diurnal and individual variations. The IOP was measured at 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 12 and 24 h after instilled.

#### Statistical Analysis

The data were expressed as mean ± SD. Statistical significance was assessed by a two-way analysis of variance using SPSS 16.0 (SPSS, Inc., Chicago, IL, United States). The p-values < 0.05 were considered statistically significant difference.

#### RESULTS

#### UV/vis Spectroscopy

The influence of the BCL on the absorption of BRZ was evaluated by UV/vis spectroscopy. The absorption spectra of BRZ, HP-β-CD, Blank LP, HP-β-CD/BRZ inclusion complex, LP/BRZ and BCL in acetonitrile were determined (**Figure 2**). The spectrum of BCL exhibited a characteristics absorption peak at 252 nm, which was similar to the peak of BRZ. These results suggested that the HPLC method developed for BRZ measurement with a detection wavelength of 252 nm could also be applied to determine the BRZ concentration in BCL.

#### Preparation of BCL

Different ratios of BRZ and HP-β-CD were used to establish the optimal preparation conditions in which BCL are formed. The drug content of BCL ranged between 84.38 and 92.50% and the size of BCL ranged between 82.29 ± 6.20 nm and 90.62 ± 11.30 nm depending on the BRZ/ HP-β-CD molar ratio (**Figure 3A**).

# Characterization of BCL

#### Particle Size, ζ Potential, and EE

The physicochemical properties of optimal BCL were evaluated. The particle size of the optimal BCL was 82.29 ± 6.20 nm with a narrow size distribution (PDI = 0.21 ± 0.01) (**Figure 3C**). The ζ potential of BCL was −3.57 ± 0.46 mV (**Figure 3D**). The appearance and morphological studies of BCL were also conducted. LP/BRZ and BCL colloidal solution were observed as a slightly blue opalescence with obvious Tyndall effect compared with water (**Figures 3E,F**). BCL were seen to be distinct lipid membrane structure with a particle size diameter of about 70 nm (**Figure 3G**).

#### Storage Stability

The particle size and EE of BCL were determined at a predetermined time of storage at 4◦C. BCL displayed good stability with no detectable changes in particle size and EE for at least 3 days (**Figure 3B**).

#### XRD Analysis

X-ray diffractograms were obtained from different samples of BRZ, HP-β-CD, HP-β-CD/BRZ and BCL to evaluate the crystal structure and entrapment of BRZ in BCL. The X-ray spectrum of BRZ showed several sharp and narrow peaks between 5 and 50◦C (2θ) with a maximal peak at 2θ = 24.97◦C, indicating their crystalline nature. HP-β-CD displayed a specific broad peak at 2θ = 20◦C in its X-ray diffractogram. Neither HP-β-CD/BRZ nor BCL showed the peak at 2θ = 24.97◦C (**Figure 4**).

#### In Vitro Release Profile

The in vitro release experiments were performed to investigate the successful inclusion and the sustained release characteristic of BCL. A sustained release phase was observed in BCL within a period of 9 h (1–10 h) as presented in **Figure 5**. BRZ release from the inclusion complex was slightly slower than free drug but faster than LP/BRZ. Moreover, BCL displayed a slower release compared to LP/BRZ.

#### Corneal Permeation and Corneal Hydration

The corneal penetration study was performed to evaluate the effect of BCL on transcorneal transportation. The corneal permeation characteristics of BRZ-Sus (10 mg/mL), LP/BRZ (1 mg/mL), and BCL (1 mg/mL) were shown in **Figure 6**. There was no significant difference in the cumulative permeation amount at 0–2 h among the three groups. At 3–6 h, BRZ-Sus and BCL showed a significant increase of BRZ cumulative permeation compared with LP/BRZ. A linear relationship was observed for all preparations over the time period from 0.5 to 6 h, indicating that the corneal integrity was maintained throughout the experiment. Moreover, the permeation coefficient of BCL was much larger than those of BRZ-Sus and LP/BRZ (**Table 1**, p < 0.001). BCL showed a 9.36- and 0.40-fold increase in Papp when compared with Brz-Sus and LP/BRZ, respectively.

Corneal HL was used to evaluate the safety of BCL (Vega et al., 2008). Its normal value is often from 75 to 80% (Vega et al., 2008). The corneal HL of BRZ-Sus, LP/BRZ, and BCL were fphar-09-00091 February 9, 2018 Time: 18:26 # 5

FIGURE 3 | Effects of various HP-β-CD/BRZ ratios on diameter and drug entrapment efficiency of BCL (A) (n = 3). Change in size and entrapment efficiency of BCL at various time points stored at 4◦C (B) (n = 3). Size distribution (C), ζ potential (D), appearance (E), and tyndall effect (F) of BCL. TEM images of BCL (G).

80.12 ± 2.43%, 79.37 ± 2.19%, and 78.56 ± 3.04%, respectively. No statistical difference in HL % was found (p > 0.05), indicating BCL might be safe for topical ocular use.

#### In Vivo IOP-Lowering Effect

fphar-09-00091 February 9, 2018 Time: 18:26 # 6

As shown in **Figure 7**, BCL (1 mg/mL BRZ) presented more effective IOP reduction with a longer term role than BRZ-Sus (10 mg/mL Brz) and LP/BRZ (1 mg/mL Brz). The novel formulation of BRZ had an onset of less than 1 h, achieved a maximum IOP reduction of 32.3% at 2 h, and sustained an effective IOP reduction until the 12th hour, while Brz-Sus began to result in an effective IOP reduction at 0.5 h and quickly reached its peak effect at 1 h (an average reduction of 16.12% from baseline). Furthermore, IOP was significantly lower in the BCL group at any time point from 2 to 12 h than Brz-Sus group. Even though the dosage of BRZ was just 10% in BCL compared with BRZ-Sus, the IOP reduction efficacy was much higher.

#### DISCUSSION

In this work, we aimed to construct a novel ocular delivery system DCL loaded with BRZ (BCL) and investigate its potential to improve BRZ local glaucomatous therapeutic effect. BCL was prepared by hydrating lipid film with inclusion complex (Lapenda et al., 2013), and HP-β-CD was selected to enhance the solubility of poorly soluble drug BRZ (Zhang et al., 2013). According to the literature, the content of HP-β-CD was critical for the physicochemical properties of DCL (Chen et al., 2014; Cutrignelli et al., 2014). Cyclodextrins would interact with the lipid membrane and influence the formation of DCL (Cutrignelli et al., 2014). Thus, the amount of HP-β-CD was optimized based on the particle size and EE of BCL in this study. As seen in **Figure 3A**, EE first increased and then reached a plateau with the decrease of the BRZ/HP-β-CD molar ratio while the average diameter of BCL first decreased and then increased. Further increase in the HP-β-CD content (when the BRZ/HP-β-CD molar ratio was set at 1:3) cannot correspondingly enhance the BRZ entrapment into BCL but reduced EE. The cholesterol in the lipid bilayer of nanoliposomes might enter the excessive HP-β-CD cavity (Morrison et al., 2013; Chen et al., 2014), thereby forming unstable nanoliposomes which cannot incorporate enough HPβ-CD/BRZ-inclusion complex. The highest EE and the smallest particle size of BCL were achieved when BRZ and HP-β-CD was fed at a 1:1 molar ratio, and thus BCL containing 1:1 BRZ/H-β-CD were chosen for the subsequent investigations.

The optimal BCL were nano-sized (less than 100 nm) with negative surface charge probably resulting from the phosphatidylcholine head group in the outer nanoliposomes surface (Ascenso et al., 2013). Interestingly, BCL colloidal suspension had a similar appearance, average size, and surface charge to LP/BRZ despite the introduction of HP-β-CD (Maestrelli et al., 2010). This phenomenon indicated that HPβ-CD might mainly distribute in the inner aqueous core of the nanoliposomes, which was consistent with the literature (Cavalcanti et al., 2011; Lapenda et al., 2013; Chen et al., 2014). Moreover, it was clear that the peaks of BRZ disappeared from

fphar-09-00091 February 9, 2018 Time: 18:26 # 7

TABLE 1 | Corneal hydration level and Papp of different formulations containing brinzolamide. Data are expressed as mean ± SD (n = 3).


the diffraction pattern of BCL, demonstrating that BRZ was completely and successfully encapsulated into BCL.

BCL could be stable for about 7 days at 4◦C, which might benefit from the negatively charged surface. However, EE sharply declined when BCL were kept for 15 days. The cholesterol in BCL might go into the hydrophobic cavity of HP-β-CD to compete with BRZ (Lapenda et al., 2013; Morrison et al., 2013), hence partially contributing to partial leakage of BRZ from the inclusion complex and the instability of nanoliposomes. HP-β-CD has previously been reported to extract cholesterol from cell membranes (Morrison et al., 2013) and decrease the integrity of liposomes composed of cholesterol and saturated phospholipids (Hatzi et al., 2007). The phenomenon found in this study was in line with these former studies. The free BRZ would easily pass through the lipid bilayer of the unstable BCL, which led to the reduced EE after relatively long-term storage. Generally, this kind of destructive role of HP-β-CD on DCL basically depends on the HP-β-CD affinity balance between the guest drug molecules and lipid components of nanoliposomes (Piel et al., 2006; Chen et al., 2014). According to the preliminary stability results, BCL would be further developed to be a freezedrying formulation for long-term storage.

BCL performed a slower release than HP-β-CD/BRZ inclusion complex and LP/BRZ. It seems that the slower release profile of BCL was partially attributed to the fact that there were more barriers to the diffusion of BRZ from the BCL than LP/BRZ. As for BCL, two definitely different routes may account for the BRZ release (Chen et al., 2014). The whole HP-β-CD/BRZ inclusion complex might release from the destructive BCL or transport to lipid phase and then release. For the other way, BRZ might release from the dissociated HP-β-CD/BRZ inclusion complex inside BCL followed by diffusion out of the nanoliposomes and dispersion through the dialysis bag (Chen et al., 2014). To date, there still remains a lot of work to do to clarify the relative contributions of the two definitely different routes in release of BCL.

BCL exhibited enhanced and sustained IOP reduction effect as shown in **Figure 7**. This promising efficacy of BCL on glaucoma treatment may be mainly explained by the combinatorial advantages of BCL with HP-β-CD and nanoliposomes. Neutral liposome with a size of about 100 nm were reported to be safe for ocular applications without ocular irritability (Taniguchi et al., 1988). Moreover, particle size < 10 µm was recommended and considered optimal for ophthalmic preparation to minimize eye irritation (Kassem et al., 2007; Yuan et al., 2009). Accordingly, BCL colloidal solution with small size rather than BRZ-Sus might attenuate ocular irritation, thereby reducing the loss of BRZ by tear flushing. In addition, BCL with higher corneal binding affinity (Natarajan et al., 2012) can increase the retention on the surface of cornea, probably extending the absorption of BRZ. BCL displayed significant improvement in corneal permeation in contrast to the conventionally prepared LP/BRZ. This was thought to be the cause of the more efficient IOP lowering effect of BCL. Moreover, the intact HP-β-CD/BRZ inclusion complex

released from BCL might lead to the enhanced penetration according to our previous study (Zhang et al., 2013). Of note, the sustained release profile of BCL as shown in the in vitro release study might be linked to the prolonged therapeutic efficacy of BCL.

#### CONCLUSION

fphar-09-00091 February 9, 2018 Time: 18:26 # 8

In this study, a novel drug-in-cyclodextrin-in-nanoliposome system containing BRZ was successfully prepared and its potential application in glaucomatous treatment was investigated. The optimized ratio of BRZ and HP-β-CD was set at 1:1 to prepare BCL with small size and high drug content. BCL incorporated BRZ very well and showed a sustained release profile and enhanced corneal permeation. Moreover, this new formulation of BRZ achieved an improved and prolonged effect of IOP reduction with similar safety to the commercially available formulation on white New Zealand rabbits. However, further studies are still required to investigate the in vitro release properties of BCL and elucidate how the release profile affects the therapeutic efficacy of BCL. In sum, all the data indicated that

# REFERENCES


DCL might be one of the potential ocular delivery systems for hydrophobic drugs and BCL were worthy of further investigation as a novel anti-glaucoma formulation candidate of BRZ.

#### AUTHOR CONTRIBUTIONS

XS conceived of the study. JW designed the experiments. XS and FW wrote the manuscript. FW and XB conducted most of the experiments. AF performed the statistical analysis and drafted the manuscript. HL and YZ performed all the preliminary experiments. YL and CJ participated in literature research and manuscript editing. All authors reviewed and approved the manuscript.

#### ACKNOWLEDGMENTS

This work was financially supported by the Sichuan Province Science and Technology Support Program (Grant Nos. 2015SZ0234 and 2017GZ0413) and the National Natural Science Foundation of China (Grant No. 81600006).


fphar-09-00091 February 9, 2018 Time: 18:26 # 9

enhancement of the antitumor activity. J. Biomed. Nanotechnol. 9, 499–510. doi: 10.1166/jbn.2013.1554


stability, anti-inflammatory efficacy, and ocular bioavailability of loteprednol etabonate. AAPS PharmSciTech 18, 1228–1241. doi: 10.1208/s12249-016- 0589-9


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Wang, Bao, Fang, Li, Zhou, Liu, Jiang, Wu and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# DACHPt-Loaded Nanoparticles Self-assembled from Biodegradable Dendritic Copolymer Polyglutamic Acid-b-D-α-Tocopheryl Polyethylene Glycol 1000 Succinate for Multidrug Resistant Lung Cancer Therapy

#### Edited by:

*Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China*

#### Reviewed by:

*Cui Cheng, Fuzhou University, China Songwei Tan, Huazhong University of Science and Technology, China Haihua Xiao, Institute of Chemistry (CAS), China*

#### \*Correspondence:

*Zihuang Li lizihuang2006@126.com Laiqiang Huang huanglq@sz.tsinghua.edu.cn Gan Liu liugan5@mail.sysu.edu.cn †Co-first authors.*

#### Specialty section:

*This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology*

> Received: *17 December 2017* Accepted: *01 February 2018* Published: *21 February 2018*

#### Citation:

*Tsai H-I, Jiang L, Zeng X, Chen H, Li Z, Cheng W, Zhang J, Pan J, Wan D, Gao L, Xie Z, Huang L, Mei L and Liu G (2018) DACHPt-Loaded Nanoparticles Self-assembled from Biodegradable Dendritic Copolymer Polyglutamic Acid-b-D-*α*-Tocopheryl Polyethylene Glycol 1000 Succinate for Multidrug Resistant Lung Cancer Therapy. Front. Pharmacol. 9:119. doi: 10.3389/fphar.2018.00119* Hsiang-I Tsai 1†, Lijuan Jiang1†, Xiaowei Zeng<sup>2</sup> , Hongbo Chen<sup>2</sup> , Zihuang Li <sup>3</sup> \*, Wei Cheng<sup>4</sup> , Jinxie Zhang<sup>1</sup> , Jie Pan<sup>5</sup> , Dong Wan<sup>5</sup> , Li Gao<sup>6</sup> , Zhenhua Xie<sup>4</sup> , Laiqiang Huang1,4 \*, Lin Mei <sup>2</sup> and Gan Liu<sup>2</sup> \*

*<sup>1</sup> School of Life Sciences, Tsinghua University, Beijing, China, <sup>2</sup> School of Pharmaceutical Sciences (Shenzhen), Sun Yat-sen University, Guangzhou, China, <sup>3</sup> Department of Radiation Oncology, Second Clinical Medicine College of Jinan University, Shenzhen Municipal People's Hospital, Shenzhen, China, <sup>4</sup> Division of Life and Health Sciences, Graduate School at Shenzhen, Tsinghua University, Shenzhen, China, <sup>5</sup> School of Environmental and Chemical Engineering, Tianjin Polytechnic University, Tianjin, China, <sup>6</sup> Department of Urology, Affiliated Hospital of Guilin Medical University, Guilin, China*

The clinical applications of platinum-based antitumor agents are still largely limited by severe side effects as well as multidrug resistance (MDR). To solve these problems, we developed an 1,2-diaminocyclohexane-platinum(II) (DACHPt)-loaded nanoparticle (NP-TPGS-Pt) by self-assembly of poly(amidoamine)-polyglutamic acid-*b*-D-α-tocopheryl polyethylene glycol 1000 succinate (PAM-PGlu-*b*-TPGS) and DACHPt. NP-TPGS-Pt showed robust stability and pH-responsive DACHPt release profile *in vitro* similar to the PEG-containing nanoparticle (NP-PEG-Pt). Meanwhile, in contrast with NP-PEG-Pt, NP-TPGS-Pt exhibited efficient nanoparticle-based cellular uptake by the Pt-resistant A549/DDP human lung cancer cells and caused much more cytotoxicity than free Oxaliplatin and NP-PEG-Pt. Finally, this NP-TPGS-Pt was proved to perform outstanding inhibition of Pt-resistant tumor growth, much superior than free Oxaliplatin and NP-PEG-Pt. Thus, this NP-TPGS-Pt provides a novel powerful nanomedicine platform for combatting multidrug resistant cancer.

Keywords: multidrug resistance, TPGS, dendritic copolymers, nanoparticles, DACHPt

# INTRODUCTION

Up to now, platinum-based antitumor agents have been widely used to treat lung cancer, bladder cancer, gastric cancer, and ovarian cancer, etc. (Graham et al., 2000; Klein and Hambley, 2009; Wheate et al., 2010; Han and Smith, 2013; Harrach and Ciarimboli, 2015; Torre et al., 2015; Ma et al., 2017). However, their clinical outcomes are still largely limited by severe side effects (Nishiyama et al., 2001; Xiao et al., 2017) and multidrug resistance (Wang and Lippard, 2005; Li et al., 2015; Yin et al., 2017). Drug nanocarriers, which can prolong the in vivo half-life of drugs and promote the enrichment in solid tumors through the enhanced permeation and retention (EPR) effect, have been verified potentially applicable to antitumor therapy (Peer et al., 2007; Oberoi et al., 2013; Salomone, 2013; Wang and Guo, 2013; Huang et al., 2015; Xu et al., 2015; Li et al., 2017; Liu et al., 2017b). We have previously reported an unimolecular micelle (UM/DACHPt) prepared by loading antitumor agent 1,2-diaminocyclohexane-platinum(II) (DACHPt) with dendritic block copolymer PAM-PGlu-b-PEG (Liu et al., 2017a).Compared to micelles self-assembled from block copolymer, this micelle showed superior stability, thereby extending the in vivo half-life, enhancing antitumor effects and reducing side effects. Nevertheless, besides relieving side effects, platinum drug-loaded nanocarriers always need to meet the requirement for overcoming tumor multidrug resistance. Since the antitumor effects of unimolecular micelles are inferior to those of free drugs in vitro, they would hardly fulfill the requirement. Therefore, it is of great significance to optimize the nanocarrier structure to be both stable in vivo and multidrugresistant.

As a soluble derivative of vitamin E, D-α-tocopheryl polyethylene glycol 1000 succinate (TPGS) is esterified from the acid group of vitamin E succinate and polyethylene glycol (PEG) 1000. It has been approved by FDA as a safe pharmaceutical excipient (Mu and Feng, 2003; Zhang et al., 2012; Guo et al., 2013; Tan et al., 2017). It is well-documented that TPGS could enhance cellular uptake (Zhang and Feng, 2006; Zeng et al., 2013), and inhibit P-glycoprotein to circumvent drug resistance via interfering with the structure and function of mitochondria (Dintaman and Silverman, 1999; Zhu et al., 2014; Wang et al., 2015; Bao et al., 2017). In addition, our previous work has reported that surface modification of PLGA nanoparticles (NPs) with TPGS prolonged the half-life of drugs in vivo and facilitated their cellular uptake (Zeng et al., 2013). Thus we ensure that conjugating TPGS with PAM-PGlu rather than PEG would not only maintain the stability of nanocarrier, but also enhance the cellular uptake and overcome drug resistance. As far as we know, no such PAM-PGlu-b-TPGS was reported to prepare DACHPt-loaded nanoparticles.

Thus, we designed a novel TPGS-containing dendritic polymer PAM-PGlu-b-TPGS to prepare DACHPt-loaded NP-TPGS-Pt. PAM-PGlu-b-TPGS consisted of dendritic molecule PAMAM-G3, DACHPt-chelating agent PGlu and polymer TPGS (**Figure 1**). We then determined the size, zeta potential, drug loading content, encapsulation efficiency, in vitro stability and drug release behaviors of NP-TPGS-Pt. Meanwhile, the cellular uptake and in vitro cytotoxicity of NP-TPGS-Pt were evaluated by using non-small cell lung cancer cell line A549 and resistant A549/DDP cell line. Finally, the antitumor effects of NP-TPGS-Pt on the A549/DDP model were assessed.

#### EXPERIMENTAL SECTION

#### Materials

Carboxyl-terminated D-α-tocopheryl polyethylene glycol 1000 succinate (cTPGS) was synthesized in our previous work [21]. BLG-NCA was purchased from Kangmanlin chemicals (Nanjing, China). PAMAM-NH2-G3 (PAM-NH2, M<sup>w</sup> = 6900 Da), CF3COOH, HBr/HAc (w/V 33%), DACHPtCl<sup>2</sup> and Oxaliplatin were purchased from Aladdin Industrial (Shanghai, China). All the chemicals were commercially available and used as received. A549 and A549/DDP cells were obtained from American Type Culture Collection (ATCC). Fetal bovine serum (FBS), RPMI1640 media and penicillin/streptomycin were both purchased from Gibco. Female Balb/C nude mice (∼18 g, 7 weeks old) were purchased from Guangdong Province Medical Animal Center and fed in a SPF (specific pathogen free) class experimental animal room. This study was carried out in accordance with the recommendations of the Care and Use of laboratory animals of Tsinghua University, the Administrative Committee on Animal Research in Tsinghua University. The protocol was approved by the Administrative Committee on Animal Research in Tsinghua University.

#### Synthesis of Dendritic Block Copolymer PAM-PGlu-b-TPGS

As shown in **Figure 2**, firstly, PAM-PBLG384-NH<sup>2</sup> was synthesized according to our previous work (Liu et al., 2017a). Then PAM-PBLG-NH<sup>2</sup> (1 g) was reacted with excess TPGS-COOH and EDC at 4◦C for 2 h to obtain the dendritic block copolymer PAM-PBLG-b-TPGS. The degree of polymerization of TPGS in PAM-PBLG-NH<sup>2</sup> was verified by <sup>1</sup>H-NMR spectroscopy (Varian UNITY-plus 400 M nuclear magnetic resonance spectrometer, solvent: CDCl3). After that, 20 mL trifluoroacetic acid dissolving PAM-PBLG-b-TPGS was added in 2 mL hydrogen bromide (HBr) (33% in acetic acid) and stirred for 1 h at room temperature. Then the solution was neutralized by sodium hydroxide (NaOH) and dialyzed against distilled water (DI water) using a dialysis membrane with molecular weight cutoff (MWCO) of 5 kDa. The aqueous solution of purified product was lyophilized to obtain PAM-PGlu-b-TPGS.

# Preparation of DACHPt-Loaded Nanoparticles

DACHPt-loaded nanoparticles were prepared similar to our previous work. Briefly, PAM-PGlu-b-TPGS solution was separately mixed with the DACHPt aqueous solution (Molar ratio Glu/ DACHPt = 3/1) and reacted for 72 h. The formed nanoparticles were purified by ultrafiltration using Centricon Plus-20 centrifugal filter units (MWCO 50 KDa, Millipore, MA, USA).

# Characterization of DACHPt-Loaded Nanoparticles

The size distribution and zeta potential of DACHPt-loaded nanoparticles were measured using Zetasizer Nano ZS90 (Malvern Instruments Ltd., UK). Before measurement, the freshly prepared nanoparticles were diluted as needed. All measurements were carried out at 25◦C. The data were obtained as the average of three measurements.

The morphology of nanoparticles were then observed by transmission electron microscopy (TEM, Tecnai G2 20, FEI Company, Oregon, USA). One microliter of the sample solution was placed on the resulting grids and dried in the air.

To measure the drug loading content (LC) and drug encapsulation efficiency (EE) of the DACHPt-loaded nanoparticles, a predefined amount of nanoparticles were dissolved in 1 mL water. The Pt concentration in the nanoparticles was quantified by Inductively coupled plasma mass spectrometry (ICP-MS, Xseries II, Thermoscientific, USA). The LC and EE of DACHPt-loaded nanoparticles were calculated according to

$$\text{LC} \left( \% \right) = \frac{\text{amount of DACHP in the nanopparticles}}{\text{amount of the nanopparticles}} \times 100\%$$

$$\text{EE} \left( \% \right) = \frac{\text{amount of DACHP in the nanopparticles}}{\text{amount of feeding DACHP}} \times 100\%$$

#### In Vitro Drug Release Study

The stability of NP-TPGS-Pt in cell culture media including 10% FBS at 25◦C was evaluated by DLS. The DACHPt release profile of the nanoparticles was monitored by the dialysis method. Firstly, 1 ml of NP-TPGS-Pt was placed into the dialysis bag (MWCO 5 kDa) and immersed in 30 ml of cell culture media including 10% FBS (pH 7.4 and 5.5) at 37◦C. The media outside was taken out at defined periods, and the concentrations of Pt were measured by ICP-MS.

#### Cellular Uptake Study

In this study, 5-Aminofluorescein-labeled NP-PEG-Pt, and NP-TPGS-Pt as the model fluorescent probe to investigate the uptake by A549 cells and A549/DDP cells. A549 cells and A549/DDP cells were cultured as described previously (Liu et al., 2017a). The cellular uptake experiments of 5-Aminofluorescein-NPs were performed using CLSM. A549 cells and A549/DDP cells (1<sup>∗</sup> 10<sup>5</sup> cells/well) were seeded in 12-well culture plates and cultured overnight in RPMI1640 with 10% FBS. On the following day, the cells were washed once with PBS and incubated with 5-AF-NP-PEG-Pt, and 5-AF-NP-TPGS-Pt in 10% serum containing the media for 3 h at 37◦C. The cells were observed using a CLSM (Olympus Fluoview FV-1000, Tokyo, Japan) using an imaging software. The images of the cells were captured with differential interference contrast channel, and the images of 5- Aminofluorescein-NPs and the nuclei of DAPI-stained cells were recorded with following channels: blue channel (DAPI) excited at 340 nm and green channel (5-Aminofluorescein) excited at 488 nm.

# MTT Assay

The 50% growth inhibitory concentrations (IC50) of free Oxaliplatin, NP-PEG-Pt, and NP-TPGS-Pt in the A549 and A549/DDP cells were measured by the MTT assay. After incubated in a 96-well culture plate (10<sup>4</sup> cells/well) for 24 h, A549 cells were then exposed to 10, 20, 40, 100, and 200µM oxaliplatin or DACHPt-loaded nanoparticles (on a platinum basis) for another 24, 48, and 72 h. At each point of time, The MTT solution was added, and cell viability was measured in a Bio-Rad 680 microplate reader by formazan absorbance at 490 nm.

# In Vivo Antitumor Efficacy

After randomly divided into four groups (n = 5), the nude mice were built with human A549 xenograft tumor model by injection of 1.5 × 10<sup>6</sup> A549/DDP cells (100 µL) subcutaneous at the right axilla of each mouse. Before initiating treatment, tumors were observed frequently and allowed to grow to ∼50 mm<sup>3</sup> in volume. Mice were injected intravenously five times via tail vein on days 0, 4, 8, 12, and 16 with saline, Oxaliplatin and DACHPt-loaded nanoparticles (6 mg/kg on a Pt basis). The antitumor efficacy was determined in accordance with the tumor volume (V), which was calculated similar to our previous work. The body weights of mice were simultaneously measured to evaluate the systemic toxicity. The results were considered statistically significant if two-tailed P-values were < 0.05.

# Statistical Analysis

All of the experiments were carried out at least three times. The Data are expressed as mean ± SD unless noted otherwise and analyzed for significance using Student's t-test. Probability value (P) < 0.05 indicates statistically significant. <sup>∗</sup>P < 0.05; ∗∗P < 0.01.

# RESULTS AND DISCUSSION

#### Synthesis and Characterizations of Dendritic Copolymer PAM-PGlu-b-TPGs

The chemical structure of dendritic copolymer PAM-PGlu-b-TPGS was validated by <sup>1</sup>H NMR spectroscopy. **Figure 3A** shows the <sup>1</sup>H NMR spectrum of PAM-PBLG-b-TPGS, in which a, b, c, d, e, and g represent the characteristic peaks of methylene group (CH2) in PAM, -C(O)CH(CH2)NH-, -CHCH2CH2C(O)-, -CHCH2CH2C(O)- and C6H5CH2- in PBLG, as well as CH<sup>2</sup> of PEG in TPGS, demonstrating the successful linking of TPGS. The number of TPGS in each dendritic molecule was 30 by calculating the areas of peaks e and g. As exhibited in **Figure 3B**, the characteristic peak of benzyl group in BLG disappears, suggesting that the deprotection was successful and PAM-PGlu-b-TPGS had been successfully prepared.

Afterwards, we obtained NP-TPGS-Pt by complexing PAM-PGlu-b-TPGS with DACHPt. The NPs had a narrow monodisperse distribution and the average hydrodynamic diameter of ∼85.3 nm (**Figure 4A**). TEM presented that NF-TPGS-Pt were uniformly distributed spherical particles with the size of about 60 nm, being consistent with the DLS results (**Figure 4B**). In the meantime, we prepared NP-PEG-Pt with PAM-PGlu-b-PEG and DACHPt as control (**Table 1**).

#### Stability and Drug Release Study

Then we determined the in vitro stability of NP-TPGS-Pt. The size of NP-TPGS-Pt remained stable after 14 consecutive days of culture in cell medium (**Figure 5A**), indicating that it holds robust stability.

As evidenced by ICP-MS, the drug loading content of NP-TPGS-Pt was 26.3% (**Table 1**). Subsequently, we tested the in vitro release of DACHPt from NP-TPGS-Pt. In a previous study, we found that Pt release was accelerated in a chloride ion-containing environment under acidic conditions. Thus, we herein compared such release from NP-TPGS-Pt at pH 7.4 and 5.5. At pH 7.4, the drug release rate from NP-TPGS-Pt was only 20% after 96 h (**Figure 5B**), revealing that these NPs were fairly stable under physiological conditions and drug hardly leaked. On the other hand, NP-TPGS-Pt managed to rapidly release DACHPt at pH 5.5, predicting that they may work so inside tumor cells.

# Cellular Uptake

It is well-established that drug delivery into cells through uptake of nanocarrier can well escape efflux protein-mediated transport, thereby overcoming multidrug resistance (Wei et al., 2015). Accordingly, we first studied the cellular uptake of NP-TPGS-Pt with confocal laser scanning microscopy (CLSM). We covalently linked NP-PEG-Pt and NP-TPGS-Pt with 5-aminofluorescein (AF), and observed their endocytosis by A549 and A549/DDP

FIGURE 3 | (A) <sup>1</sup>H-NMR spectra of PAM-PBLG384-NH<sup>2</sup> in CDCl3, and (B) PAM-PGlu384-*b*-(TPGS)<sup>30</sup> in D2O.

cells. After 3 h, A549 cells emitted obvious fluorescence of NP-TPGS-Pt-AF (green), but the NP-PEG-Pt-AF group barely showed signal (**Figure 6A**), because intracellular fluorescence originated from NPs to which AF was covalently linked. Collectively, TPGS modification substantially promoted the cellular uptake of nanocarrier, whereas PEG hardly underwent uptake due to the apparent shielding effect. Given that the fluorescence signal of AF (green) in A549/DDP cells was almost equivalent to that of A549 cells, NP-TPGS-Pt was also subjected to uptake by A549/DDP cells (**Figure 6B**).

Then we further quantified the cellular uptake efficiency of NP-TPGS-Pt at 3 h. As presented in **Figure 6C**, the NP-PEG-Pt uptake efficiencies of A549 and A549/DDP cells are lower than 20%, while those for NP-TPGS-Pt reach as high as 60%. In short, drug-resistant cells were still capable of efficient uptake of NP-TPGS-Pt (**Figure 6D**).

#### In Vitro Cytotoxicity

To demonstrate the ability of NP-TPGS-Pt to kill tumor cells, especially the drug-resistant ones, we performed

TABLE 1 | Characterization of DACHPt-loaded nanoparticles.


*PDI, polydispersity index; ZP, zeta potential; LC, loading content; EE, encapsulation efficiency.*

MTT assay using A549 and A549/DDP cells. Meanwhile, free Oxaliplatin, NP-PEG-Pt and drug-free NP-TPGS were employed as controls. A549 and A549/DDP cells were treated by different concentrations of drugs for 24 and 48 h. After 24 h of culture, Oxaliplatin significantly killed A549 cells, whereas NP-PEG-Pt also exerted much weaker effects (**Figure 7A**), being in agreement with the above cellular uptake efficiencies. However, NP-TPGS-Pt worked more effectively than Oxaliplatin did. Moreover, the IC<sup>50</sup> values of Oxaliplatin and NP-PEG-Pt were 98.8 and 214.0µM, respectively, but that of NP-TPGS-Pt was merely 45.4µM (**Table 2**). With increasing culture time, although the antitumor effects of the

FIGURE 5 | (A) DACHPt-loaded NPs incubated in medium containing 10% FBS maintained their sizes for 14 days. (B) Accumulative release of DACHPt-loaded NPs in media containing 10% FBS with different pH-values. The result was reported as the average of three measurements.

three Oxaliplatin formulations were significantly boosted, the outcomes of NP-TPGS-Pt were significantly superior to those of Oxaliplatin and NP-PEG-Pt (**Figure 7B**). Taken together, modifying NP surface with TPGS increased the cytotoxicity through efficient uptake and persistent drug release.

On the other hand, Oxaliplatin killed A549/DDP cells far less effectively than A549 cells. Since the IC<sup>50</sup> values of

Oxaliplatin dose and that of the drug-free TPGS with the same polymer concentrations: (A,B) A549 cells for 24 and 48 h. (C,D) A549/DDP cells for 24 and 48 h.

Oxaliplatin at 24 and 48 h were 347.4 and 118.6µM, respectively, A549/DDP cells were indeed strongly drug-resistant. Likewise, the antitumor effects of NP-PEG-Pt were inferior to those of Oxaliplatin. In contrast, the IC<sup>50</sup> values of NP-TPGS-Pt at 24 and 48 h were 85.4 and 23.1µM, respectively which were much lower than those of Oxaliplatin (**Figures 7C,D**), and slightly lower than those against A549 cells. Hence, NP-TPGS-Pt could overcome the drug resistance of cells. Additionally, MTT assay disclosed no influence of DACHPt-free NP-TPGS on cell growth, so they were highly biosafe and thus applicable to practice.

#### In Vivo Antitumor Effects

Given the positive results above, we ultimately evaluated the in vivo antitumor effects of NP-TPGS-Pt. The mice xenografted with A549/DDP cells were injected with normal saline, Oxaliplatin, NP-PEG-Pt and NP-TPGS-Pt respectively every 4 days, five times in total. After 20 days of treatment, the average tumor volumes of saline, Oxaliplatin and NP-PEG-Pt groups grew to 408.8, 337.3, and 278.8 mm<sup>3</sup> , respectively, but that of the NP-TPGS-Pt group only increased to 188.7 mm<sup>3</sup> (**Figure 8A**). Clearly, NP-TPGS-Pt TABLE 2 | IC50 values of Oxaliplatin, NP-PEG-Pt and NP-TPGS-Pt against A549 and A549/DDP cells following 24 and 48 h of incubation.


markedly inhibited tumor growth, with better outcomes than those of Oxaliplatin and NP-PEG-Pt. Besides, the body weights of NP-TPGS-Pt-treated mice and the other three groups were similar, without obvious changes (**Figure 8B**). After 20 days of treatment, all mice were sacrificed, from which tumors were collected. **Figures 8C,D** show the morphologies and average body weights of all tumors, both confirming the outstanding antitumor effects of NP-TPGS-Pt. Overall, NP-TPGS-Pt were able to solve tumor multidrug resistance.

# DISCUSSION AND CONCLUSIONS

In our previous research (Liu et al., 2017a), we developed the DACHPt-loaded UM (UM/DACHPt) dendritic block copolymer for micelles formulation of small molecular anti-tumor drugs and characterized the properties of the nanoparticle in vitro. Since the antitumor effects of unimolecular micelles are inferior to those of free drugs in vitro, they would hardly fulfill the requirement. Therefore, to optimize the nanocarrier structure to be both stable in vivo and multidrug-resistant. Indeed, TPGS has been reported to inhibit P-glycoprotein mediated multi-drug resistance (MDR) in tumor cells, which may reduce the excretion of drugs (Collnot et al., 2007; Zeng et al., 2013).

weights (D) of tumors resected from each group of sacrificed mice on the last day.

From the DLS results (**Figure 4A** and **Table 1**), we obtained NP-TPGS-Pt by complexing PAM-PGlu-b-TPGS with DACHPt. The NPs had a narrow monodisperse distribution and the average hydrodynamic diameter of ∼85.3 nm. The size of NP-TPGS-Pt was much larger than which of NP-PEG-Pt. Because of PEG is completely hydrophilic polymers, PEG-b-PGlu could interact with DACHPt to be unimolecular micelles, therefore the size and the dendritic block copolymers itself size didn't big changed. But TPGS has certain hydrophobic, after PEG-b-PGlu interact with DACHPt, TPGS would induce the accumulation of the dendritic block copolymers, so the final particle size of NP-TPGS-Pt was bigger than NP-PEG-Pt. In other reports also showed TPGS dendritic block copolymers gathered into 100 nanometers in micelle (Zeng et al., 2013). And in TEM presented that NF-TPGS-Pt were uniformly distributed spherical particles with the size of about 60 nm, being consistent with the DLS results (**Figure 4B**). The particle size measured by TEM was smaller than that by DLS, because DLS detected the hydrodynamic diameter in aqueous solution in which the hydrophilic TPGS layer expanded while TEM was performed by using dry NPs.

Previous reports found that Pt release was accelerated in a chloride ion-containing environment under acidic conditions (Song et al., 2012). In **Figure 5**, the drug release rate from

NP-TPGS-Pt was only 20% after 96 h, revealing that these NPs were fairly stable under physiological conditions and drug hardly leaked. We observed the acid environment could accelerate the Pt release, because the accelerated release at acidic pH may be due to the protonation of carboxylic groups of PGlu, which weakens the drug and micelles coupling.

In conclusion, we successfully synthesized the dendritic block copolymer PAM-PGlu-b-TPGS which was thereafter selfassembled into nanoparticles by chelating the potent antitumor agent DACHPt. This TPGS-coated nanocarrier had robust stability and underwent nanoparticle-based cellular uptake by drug-resistant cancer cells, eventually remarkably suppressing the growth of these cells and drug-resistant tumors in vivo. Hence, the nanocarrier provides a novel strategy for treating multidrug-resistant tumors.

#### AUTHOR CONTRIBUTIONS

GL, LH, and ZL designed the research project; H-IT and LJ had full controlled the experiments, data analysis, and preparation of

#### REFERENCES


article; XZ, HC, WC, JZ, JP, DW, LG, ZX, and LM were involved in planning the analysis and drafting the article. The final draft article was approved by all the authors.

#### ACKNOWLEDGMENTS

The authors are grateful for the financial support from National Natural Science Foundation of China (81771966, 51703258, 31270019,21506161, 21646010, 81660425, 81670141), Guangdong Natural Science Funds for Distinguished Young Scholar (2014A030306036), Natural Science Foundation of Guangdong Province (2015A030313848, 2015A030313846), Guangdong Special Support Program, Science and Technology Planning Project of Guangdong Province (2016A020217001), Science, Technology & Innovation Commission of Shenzhen Municipality (JCYJ20160301152300347, JCYJ20160428182427603, JCYJ20150518162154828), the Fundamental Research Funds for the Central Universities (No. 17ykjc05), and Project of scientific research and technology development of Guilin City (20170226).


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Tsai, Jiang, Zeng, Chen, Li, Cheng, Zhang, Pan, Wan, Gao, Xie, Huang, Mei and Liu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Polydopamine-Functionalized CA-(PCL-ran-PLA) Nanoparticles for Target Delivery of Docetaxel and Chemo-photothermal Therapy of Breast Cancer

#### Na Kong1,2† , Mei Deng<sup>3</sup>† , Xiu-Na Sun1,2, Yi-Ding Chen<sup>3</sup> \* and Xin-Bing Sui4,5,6 \*

#### Edited by:

Chao Wang, University of North Carolina at Chapel Hill, United States

#### Reviewed by:

Jun Wu, Sun Yat-sen University, China Xiaoxue Zhang, Stony Brook University, United States Yingjie Yu, Tufts University, United States

#### \*Correspondence:

Xin-Bing Sui hzzju@zju.edu.cn Yi-Ding Chen ydchen@zju.edu.cn †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 19 January 2018 Accepted: 05 February 2018 Published: 21 February 2018

#### Citation:

Kong N, Deng M, Sun X-N, Chen Y-D and Sui X-B (2018) Polydopamine-Functionalized CA-(PCL-ran-PLA) Nanoparticles for Target Delivery of Docetaxel and Chemo-photothermal Therapy of Breast Cancer. Front. Pharmacol. 9:125. doi: 10.3389/fphar.2018.00125 <sup>1</sup> Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, China, <sup>2</sup> Institute of Translational Medicine, Zhejiang University, Hangzhou, China, <sup>3</sup> Department of Surgical Oncology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China, <sup>4</sup> Department of Medical Oncology, Holistic Integrative Oncology Institutes and Holistic Integrative Cancer Center of Traditional Chinese and Western Medicine, The Affiliated Hospital of Hangzhou Normal University, College of Medicine, Hangzhou Normal University, Hangzhou, China, <sup>5</sup> Department of Cancer Pharmacology, Holistic Integrative Pharmacy Institutes, College of Medicine, Hangzhou Normal University, Hangzhou, China, <sup>6</sup> Key Laboratory of Elemene Class Anti-cancer Chinese Medicine of Zhejiang Province and Engineering Laboratory of Development and Application of Traditional Chinese Medicine from Zhejiang Province, Hangzhou Normal University, Hangzhou, China

Current limitations of cancer therapy include the lack of effective strategy for target delivery of chemotherapeutic drugs, and the difficulty of achieving significant efficacy by single treatment. Herein, we reported a synergistic chemo-photothermal strategy based on aptamer (Apt)-polydopamine (pD) functionalized CA-(PCL-ran-PLA) nanoparticles (NPs) for effective delivery of docetaxel (DTX) and enhanced therapeutic effect. The developed DTX-loaded Apt-pD-CA-(PCL-ran-PLA) NPs achieved promising advantages, such as (i) improved drug loading content (LC) and encapsulation efficiency (EE) initiated by star-shaped copolymer CA-(PCL-ran-PLA); (ii) effective target delivery of drugs to tumor sites by incorporating AS1411 aptamers; (iii) significant therapeutic efficacy caused by synergistic chemo-photothermal treatment. In addition, the pD coating strategy with simple procedures could address the contradiction between targeting modification and maintaining formerly excellent bio-properties. Therefore, with excellent bio-properties and simple preparation procedures, the DTX-loaded Apt-pD-CA-(PCL-ran-PLA) NPs effectively increased the local drug concentration in tumor sites, minimized side effects, and significantly eliminated tumors, indicating the promising application of these NPs for cancer therapy.

Keywords: polydopamine coating, star-shaped copolymer, aptamer, target delivery, chemo-photothermal therapy

# INTRODUCTION

Breast cancer, the leading type of cancer for deaths among women worldwide, has caused over an estimated 2,421,698 deaths for the year 2015 (Global Burden of Disease Cancer Collaboration, 2017). Current approaches for breast cancer therapy are still limited by the level of our medical technology. For example, chemotherapy, which is one of the most common clinical methods for

breast cancer therapy, is non-specific and always accompanied with serious side effects because of lacking effective carriers (Tao et al., 2013, 2015a; Shi J. et al., 2017; Zeng et al., 2017). Nanobiotechnology has shown promising prospects for changing the landscape of pharmaceutical and biomedical industries fundamentally (Shi et al., 2010; Tao et al., 2015b, 2017b,d; He et al., 2016; Wang et al., 2016, 2017; Hofmann et al., 2017; Niu et al., 2017; Qi et al., 2017; Shi B. et al., 2017; Shi J. et al., 2017; Xu et al., 2017a,b). The key to solve the side effects caused by cytotoxic anticancer drugs is to increase the local effective drug concentrations in the tumor sites, which could be addressed by employing polymeric nanoparticles (NPs) as highly promising target drug delivery systems (Kamaly et al., 2016; Kakkar et al., 2017; Nie et al., 2017; Tao et al., 2017c; Xu et al., 2017). In addition, due to the molecular complexity of cancers, treatment based on monotherapies is suboptimal and can not effective eliminated tumors (Greco and Vicent, 2009; Sun et al., 2011). By combining two or more therapies with different therapeutic mechanisms in one nano-system (e.g., chemotherapy and photothermal therapy), a promising strategy could be achieved to significantly enhance the therapeutic efficacy and have a better long-term prognosis (Cheng et al., 2017; Tao et al., 2017d). Thus, developing multifunctional target delivery platforms that intelligently integrated with different therapeutic approaches is urgently needed for effective cancer therapy.

Aptamers, which are essentially single-stranded DNA or RNA oligonucleotides and have specific 3D structures, are able to bind various biological targets on cancer cells with unique specificity and high affinity (Farokhzad et al., 2004; Dhar et al., 2008; Xu et al., 2016). Aptamers possess distinct advantages over antibodies. Examples include easier production without the use of animal, smaller size, readily availability, lower molecular weights, non-immunogenicity, higher targeting efficacy and in vivo stability (Jiang et al., 1996; Sun et al., 2014). Therefore, aptamers have been widely used as effective targeting agents for cancer therapy and diagnosis (Zhu D. et al., 2016). As one of the most commonly used DNA aptamers, AS1411 is able to effectively target a wide variety of cancer cells including breast cancer cells (MCF-7) (Keefe et al., 2010). Because of the excellent performance on the targeting efficacy, AS1411 can be incorporated in the surface modification of NPs as an effective targeting ligand for breast cancer therapy. Nevertheless, major concerns still exist for NPs prepared by pre-functionalized polymers. For example, the synthesis of AS1411 aptamer-polymer conjugates could be inefficient, possibly altering the chemical properties, lengthy with high cost, and at a risk of compromising the ability for drug encapsulation of polymers (Ho and Leclerc, 2004). Another method to conjugate AS1411 on the surface of NPs is surface modification of prepared NPs, but this will be cumbersome for NPs lacking reactive functional groups, which still require reacting with reactive linkers or coupling agents followed by exhaustive purification processes in order to remove excess reactants and catalysts (Takahashi et al., 2010).

In order to address these issues, a simple and versatile surface modification strategy based on polydopamine (pD) was adopted in this study. As reported by Park et al., surface modification with pD is applicable to various NP-based drug carriers no matter what the types of ligands (e.g., small molecules, peptides, or polymers) and how is the chemical reactivity of NPs (Park et al., 2014). Briefly, dopamine catechol could be oxidized to quinone followed by reactions with other quinones or catechols to form pD in weak alkaline conditions, gradually forming a water-insoluble polymeric film on NP surface during the process (Jiang et al., 2011; Chang et al., 2016). Afterward, ligands possessing amine or thiol groups could be simply conjugated on the surface of pD-coated NPs via Michael addition or Schiff base reactions (Lee et al., 2007, 2009; Tao et al., 2016). Besides the function of surface modification, pD and its derivatives have been widely reported as effective photothermal agents, which could contribute to phtotothermal therapy of breast cancer in the meantime. As phtotothermal therapy is a potent technique for cancer therapy because of its high selectivity and minimal invasiveness (Huang et al., 2013; Miao et al., 2016; Song et al., 2016; Tao et al., 2017a), it will be rather promising to initiate a synergistic target chemo-photothermal strategy (Tao et al., 2017d).

It has been reported that star-shaped copolymers possess many unique advantages for developing NP-based drug delivery platforms (Mei et al., 2013). To build the core of the NPs, a star-shaped co-polymer cholic acid functionalized poly(ε-caprolactone-ran-lactide) [CA-(PCL-ran-PLA)] was successfully synthesized according to a previous report (Tao et al., 2014), which has been demonstrated with many advantages such as improved bioavailability (i.e., solubility, stability, and permeability) of anticancer drugs, increased drug loading content (LC) and encapsulation efficiency (EE), controlled drug release profiles, and excellent antitumor efficacy. Herein, we used a nano-precipitation method to prepare CA-(PCL-ran-PLA NPs, coated the developed NPs with pD layer [pD-CA-(PCLran-PLA) NPs], and target functionalizing the pD-coated NPs with AS1411 aptamers [Apt-pD-CA-(PCL-ran-PLA) NPs]. The newly developed NPs were characterized by surface morphology, LC and EE, stability, photothermal properties and drug release profiles. The in vitro and in vivo targeting effect of these NPs were also accessed. With excellent biocompatibility, the docetaxel (DTX)-loaded Apt-pD-CA-(PCL-ran-PLA) NPs were demonstrated with a significant antitumor efficacy through a target chemo-photothermal strategy.

# MATERIALS AND METHODS

#### Materials

Colic acid, D,L-Lactide (3,6-dimethyl-1,4-dioxane-2,5 dione, C6H8O4), 4-(dimethylamino)pyridine (DMAP), 1,3-diisopropylcarbodiimide (DCC), 2-(3,4-dihydroxyphenyl) ethylamine (dopamine) hydrochloride, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), and Stannous octoate [Sn(Oct)2] were purchased from Sigma (St. Louis, MO, United States). ε-caprolactone (CL) was purchased from Acros Organics (Geel, Belgium). Commercial Taxotere <sup>R</sup> and docetaxel (DTX) were provided by Shanghai Jinhe Bio-tech Co., Ltd (Shanghai, China). Methanol and acetonitrile were provided by EM Science (ChromAR, HPLC grade, Mallinckrodt Baker, United States). All other agents used were of analytical reagent

grade. Boon Environmental Tech. Industry Co., Ltd (Tianjin, China) provided the ultrahigh pure water utilized throughout all the studies. Human breast cancer cell line MCF-7 was purchased from American Type Culture Collection (ATCC, Rockville, MD, United States).

#### Synthesis of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs Preparation of DTX/CA-(PCL-ran-PLA) NPs

The star-shaped random copolymer CA-(PCL-ran-PLA) was synthesized through the ring opening copolymerization, and provided by Prof. Lin Mei's group at Tsinghua University (Tao et al., 2014). We adopted a modified nanoprecipitation method with water/acetone system to develop DTX-loaded CA-(PCL-ran-PLA) NPs (DTX/CA-(PCL-ran-PLA) NPs) (Zeng et al., 2013a, 2015a; Zhang et al., 2014a,b, 2015; Wang et al., 2015; Wu et al., 2015). In brief, 200 mg of CA-(PCL-ran-PLA) copolymers and 20 mg of DTX were dissolved in 16 mL of acetone. The solution was then titrated into 200 mL of 0.03% (w/v) TPGS aqueous solution with continuous stirring. After stirring overnight at a speed of 800 rpm to remove acetone, the pure DTX/CA-(PCL-ran-PLA) NPs were obtained by centrifugation at a speed of 20,000 rpm (4◦C) for 20 min, followed by three times washing in 20 mL of deionized water to remove TPGS emulsifier and unencapsulated DTX. Finally, pre-calculated amount of the obtained DTX/CA-(PCL-ran-PLA) NPs were weighted for pD coating, while the left NPs were dispersed in deionized (DI) water and lyophilized 2 days for further use.

#### Surface Modification with pD Layers

The DTX/pD-CA-(PCL-ran-PLA) NPs were prepared by incubating DTX/CA-(PCL-ran-PLA) NPs in 0.1 mg/mL dopamine hydrochloride/10 mM Tris buffer solution (pH 8.5) under magnetic stirring at room temperature. After 6 h of reaction, the suspensions gradually turned darker, indicating that dopamine was successfully polymerized. The obtained DTX/pD-CA-(PCL-ran-PLA) NPs were centrifuged (12,000 rpm, 30 min) and lyophilized for conjugation of targeting AS1411 aptamers.

#### Conjugation of Targeting AS1411 Aptamers

The sequence of AS1411 aptamers used in this study contains 10 extra T bases at the 3-terminus (5<sup>0</sup> -GGT GGT GGT GGT TGT GGT GGT GGT GGT TTT TTT TTT-thiol-3<sup>0</sup> ). The SH-terminated aptamers were conjugated on the surface of pD-coated NPs through a Michael addition reaction (Park et al., 2014). Briefly, the DTX/pD-CA-(PCL-ran-PLA) NPs were dissolved in the Tris buffer (pH 8.0) containing aptamers with final concentrations of NPs and AS1411 aptamers at 1 and 0.5 mg/mL, respectively. After 2 h of magnetic stirring at room temperature, the resulting DTX/Apt-pD-CA-(PCL-ran-PLA) NPs were centrifuged (12,000 rpm, 30 min), washed three times with deionized water and lyophilized 2 days for further use.

All the fluorescent coumarin-6 (C6)-loaded NPs (i.e., C6/CA-(PCL-ran-PLA) NPs, C6/pD-CA-(PCL-ran-PLA) NPs, and C6/Apt-pD-CA-(PCL-ran-PLA) NPs) were prepared with the same method at each of the 3 steps described above.

# Characterization of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs

The size of DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCLran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs were performed by Malvern Mastersizer 2000 (Zetasizer Nano ZS90, Malvern Instruments Ltd., United Kingdom). The transmission electron microscopy (TEM, Tecnai G2 F30, FEI Company, Hillsboro, OR, United States) was used to access the surface morphology of these NPs. The photothermal properties of these NPs were determined by recording the temperature changes of various solutions with different concentrations of NPs (62.5, 125, 250, and 500 µg/mL) under the irradiation of an 808 nm NIR laser (Shanxi Kaisite Electronic Technology Co., Ltd., Xi'an, China) at a power density of 1.0 W/cm<sup>2</sup> . To study the effect of power density on the photothermal effect of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs, NP solution at the concentration of 500 µg/mL was irradiated by different power densities (0.5–2.0 W/cm<sup>2</sup> ) and the changes of temperature were recorded. The temperatures were monitored by an infrared thermal imaging camera (TI100 Infrared Camera FLK-TI100 9HZ, FLUKE).

# Drug Loading Content (LC) and Encapsulation Efficiency (EE)

The LC and EE of the DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs were determined by HPLC (LC 1200, Agilent Technologies, Santa Clara, CA, United States). In brief, 5 mg of NPs were dissolved in 1 mL of DCM under vigorous vortexing, and the prepared solution was transferred to 5 mL of mobile phase containing acetonitrile and DI water (50:50, v/v). In order to get a clear solution for HPLC, a nitrogen stream was used to evaporate DCM for 15 min. A reverse-phase C-18 column (150 × 4.6 mm, 5 µm, C18, Agilent Technologies, Santa Clara, CA, United States) was utilized at 35◦C, and the flow rate of mobile phase was set at 1.0 mL/min. A UV/VIS detector was used to detect the column effluent at 227 nm. The drug LC and EE of these NPs were calculated by the following equations respectively (n = 3).

$$\text{LC}(\%) = \frac{\text{Weight of DTX in NPs}}{\text{Weight of NPs}} \times 100\%$$

$$\text{EE}(\%) = \frac{\text{Weight of DTX in NPs}}{\text{Weight of the feeding DTX}} \times 100\%.$$

#### In Vitro Drug Release Profiles

In order to study the in vitro DTX release profiles, 5 mg of the lyophilized NPs were dispersed in 5 mL of PBS (pH 7.4, containing 0.1% w/v Tween 80). Tween 80 was used to increase the solubility of DTX while avoid the binding of DTX on the tube wall. Afterward, the NP suspension was transferred into a dialysis membrane bag (MWCO = 3,500, Shanghai Sangon, China) which immersed in 15 mL of fresh PBS in a centrifuge tube. The whole tube was then transferred into an orbital water bath and shaken at a speed of 120 rpm (37◦C). At designated time points, 10 mL of release medium was picked out for HPLC analysis. After changing 15 mL of fresh PBS solution, the tuber was transferred

back to the shaker for continuous tests. The cumulative release of DTX from the DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA- (PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs was plotted against time.

#### Cellular Uptake of Fluorescent NPs

MCF-7 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 mg/mL streptomycin, and 100 U/mL penicillin in 5% CO<sup>2</sup> at 37◦C. The cell culture was stayed in 95% air humidified atmosphere. The cells were incubated with 250 µg/mL C6/CA-(PCL-ran-PLA) NPs, C6/pD-CA-(PCL-ran-PLA) NPs, or C6/Apt-pD-CA-(PCL-ran-PLA) NPs for 2 h, washed with cold PBS three times, and fixed by cold methanol for 20 min. After that, the nuclei were counterstained with DAPI for 10 min and washed twice with PBS. In order to visualize the cells, the chambers were mounted onto the confocal laser scanning microscope (CLSM, Olympus Fluoview FV-1000, Tokyo, Japan) with a blue channel excited at 340 nm and a green channel excited at 485 nm.

For quantitative analysis, MCF-7 cells were plated in 96-well black plates and incubated overnight at its initial density of 1 × 10<sup>4</sup> cells/well. Then the cells were equilibrated with Hank's buffered salt solution (HBSS) for 1 h at 37◦C, and C6/CA-(PCLran-PLA) NPs, C6/pD-CA-(PCL-ran-PLA) NPs, or C6/Apt-pD-CA-(PCL-ran-PLA) NPs were added at concentrations of 100, 250, and 500 µg/mL, respectively. After 2 h of incubation, the medium was removed and the wells were washed three times with 50 µL of cold PBS. Finally, 50 µL of 0.5% Triton X-100 in 0.2 N sodium hydroxide was added to each sample to lyse the MCF-7 cells.

For flow cytometric (FCM) experiments, MCF-7 cells were seeded in 6-well plates (at a density of 1 × 10<sup>5</sup> cells/well), and treated with 500 µg/mL of C6/CA-(PCL-ran-PLA) NPs, C6/pD-CA-(PCL-ran-PLA) NPs, or C6/Apt-pD-CA-(PCL-ran-PLA) NPs for 1 h at 37◦C, respectively. After removing the cell culture medium, the cells were washed with PBS twice, digested by trypsin, and harvested by centrifugation. Finally, the fluorescence intensity of C6 was detected by a flow cytometer (BD Biosciences, San Jose, CA, United States) at an excitation wavelength of 488 nm and an emission wavelength of 530 nm.

#### In Vitro Cell Viability Studies

In order to evaluate the biocompatibility of drug-free NPs and the in vitro antitumor efficacy of the DTX-loaded NPs, MCF-7 cells were seeded in 96-well plates (at a density of 5,000 cells/well) and incubated overnight.

For the biocompatibility studies, the MCF-7 cells were incubated with drug-free CA-(PCL-ran-PLA) NPs, pD-CA- (PCL-ran-PLA) NPs, or Apt-pD-CA-(PCL-ran-PLA) NPs at different NP concentrations (25, 125, 250, and 500 µg/mL) and incubation time (24 and 48 h). Afterward, the previous mediums were changed with MTT-contained DMEM (5 mg/mL) and the cells were incubated for an additional 4 h. Then MTT was removed and DMSO was added to dissolve the formazan crystals (2 h, dark, 37◦C). The absorbance at 570 nm was measured using a microplate reader (Bio-Rad Model 680, United Kingdom). Control group with untreated cells was used as 100% of viability, and cells incubated with MTT-free medium were utilized as blank to calibrate the spectrophotometer to zero absorbance.

For the in vitro antitumor efficacy studies, (i) Chemotherapy groups: the MCF-7 cells were incubated with commercial Taxotere <sup>R</sup> , DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCLran-PLA) NPs, or DTX/Apt-pD-CA-(PCL-ran-PLA) NPs at 0.25, 2.5, 12.5, and 25 µg/mL equivalent DTX concentrations for 24 h and 48 h, respectively. (ii) Irradiation groups: DTX/CA-(PCLran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, or DTX/AptpD-CA-(PCL-ran-PLA) NPs at 0.25, 2.5, 12.5, and 25 µg/mL equivalent DTX concentrations were added into the medium of MCF-7 cells and immediately irradiated with an 808 nm NIR laser for 10 min at a power density of 1.5 W/cm<sup>2</sup> . Afterward, MCF-7 cells from all the groups were further cultured for 24 and 48 h, respectively. Untreated cells only received NIR laser irradiation were used as controls. The same MTT method was used to test the cell viability of cells from all these groups in order to access the in vitro antitumor efficacy.

# Xenograft Breast Tumor Model

The 4–5 weeks old of female severe combined immunodeficient (SCID) nude mice were purchased from the Institute of Laboratory Animal Sciences, Chinese Academy of Medical Science. The Administrative Committee on Animal Research in Zhejiang University approved the protocols for animal studies. A total volume of 200 µL MCF-7 cells in PBS were implanted subcutaneously on the backs of mice at a dosage of 2 × 10<sup>6</sup> cells per mouse. The tumor growth in each mouse was observed at frequent intervals. The tumor size was measured by a vernier caliper, and tumor volume (V) was calculated by this formulation: 4π/3 × (length/2) × (width/2)<sup>2</sup> . About 95% of the mice developed a tumor with an average volume of about 100 mm<sup>3</sup> after 2 weeks.

# In Vivo Photothermal Imaging and in Vivo Antitumor Efficacy

The photothermal imaging studies were performed when tumor size reached about 400 mm<sup>3</sup> . The mice were intravenously (i.v.) injected with Saline, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs at the same volume and dose. After 12 h of injection, NIR laser irradiation was performed (808 nm, 1.5 W/cm<sup>2</sup> , 10 min) at tumor sites. The temperature changes of tumor sites, as well as infrared thermographic maps, were recorded by an infrared thermal imaging camera (TI100 Infrared Camera FLK-TI100 9HZ, FLUKE).

For in vivo antitumor efficacy studies, Apt-pD-CA-(PCL-ran-PLA) NPs were chosen based on their excellent performance on in vitro cellular targeting, in vitro antitumor efficacy, and in vivo tumor targeting. Different treatments were performed when the tumor volume reached approximately 100 mm<sup>3</sup> . The tumor-bearing mice were randomly divided into five groups

(n = 5), and different treatments were performed on Day 0, 4, 8, and 12: (1) Saline, (2) Taxotere <sup>R</sup> , (3) DTX/Apt-pD-CA-(PCLran-PLA) NPs, (4) Apt-pD-CA-(PCL-ran-PLA) NPs + NIR, and (5) DTX/Apt-pD-CA-(PCL-ran-PLA) NPs + NIR. All the formulations were i.v. injected. Taxotere <sup>R</sup> and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs were injected at a same DTX dose of 10 mg/kg in 100 µL PBS. Apt-pD-CA-(PCL-ran-PLA) NPs were injected at a same NP dose with DTX/Apt-pD-CA-(PCLran-PLA) NPs. NIR laser irradiation was performed (808 nm, 10 min, 1.5 W/cm<sup>2</sup> ) after 12 h injection of NPs. The tumor size and body weights were recorded every 2 days. After 14 days of treatment, the mice were humanely executed. Tumor growth profiles were recorded to evaluate the antitumor efficacy in this study.

Another batch of SCID mice bearing MCF-7 tumor xenograft were further used to access the antitumor efficacy of the mice through recording their survival time after receiving the same treatments mentioned above. The mice with similar physical status (i.e., age, body weight, and 100 mm<sup>3</sup> of tumor volume) were randomly divided into five groups (n = 5). The survival results were presented as Kaplane–Meier plots and evaluated using a log-rank test.

# Statistical Methodology

All tests were performed at least three times in all studies unless otherwise stated. The results are expressed as mean ± SD, and the statistical significance of the results was determined by the Student's t-test. The results were considered to be statistically significant if P < 0.05.

# RESULTS AND DISCUSSION

# Preparation of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs

The star-shaped random copolymer CA-(PCL-ran-PLA) was synthesized through a the ring opening copolymerization of ε-caprolactone (CL) and D,L-lactide (LA) with colic acid (CA) as an initiator (**Figure 1A**) (Tao et al., 2014). The fabrication of DTX-loaded Apt-pD-CA-(PCL-ran-PLA) NPs (DTX/Apt-pD-CA-(PCL-ran-PLA) NPs) was illustrated in **Figure 1B**: (i) First, a modified nano-precipitation method, which is a mild and effective way to prepare drug-loaded NPs (Zeng et al., 2013a, 2015b), was applied to prepare DTX/CA-(PCL-ran-PLA) NPs. (ii) Second, the prime-coating of pD layer was achieved through an oxidative polymerization reaction (pH 8.5) to prepare DTX/pD-CA-(PCL-ran-PLA) NPs. (iii) Finally, conjugation of aptamers on the surface of DTX/pD-CA-(PCL-ran-PLA) NPs could be fulfill via a Michael addition reaction in a weak alkaline solution (pH 8.0). The suspensions gradually turned darker in the second step after dopamine hydrochloride was added, indicating the formation of pD layer. A 26-mer SH-terminated DNA aptamer-AS1411, which possesses a thiol group at the 3<sup>0</sup> - end for effective conjugation to the pD layer, was applied to provide high affinity to interact with breast cancer cells (Cao et al., 2009).

# Characterization of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs

The size and surface properties of polymeric NPs, which play a vital role on their drug release, endocytosis and in vivo pharmacokinetics (Tao et al., 2016; Behzadi et al., 2017; Ding et al., 2017), have been well-characterized in this study. As shown in **Figures 2a–c**, the surface morphology of DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/AptpD-CA-(PCL-ran-PLA) NPs was characterized through TEM images. It could be obviously observed that spherical films have been deposited on the surface of DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs, which is quite different from bare DTX/CA-(PCL-ran-PLA) NPs and gives direct evidences on the formation of pD layers. The average size distribution in TEM images is about 80 nm for DTX/CA-(PCL-ran-PLA) NPs, and about 95 nm for DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs (**Figures 2d–f**). As presented in **Table 1**, the average DLS size of DTX/CA-(PCL-ran-PLA) NPs was 103.4 ± 3.3 nm (PDI 0.126), while that of DTX/pD-CA-(PCL-ran-PLA) NPs was 120.3 ± 4.6 nm (PDI 0.115) and that of DTX/Apt-pD-CA-(PCLran-PLA) NPs was 124.6 ± 5.1 nm (PDI 0.123). The increased DLS size of DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/AptpD-CA-(PCL-ran-PLA) NPs could be attributed to the thickness of pD layers formed on the surface of DTX/CA-(PCL-ran-PLA) NPs, which also indicates the successful modification of pD layers through the oxidative polymerization reaction. With the mean hydrodynamic size of NPs ranging from about 100–120 nm in diameter, they are perfectly fit for high accumulation in tumor vasculature driven by the influence of the enhanced permeability and retention (EPR) effect (Zhu et al., 2015; Zhu X. et al., 2016, 2017; Liu et al., 2017). The smaller size obtained from TEM images (compared with DLS testing) may be contributed to the shrink and collapse of NPs when in the dry state (Tao et al., 2013). The stability of NPs was accessed by monitoring particle size in PBS over a span of 2 weeks (**Figure 3A**). The average size of all the NPs hardly changed during the testing period, indicating a great stability of the NPs.

As shown in **Table 1**, the absolute value of zeta potential of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs (−19.2 ± 5.2 mV) increased slightly or kept similarly compared to DTX/pD-CA- (PCL-ran-PLA) NPs (−18.6 ± 3.6 mV) and DTX/CA-(PCLran-PLA) NPs (−17.8 ± 3.9 mV). The negative charge of NPs could benefit the blood compatibility and passive accumulation of the NPs in tumor sites through reducing clearance by the reticuloendothelial system (RES) including liver. Moreover, **Table 1** also showed that the LC and EE of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs (9.73 ± 0.46%, 94.18 ± 2.76%) were almost the same as those of DTX/pD-CA-(PCL-ran-PLA) NPs (9.98 ± 0.39%, 94.31 ± 1.98%) and DTX/CA-(PCL-ran-PLA) NPs (10.02 ± 0.28%, 95.01 ± 2.16%), indicating this strategy is facile and does not affect the effectiveness of drug LC and EE.

# In Vitro Drug Release Kinetics

**Figure 3B** showed the in vitro DTX release from DTX/CA- (PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and

TABLE 1 | Characterization of DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs.


PDI, polydispersity index; ZP, zeta potential; LC, loading content; EE, encapsulation efficiency, n = 3.

DTX/Apt-pD-CA-(PCL-ran-PLA) NPs in the release medium (pH 7.4, PBS containing 0.1% w/v Tween 80) at 37◦C. An initial burst release of DTX could be observed within the first 48 h, as about 40% of the loaded drugs were released in all kinds of NPs. At the end of our release studies (Day 15), about 66.8, 63.2, and 62.8% of DTX were released from DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs, respectively. Notably, the release rates from DTX/CA- (PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs were quite similar, which means the pD modification and AS1411 functionalization did not actually change the drug release properties of the prepared NPs. Therefore, Apt-pD-CA-(PCL-ran-PLA) NPs may be a promising drug delivery platform in nano-biotechnology and nanomedicines.

# In Vitro Photothermal Properties of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs

The photothermal properties of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs was accessed by testing the temperature changes under the irradiation of an 808-nm NIR laser. **Figures 4A,E** showed a rapid temperature increase of ∼20◦C was achieved by DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/Apt-pD-CA- (PCL-ran-PLA) NPs (500 µg/mL) after 10 min of irradiation with the NIR laser (1.0 W/cm<sup>2</sup> ). However, DI water and the DTX/CA-(PCL-ran-PLA) NPs showed little temperature change at totally the same conditions, demonstrating the superior photothermal effect of these NPs was majorly attributed to the pD layers. As shown in **Figure 4B**, the photothermal effect of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs had a concentrationdependent relationship. The solutions of DTX/Apt-pD-CA- (PCL-ran-PLA) NPs also exhibited a laser power intensitydependent manner as the temperatures increased with the rise of laser power intensity (**Figure 4C**). Moreover, the temperature changes did not exhibit noticeable difference during five cycles of "laser on"-"laser off " (**Figure 4D**), indicating an excellent photo-stability of DTX/Apt-pD-CA- (PCL-ran-PLA) NPs. Therefore, DTX/Apt-pD-CA-(PCL-ran-PLA) NPs have a good photothermal effect and a superior photostability, which could be applied in cancer photothermal therapy.

#### Endocytosis of Fluorescent NPs and in Vitro Cellular Targeting

Since the endocytosis and sustained retention of NPs acts as a critical and final step on the therapeutic efficacy of NPs (Xu et al., 2012; Behzadi et al., 2017), the endocytosis of Apt-pD-CA-(PCL-ran-PLA) NPs was carefully investigated. We replaced the loaded DTX with a fluorescent molecule, coumarin-6 (C6), in order to study the endocytosis of all the developed NPs in MCF-7 cells. The CLSM images of MCF-7 cells after 2 h of incubation with 250 µg/mL of C6/CA-(PCL-ran-PLA) NPs, C6/pD-CA-(PCL-ran-PLA) NPs, and C6/Apt-pD-CA-(PCL-ran-PLA) NPs in DMEM are provided in **Figure 5A**. As could be observed from this figure, the fluorescence intensity (i.e., representing cellular uptake efficiency) within the MCF-7 cells did not show much difference between C6/CA-(PCLran-PLA) NPs and C6/pD-CA-(PCL-ran-PLA) NPs. Notably, the fluorescence intensity increased significantly within the MCF-7 cells incubated with C6/Apt-pD-CA-(PCL-ran-PLA) NPs, demonstrating the presence of AS1411 aptamers on NP surface may lead to high levels of in vitro cellular targeting efficacy.

In order to quantitatively verify the significant role of AS1411 aptamers during the endocytosis of NPs, MCF-7 cells were incubated with C6/CA-(PCL-ran-PLA) NPs, C6/pD-CA- (PCL-ran-PLA) NPs, and C6/Apt-pD-CA-(PCL-ran-PLA) NPs at the concentration of 100, 250, and 500 µg/mL, respectively. As shown in **Figure 5B**, the cellular uptake efficiency of C6/CA-(PCL-ran-PLA) NPs did not show much difference compared to that of C6/pD-CA-(PCL-ran-PLA) NPs at all concentrations. However, the cellular uptake efficiency of C6/Apt-pD-CA-(PCL-ran-PLA) NPs was 1.37-, 1.53-, and 1.91 fold of that of C6/CA-(PCL-ran-PLA) NPs at the concentration of 100, 250, and 500 µg/ml, respectively. In addition, similar results could also be confirmed by FCM assays (**Figure 5C**). Taken all together, the specific interactions between AS1411 aptamers and MCF-7 cells reinforcing the endocytosis of targeted C6/Apt-pD-CA-(PCL-ran-PLA) NPs compared to

non-targeted C6/CA-(PCL-ran-PLA) NPs and C6/pD-CA-(PCLran-PLA).

# Effects of the Developed NPs on Cell Viability

MTT assays were carried out to evaluate the cytotoxicity of all the developed NPs. We first checked the in vitro toxicity of drug-free CA-(PCL-ran-PLA) NPs, pD-CA-(PCL-ran-PLA) NPs, and Apt-pD-CA-(PCL-ran-PLA) NPs at different concentrations (25, 125, 250, and 500 µg/mL) and after different incubation time (24 and 48 h). As shown in **Figures 6A,B**, none of these drug free NPs showed any functionalization of NPs (pD modification and aptamer conjugation) were biocompatible and non-toxic. Furthermore, the in vitro therapeutic efficacy of DTX/CA-(PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs without (i.e., chemotherapy groups) and with (i.e., chemo-photothermal therapy groups) the irradiation of NIR laser (808 nm, 10 min, 1.5 W/cm<sup>2</sup> ) was performed in MCF-7 cells. Taxotere <sup>R</sup> , which is a clinical DTX formulation, was chosen as a reference. Group treated with only NIR laser were also chosen as a reference. MCF-7 cells were treated at 0.25, 2.5, 12.5, and 25 µg/mL of equivalent DTX concentrations for 24 and 48 h. For the chemo-photothermal therapy groups, NIR laser irradiation was performed immediately after adding different NPs before 24 or 48 h of incubation. It could be concluded from **Figure 6C**: (i) the cell viability decreased as the incubation time increased for both Taxotere <sup>R</sup> and NP groups, indicating a time-dependent and dose-dependent effect; (ii) the DTX/Apt-pD-CA-(PCL-ran-PLA) NPs showed better in vitro therapeutic efficacy than Taxotere <sup>R</sup> , DTX/CA- (PCL-ran-PLA) NPs, DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA-(PCL-ran-PLA) NPs. For instance, the cell viability of MCF-7 cells (48 h, at the DTX concentration of 12.5 µg/mL) was 48.2% for Taxotere <sup>R</sup> , 42.3% for DTX/CA- (PCL-ran-PLA) NPs, 39.4% for DTX/pD-CA-(PCL-ran-PLA) NPs, and 25.9% for DTX/Apt-pD-CA-(PCL-ran-PLA) NPs; (iii) the NIR laser irradiation did not change the viability of MCF-7 cells treated with DTX/CA-(PCL-ran-PLA) NPs. However, the cell viability significantly decreased for groups treated with DTX/pD-CA-(PCL-ran-PLA) NPs and DTX/AptpD-CA-(PCL-ran-PLA) NPs after the irradiation of NIR laser. For example, the cell viability of MCF-7 cells (48 h, at the DTX concentration of 12.5 µg/mL) was 42.8% for DTX/CA-(PCLran-PLA) NPs + NIR, 27.9% for DTX/pD-CA-(PCL-ran-PLA) NPs + NIR, and 16.2% for DTX/Apt-pD-CA-(PCL-ran-PLA)

NPs + NIR. (iv) the DTX/CA-(PCL-ran-PLA) NPs and DTX/pD-CA-(PCL-ran-PLA) NPs showed similar viability at various concentrations, further demonstrating the surface coating of pD layers is biocompatible and non-toxic. In a word, AS1411 aptamers may have an effective targeting efficacy for NPs and the synergistic chemo-photothermal strategy showed the most significant therapeutic efficacy. Therefore, DTX/Apt-pD-CA- (PCL-ran-PLA) NPs could be employed as promising targeted drug delivery systems for effective chemo-photothermal cancer therapy.

# In Vivo Photothermal Imaging of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs

Based on the promising in vitro targeting and therapeutic efficacy, we performed in vivo photothermal imaging of mice after intravenous (i.v.) injection of saline,

DTX/pD-CA-(PCL-ran-PLA) NPs, and DTX/Apt-pD-CA- (PCL-ran-PLA) NPs respectively to further access the in vivo targeting efficacy. After 12 h of i.v. injection, NIR laser irradiation (808 nm, 10 min, 1.5 W/cm<sup>2</sup> ) was performed at tumor sites. As shown in **Figure 7**, the temperature increments in the tumor sites of the mice treated with DTX/Apt-pD-CA-(PCL-ran-PLA) NPs and DTX/pD-CA-(PCL-ran-PLA) NPs were 20.8◦C and 11.1◦C respectively, reaching a temperature of 53.8◦C and 44.5◦C after irradiation. DTX/Apt-pD-CA-(PCL-ran-PLA) NPs group exhibited a higher temperature increment than DTX/pD-CA-(PCL-ran-PLA) NPs group, which could be explained by the active tumor targeting effect of AS1411 aptamers and higher accumulation of DTX/Apt-pD-CA-(PCL-ran-PLA) NPs in tumor sites. However, the tumor site temperature of mice

received i.v. injection of saline did not show much increment with a highest temperature of 36.8◦C, which was not enough for tumor ablation.

# In Vivo Antitumor Effects of NPs

The blood circulation time was firstly accessed by in vivo pharmacokinetic studies before in vivo antitumor studies. As shown in Supplementary Figure S1, all the prepared NPs have extended the blood circulation time of the DTX compared to Taxotere <sup>R</sup> , which may contribute to effective delivery of DTX in vivo. On the basis of the above in vitro cell experiments and in vivo targeting imaging studies, DTX/Apt-pD-CA-(PCLran-PLA) NPs were chosen to further verify the in vivo antitumor efficacy of this targeting chemo-photothermal strategy. As shown in **Figures 8A,B**, both commercial Taxotere <sup>R</sup> and NP treatment groups could inhibit the tumor growth. In detail, all the NP treatment groups, i.e., DTX/Apt-pD-CA-(PCL-ran-PLA) NPs (target chemotherapy group), Apt-pD-CA-(PCL-ran-PLA) NPs + NIR (target photothermal group), and DTX/Apt-pD-CA- (PCL-ran-PLA) NPs + NIR (target chemo-photothermal therapy group), showed better in vivo antitumor efficacy than Taxotere <sup>R</sup> . The synergistic chemo-photothermal therapy group showed the highest therapeutic efficacy which significantly eliminated the tumors. These results were consistent with the above cellular experiments and in vivo targeting studies. Moreover, the body weight of the mice in all the NP treatment groups did not show any difference compared to the control group receiving saline, which that of Taxotere <sup>R</sup> showed slightly weight loss, indicating the potential in vivo biocompatibility of DTX/AptpD-CA-(PCL-ran-PLA) NPs (**Figure 8C**). Another batch of nude mice were used to access the survival data of nude mice receiving different treatments (**Figure 8D**). As could be concluded from this figure, the survival time of "DTX/Apt-pD-CA-(PCL-ran-PLA) NPs + NIR" group was at least extended by 35, 28, 14, and 11 days compared to "Saline" group, "Taxotere <sup>R</sup> " group, "DTX/Apt-pD-CA-(PCL-ran-PLA) NPs" group, and "Apt-pD-CA-(PCL-ran-PLA) NPs + NIR" group. These results further demonstrated the significant therapeutic efficacy of this target chemo-photothermal therapy strategy based on DTX/Apt-pD-CA-(PCL-ran-PLA) NPs, indicating these NPs are able to be applied as promising targeting drug delivery systems for synergistic chemo-photothermal therapy of breast cancer.

# CONCLUSION

In this study, we for the first time reported the successful synthesis of robust DTX/Apt-pD-CA-(PCL-ran-PLA) NPs with star shaped CA-(PCL-ran-PLA) copolymers, which could be applied as promising targeting drug delivery systems for synergistic chemo-photothermal therapy of breast cancer. By surface modification with facile dopamine polymerization method, AS1411 aptamers were able to be simply conjugated on the surface of NPs for target delivery of drugs. The DTX/AptpD-CA-(PCL-ran-PLA) NPs were well characterized by surface morphology, LC and EE, stability, photothermal properties and drug release profiles. The in vitro and in vivo targeting effect of these NPs were also accessed. The in vitro cytotoxicity assays by MTT showed that DTX/Apt-pD-CA-(PCL-ran-PLA) NPs together with NIR laser irradiation (target chemo-photothermal therapy) could significantly inhibit cell proliferation compared with all other groups. The in vivo antitumor assays, as well as the improved survival time and reduced side effects, further confirmed the significant therapeutic effects of this target chemophotothermal therapy strategy. All the results observed from the in vivo studies were consist with the in vitro assays, suggesting the robust DTX/Apt-pD-CA-(PCL-ran-PLA) NPs are promising in target delivery of drugs and synergistic chemo-photothermal therapy of breast cancer.

### AUTHOR CONTRIBUTIONS

fphar-09-00125 February 20, 2018 Time: 16:54 # 12

NK, MD, Y-DC, and X-BS conceived the idea and designed the study. NK and MD performed all the experiments and analyzed the data. X-NS helped in nanoparticle preparation and in vitro experimental assays. Y-DC and X-BS provided the technical support and corrections of the manuscript. NK wrote the manuscript and revised it according to the comments of Y-DC and X-BS.

# REFERENCES


#### FUNDING

This research was supported by grants from National Natural Science Foundation of China (Grant No. 81672932), Zhejiang Provincial Natural Science Foundation of China for Distinguished Young Scholars (Grant No. LR18H160001), Zhejiang Province Medical Science and Technology Project (Grant No. 2017RC007), Talent Project of Zhejiang Association for Science and Technology (Grant No. 2017YCGC002), and Zhejiang Province Science and Technology Project of TCM (Grant No. 2015ZB033).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2018.00125/full#supplementary-material

and minimally invasive pharmacokinetic analysis. Nanoscale 9, 13465–13476. doi: 10.1039/c7nr02363d


microfilament imaging in living cells. Chem. Commun. 53, 7541–7544. doi: 10.1039/C7CC02555F



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Kong, Deng, Sun, Chen and Sui. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Significant Suppression of Non-small-cell Lung Cancer by Hydrophobic Poly(ester amide) Nanoparticles with High Docetaxel Loading

Xing Chen<sup>1</sup>† , Lili Zhao<sup>2</sup>† , Yang Kang<sup>3</sup>† , Zhiyu He<sup>1</sup> , Fei Xiong<sup>1</sup> , Xiang Ling<sup>1</sup> \* and Jun Wu<sup>1</sup> \*

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Guoming Sun, Sunogel Biotechnologies, Inc., United States Yue Pan, Soochow University, China Weiping Wang, The University of Hong Kong, Hong Kong

#### \*Correspondence:

Jun Wu wujun29@mail.sysu.edu.cn Xiang Ling lingx23@mail.sysu.edu.cn

†These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 18 January 2018 Accepted: 01 February 2018 Published: 28 February 2018

#### Citation:

Chen X, Zhao L, Kang Y, He Z, Xiong F, Ling X and Wu J (2018) Significant Suppression of Non-small-cell Lung Cancer by Hydrophobic Poly(ester amide) Nanoparticles with High Docetaxel Loading. Front. Pharmacol. 9:118. doi: 10.3389/fphar.2018.00118 <sup>1</sup> Key Laboratory of Sensing Technology and Biomedical Instrument of Guangdong Province, School of Engineering, Sun Yat-sen University, Guangzhou, China, <sup>2</sup> Digestive Endoscopy Center, Jiangsu Province Hospital, Nanjing, China, <sup>3</sup> Key Laboratory of Mountain Ecological Restoration and Bioresource Utilization, Chengdu Institute of Biology, Chinese Academy of Sciences, Chengdu, China

Non-small-cell lung cancer (NSCLC) accounts for over 85% of clinical lung cancer cases, which is the leading cause of cancer-related death. To develop new therapeutic strategy for NSCLC, a library of L-phenylalanine-based poly(ester amide) (Phe-PEA) polymers was synthesized and assembled with docetaxel (Dtxl) to form Dtxl-loaded Phe-PEA nanoparticles (NPs). The hydrophobic Phe-PEA polymers were able to form NPs by nanoprecipitation method and the characterization results showed that the screened Dtxl-8P4 NPs have small particle size (∼100 nm) and high Dtxl loading (∼20 wt%). In vitro experiments showed that Dtxl-8P4 NPs were rapidly trafficked into cancer cells, then effectively escaped from lysosomal degradation and achieved significant tumor cell inhibition. In vivo results demonstrated that Dtxl-8P4 NPs with prolonged blood circulation could efficiently deliver Dtxl to A549 tumor sites, leading to reduced cell proliferation, block metastasis, and increase apoptosis, then persistent inhibition of tumor growth. Therefore, Phe-PEA NPs are able to load high amount of hydrophobic drugs and could be a promising therapeutic approach for NSCLC and other cancer treatments.

#### Keywords: cancer, docetaxel, hydrophobicity, nanoparticle, poly(ester amide)

#### INTRODUCTION

Being ranked the majority of lung cancer, non-small-cell lung cancer (NSCLC) causes human death due to its relative insensitiveness to chemotherapy (Jemal et al., 2008; Umar et al., 2012). More than 60% of NSCLC patients are diagnosed to have advanced or metastatic tumors, which are unsuitable for surgical resection with curative intent (Tong, 2006; Pao and Chmielecki, 2010). Thus, alternative therapeutic platforms to control or inhibit tumor development are highly desired. Docetaxel (Dtxl) is of the chemotherapy drug class taxane, structurally similar to paclitaxel, but more effective as the inhibitor of microtubule depolymerization (Bissery, 1995). In the past decades, Taxotere has emerged as one of the most important cytotoxic agents, with proven clinical efficacy against many cancers including NSCLC (Kintzel et al., 2006; Baker et al., 2009). However, the use of

**52**

Dtxl in this formulation with non-ionic surfactant Tween 80 and 13% ethanol leads to several well-known adverse reactions due to either the agents itself (e.g., neutropenia, anemia, nephrotoxicity, neurotoxicity, and musculoskeletal toxicity) or the solvent system (e.g., hypersensitivity and fluid retention) (Persohn et al., 2005). Side effects of commercial Taxotere have considerably overshadowed its clinical application.

Nanoparticles (NPs) have been extensively reported for their prominent superiorities that can be delivered to specific sites by size-dependent passive targeting (Matsumura and Maeda, 1986; Pan et al., 2011a,b, 2012; Bertrand et al., 2014; Deng et al., 2016; Lu et al., 2016; Zhao et al., 2016; Li et al., 2017; Lin et al., 2017; Liu et al., 2017; Shi et al., 2017). Among them, numerous nanoplatforms have been utilized to deliver Dtxl for improved cancer treatment (Tao et al., 2014; Bowerman et al., 2017; Laura et al., 2017). However, the clinical application of most reported platforms is hindered by the low loading capacity of Dtxl (Guo and Huang, 2014). Thus, the carrier material with reasonable hydrophobicity is urgently needed (Chu et al., 2013). Amino acid-based poly(ester amide) polymers with both ester and amide blocks on their backbones have been widely studied over many years (Wu et al., 2012a,b; Wu and Chu, 2013), as they possess not only good biodegradability and biocompatibility but also tunable physicochemical properties, especially hydrophobicity (Wu et al., 2011, 2015; Yu et al., 2014), which may be promising for developing NPs with high Dxtl loading.

Hence, we postulated that the novel design of hydrophobic L-phenylalanine-poly(ester amide) (Phe-PEA) polymer NPs with higher Dtxl loading could bring about more effective antitumor efficiency with better in vivo tolerance. In this paper, Phe-PEA polymers comprised of phenylalanine, diacid, and diol were synthesized and used because of their excellent NP formation and drug loading capability. Dtxl-loaded Phe-PEA polymer NPs were prepared by nanoprecipitation and the physicochemical characteristics were determined. After optimization, Dtxl-8P4 NPs with attractive uptake kinetics and strong cytotoxicity were found to greatly improve circulation retention, then enhance therapeutic effects for A549 tumors with less systemic toxicity (**Figure 1**).

# MATERIALS AND METHODS

#### Materials and Reagents

L-Phenylalanine, 1,4-butanediol, 1,6-hexanediol, adipoyl dichloride, sebacoyl dichloride, toluene-4-sulfonic acid monohydrate, and p-nitrophenol were purchased from Sigma–Aldrich and used without further purification. DMSO, acetone, toluene, triethylamine, ethyl acetate, and methanol were purchased from Aladdin. Dtxl was purchased from LC Laboratories. Taxotere <sup>R</sup> was purchased from Sanofi Aventis. Dil was purchased from Thermo Fisher Scientific, United States. 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy- (polyethylene glycol)-3000] (ammonium salt) (DSPE-PEG 3000) was purchased from Avanti Polar Lipids. F-12K, EMEM, trypsin-EDTA, fetal bovine serum (FBS), penicillin–streptomycin solution, phosphate-buffered saline (PBS), and water were provided by Gibco <sup>R</sup> .

#### Synthesis of Monomers and Polymers

The synthesis of Phe-PEA polymers was divided into following steps (Katsarava et al., 1999): (i) preparation of di-p-nitrophenyl esters of dicarboxylic acid (monomer I) by condensation reaction; (ii) preparation of toluene-4-sulfonic acid salts of bis(Phe) alkylene diesters (monomer II) via solid–liquid reaction; (iii) preparation of Phe-PEA polymers through solution polycondensation (**Figure 2**). Synthetic details of monomers I and II could be found in previous reports (Wu and Chu, 2012), while Phe-PEA polymers (yield > 80%) were obtained by optimized protocols (Fonseca et al., 2014): monomers I (5 mmol) and II (5 mmol) in dry DMSO (8 ml) were mixed well by vortex and kept at 120◦C under vigorous stirring, then triethylamine (15 mmol) was drop-wisely added to get a uniform yellow solution. The mixture was kept at 80◦C overnight without stirring and resulting polymers were precipitated by adding cold ethyl acetate, washed with methanol, and dried under vacuum.

Here, two kinds of monomers I were prepared: di-pnitrophenyl adipate (N4, x = 4) and di-p-nitrophenyl sebacate (N8, x = 8). Two kinds of monomers II were prepared: toluene-4-sulfonic acid salts of bis(Phe) butane diesters (Phe-4, y = 4) and toluene-4-sulfonic acid salts of bis(Phe) hexane diesters (Phe-6, y = 6). Phe-PEA polymers (x-Phe-y) were prepared by solution polycondensation of monomers I and II at various combinations and summarized in **Table 1**, where x was the numbers of methylene in diacid and y was the numbers of methylene in diol. Chemical structures of above monomers and polymers were confirmed by <sup>1</sup>H-NMR (Avance III, Bruker, Switzerland). All the spectra were the same as previously reported (**Supplementary Figure S1**) (Katsarava et al., 1999).

For measuring of molecular weight (MW) of Phe-PEAs, gel permeation chromatography (GPC) method was used and the PEA solutions were prepared at a concentration of 1 mg/ml in a tetrahydrofuran (THF) solution. The MWs of Phe-PEAs were determined from a standard curve generated from polystyrene standards with MWs ranging from 841.7 to 2.93 kDa that were chromatographed under the same conditions as the samples. The standard curve was generated from a third-order polynomial fit of the polystyrene standard MWs (**Table 2**).

#### Preparation and Characterization of Nanoparticles (NPs)

Dtxl-8P4 NPs were prepared by nanoprecipitation method: 6 mg of 8-Phe-4 polymer and a certain amount of Dtxl (10, 20, or 30 wt% of NPs) was dissolved in 0.2 ml of DMSO. Next, the mixture was dropwise added to 10 ml of aqueous solution containing DSPE-PEG 3000 (20 wt% of

TABLE 1 | Phe-PEA polymers and their corresponding NPs.


TABLE 2 | Molecular mass characteristics of Phe-PEA polymers.


NPs) under vigorously stirring. The remaining free molecules and organic solvent were removed by washing with PBS twice using Amicon Ultra-15 centrifugal filters (MWCO 100 KDa, Millipore, United States). Finally, Dtxl-8P4 NPs were dispersed in 1.0 ml of PBS for further use. 8P4 NPs were prepared without Dtxl and used as blank control. Dtxl-4P4, Dtxl-4P6, or Dtxl-8P6 NPs were prepared by the similar procedure using 6 mg of 4-Phe-4, 4-Phe-6, or 8-Phe-6 polymer and fixed feeding of Dtxl (10 wt% of NPs). Dil-8P4 NPs were prepared by mixing pre-determined amounts of 8-Phe-4 polymer, Dtxl, and Dil (3 wt% of NPs) in DMSO, then following above nanoprecipitation procedure. Particle size and zeta potential of NPs were measured by dynamic light scattering (DLS, Zetasizer Nano-ZS90, Malvern, United Kingdom). Morphology of NPs was visualized by transmission electron microscopy (TEM, Tecnai G2 Spirit, FEI, United States).

Docetaxel loading capacity of NPs was determined by Agilent 1260 HPLC with a ZORBAX Extend-C18 column at the temperature of 30◦C and a flow rate of 1.0 ml/min (mobile phase, water:acetonitrile = 50:50). The injection volume was 20 µl for each sample. The UV detection wavelength was 232 nm and Dtxl loading was calculated using following equation:

Loading capacity % = Weight of loaded drugs Weight of polymers <sup>+</sup> Weight of loaded drugs <sup>×</sup> <sup>100</sup>.

#### In Vitro Release Profiles

Dtxl-8P4 NPs were transferred to dialysis bags (MWCO 3500 Da, Spectrum, United States) and immersed in PBS (pH 5.0 or 7.4). Dtxl release was conducted at 37◦C with constant stirring at 100 rpm. At specific time intervals, 1 ml of the sample solution was collected and replaced with equal volume of fresh PBS.

The amounts of Dtxl were analyzed by UPLC-MS/MS (TSQ Quantum Access Max, Thermo Fisher Scientific, United States) with following MS ionization parameters: positive ESI mode; spray voltage, 3500 V; ion source temperature, 300◦C; collision energy, 0 eV. The analytes were quantified by using Multiple Reaction Monitoring (MRM) to monitor ion transitions m/z of 830.2–303.7. Chromatography was performed via Agilent 1100 HPLC with an Ultimate XB-C18 column at the temperature of 40◦C and a flow rate of 0.2 ml/min (mobile phase, 0.1% formic acid:methanol = 40:60). The gradient elution was 60% methanol at 0–0.30 min, 60–100% methanol at 0.30–0.50 min, 100% methanol at 0.50–2.00 min, 100–60% methanol at 2.00– 2.50 min, and 60% methanol at 2.50–5.00 min.

#### Cell Culture

A549, PC3, and DU145 cells were purchased from American Type Culture Collection (ATCC) and cultured by recommended protocols from the manufacturer. Cells were grown in the corresponding medium, supplemented with 10% FBS and 1% penicillin–streptomycin solution, maintained at 37◦C and 5% CO2.

#### Cellular Uptake

A549 cells were seeded in six-well plates (20,000 cells per well) and incubated with 1 ml of complete medium for 24 h. Dil-8P4 NPs at different concentrations were added. At selected time points, cells were washed with cold PBS twice, harvested by trypsinization, centrifuged, and resuspended in 4% formaldehyde, then analyzed by flow cytometer (FACSCalibur, BD, United States).

#### Cellular Internalization

A549 cells were seeded in 35-mm dishes (20,000 cells per well) and incubated with 1 ml of complete medium for 24 h. Dil-8P4 NPs were added. At selected time points, cells were washed with cold PBS twice and fixed with 4% formaldehyde at 37◦C for 15 min. Subsequently, cells were washed with PBS twice again and stained with LysoTracker green and Hoechst 33342, then observed under an FV3000 confocal laser scanning microscope (CLSM, Olympus, Japan).

#### Cytotoxicity

Cytotoxicity was evaluated with AlamarBlue Cell Viability Assay (Thermo Fisher Scientific, United States) against A549, PC3, and DU145 cells. Cells were seeded in 96-well plates (5000 cells per well) and incubated with 0.1 ml of complete medium for 24 h. Cells were incubated with different concentrations of 8P4 NPs, Taxotere, or Dtxl-8P4 NPs. At selected time points, cells were treated according to manufacturer's protocol using a microplate reader (Synergy4, Bio Tek Instruments, United States).

#### Animals

BALB/c mice (male, 4–5 weeks old), nude mice (female, 4 weeks old), and SD rats (male, 200–220 g) were provided by the Laboratory Animal Center of Sun Yat-sen University. This animal study was carried out in accordance with the recommendations of "the guidelines of the Experimental Laboratory Animal Committee of Sun Yat-sen University and the National Institutes of Health's Guide for the Care and Use of Laboratory Animals."

TABLE 3 | Dtxl-8P4 NPs at different Dtxl feeding.


The animal protocol was approved by the "Experimental Laboratory Animal Committee of Sun Yat-sen University."

The human pulmonary carcinoma xenograft model was established by subcutaneously injecting A549 cell suspension (2,000,000 cells in medium and Matrigel) into the back region of nude mice. As the volume of xenograft tumor reached ∼100 mm<sup>3</sup> , mice were used for following experiments.

#### Pharmacokinetics

SD rats (n = 3 per group) were intravenously injected with either (i) PBS, (ii) Taxotere, or (iii) Dtxl-8P4 NPs at a dose of 5 mg Dtxl/kg, respectively. At pre-determined time points, blood was withdrawn from retro-orbital plexus and plasma was collected. Dtxl concentrations were analyzed by UPLC-MS/MS with the same chromatographic condition as described above. Hundred microliters of plasma was mixed with 10 µl of Dtxl, followed by adding 500 µl of methyl tert-butyl ether, vortexed, and centrifuged at 12,000 rpm for 10 min. The supernatant was evaporated and re-constituted with mobile phase. Pharmacokinetic parameters were calculated with Phoenix WinNonlin 6.3 program (Pharsight Corporation, St. Louis, MO, United States).

#### Histology Analysis

BALB/c mice (n = 5 per group) was intravenously injected with either (i) PBS, (ii) 8P4 NPs, (iii) Taxotere, or (iv) Dtxl-8P4 NPs at a dose of 5 mg Dtxl/kg every 7 days. After 39 days, all the mice were sacrificed and major organs (heart, liver, spleen, lung, kidneys) were excised for hematoxylin and eosin (HE) staining. Slides were observed under a fully automated upright microscope (DM6000 B, Leica, Germany).

#### In Vivo Antitumor Efficacy

A549 tumor-bearing nude mice (n = 5 per group) was intravenously injected with either (i) PBS, (ii) 8P4 NPs, (iii) Taxotere at a maximum tolerated dosage (MTD) of 5 mg Dtxl/kg, (iv) Dtxl-8P4 NPs at a dose of 5 mg Dtxl/kg, or (v) Dtxl-8P4 NPs at a dose of 10 mg Dtxl/kg every 7 days. Body weights and tumor volumes were recorded every 2 days. Tumor volume was calculated as follows:

$$\text{Volume} = \text{length} \times \text{width}^2 / 2.$$

#### HE, IHC, and TUNEL

A549 tumor-bearing nude mice (n = 5 per group) was intravenously injected with either (i) PBS, (ii) 8P4 NPs, (iii) Taxotere, or (iv) Dtxl-8P4 NPs at a dose of 5 mg Dtxl/kg every 7 days. After 39 days, tumors were quickly excised for HE and IHC. Slices were incubated with primary antibodies of CD31 and MMP2 (Cell Signaling) and HRP/DAB Detection IHC Kit (Abcam) according to the manufacturers' instructions.

(D) PC3 cells after exposure for 48 h.

Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) was performed according to the instruction of In Situ Cell Death Detection Kit (Roche).

#### Statistical Analysis

Results were expressed as mean ± SD and repeated at least three times. Two-tailed Student's t-test was applied to analyze the statistical significance of difference between two groups, oneway analysis of variance (ANOVA) was used for multiple groups. Statistical significance was set at <sup>∗</sup>p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

#### RESULTS

#### The Hydrophobic Nature of Phe-PEA Polymers Induced Formation of Dtxl-8P4 NPs with Small Particle Size and High Dtxl Loading

L-Phenylalanine-based poly(ester amide) polymers were prepared by solution polycondensation of monomers I and II at various combinations. By introducing diacid or diol segments with different alkyl chains, the hydrophobic nature of Phe-PEA polymers changed accordingly. As the length of alkyl chain increased, the hydrophobicity as well as loading capacity of polymers enhanced, while the formed NPs tended to possess smaller size, which might be caused by the formation of more dehydrated and compacted cores via hydrophobic force. Due to high drug loading, satisfying entrapment efficiency and reliable stability, Dtxl-8P4 NPs were selected for the following experiments (**Table 3**).

Transmission electron microscopy images showed that the loading of a relatively higher amount of Dtxl (∼16 wt%) into 8P4 NPs did not significantly alter the spherical morphology of NPs, but slightly increased their particle size (**Figures 3A–C**).

#### Dtxl-8P4 NPs Exhibited Attractive Uptake Kinetics and Strong Cytotoxicity in Vitro

In vitro release profiles were obtained by representing the percentage of Dtxl released with respect to the amount of drug loaded into NPs. **Figure 3D** demonstrated a sustained-release phase, in which ca. 82.3 and 96.5% of Dtxl were released from Dtxl-8P4 NPs in 144 h at pH 7.4 and 5.0, respectively. This

sustained release could mainly result from the erosion and degradation of the components of NPs. Importantly, no burst effect was observed, further confirming that Dtxl incorporated into 8P4 NPs was likely to remain association with NPs and be taken up into cells as the form of particles rather than free drugs.

Cellular internalization of NPs was performed by labeling A549 cells with a specific fluorescent probe, Dil, which was entrapped into 8P4 NPs at a minor amount to minimize the effect on intracellular trafficking. As displayed in **Figure 4A**, prolonging incubation time or increasing incubation concentration of Dil-8P4 NPs resulted in higher internalization, suggesting timeand concentration-dependent manners. After trans-membrane transport, Dil-8P4 NPs were found to co-localize with endosomes within 1 h, confirming a relatively fast cellular uptake (**Figure 5**). After that, the red fluorescence was mainly observed in cytoplasm, effectively avoiding the fate of lysosomal degradation. As a next step, in vitro cytotoxicity of Dtxl-8P4 NPs was tested against A549, PC3, and DU145 cells by AlamarBlue (**Figures 4B–D**). The viability of cells treated with Dtxl-8P4 NPs did not demonstrate a significant difference compared with Taxotere at low drug levels, but the inhibiting activity of NPs increased at high drug levels. In addition, blank 8P4 NPs without Dtxl had a negligible toxic effect at all test concentrations.

### Dtxl-8P4 NPs Enhanced Therapeutic Effects for A549 Tumors with Less Systemic Toxicity

In order to verify whether Dtxl-8P4 NPs impaired major organs, pathological examination was evaluated in healthy BALB/c mice (**Figure 6**). Taxotere caused severe hepatotoxicity with several structural and metabolic changes, i.e., vacuolar degeneration

TABLE 4 | Pharmacokinetic parameters of Dtxl after intravenous injection of Dtxl-8P4 NPs or Taxotere.


and inflammatory cell infiltration; splenic nodule atrophy; pulmonary hemorrhage; and tubular dilation with flattening of renal epithelium cells. However, no any noticeable histological alternations were captured in Dtxl-8P4 NPs as well as PBS and 8P4 NPs groups, confirming good biocompatibility.

The plasma Dtxl concentrations vs. time profiles were shown in **Figure 7A**. Pharmacokinetics presented the remarkably enhanced retention of Dtxl-8P4 NPs in blood circulation, whereas Taxotere exhibited the rapid elimination from circulation system. Non-compartmental and twocompartmental analysis showed significant changes in pharmacokinetic parameters of Dtxl (**Table 4**). Area under the curve (AUC0→inf), area under the first moment curve (AUMC0→inf), and mean residence time (MRT0−inf) of NPs

FIGURE 7 | (A) Pharmacokinetics after intravenous injection of Dtxl-8P4 NPs and Taxotere at a dose of 5 mg Dtxl/kg (n = 3 per group). (B) Tumor growth and (C) body weight curves of A549 tumor-bearing nude mice (n = 5 per group) treated with PBS, 8P4 NPs, Dtxl-8P4 NPs, and Taxotere.

were 4.7-, 7.6-, and 1.6-fold higher than Taxotere, while clearance (CL) and volume of distribution (Vss) were reduced by 78.6 and 68.1%, respectively. t1/<sup>2</sup> of distribution and elimination phase was all dramatically extended compared with Taxotere.

To prove the potential of Dtxl-8P4 NPs for tumor growth suppression, a schedule of multiple dosing was applied since day 9 after A549 tumor implantation (**Figures 7B,C**). PBS and 8P4 NPs groups exhibited rapid tumor growth, whereas MTD for weekly dosing of Taxotere significantly delayed tumor growth. In comparison, a better tumor inhibition with sustaining weight gain was observed in mice receiving equal dosing of Dtxl-8P4 NPs. What's more, the most aggressive treatment with double dosing of chemotherapy suppressed tumor growth for longer but barely induced weight loss, probably due to biocompatible Dtxl-8P4 NPs prevented the random drug release in the body and enhanced therapeutic efficacy of Dtxl.

#### Dtxl-8P4 NPs Effectively Suppressed Proliferation, Metastasis, and Apoptosis of Tumors

The pathology of tumor tissues revealed coincident results (**Figure 8**). For the vigorous growth of tumors in PBS and 8P4 NPs groups, nuclei and cytoplasm presented a blue– pink interlaced image on the whole section. Once tumors underwent apoptosis, nuclei disappeared and cytoplasm became

an amorphous mass of necrotic material. The destructed tumor area of Dtxl-8P4 NPs accounted for the highest percentage among all the groups, further revealing the enhanced chemotherapeutic efficiency of NPs. Besides, the histological analysis of proliferation, metastasis, and apoptosis for tumors treated with rounds of chemotherapy was carried out through IHC and TUNEL. Dtxl-8P4 NPs greatly suppressed the expression of CD31 and MMP2, compared with other groups, verifying that tumor proliferation and metastasis were effectively restricted (Giatromanolaki et al., 1997; Bjorklund and Koivunen, 2005; Kim et al., 2009; Jacob and Prekeris, 2015). The administration of PBS or 8P4 NPs caused negligible TUNEL-positive staining, while Dtxl-8P4 NPs resulted in the most remarkable apoptosis of tumors, emphasizing the great efficacy of NPs.

#### DISCUSSION

In summary, amino acid-based Phe-PEA polymers were synthesized and formulated with Dtxl to construct Dtxl-loaded Phe-PEA polymer NPs. The hydrophobic nature of polymers contributed to the installation of high hydrophobic payloads. Dtxl-8P4 NPs showed the small particle size ∼100 nm with high loading capacity ∼20 wt%, a low burst effect, and a sustained drug release in vitro. Cytotoxicity of Dtxl-8P4 NPs against tumor cells was superior to Taxotere would be attributed to the rapid cellular uptake and effective lysosomal escape. The better antitumor efficacy and less systemic toxicity of Dtxl-8P4 NPs might be attributed to extended blood circulation and high Dtxl loading. Thus, Dtxl-8P4 NPs could be promising as a novel formulation of Dtxl in cancer chemotherapy to fight against NSCLC.

#### REFERENCES


#### AUTHOR CONTRIBUTIONS

XC and LZ conceived and directed the study. YK, ZH, and FX performed syntheses and spectroscopic studies. XC and YK co-wrote the paper. XL and JW oversaw the project and contributed to the execution of the experiments and interpretation of the results. All authors contributed to the characterizations and discussion and reviewed and approved the final paper.

#### FUNDING

This work was supported by the Thousand Talents Plan for Young Professionals, the National Natural Science Foundation of China (21704104), West Light Foundation of the Chinese Academy of Sciences (Y7C1021100), the Science and Technology Planning Project of Guangdong Province (2016A010103015), and the Science and Technology Program of Guangzhou (201707010094).

#### ACKNOWLEDGMENTS

We appreciate the valuable and insightful thoughts from all the members of the Wu Laboratory.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2018.00118/full#supplementary-material

FIGURE S1 | <sup>1</sup>H-NMR of Phe-6 (solvent: DMSO-d6).


chronic angiogenesis. Am. J. Pathol. 174, 1972–1980. doi: 10.2353/ajpath.2009. 080819


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Chen, Zhao, Kang, He, Xiong, Ling and Wu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Poly(Ethylene Glycol)–Polylactide Micelles for Cancer Therapy

Jixue Wang1,2, Shengxian Li 1,2, Yuping Han<sup>3</sup> \*, Jingjing Guan<sup>1</sup> \*, Shirley Chung<sup>4</sup> , Chunxi Wang<sup>1</sup> and Di Li <sup>2</sup> \*

*<sup>1</sup> Department of Urology, The First Hospital of Jilin University, Changchun, China, <sup>2</sup> Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, China, <sup>3</sup> Department of Urology, China-Japan Union Hospital of Jilin University, Changchun, China, <sup>4</sup> Department of Biology, University of Waterloo, Waterloo, ON, Canada*

#### Edited by:

*Wei Tao, Harvard Medical School, United States*

#### Reviewed by:

*Yingjie Yu, Tufts University, United States Huihui Kuang, Independent Researcher, United States Mahavir Bhupal Chougule, University of Mississippi, United States*

#### \*Correspondence:

*Yuping Han hyp181818@126.com Jingjing Guan guanjingjing99@126.com Di Li lidi@ciac.ac.cn*

#### Specialty section:

*This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology*

> Received: *12 January 2018* Accepted: *22 February 2018* Published: *08 March 2018*

#### Citation:

*Wang J, Li S, Han Y, Guan J, Chung S, Wang C and Li D (2018) Poly(Ethylene Glycol)–Polylactide Micelles for Cancer Therapy. Front. Pharmacol. 9:202. doi: 10.3389/fphar.2018.00202* For the treatment of malignancy, many therapeutic agents, including small molecules, photosensitizers, immunomodulators, proteins and genes, and so forth, have been loaded into nanocarriers for controllable cancer therapy. Among these nanocarriers, polymeric micelles have been considered as one of the most promising nanocarriers, some of which have already been applied in different stages of clinical trials. The successful advantages of polymeric micelles from bench to bedside are due to their special core/shell structures, which can carry specific drugs in certain disease conditions. Particularly, poly(ethylene glycol)–polylactide (PEG–PLA) micelles have been considered as one of the most promising platforms for drug delivery. The PEG shell effectively prevents the adsorption of proteins and phagocytes, thereby evidently extending the blood circulation period. Meanwhile, the hydrophobic PLA core can effectively encapsulate many therapeutic agents. This review summarizes recent advances in PEG–PLA micelles for the treatment of malignancy. In addition, future perspectives for the development of PEG–PLA micelles as drug delivery systems are also presented.

Keywords: polylactide, poly(ethylene glycol), micelle, nanocarrier, controlled drug release, antitumor treatment

# INTRODUCTION

Cancer is one of the major health problems that threaten human life. According to the worldwide statistic, there were 14.1 million new tumor incidences and 8.2 million cancer-related deaths in 2012 (Scsukova et al., 2015). Cancer cells proliferate uncontrollably and rapidly, thus they are characterized by the development of abnormalities and the combination of mutagenic stages. Moreover, they can realize self-sufficiency in growth signals, resistance to growth inhibition and evasion of apoptotic cues (Luo et al., 2009). Furthermore, tumors can induce angiogenesis, evasion from immune surveillance, and metastasis to distant sites through interactions with surrounding stromal cells (Mohme et al., 2017). All of these reasons have led to the refractoriness of cancer. To date, chemotherapy, radiation, surgery, and hormonal therapies are still the major treatment methods for cancer in clinics. Whereas in the research industry, many other treatments, such as photodynamic therapy, photothermal therapy, gene therapy, immunotherapy, and so forth, are being studied. Despite the relatively satisfactory results these therapeutic agents exhibit, they also possess many disadvantages, including poor pharmacokinetics, unspecific bio-distribution, and low targeting ability. The poor solubility and hydrophobicity are considered the major hurdles when therapeutic agents are applied in cancer therapy. Therefore, it is urgent and necessary to overcome these shortcomings to enhance the anti-tumor efficiency.

With the development of nanotechnology, nanomaterials have been widely used in biological application, such as biosensor, tissue engineering as well as drug delivery (Yu et al., 2016a,b,c; Zhang et al., 2017). Nanocarriers have attracted more and more attention in cancertherapy owing to their unique properties, such as nanoscaled size, high surface-to-volume ratio, and favorable physico-chemical characteristics. Various nanocarriers, including liposomes, micelles, and nanocapsules, have been studied in anticancer trials (He et al., 2016; Hofmann et al., 2017; Niu et al., 2017; Tao et al., 2017). They have the capacity to modulate both pharmacokinetic and pharmacodynamic properties, thereby improving their therapeutic index. Among these nanocarriers, polymeric micelles have gained considerably more attention as a multifunctional drug delivery system for poorly water-soluble agents.

Polymeric micelles are the nano-scaled sized particles (5–200 nm) which are self-assembled by amphiphilic polymers. They consist of two parts: the hydrophobic part on the inside (core) and hydrophilic part on the outside (shell). Therefore, the hydrophobic core can serve as a solubilization depot for agents with poor aqueous solubility. The hydrophilic shell provides advantages including longer blood circulation time and increased stability in the blood. In addition, polymeric micelles can be functionalized with targeting ligands to enhance tumor accumulation. As a result, the role of polymeric micelles in delivery of hydrophobic therapeutic agents for anticancer therapy is promising and opportunistic.

Polyethylene glycol (PEG)-polylactide (PLA) is one of the most prominent amphipathic polymers, therefore it is very suitable for constructing micelles. PLA is a form of biodegradable and biocompatible polyester derived from renewable resources and approved by the Food and Drug Administration (FDA) for clinical use. The hydrophobicity of PLA makes it suitable for the hydrophobic portion of micelles. PLA has three types of stereoisomers: poly(L-(-)-S-lactide) (PLLA), poly(D-(+)-Rlactide) (PDLA), and racemic PDLLA. Interestingly, PLLA and PDLA can form stereocomplexes through physical association of PLLA and PDLA chains (Ikada et al., 1987). PLA can interact with different hydrophilic agents, such as PEG (Danafar et al., 2017), poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC) (Long et al., 2016), poly(ethylene oxide) (PEO) (Fang et al., 2015), and poly(N-isopropylacrylamide) (PNIPAAm) (Wei et al., 2009), to form amphiphilic block copolymers and further selfassemble into micelles. Among them, PEG is the most popular hydrophilic agent due to its various advantages, including linearity, lack of charge, immunogenicity, low polydispersity, and easy activation for conjugation. PEG–PLA micelles have been widely used as drug delivery systems for cancer therapy owing to the excellent physicochemical and biological properties, namely nontoxicity, non-protein adsorption, and weakened uptake by the reticulo-endothelial system (RES) after intravenous injection (Wang et al., 2015, 2017). Notably, Genexol <sup>R</sup> - PM, a paclitaxel formulation based on PEG–PLA copolymer micelles, was approved in Korea in 2007 for the treatment of breast, lung, and ovarian cancers (Luo et al., 2012). Moreover, it is currently under clinical development in the USA (Lee et al., 2008).

In this review paper, an overview of PEG–PLA-based micelles utilized for the effective delivery of therapeutic agents possessing varying mechanisms for cancer treatment is discussed, as shown in **Scheme 1**. Additionally, the applications of different treatment modalities are described in detail. Features of nanocarriers in references are shown in **Table 1**. In particular, the development of stimuli-responsive, targeted-modified, and multifunctional PEG–PLA micelles are also highlighted.

# CHEMOTHERAPEUTICS

In clinical settings, surgery and radiotherapy are the most commonly used and effective therapeutic means for local and non-metastatic tumors. However, they are inefficient for metastatic tumors. Currently, application of anti-cancer drugs, such as chemotherapeutic drugs, hormone drugs, and biological drugs, has become the main method of treatment. These anticancer drugs are able to reach all parts of the body via the bloodstream and primarily inhibit the rapid replication of tumor cells. Unfortunately, they also inhibit the rapid growth of healthy cells which are crucial to maintaining normal function of the organism, such as hair follicles, bone marrow, and gastrointestinal tract cells (Chabner and Roberts, 2005). As a result, serious side effects are apparent in chemotherapeutic drug treatments. Due to the severe toxicity and concomitant multidrug resistance of conventional chemotherapeutic drugs, it is urgent to find new effective drug carriers to solve these problems. PEG–PLA micelles provide significant advantages over standard treatments. Drug-loaded micelles exhibit a huge potential for tumor drug delivery to overcome the limitations of chemotherapeutic agents.

As generally known, many chemotherapeutic drugs are not highly water-soluble. Hydrophobic agents are associated with several problems in therapeutic applications, such as poor absorption, bioavailability, and drug aggregation-related complications. Fortunately, polymeric micelles constructed from amphiphilic copolymers can promisingly increase the water solubility of such hydrophobic chemotherapeutic drugs by 10–5000 folds (Savic et al., 2006 ´ ). Puntawee et al. improved the aqueous solubility and bioavailability of semisynthetic andrographolide analog (19-triphenylmethyl ether andrographolide, AG 050) by utilizing the PEG-b-PLA micelles. As a result, PEG-b-PLA micelle was able to significantly increase the encapsulation efficiency of hydrochloride salt of AG 050 (AG 050-P) in aqueous solution (280-fold) (Puntawee et al., 2016).

It is worth noting that conventional chemotherapeutic drugs are quite small in size. As a result, they are rapidly cleared from the bloodstream, thus leading to decreased concentration within the tumor (Allen and Cullis, 2004). When chemotherapeutic drugs are loaded into PEG–PLA micelles, their circulation time in the bloodstream will be prolonged, therefore allowing adequate amounts of drugs to reach the target site (Jin et al., 2017). Furthermore, compared to other drug carriers, micelles have the advantage of possessing a very small size (10– 100 nm). This is critical for passive accumulation through the leaky vasculature and into the tumor site via the enhanced

permeability and retention (EPR) effect to solid tumors, particularly to poorly vascularized tumors (Davis and Shin, 2008; Feng et al., 2017). Two main factors are responsible for the EPR effect: (1) the angiogenic vasculature of tumor sites have higher permeability than normal vasculature because of their discontinuous endothelia and (2) the lymphatic drainage is not fully developed in tumors. These conditions result in the exudation of colloidal particles through the "leaky" endothelial layer of tumor vascular and subsequently accumulating in tumor tissues. Majority of vascular pore cutoff size of tumor sites is between 380 and 780 nm (Hobbs et al., 1998). Shi et al. synthesized a docetaxel (DTX)-conjugated mPEG–PLA micelle via an ester linkage (DTX-PM). The average size of DTX-PM was 58.2 ± 2.3 nm, which was suitable for passive targeting via an EPR effect for anti-cancer drug delivery. More importantly, the DTX-PM significantly induced apoptosis of HSC-3 cancer cells and effectively suppressed the tumor progression in the HSC-3 xenograft model in vivo (Shi et al., 2016) (**Figure 1**). Particle size is a critical factor in the efficacy of drug-loadedmicelles. It is of fundamental importance to understand the association between the size of drug-loaded nanocarriers and their fate in biological systems in vivo for rational design of drug delivery systems. Yu et al. prepared five different sized (30–230 nm) DTX-loaded methoxy poly(ethylene glycol) poly(lactide) micelles. They studied the permeation effect of micelles within both multicellular tumor spheroids and tumor xenografts in mice. The results indicated that small micelles could enhance the tissue penetration however, the blood half-life was short. Micelles of 100–200 nm showed prolonged circulation time compared with that of smaller or larger micelles, but the


#### TABLE 1 | Features of nanocarriers in references.

penetration was limited. Large micelles (>200 nm) had both short half-life and weak penetration ability, leading to limited tumor accumulation (Yu and Qiu, 2016). Through the EPR effect, drug-loaded PEG–PLA micelles can selectively kill tumor cells and reduce the damage to normal tissues.

Active targeting is another effective method to reduce the damage of chemotherapeutic drugs to normal tissues (Allen, 2002; Szakács et al., 2006). Ligands are modified on the surface of nanocarriers to bind to particular receptors overexpressed by tumor cells or tumor vasculature. Many kinds of targeted ligands have been researched, such as small organic molecules, sugar moieties, peptides, and monoclonal antibodies (Allen, 2002; Brannon-Peppas and Blanchette, 2012). Folate is an example of a small organic molecule that is a cancer targeting ligand because the folate receptor is frequently overexpressed on tumors (Cho et al., 2016). Xiong et al. designed folate-conjugated interfacially crosslinked biodegradable micelles consisting of poly(ethylene glycol)-block-poly(acryloyl carbonate)-blockpoly(D,L-lactide) (PEG–PAC–PLA) and folate-PEG–PLA (FA-PEG–PLA) block copolymers for receptor-mediated delivery of paclitaxel (PTX). Remarkably, folate-decorated PTX-loaded crosslinked micelles displayed significantly higher toxicity to KB cells than free PTX. This is most likely due to their much more efficient cellular uptake through FA receptor-mediated endocytosis. Flow cytometry studies also showed that folate-decorated FITC-labeled crosslinked micelles were much more efficiently taken up by KB cells than that of controls without folate ligands. All these results indicated that ligand-conjugated interfacially crosslinked PEG–PLA micelles have great potential in targeted cancer therapy (Xiong et al., 2011). Peptides are another form of actively used ligands in anticancer drug delivery. Li et al. developed DTX-loaded target micelles (c(RGDfK)-PEG–PLA/PEG–PLA/DTX) that were able to targeted delivery DTX to the tumor cells. Cellular uptake and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) studies revealed that the target micelles were more efficiently taken up by HeLa cells in the presence of c(RGDfK) and significantly improved the cytotoxicity compared to that of non-target micelles (Li et al., 2016a). In another study, researchers constructed somatostatin analog octreotide (OCT)-modified, DTX loaded PEG-b-PLA micelles that bind to somatostatin receptors (SSTRs) overexpressed on tumor cells. The results showed that OCT-PM-DTX enhanced intracellular delivery efficiency in cells and exhibited

higher retardation of tumor growth in vivo (Zhang et al., 2011).

In another study, researchers synthesized a folate modified pH sensitive targeted polymeric micelle to reduce the systemic toxicity of doxorubicin (DOX) and to increase the antitumor efficacy in a multi-drug resistant tumor model (Li et al., 2015). A pH sensitive targeted strategy is derived from the pH of the tumor microenvironment which is slightly lower (6.5–7.2) compared to physiological pH (7.4) (Tannock and Rotin, 1989; Gerweck et al., 2006). The triggering mechanisms of most pH sensitive micelles relate to the endosome and release of the loaded drugs into the cytoplasm. This is because the pH significantly drops to 5.0–6.5 in the endosome and to 4.5 in primary and secondary lysosomes (Hubbell, 2003; Schmaljohann, 2006). Therefore, based on the pH of endosomes/lysosomes, pH sensitive PEG– PLA-based micelles loaded with chemotherapeutic drugs have been widely studied to improve the efficiency of cancer therapy. The major strategies used for inducing pH-sensitive behaviors are changing the charges in the micellar system and the dissociating pH-dependent drug binding linkers. Liang et al. constructed size shifting micelle nanoclusters (MNC) based on a cross-linked framework interspersed with PEG–PLA micelles. After being internalized into tumor cells, the framework of MNC became swollen and disintegrated within the acidic environment of lysosomes via the proton sponge effect of PEI resulting in the release of nanosized micelles for further penetration (Liang et al., 2016). Wu et al. fabricated tumor-targeted and pH-responsive polymeric micelles by mixing AP peptide (CRKRLDRN) conjugated PEG-poly(D,L-lactic acid) block copolymer (AP-PEG–PLA) into the pH-responsive micelles of methyl ether poly(ethylene glycol) (MPEG)-poly(-amino ester) (PAE) block copolymer (MPEG-PAE). This micelle showed a sharp pHdependent micellization/demicellization transition at the acidic environment of the tumor. When loaded with DOX, the DOX-loaded micelles exhibited excellent anticancer therapeutic efficacy (Wu et al., 2010).

Redox-responsive nanocarriers are another captivating direction of research for effective intracellular anticancer drug release (Meng et al., 2009; Cheng et al., 2011). Unlike pH-sensitive strategies which usually release drugs in the endo/lysosomal compartments, redox-responsive strategies aim to disassemble and release drugs in the cytosol and cell nucleus where many chemotherapeutic drugs, such as DOX and PTX, elicit their therapeutic effects (Zhang et al., 2012; Chuan et al., 2014). Since the amounts of glutathione (GSH) tripeptide in the cytosol and cell nucleus (∼2–10 mM) are confoundedly higher than that of extracellular fluids and circulation (∼2–20µM), the intracellular redox-response is exceedingly fast and efficient. Yang et al. constructed a type of redox-responsive micelle which could self-assemble from dynamic covalent PEG–PLA block copolymers that contained a double disulfide linkage in the backbone. The in vitro drug release analyses indicated that a reductive environment could result in triggered drug release profiles. A cytotoxicity assay of DOX-loaded micelles indicated higher cellular proliferation inhibition against HeLa cells pretreated with 10 mM GSH monoester (GSH-OEt) for 2 h than that of nonpretreated ones (Yang et al., 2015). In order to further improve the anticancer efficiency of chemotherapeutic drug-loaded PEG–PLA micelles, dual or multi-stimuli responsive micelles that respond to a combination of two or more signals mentioned above have also been developed.

### PHOTOTHERMAL THERAPY COMPONENTS

Photothermal therapy (PTT) has become an effective alternative in cancer therapy because of the advantages over other treatment methods, such as minimal invasiveness, low toxicity, and high specificity to tumor sites (Geng et al., 2015; Mebrouk et al., 2015). In clinical practice, to maximize the light penetration depth and minimize the influence of biological chromophores, photothermal agents should have strong optical absorbance in the near-infrared (NIR) window (700–900 nm) (Jaque et al., 2014). In recent years, a variety of nanomaterials, such as gold nanostructures (nanoshells, nanorods, nanostars and nanocages) (Chen et al., 2010; Zhang et al., 2012; Ma et al., 2013; Wang et al., 2013), carbon nanomaterials (carbon nanotube and graphene) (Liu et al., 2011; Yang et al., 2012a), and various other inorganic (Tian et al., 2011) and organic nanoparticles (Yang et al., 2012b) have been used for PTT. Although most inorganic nanomaterials have shown effective therapeutic effect for cancers, majority of them are non-biodegradable and will retain in the body for a long time (Sharifi et al., 2012; Zhang et al., 2013). Compared to inorganic nanomaterials, polymeric NIR-absorbing nanomaterials agents have shown great superiority for PTT. Among PTT agents, small molecular organic dyes, such as (ICG) (Fang et al., 2017) and prussian blue (PB) (Hoffman et al., 2014), have been used not only as fluorescent probes in optical imaging, but also as PTT agents. Sun et al. fabricated a new photothermal nano-agent by coprecipitation of 2,5-Bis(2,5-bis(2-thienyl)- N-dodecyl pyrrole) thieno[3,4-b][1,2,5] thiadiazole (TPT-TT) and a biodegradable amphiphilic block copolymer, methoxy poly(ethylene glycol)2K-block-poly(D,L-lactide)2K (mPEG2K– PDLLA2K). As a result, TPT-NPs showed high photothermal conversion efficiency, excellent photostability, and heating reproducibility. The photostability of TPT-TT NPs was much better than that of ICG. Besides, TPT-TT NPs exhibited significant photothermal therapeutic effects toward human cervical carcinoma (HeLa) and human liver hepatocellular carcinoma HepG2 cells (Sun et al., 2015) (**Figure 2**).

# PHOTODYNAMIC THERAPY COMPONENTS

Photodynamic therapy (PDT) is a light triggered method for cancer treatment. Due to the minimal aggressiveness and harmlessness to healthy tissue, PDT can avoid the disadvantages of conventional chemotherapeutic agents, such as serious side

FIGURE 2 | (A) Schematic illustration of the preparation and their cellular action process of TPT-TT NPs. (B) TEM image (Scale bar 500 nm) and size distribution determined by DLS of TPT-TT NPs. (C) Photothermal conversion behavior of TPT-TT NPs at various concentrations. (D) Relative cell viabilities of HepG2 cells incubated with different concentrations of TPT-TT NPs. Reproduced with permission from Sun et al. (2015).

effects or multidrug resistances. PDT can kill tumor cells mainly through the generation of singlet oxygen (1O2) or free radicals. The reactive oxygen species (ROS) can cause significant cellular damage, destruction of tumor blood vessels, and stimulation of antineoplastic immunity (Juarranz et al., 2008). PDT depends on three factors to perform its anticancer effect: (i) a photosensitizer (PS) agent; (ii) irradiation of the affected region using light of an appropriate wavelength; and (iii) presence of oxygen. During irradiation, highly reactive species, such as ROS, which are capable of causing direct damage to biomolecules and triggering the death of tumor cells, are generated in situ (Paszko et al., 2011). Recently, the combination of photodynamic therapeutic agents and nanosystems has proven to be promising. Nanosystems, such as liposomes, micelles, and polymeric nanoparticles, have been widely studied for PDT application, as well as for combination of PDT agents with chemotherapeutic agents or with other forms of therapies. For example, Ogawara et al. encapsulated the photoprotoporphyrin IX dimethyl ester (PppIX-DME) into PEG–PLA nanoparticles (PN-Por). The PN-Por showed significant phototoxicity in vitro and effective antitumor effect in C26 tumor-bearing mice in vivo. All results exhibited the potency of PN-Por for PDT-based cancer treatments (Ogawara et al., 2016). In another study, in order to increase the solubility and delivery efficiency of PpIX, Ding et al. prepared PpIX loaded polymeric micelles using non-covalent encapsulation and covalent conjugation methods. Micelles with lower PpIX loading density (e.g., 0.2%) showed brighter fluorescence and higher <sup>1</sup>O<sup>2</sup> yield than higher PpIX loading density (e.g., 4%) in solution. However, 4% PpIXconjugated micelles demonstrated better antitumor efficiency in vivo (Ding et al., 2011a). Moreover, they further studied the effect of carrier microenvironment on photophysical properties of 5,10,15,20-tetrakis(meso-hydroxyphenyl)porphyrin (mTHPP) and its biological efficacy in tumor inhibition. The electrondeficient PEG–PLA micelle was used as a control. Results displayed that the photophysical and photodynamic properties of mTHPP were highly related to the micelle core environment (Ding et al., 2011b).

PDT can also be combined with active targeting strategies to enhance the antitumor efficiency. Cyclic RGD (cRGD) is a type of peptide that can target the αvβ<sup>3</sup> integrin-rich tumor cells. Tian et al. encapsulated NEt2Br2BDP (a trifunctional photosensitizer) into a cRGD peptide-poly(ethylene glycol)-block-poly(lactic acid) (cRGD-PEG–PLA) and methoxyl poly(ethylene glycol)- blockpoly(lactic acid) (mPEG–PLA) nanomicelle. Under the acidic tumor environment (pH 4.5–5.0), the nanoprobe could be activated to produce fluorescence for tumor detection and <sup>1</sup>O<sup>2</sup> for effective tumor therapy (Tian et al., 2015) (**Figure 3**).

Although there are few related studies, the PEG–PLA-micellebased PDT represents a new kind of therapeutic method, which may be used in clinics in the future.

#### IMMUNE PREPARATION

Immunotherapy is a method of treatment that utilizes the patients' own immune system to treat their illness. Recent strategies for cancer immunotherapy mainly focus on tumorassociated antigens (TAAs), known as a tumor vaccine, and the induction of antigen-specific T cell-mediated immune responses (Cheever and Higano, 2011; Tefit and Serra, 2011; Ledford, 2014). With the thriving progress of genomics and proteomics, various potential target antigens, such as recombinant proteins, synthetic peptides, and DNA, have been studied (De Gregorio and Rappuoli, 2014). However, when administered alone in vivo, they often suffer from short half-life and ineffectively activate the immune system. Hence, adjuvants are required to elicit effective immune responses against tumor cells (Brichard and Lejeune, 2007). Based on the mechanisms of action, vaccine adjuvants can be divided into two categories: immunomodulatory adjuvants and delivery systems (Huang et al., 2011). Coumes et al. synthesized a peptide/polymer conjugate copolymer by conjugation of the amine end-group of LD-indolicidin to the Nhydroxysuccinimide-activated carboxyl end-group of PEG. When the TAA vaccine candidate was formulated with LD-indolicidin-PEG–PLA-stabilized squalene-in-water emulsion, they could effectively elicit a T helper (Th)1-dominant antigen-specific immune response and exhibit satisfactory antitumor activity (Coumes et al., 2015) (**Figure 4**). To induce T cell-mediated immune responses, Li et al. used PEG–PLA based nanoparticles to deliver cytotoxic T lymphocyte-associated molecule-4 (CTLA−4)-siRNA (NPsiCTLA-4). Both the in vitro and in vivo studies showed that this nanoparticle delivery system could effectively deliver CTLA-4-siRNA into both CD4+ and CD8+ T cells, and could significantly increase the percentage of anti-tumor CD8+ T cells, therefore enhancing the antitumor immune responses (Li et al., 2016b).

#### PROTEINS

Protein therapeutics, especially applying cytokines and antibodies, have attracted increasing attention in recent clinical cancer treatments. Compared to chemotherapeutic drugs, proteins have many unique advantages including high specificity, incorporation of diversified functions, and minor side effects on normal tissue (Leader et al., 2008). The mechanism of protein therapy for cancer treatment typically depends on the direct induction of apoptosis in tumor cells and indirect tumor inhibition by activating an immune response or targeting the tumor vasculature and stroma. In order to overcome the shortcomings of proteins in cancer treatment, such as rapid in vivo degradation, poor pharmacokinetics, and instability, combination of proteins with nano delivery systems has been intensively researched.

Cytokines are a class of secreted or membrane-bound proteins. They play an important role in regulating the growth, differentiation, and activation of immune cells (Dranoff, 2004). Many kinds of cytokines, such as tumor necrosis factors (TNFs), interleukins (ILs), and interferons (INFs), have been widely used in clinical cancer treatment. These cytokines can be targeted through the use of antibodies. Antibodies are one of the most efficient and promising approaches for the treatment of hematological malignant neoplasms and solid

tumors (Scott et al., 2012). A series of monoclonal antibodies (mAbs) have been approved by the FDA or are being assessed in clinical trials for cancer therapy (Scott et al., 2012). Antibodies can also act as target groups to enhance the uptake of drugs in tumor cells and can be combined with a nano drug delivery system. In order to achieve targeted delivery of siPlk1 to Her2<sup>+</sup> breast cancer, Dou et al. designed an anti-Her2 single-chain variable fragment antibody (ScFv Her2) decorated PEG–PLA-based nanoparticles encapsulated with siPlk1 (ScFv Her2 -NP si Plk1). This ScFv Her2 -NP siRNA could specifically bind to the Her2 antigen overexpressed on the surface of Her2<sup>+</sup> breast cancer cells. Therefore, the antitumor efficiency was evidently improved (Dou et al., 2014). Yue et al. prepared multifunctional hybrid micelles based on amphiphilic mal-PEG-b-PLA and mPEG-b-P(LA-co-DHC/RhB) block copolymers. A specific anti-transferrin receptor antibody, OX26, was then linked onto the surface of the micelles. The results showed that OX26 conjugation visibly increased the uptake efficiency of micelles by target cell lines (C6) and effectively passed through the blood-brain barrier (Yue et al., 2012).

# GENE

Nucleic acid (such as plasmid DNA, antisense oligonucleotides, and siRNA)-based gene therapeutics have received an increased amount of attention in the last few decades because of their unique advantages for attacking critical cancer hallmarks (Das and Verma, 2016; Jiang et al., 2017). Even though the use of viral vectors for gene therapy offers highly efficient gene

transfer and has shown satisfactory results in clinical trials (Alton, 2007), they possess various unwanted effects, such as immune stimulation and the potential for mutagenesis. Therefore, it is clear there is still need for further investigation to improve the delivery mechanism (Thomas et al., 2003; Ginn et al., 2013). These unwanted effects have led researchers to turn to nonviral vector solutions with high efficiency, stability, and minimal toxicity (Mykhaylyk et al., 2007; Sun et al., 2008). Among them, cationic polymers have been widely studied due to its unique advantages (Merdan et al., 2002). Polyethyleneimine (PEI), a synthetic polymer which has shown high condensation and superior transfection efficiencies, is one of the most used cationic polymers (Wiseman et al., 2003). In order to achieve high efficiency for either DNA or siRNA loading and transfection efficiency, the molecular weight (MW) of the PEI block must be high (e.g., 25 k Da). However, high MW PEI is known to cause systemic toxicity upon intravenous administration (Moghimi et al., 2005). Moreover, another shortcoming is that polycation/DNA or siRNA complexes have the tendency to aggregate when induced by salt- and serum proteins in vivo and can be rapidly cleared by the immune system (Moret et al., 2001). Although PEI can be conjugated with PEG to form block copolymer assemblies and has shown increased stability, reduced toxicity, and lower immunogenicity (Kim et al., 2005; Sato et al., 2007), block copolymers still limit timely release of the internalized gene from the complex due to steric interferences of the associated PEG chains (Grigsby and Leong, 2010; Salcher and Wagner, 2010). Therefore, developing efficient delivery systems based on biomaterials may be an effective means for nucleic acid delivery in therapeutic applications. PEG– PLA micelles are able to resolve parts of the aforementioned problems of gene delivery. In order to systematically deliver DNA expression vectors to tumors, Shukla et al. developed a novel dual nanoparticle (DNP) system to deliver DNA expression vectors to tumor cells. The DNP system was consisted of a DNA expression vector–cationic peptide nanocomplex (NC) surrounded by a PEG–PLA nanoparticle. The results showed that the DNP system could effectively induce apoptosis of tumor cells and possess highly anti-tumor efficacy in vivo (Shukla et al., 2017).

RNA interference (RNAi) is a post-transcriptional gene silencing phenomenon. The discovery of RNAi was through the mechanism of small interfering RNA (siRNA). The mechanism is that siRNA can be incorporated into the RNAinduced silencing complex (RISC) and specifically degrades the target messenger mRNA through complementary base pairing, thereby prohibiting the translation into target proteins (Amjad et al., 2017). Zhao et al. synthesized a biodegradable amphiphilic tri-block copolymer (mPEG2000-PLA3000 b-R15) composed of monomethoxy poly(ethylene glycol), poly(D,L-lactide), and polyarginine. This copolymer can further self-assemble into cationic polymeric nanomicelles for in vivo siRNA delivery. The polymeric nanomicelles showed excellent haemocompatibility and higher cell growth inhibition toward EGFR expressed MCF-7 cells. In vivo experiments further proved the effective tumor growth inhibition effect of the polymeric nanomicelles (Zhao et al., 2012) (**Figure 5**).

In addition to the therapeutic application of genetic material, genes can also be delivered alongside traditional chemotherapy drugs. Zhan et al. investigated the antiglioblastoma effects of RGD-PEG-PEI/pORF-hTRAIL nanoparticle combined with CDX-PEG–PLA-PTX micelle (paclitaxel loaded CDX-poly(ethylene glycol)–block-poly(lactic acid) micelle). When administered the same dosages, the survival time of the intracranial glioblastoma-bearing model mice was significantly longer in the co-delivery (33.5 days) treated group than that of the groups solely treated with CDX-PEG–PLA-PTX (25.5 days), RGD-PEG-PEI/pORFhTRAIL [24.5 days), or physiological saline (21.5 days)]. This research proved the high efficacy for co-delivery of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and PTX in the intervention of intracranial glioblastoma by employing tumor-targeted gene carrier RGD-PEG-PEI and brain-targeted micelle CDX-PEG–PLA, respectively (Zhan et al., 2012).

# OTHERS

Curcumin (Cur), a natural polyphenol of Curcuma longa, has been widely researched for its antitumor activities. However, the poor aqueous solubility and low biological availability have limited its further application. Zheng et al. fabricated Curloaded PEG–PLA micelles. The preparation of Cur-MPEG-PLA was very simple and fast. Besides, the micelle group showed a sustained release behavior of Cur and an enhanced toxicity on C6 and U251 glioma cells in vitro. Moreover, compared to free Cur, they induced more apoptosis on C6 glioma cells.

The Cur-loaded micelles also effectively improved the antiglioma activity of Cur in vivo (Zheng et al., 2016) (**Figure 6**). In another study, researchers constructed Cur-loaded pH-sensitive MPEG–PLAPAE micelles. These micelles could shrink from 171.0 nm to 22.6 nm and could increase their surface charge to 24.8 mV, which significantly improved the cell uptake of Cur by MCF-7 cells. Moreover, these micelles also exhibited excellent antitumor efficiency in vivo (Yu et al., 2014). The antitumor activity of Cur-loaded PEG–PLA micelles were also reported in other studies (Yang et al., 2012c; Kumari et al., 2016, 2017).

Antivascular therapy is a distinctive form of cancer treatment. It can cause a selective and rapid shutdown of the tumor vasculature, thus resulting in extensive cancer cell death. Wang et al. developed αvβ<sup>3</sup> integrin-targeting peptide (RGD) functionalized polymeric micelles (RFPMs) based on the use of poly(ethylene glycol)-block-poly(D,L-lactide) (PEG–PLA). DOX was conjugated to the PEG–PLA micelle core and combretastatin A4 was physically encapsulated into the micelles (RFPMs-DOX-CA4). The micelles exhibited sequential release kinetics, resulting in sequential killing of endothelial cells and tumor cells in vitro. In B16-F10 tumor-bearing mice, RFPMs-DOX-CA4 showed stronger tumor growth inhibition and significantly higher survival rate compared to other treatment groups (Wang et al., 2011).

# CONCLUSIONS

In the last few decades, polymeric micelles have become one of the most promising nano-delivery systems for the treatment of cancers have been used for the delivery of a variety of cargoes, from conventional chemotherapeutic drugs to specific therapeutic agents and biological macromolecules. Among various polymeric micelles, PEG–PLA based micelles have been intensively studied because of their excellent biodegradability and biocompatibility. Genexol-PM has already been approved for breast cancer treatment in South Korea. With the progressive development of cancer treatment methods, such as photodynamic therapy, photothermal therapy, immunotherapy, and gene therapy, PEG–PLA micelles are being increasingly applied in combination with these treatments. It is worth noting that besides primary passive targeting through the EPR, there is a clear shift toward the utilization of micelles which can be modified for active targeting, controlled delivery of therapeutic agents reliant on the unique tumor microenvironment or external environment, and combination of more than one type of therapeutic payload.

This review has discussed various examples of PEG–PLA micelles being applied in a variety of therapeutic applications for the treatment of cancers. These micelles contain a wide range of modifications including primary modification for passive

Reproduced with permission from Zheng et al. (2016).

targeting, incorporation of targeting ligands, responsiveness to the tumor microenvironment, and the mixing of micelles with drugs and other therapeutic agents. Among them, multifunctional PEG–PLA micelles have gained immense attention due to their versatility in simultaneously incorporating various agents (e.g., chemotherapeutic drugs and RNAi) and their ability to achieve multiple modifications (e.g., active targeting, passive targeting, and response to stimuli) to enhance cancer therapy.

It is promising to be hopeful about the future of PEG–PLA micelles given their inherent advantages. However, it should be noted that the safety of these novel concepts is a major concern. Rollerova et al. validated that PEG-b-PLA NPs might interfere with the activation and function of the hypothalamic-pituitarygonadal (HPG) axis, which might relate to the nanoreprotoxicity of PEG-b-PLA NPs at both the central neuroendocrine and gonadal levels (Rollerova et al., 2015). In addition, PEG-b-PLA NPs also could cause neuroendocrine disrupting effect in the neonatal female rats (Scsukova et al., 2015). Dvoráková et al. indicated a possible age-related association between the oxidative stress and neonatal PEG-b-PLA administration (Dvoráková et al., 2017). Therefore, the PEG–PLA micelles still need to be carefully studied in several animal models and eventually

#### REFERENCES


in human patients.It is known that the simpler the structure of a system, the easier it is to translate. In the future, we should bear this in mind to design more "active" targeting and simple structures of PEG–PLA micellar drug delivery systems. It is promising that PEG–PLA micelles will play an increasingly important role in the wide range of treatment methods against cancer.

# AUTHOR CONTRIBUTIONS

JW and SL produced the first draft. SC and DL revised the manuscript. DL, YH, and JG proposed the outline of the article and revised the draft before submission. CW revised the draft carefully. In additional all authors provided final approval of the manuscript.

#### ACKNOWLEDGMENTS

This work was financially supported by the National Natural Science Foundation of China (Grant Nos. 51673190, 51603204, 51303174, 51473165, 51390484, and 51520105004) and the Science and Technology Development Program of Jilin Province (Grant Nos. 20160204015SF and 20160204018SF).


micelles: an in vitro anticancer study. Pharm. Res. Dordr. 33, 826–841. doi: 10.1007/s11095-015-1830-z


copolymers for intracellular drug delivery. Acta Biomater. 17, 193–200. doi: 10.1016/j.actbio.2015.01.044


and in vivo. J. Control. Release 160, 630–636. doi: 10.1016/j.jconrel.2012. 02.022


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Wang, Li, Han, Guan, Chung, Wang and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# H2O2-Responsive Nanoparticle Based on the Supramolecular Self-Assemble of Cyclodextrin

#### Zhenqiang Dong1,2† , Yang Kang<sup>3</sup>† , Qijuan Yuan<sup>1</sup> , Manli Luo<sup>4</sup> and Zhipeng Gu<sup>1</sup> \*

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Peng Miao, Suzhou Institute of Biomedical Engineering and Technology (CAS), China Pedro Fonte, CBIOS – Universidade Lusófona, Portugal Xiaolong Li, Dana–Farber Cancer Institute, United States

\*Correspondence:

Zhipeng Gu guzhp@mail.sysu.edu.cn †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 13 March 2018 Accepted: 08 May 2018 Published: 28 May 2018

#### Citation:

Dong Z, Kang Y, Yuan Q, Luo M and Gu Z (2018) H2O2-Responsive Nanoparticle Based on the Supramolecular Self-Assemble of Cyclodextrin. Front. Pharmacol. 9:552. doi: 10.3389/fphar.2018.00552 <sup>1</sup> Key Laboratory of Sensing Technology and Biomedical Instrument of Guangdong Province, School of Engineering, Sun Yat-sen University, Guangzhou, China, <sup>2</sup> College of Chemical and Material Engineering, Quzhou University, Quzhou, China, <sup>3</sup> Key Laboratory of Mountain Ecological Restoration and Bioresource Utilization, Chengdu Institute of Biology, Chinese Academy of Sciences, Chengdu, China, <sup>4</sup> Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China

Designing stimuli responsive, controllable and biocompatible multifunctional nanoparticles is an important progress in the current quest for drug delivery systems. Herein, we devoted to developing a β-cyclodextrin (β-CD) based drug delivery nanoparticles (NPs) that release Bovine serum albumin (BSA) via glucose-responsive gate. The design involves synthesis of sodium alginate with β-CD modified (Alg-β-CD) and methoxypolyethylene glycol (mPEG-Fc) containing ferrocene (Fc) uncharged end-capping. When α-cyclodextrin (α-CD) was added with these two segments, the stable non-covalent supramolecular structure of Alg-β-CD/mPEG-Fc/α-CD can be selfassembled into NPs in aqueous solution. BSA loaded Alg-β-CD/mPEG-Fc/α-CD also has been prepared. Interestingly, these supramolecular Alg-β-CD/mPEG-Fc/α-CD/BSA NPs showed uniform sphere structure and constant BSA loading content. Also, this new kind of NPs can disassemble in the present of hydrogen peroxide (H2O2). Since glucose oxidase (GOD) can oxidize glucose and produce H2O2, so this kind of polymeric NPs can also have glucose responsive behavior in the GOD containing environment. Developed functional Alg-β-CD/mPEG-Fc/α-CD might be a promising drug delivery strategy for diabetes or immunotherapy with more efficiency.

Keywords: responsive, cyclodextrin, self-assemble, nanoparticle, BSA

# INTRODUCTION

For the past decades, stimuli responsive nanoparticles (NPs) based delivery technique have emerged as a promising strategy for controllable drug delivery, especially for protein delivery due to ingenious design (Tao et al., 2013; Wu et al., 2014; Kuang et al., 2016; Ding et al., 2017; Kang et al., 2017b). As smart vehicles, these ideal NPs can undergo reversible physical or chemical changes to control drug release in response to external stimuli such as pH, light, temperature, redox and molecules (Kang et al., 2015; Tao et al., 2016, 2017a,b; Rosenblum et al., 2018; Zhu et al., 2018). For cancer therapy, diverse intracellular and extracellular endogenous stimuli within

FIGURE 1 | Schematic illustration of the formation of Alg-β-CD/mPEG-Fc/α-CD nanoparticles and their enzyme encapsulation and controlled release behavior.

tumor microenvironment, such as low pH value, hypoxia, enzymes and reactive oxygen species (ROS) have been widely investigated in stimuli responsive NPs and explored as the stimuli to selectively trigger drug control release or transfer prodrug systems into active form (Chen et al., 2015; Ruan et al., 2015, 2016; Liao et al., 2016; Wang et al., 2017b).

As an important hallmark, it should be noticed that various studies have demonstrated that hydrogen peroxide (H2O2) could be used as a second messenger in intracellular signaling cascades and would be generated in multiple skin lesions, initiating aging cells and the tumoral tissue (Miao et al., 2015). In this regard, H2O2-responsive NPs would be a promising therapeutic strategy for the treatment of specific disease.

At present, block copolymer approaches for H2O2-responsive NPs has been widely used and rapidly developed. For instance, a tyrosol incorporated copolyoxalate (TPOX) NPs have been developed by Kim et al. (2017), which demonstrated that TPOX NPs released entrapped nile red through the degradation under H2O<sup>2</sup> stimulation and could be used as H2O2-responsive therapeutic NPs for the treatment and inflammatory. Recently, Gu group have designed a glucose-responsive mechanism directly utilizing H2O2-sensitive polymeric vesicles by using polyethylene glycol (PEG) and phenylboronic ester (PBE)-conjugated polyserine for smart insulin delivery (Hu et al., 2017). Apart from stimuli-responsive polymeric NPs via cleavable covalent bonds to control their properties, supramolecular based NPs via non-covalent bonds have also been widely studied due to that functional ligands and environmental responsive bonds can be conveniently incorporated through host-guest interaction, π–π interaction, hydrogen bonding interaction and so on (Wang et al., 2015, 2017a). Cyclodextrin (CD), as a most common case of supermolecules, which has been demonstrated that can bind some specific hydrophobic guest molecules into its cavities to construct stable host-guest complex (Yuan et al., 2013). CDs are also considered as a good choice for drug delivery strategy due to their adjustable water solubility, good biocompatibility, and non-toxicity toward biological systems (Kang et al., 2014). In the previous studies, various stimuli-responsive CDs NPs based on host-guest interactions between CDs and different guest molecules [such as azobenzene, ferrocene (Fc), and adamantine] have been widely developed. For instance, Guo and Lei (2015) have developed a kind of redox-responsive cationic supramolecular polymers based on the host-guest interaction between β-CD and Fc, which indicated the potential application for gene nanocarrier. To conduct the redox-stimuli in an effect way, Kang et al. (2017a) have designed a kind of dual redox-responsive micelles which fabricated by methoxypolyethylene glycol conjugated β-CD (mPEG-β-CD) and Fc conjugated camptothecin (Fc-SS-CPT)

segments for achieving dual ROS and glutathione responsive. It is reasonable to design H2O2-sensitive NPs based on CD, which can load and controllable releasing drugs for treatment of specific disease.

In this paper, a novel oxidation responsive rod-coil polymeric NPs have been prepared using water as the self-assemble solvent (**Figure 1**) Firstly, sodium alginate with β-CD end-decoration (Alg-β-CD) and mPEG conjugated with Fc (mPEG-Fc) were designed and synthesized, and then the self-assembly behavior of them in water was studied in detail. This supramolecular copolymer grows up to form comb shaped polymer in aqueous solution by the terminal host-guest inclusion. When adding certain amount of α-CD to this solution, the PEG will form inclusion complex with α-CD to form the rod section while sodium alginate form the coil section. This kind of structure will self-assemble into rod-coil NPs in water and load Bovine serum albumin (BSA). This new kind of NPs can disassemble in the present of H2O2. Since glucose oxidase (GOD) can oxidize glucose and produce H2O2, so this kind of polymeric nanosphere can also have glucose responsive behavior in the GOD containing environment. This will make it very promising to use in biomedical application field in body environment.

# MATERIALS AND METHODS

#### Materials and Reagents

β-cyclodextrin (β-CD), α-Cyclodextrin (α-CD) was purchased from Sigma and dried in a vacuum before use. Sodium alginate (Alg) was purchased from Tianjin Yuanhang Chemicals Co., Ltd., China. Methoxypolyethylene glycol (mPEG, Mw = 2000) was purchased from Shanghai Jingchunshiye Co., Ltd., China. Glucose oxidase was purchased from Aspergillus Niger (GOD, lyophilized powder protein, Mw = 186 000, catalase <4%). BSA was purchased from Aoke Biological Co., Ltd., China. Carboxyferrocene (Fc-COOH), morpholinoethanesulfonic acid (MES), dicyclohexylcarbodiimide (DCC), 4-(dimethylamino) pyridine (DMAP), N-hydroxysuccinimide (NHS), 1-ethyl-3-[3- (dimethylamino)propyl]carbodiimide hydrochloride (EDCI), and other reagents were purchased from Chengdu Kelong Co., Ltd., China.

# Synthesis of Alg-β-CD and mPEG-Fc

Sodium alginate with β-CD was synthesized in two steps with slight modification. Firstly, the mono-6-(p-tolylsulfonyl)-β-cyclodextrin (6-β-CD-OTs) and

mono-6-deoxy-6-hexamethylenediamine-β-CD (6-β-CD-HDA) were synthesized according to the method reported in the previous study (Quan et al., 2010). Briefly, β-CD was grafted onto the sodium alginate backbone via amido link condensation reaction. 1% sodium alginate solution (w/v) was prepared in MES buffer solution (0.1 M) and NaCl (0.5 M), and the pH was adjusted to 6.0. After 29 mg NHS and 95 mg EDC (molar ration of EDC: NHS: COO- = 1: 0.5: 1) were added to 40 mL above alginate solution, the solution was then agitated for 30 min to obtain a homogeneous solution followed by the addition of 10 mL 3.08 % 6-β-CD-HDA MES solution (w/v). The reaction was then continuing at 4◦C for 24 h. Alg-β-CD was obtained after the resulting mixture dialyzed against pure water and then freeze dried. The degree of β-CD substitution (DS) to alginate was calculated according to the followed equation:

$$\frac{I\_1}{I\_2 - 6I\_1 - \frac{2}{7}I\_2} = \frac{7}{4} \text{DS}$$

where I<sup>1</sup> is the digital integration of anomeric protons of β-CD, I<sup>2</sup> is the total digital integration of 3.4–4.2 ppm. The final degree of substitution (DS) of Alg-β-CD was 25.6 % by this calculated method. <sup>1</sup>HNMR (400 MHz, D2O) were shown in **Figure 2**.

mPEG-Fc was synthesized with one pot method. DMAP (0.183 g) and DCC (1.12 g) were added successively to a solution of mPEG (3.0 g) and carboxyferrocene (0.69 g) in dry chloroform (CHCl3, 100 mL) for 24 h at room temperature. After removing dicyclohexylurea (DCU) by filtration, the filtrate was concentrated in a vaccum and purified by dissolving in deionized water and extracting with diethyl ether four times. Subsequently, the product was further purified by using gel column with methanol and water as the solvent to remove the unreacted carboxyferrocene and catalyst residue. Finally, the dark red product was concentrated in a vaccum and dried in the vacuum oven for 2 days. The final DS of PEG-Fc used was 95.4% which was calculated though the integral ratio between H<sup>d</sup> and Hf . <sup>1</sup>HNMR (400 MHz, CDCl3) were shown in **Figure 3**.

#### Preparation of Alg-β-CD/mPEG-Fc/α-CD and Alg-β-CD/PEG-Fc/α-CD/BSA NPs

PEG-Fc (6.6 mg) and Alg-β-CD (8.4 mg) were added to 3 mL deionized water and kept stirring for 5 h to get a transparent solution. Then α-CD (0.2 g) was added to the above solution and continued to stir over night. For BSA loading, 1.5 mg BSA was added with stirring for 10 min before the addition of α-CD. The solution changes from transparent to turbid gradually indicating the formation of the Alg-β-CD/PEG-Fc/α-CD and Alg-β-CD/PEG-Fc/α-CD/BSA NPs.

### Characterization of the Alg-β-CD/mPEG-Fc/α-CD and Alg-β-CD/mPEG-Fc/α-CD/BSA NPs

The product was characterized by <sup>1</sup>H–1H (2D) NOESY NMR spectroscopy which recorded on an Advance Bruker 600 NMR spectrometer. The transmission electron microscopy (TEM)

images were taken by using a JEOL JEM-100CX instrument operating at an accelerating voltage of 80 kV. Each sample were prepared and imaged by Atomic force microscopy (AFM) in tapping mode with a Nanoscope IIIa Digital Instrument. X-ray powder diffraction (XRD) patterns of the samples were recorded by using Cu-Kα irradiation source with X' Pert MPD (20 kV; 35 mA; 2◦ theta/min) to determine the crystalline structures.

#### Self-Assembly NPs Encapsulation and Leakage of BSA

Self-assembly NPs encapsulated BSA were prepared and then their encapsulation efficiency was determined as follows. The BSA loaded NPs (0.5 mg BSA, 0.5%) was centrifuged for 20 min at 12,000 rpm to completely separate the free and encapsulated BSA and the supernatant was analyzed by determined by UV spectrophotometry. The BSA encapsulation efficiency was

FIGURE 6 | X-ray diffraction patterns of (a) α-CD, (b) PEG/α-CD inclusion complex, (c) Alg-β-CD/PEG-Fc/α-CD particles, (d) Alg-β-CD/PEG-Fc.

FIGURE 8 | Bovine serum albumin release from Alg-β-CD/PEG-Fc/α-CD/BSA and Alg-PEG/α-CD/BSA nanoparticles at 37◦C in 12 mL 0.1 M glucose and 10 mg GOD containing solution.

calculated as follows (1):

Bovine serum albumin encapsulation efficiency (%) = mass of BSA used in formulation − mass of free BSA mass of BSA used in formulation <sup>×</sup> <sup>100</sup>

Then the NPs were continuing stored in aqueous solution for 48 h to evaluate leakage of BSA. The mixture was again separated by centrifugation and the supernatant was analyzed for determination of BSA leakage efficiency as above method (2):

Bovine serum albumin leakage efficiency (%) =

$$\frac{\text{mass of BSA in the supernovaantant}}{\text{mass of BSA encapsolated}} \times 100$$

#### In Vitro Release Studies

The Alg-β-CD/PEG-Fc/α-CD self-assembly NPs encapsulated BSA was prepared and centrifuged for in vitro BSA release study (12,000 rpm, 20 min). The lower sediment was added 12 mL 0.1 M glucose solution and 10 mg GOD and kept stirring for 10 min. The solution was placed in a constant temperature incubator and kept the temperature at 37◦C during the whole release process. 2 mL solution had been taken out and centrifugation at 12,000 rpm for 20 min every 6 h. The amount of BSA released in the supernate was collected and determined by UV spectrometry at specified time. The BSA release was quantified when the standard curve was plotted using a standard BSA solution.

#### In Vitro Cell Studies

Methyl tetrazolium (MTT) assay was carried out on murine colon carcinoma cells (CT26) as the cell line for biocompatibility evaluation of the self-assembly NPs. 5 × 10<sup>3</sup> cells were culture for

24 h to attach in 96-well plates. Then the cells were treated with fresh medium containing 1 mg/mL Alg-β-CD/mPEG-Fc/α-CD NPs and Alg-β-CD/mPEG-Fc/α-CD/BSA NPs. The cell viability was measured after incubation for desired time duration (48 h). The cells without treatment were used as the control. Each group was repeated in triplicate.

Further, confocal laser scanning microscopy (CLSM) was used for monitoring the in vitro endocytosis. Briefly, selfassembly NPs encapsulated Dil were prepared and incubated with CT26 cells at 37◦C for 4 h. After incubation, all samples were stained with LysoTracker Green (Life Technologies, Carlsbad, CA, United States) and 10 µg/mL of Hoechst 33342 (Life Technologies, Carlsbad, CA, United States).

#### Statistical Analysis

All the results obtained in this study were expressed as mean ± standard deviation unless otherwise noted. Comparisons between two groups were made using two-tailed student's t-test and between multiple groups were made using one-way analysis of variance (ANOVA). Statistical analyses were performed using SPSS 21.0 and considered statistically significance when <sup>∗</sup>p < 0.05.

#### RESULTS AND DISCUSSION

According to the previous, it has demonstrated that inclusion complexes would form when β-CD interact with Fc group (Tomatsu et al., 2010). Further, the formation of complexation between the Fc groups on mPEG-Fc and β-CD groups on Algβ-CD could be observed by using 2D NMR NOESY spectra. As shown in **Figure 4**, the peaks of β-CD group were in accordance with the resonance of the Fc groups could be found in the 2D NOESY spectrum of Alg-β-CD/PEG-Fc, which indicating that the Fc groups threaded through the cavity of β-CD to form an inclusion complex.

The resulting Alg-β-CD/PEG-Fc was hydrophilic copolymers, which could form transparency solution in water. By adding α-CD to the aqueous solution of Alg-β-CD/PEG-Fc at room temperature, the solution gradually became turbid gradually when the aggregates were formed. The morphology of these aggregates has been observed by TEM and atomic force microscope (AFM). TEM micrographs of **Figure 5a** revealed that the Alg-β-CD/PEG-Fc/α-CD aggregates appeared round line shapes and their diameters were about 100 nm. In the AFM micrographs of **Figure 5c**, Alg-β-CD/PEG-Fc/α-CD selfassembly NPs also showed rounded line shapes with a horizontal length about 100 nm. Particularly, it is noted that most height of the aggregates (<5 nm) was about 20 times smaller than horizontal length. The main reason for such size differences between horizontal and vertical direction might be the collapsing of this NPs. Further, the height of NPs increased about threefold when BSA was encapsulated, which demonstrated that the BSA has been successfully encapsulated into the Alg-β-CD/PEG-Fc/α-CD NPs (**Figures 5b,d**).

The XRD patterns of the α-CD (**Figure 6a**), PEG/α-CD inclusion complex (**Figure 6b**), Alg-β-CD/PEG-Fc/α-CD (**Figure 6c**), and Alg-β-CD/PEG-Fc (**Figure 6d**) has been shown in **Figure 6**. As an important index to evaluate the inclusion structures formation of CD based complex. The pattern of Alg-β-CD/PEG-Fc/α-CD showed in the figure was differs from that of free Alg-β-CD/PEG-Fc or α-CD. It was also found that the homo-PEG crystalline peaks (2 theta = 19.2◦ and 23.3◦ ) and α-CD crystalline peak (2 theta = 21.5◦ ) were absent. According to the previous study, the peak at 2 theta = 19.9◦ is a typical peak of PEG-α-CD channel-type crystallites that demonstrated the inclusion structures of Alg-β-CD/PEG-Fc/α-CD contained PEG-α-CD (Lautens et al., 1990). Meanwhile, the NPs would

form an inner rod-like PEG-α-CD inclusion block surrounded by a protonated coil-like Alg shell when insoluble PEG-α-CD inclusion blocks formed and water became a selective solvent for the Alg-β-CD/PEG-Fc/α-CD. This rod-like block in rodcoil system preferred parallel packing which may result in the formation of hollow NPs promoted by efficient space-filling has been investigated (Jenekhe and Chen, 1999).

As mentioned in the literature and our previous study, the Alg-β-CD/PEG-Fc/α-CD self-assemble NPs were connected by the Fc-β-CD inclusions. These NPs could release BSA through their collapse after the dissociation of Fc-β-CD. To achieve this goal, the oxidization Fc to Fc<sup>+</sup> is an efficacy way while β-CD could not form complex with Fc derivatives in their oxidized state (Matsue et al., 1985). To get a clear vision, hydrogen peroxide (H2O2) was chosen as the oxidant in this study. After the addition of H2O<sup>2</sup> (1 mL, 30%, H2O<sup>2</sup> solution) to Alg-β-CD/PEG-Fc/α-CD hollow NPs (10 mL, 1%), the system became unstable and produced polyrotaxane precipitate in 5 min. The possible mechanism of this phenomenon can be seen in **Figure 7**. The addition of H2O<sup>2</sup> dissociated Fc-β-CD inclusion complex and leaving the rod-like hydrophobic PEG-Fc/α-CD part precipitated out of water.

Nowadays, one of the most widely used applications of the stimuli-responsive NPs is in the on demand drug-release field (Tao et al., 2015). It requires the micelle system release the aim drugs gradually to achieve the long-term drug action. Since H2O<sup>2</sup> could be produce in a rather sustained way through oxidization of glucose by glucose oxidase (GOD) so that can oxidize Fc to its oxidation state gradually. **Figure 7** shows the UV-vis spectra of the PEG-Fc in 0.1 M glucose before and after 10 mg GOD was added. After adding GOD for 12 h, the absorption peak around 450 nm due to Fc groups d–d transition peak (in reduced form) almost disappeared indicating that the Fc groups in the PEG-Fc could be oxidized to ferrocenium as anticipated (Yamamoto et al., 2000). Tan et al. (2012) have been developed the glucose sensitive hydrogel using the GOD and glucose interaction to produce hydrogen peroxide to disintegrate the ICs of Fc and β-CD. Therefore, it is possible to get glucose sensitive Alg-β-CD/PEG-Fc/α-CD micelles by adding GOD to the Alg-β-CD/PEG-Fc/α-CD system.

Otherwise, serum albumins, as a major soluble protein, have got much attention and been widely used due to their physiological functions. It has been applied in the nanomedicine field because of they serve as a depot protein and as a transport protein for a variety of compounds. BSA has been one of the most extensively member used as a protein drug model, particularly because of its structural homology with human serum albumin (Papadopoulou et al., 2005). The incorporation of BSA into Alg-β-CD/PEG-Fc/α-CD particles had been studied. The encapsulation efficiency and leakage efficiency was 71.4 and 6.6%, respectively. The in vitro release profile of BSA from Algβ-CD/PEG-Fc/α-CD/BSA and Alg-PEG/α-CD/BSA particles was presented in **Figure 8**. In Alg-β-CD/PEG-Fc/α-CD/BSA particles group, the BSA appeared to be released in a continuous way. After 48 h, about 80% of BSA had been released. While as a counterpart, the Alg-PEG/α-CD/BSA group (BSA encapsulation efficiency was 66.7%), under the same condition, there was less than 20% of BSA had been released which is much less than the Alg-β-CD/PEG-Fc/α-CD/BSA group. This result further confirmed that in glucose and GOD containing environment the Alg-β-CD/PEG-Fc/α-CD/BSA particle achieved BSA delayed release behavior through the β-CD-Fc inclusion complex as redox trigger.

The biocompatibility of NPs is of critical importance to their eventual success (Tao et al., 2014). As shown in **Figure 9A**, the results demonstrate the non-toxic of Alg-β-CD/PEG-Fc/α-CD and Alg-β-CD/PEG-Fc/α-CD/BSA NPs, which could be used as the biocompatible drug delivery system. The performance of NPs cellular internalization through the observation of NPs encapsulated Dil interaction with labeling CT26 cells (**Figure 9B**). The results indicated that large numbers of Alg-β-CD/PEG-Fc/α-CD NPs entered the cells after 4 h and some of them began to enter into cell nucleus.

# CONCLUSION

In summary, Alg-β-CD/PEG-Fc/α-CD rod-coil hollow NPs have been prepared through the supramolecular self-assemble. The rod-part PEG-α-CD and the coil-part Alg were connected through β-CD-Fc inclusion complex. The BSA enzyme was successfully encapsulated in Alg-β-CD/PEG-Fc/α-CD hollow sphere in aqueous solution. These Alg-β-CD/PEG-Fc/α-CD hollow NPs showed good H2O<sup>2</sup> sensitivity. Further study showed that this nanoparticle encapsuled BSA can achieve BSA delayed drug-release behavior in the glucose containing environment at the existence of GOD. Therefore, this study exhibits that using the host-guest interactions would be a general, useful and effective way for fabricating stimuli-responsive drug delivery NPs. Furthermore, the glucose mediated H2O<sup>2</sup> sensitive behavior of the Alg-β-CD/PEG-Fc/α-CD NPs may benefit its application in more and more fields such as enzyme immobilization, drug delivery and micro-reactors.

# AUTHOR CONTRIBUTIONS

ZG designed experiments, supervised the experiments, and revised and finalized the manuscript. ZD and YK performed the experiments, analyzed the data, and prepared the figures and the manuscript. QY performed the cell culture. ML contributed to the results discussion and paper writing. All authors reviewed and approved the final paper.

# FUNDING

This study was supported by the funding and generous support of the National Natural Science Foundation of China (21704104), West Light Foundation of the Chinese Academy of Sciences (2016XBZG\_XBQNXZ\_B\_003). Meanwhile, we also want to sincerely acknowledge the funding from the Guangdong Province Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation (2017B030314026).

### REFERENCES

fphar-09-00552 May 24, 2018 Time: 15:49 # 9


the π-conjugated main chain. preparation, optical properties, redox behavior, and mössbauer spectroscopic analysis. Macromolecules 30, 5390–5396. doi: 10.1021/ma9704368


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Dong, Kang, Yuan, Luo and Gu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Icariin Activates Autophagy via Down-Regulation of the NF-κB Signaling-Mediated Apoptosis in Chondrocytes

Bobin Mi<sup>1</sup> , Junqing Wang<sup>2</sup> , Yi Liu<sup>1</sup> , Jing Liu<sup>1</sup> , Liangcong Hu<sup>1</sup> , Adriana C. Panayi<sup>3</sup> , Guohui Liu<sup>1</sup> \* and Wu Zhou<sup>1</sup> \*

<sup>1</sup> Department of Orthopaedics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China, <sup>2</sup> State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics & Center for Molecular Imaging and Translational Medicine, School of Public Health, Xiamen University, Xiamen, China, <sup>3</sup> Addenbrooke's Hospital, University of Cambridge School of Clinical Medicine, Cambridge, United Kingdom

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Fang Chen, Harvard Medical School, United States Chun Yao, Harvard University, United States

#### \*Correspondence:

Guohui Liu liuguohui@hust.edu.cn Wu Zhou wuzhoutjmu1986@163.com

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 03 April 2018 Accepted: 21 May 2018 Published: 06 June 2018

#### Citation:

Mi B, Wang J, Liu Y, Liu J, Hu L, Panayi AC, Liu G and Zhou W (2018) Icariin Activates Autophagy via Down-Regulation of the NF-κB Signaling-Mediated Apoptosis in Chondrocytes. Front. Pharmacol. 9:605. doi: 10.3389/fphar.2018.00605 Osteoarthritis (OA) is a common chronic and degenerative joint condition that is mainly characterized by cartilage degradation, osteophyte formation, and joint stiffness. The NF-κB signaling pathway in inflammation, autophagy, and apoptosis plays a prominent role in the progression of OA. Icariin, a prenylated flavonol glycoside extracted from Epimedium, have been proven to exert anti-osteoporotic and anti-inflammatory effects in OA. However, the action mechanisms of its effect on chondrocytes have yet to be elucidated. In the present study, we demonstrated that the in vitro therapeutic effects of icariin on rat chondrocytes in a dose-dependent manner. We found that TNF-α induced the production of IL-1, IL-6, IL-12, reactive oxygen species (ROS), nitric oxide (NO), Caspase-3, and Caspase-9 in chondrocytes. We also provided evidence that TNF-α inhibited autophagy markers (Atg 5, Atg 7) and prevented LC3 I translate to LC3 II. Furthermore, TNF-α induced matrix metalloproteinase (MMP)3 and MMP9 expression. The negative effects of TNF-α on chondrocytes can be partially blocked by treating with icariin or ammonium pyrrolidinedithiocarbamate (PDTC, an NF-κB inhibitor). The present study data also suggested that icariin suppressed both TNF-α-stimulated p65 nuclear translocation and IκBα protein degradation. These results indicated that icariin protected against OA by suppressing inflammatory cytokines and apoptosis, through activation of autophagy via NF-κB inhibition. In conclusion, icariin appears to favorably modulate autophagy and apoptosis in chondrocytes making it a promising compound for cartilage tissue engineering in the treatment of OA.

Keywords: icariin, tissue engineering, autophagy, apoptosis, chondrocytes

# INTRODUCTION

Osteoarthritis (OA) is a chronic degenerative joint disease that occurs mostly in the elderly (Glyn-Jones et al., 2015). As the population ages, the occurrence of OA increases and, consequently, finding an effective treatment is imperative. The capacity of chondrocytes to regenerate normal cartilage matrix architecture declines with aging, resulting in cartilage degradation and erosion

(Blanco et al., 2011). In the cases of OA, various changes occur in the cartilage including inflammation, oxidative stress, loss of cartilage matrix, autophagy, and apoptosis.

During the progression of OA, the increased expression of inflammatory cytokines (such as TNF-α and IL-1β) in articular cartilage and synovium contribute to the degradation and erosion of cartilage (Raman et al., 2018). At the same time, autophagy and apoptosis occur in the progression of OA. Autophagy is necessary for maintaining the cell's metabolism and homeostasis, and for cellular quality control by clearing waste or damaged proteins and organelles (Galluzzi et al., 2014). The dysregulation of autophagy that happens with OA contributes to the degeneration of the articular cartilage (Vasheghani et al., 2015). The importance of autophagy in preventing age-related OA has been demonstrated by the increasing number of studies on the topic (Duarte, 2015; Loeser et al., 2016). In addition, more and more evidences suggest that increased chondrocyte apoptosis induces cartilage degeneration in OA (Kobayashi et al., 2016). Thus, activation of autophagy and inhibition of apoptosis in chondrocytes may limit OA progress.

Recently, various biomaterials, such as hydrogels, have been used as drug delivery systems to regulate chondrocyte autophagy (Chen et al., 2016). Driven by the rapid progression of nanomedicine and nanotechnology (Tao et al., 2015, 2017a; Ding et al., 2017; Rosenblum et al., 2018), compounds have been increasingly studied in the context of regeneration medicine and tissue engineering. However, few compounds are reported to be useful for tissue engineering in cartilage repair. Icariin is a well-known compound extracted from Herba Epimedil, with a wide range of pharmacological effects, including antiinflammatory, anti-atherosclerotic, and anti-oxidative properties (Fang and Zhang, 2017; Wang G.Q. et al., 2017; Xiong et al., 2017). Recently, icariin-mediated chondroprotective effects have attracted growing attention. Icariin protects against OA by inhibiting overexpression of metalloproteinase 13 (MMP-13) and proinflammatory cytokines in chondrocytes (Zeng et al., 2014). In addition to inhibiting H2O2-induced human umbilical vein endothelial cell apoptosis, icariin suppresses NF-κB signaling in macrophages (Wang and Huang, 2005; Chen et al., 2010). The effect of icariin on chondrocyte autophagy and apoptosis, however, remains unclear.

In this study, we investigate whether icariin has chondroprotective effects against TNF-α-induced cell death. These effects might be closely related with autophagy activation and apoptosis inhibition. In addition, we explore the underlying mechanisms of icariin-mediated cell autophagy and apoptosis.

# MATERIALS AND METHODS

#### Reagent, Antibodies, and Ethics Statement

Icariin was purchased from Sigma-Aldrich (MO, United States). Cell Counting Kit-8 (CCK8) was purchased from MedChemExpress (NJ, United States). Antibodies against ATG5, ATG7, LC3, p65, phosphorylated p65 (p-p65), IκBα, MMP3, MMP9 were purchased from Abcam (Cambridge, United Kingdom). Antibody against GAPDH was purchased from Cell Signaling Technology, Inc. (MA, United States). The NF-κB inhibitor pyrrolidinedithiocarbamate (PDTC) was purchased from Abcam (Cambridge, United Kingdom). Enzymelinked immunosorbent assay (ELISA) kits were purchased from Bio-Swamp Life Science (Shanghai, China). The Sprague-Dawley (SD) rat was purchased from The Center of Experimental Animal, Tongji Medical College, Huazhong University of Science and Technology. All experimental procedures were approved by the Institutional Animal Care and Use Committee at Tongji Medical College, Huazhong University of Science and Technology.

#### Cell Isolation and Culture

Chondrocytes were isolated from the articular cartilage of 7–10 day-old male SD rats. Briefly, pieces of articular cartilage were digested with 0.25 mg/mL trypsin for 30 min and 2 mg/mL collagenase type II for 8 h at 37◦C. After digestion, isolated chondrocytes were passed through a 180-µm filter and the cells were centrifuged and washed with PBS several times. The cells were then isolated and stained with trypan blue to evaluate cell viability. Chondrocytes with viability greater than 85% were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 100 IU/ml penicillin, 100 µg/ml streptomycin at 37◦C, 5% CO2. When the cells were treated with 50 ng/ml TNF-α, 10 µM icariin, 10 µM PDTC, icariin or PDTC were added to the medium 2 h prior to TNF-α addition. In the autophagic flux assay, 0.1 µm bafilomycin A1 was added to the medium 1 h prior to icariin or PDTC addition. In the experiments, chondrocytes were treated with PBS (control), TNF-α, icariin with TNF-α, or PDTC with TNF-α for 24 h. All experiments were conducted in triplicate.

# Cytotoxicity Assay

Rat chondrocytes were seeded in 96-well plates (5 × 10<sup>3</sup> cells/well) overnight, followed by treatment with various concentrations of icariin for 24 h, 48 h, and 72 h at 37◦C in an atmosphere containing 5% CO2. Following this, 10 µl CCK8 solution was added to each well and the cells were cultured at 37◦C for 2 h. The OD value was then measured with a microplate reader (Thermo Fisher Scientific, United States) at 450 nm.

# ELISA

Chondrocytes were seeded onto a 24-well plate at 2 × 10<sup>4</sup> cells per well. Following 24 h incubation, cells were treated with PBS, TNFα, icariin with TNF-α, or PDTC with TNF-α. The supernatants were collected after 24 h incubation and the levels of IL-1, IL-6, and IL-12 were quantified using the ELISA kits.

#### Western Blot

Total proteins were extracted using cold RIPA buffer containing protease inhibitor (Boster Biological Technology, Ltd., Wuhan, China). Proteins were quantified using the bicinchoninic acid protein assay kit (BCA kit) according to the manufacturer's

instructions. Equal amount of proteins (10 µg) were separated by 10% SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, United States). Membranes were blocked in 5% bull serum albumin (BSA, Sigma, United States) in TBST for 2 h, then incubated with primary antibody at 4 ◦C overnight. The primary antibodies were as follows: p65 (1:2000), p-p65 (1:2000), IκBα (1:5000), ATG5 (1:5000), ATG7 (1:10000), LC3 (1:3000), MMP3 (1:10000), MMP9 (1:10000), and GAPDH (1:1000). After washing with TBST (50 mM Tris PH 8.0, 150 mM NaCl, 0.01% Tween-20) three times, the membrane was incubated with HRP-conjugated corresponding secondary antibodies (Goat Anti-Rabbit IgG, 1:10000) for 1 h at room temperature. Following washing, the immunoreactive proteins were visualized with the ECL western detection kit (Thermo Fisher Scientific, United States). The band density was quantified using TANON GIS software (Tanon, Shanghai, China).

#### Reverse Transcription-Quantitative (RT-q) PCR

Total RNA from chondrocytes was extracted using Trizol reagent (Thermo Fisher Scientific, United States). Total RNA was reverse-transcribed to cDNA with RT Master Mix (Takara Japan). The RT-PCR was performed with SYBR Master Mix using StepOne-Plus system (ABI, United States) under the following conditions: denature at 95◦C for 30 s, anneal at 60◦C for 1 min and extend at 95◦C for 5 s. The gene expression was analyzed by 2−11Ct method using GAPDH as the internal control. The primer sequences were as follows: ATG5, forward (50 -AA CGAGAAGCAGAGCCA-3<sup>0</sup> ) and reverse (5<sup>0</sup> -ATGCCAT TTCAGGGGTG-3<sup>0</sup> ); ATG7, forward (5<sup>0</sup> -GAAGAACCAGAAA GGAGG-3<sup>0</sup> ) and reverse (5<sup>0</sup> -CAGGCACTTGACAGACAC-3<sup>0</sup> ); Bax, forward (5<sup>0</sup> -TGGTTGCCCTCTTCTA-3<sup>0</sup> ) and reverse (50CACCCTGGTCTTGGAT-3<sup>0</sup> ); Bcl-2 forward (5<sup>0</sup> -CACAG AGGGGCTACGAGT-3<sup>0</sup> ) and reverse (5<sup>0</sup> -CAGGCTGGAAGG AGAAGA-3<sup>0</sup> ); GAPDH, forward (5<sup>0</sup> -CAAGTTCAACGGCA CAG-3<sup>0</sup> ) and reverse (5<sup>0</sup> -CCAGTAGACTC CACGAC AT-3<sup>0</sup> ).

#### Transmission Electron Microscopy (TEM)

Collected cells were fixed with 2.5% glutaraldehyde for 24 h, which followed by fixed with 1% osmium tetroxide for 1 h at 4 ◦C. After dehydrated with a series of ethanol concentrations (50, 70, 80, 90, and 100%) for 10-min intervals, the samples were incubated in a mixture of acetone and epoxy resin (v:v = 1:1) for 6 h, followed by incubation with pure epoxy resin for 4 h. After semi-thin sectioning, cells were stained with 0.5% toluidine blue and observed under the microscope. Finally, the ultra-section sections were observed using a TEM (Hitachi, Japan).

# Cell Cycle Assay

Collected cells were washed and suspended in 0.3 mL PBS containing 10% FBS and 0.7 mL ethyl alcohol for 24 h at −20◦C. The cells were then suspended with 0.1 mL RNase A (1 mg/mL) and 0.4 mL propidium iodide (PI) (50 µg/mL) for 10 min. The percentage of cells in the different stages was measured using Flow Cytometry.

# Cell Apoptosis Assay

Collected cells were stained with a mixture of calcein-AM and PI solution for 20 min. Fluorescence images of cells were then recorded using an inverted fluorescent microscope. The percentages of cell death were evaluated by calculating the number of PI-stained (dead, red) and calcein-AM-stained (live, green) cells. The number of cells were counted in five random fields by three independent authors. The mean value of each measurement was used for analysis.

# Measurement of ROS Production

The ROS level was measured using Reactive oxygen species assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufactures' instructions. Briefly, treated cells were incubated with 10 µM dichlorodihydrofluorescein diacetate (DCFH-DA) in the dark for 30 min. Then, the cells were rinsed with DMEM and analyzed on a flow cytometer with excitation wavelength of 500 nm and emission wavelength of 525 nm. ROS level in the experimental group was normalized to the control group.

#### Measurement of NO Production

NO production was detected with nitrate/nitrite colorimetric assay kit of Griess reaction (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's protocol. In brief, cultured medium was collected from treated cells and mixed with equal volume of Griess reagent. Following a 10-min incubation, absorbance was measured at 550 nm on a microplate reader (Thermo Fisher Scientific, United States).

# Caspase-3 and Caspase-9 Activity Assay

Caspase-3 and caspase-9 was detected with Caspase colorimetric assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Briefly, collected cells were lysed in the provided lysis buffer. The absorbance was measured at 405 nm on a microplate reader (Thermo Fisher Scientific, United States).

# Immunocytochemical Staining of p65

Cells were collected and washed with PBS, and then fixed with 4% paraformaldehyde for 15 min. After washing with PBS, the cells were blocked with 5% BSA for 2 h at 37◦C. Then, p-p65 antibody (1:150) was added and incubated overnight at 4◦C, followed by adding Max Vision TMHRP-polymer for 1 h, and followed by incubation with DAB for 5 min. The number of p65 in the nucleus was counted in five random fields by three independent authors. The mean value of each measurement was taken for analysis.

# Statistical Analysis

Data were presented as means ± standard deviation (SD). Statistical analysis was performed in GraphPad Prism 6.0 (GraphPad Software, United States) using Student's t-test and

one-way analysis variance (ANOVA). Statistical significance was considered when P < 0.05.

# RESULTS

# Cytotoxicity of Icariin on Chondrocytes

To determine whether icariin was toxic to chondrocytes, CCK8 assay was used after the cells had been treated with increasing concentration of icariin (0, 3, 5, 7, 10, 20 µM) for 24, 48, and 72 h. As shown in **Figure 1**, icariin promoted cell viability in a dose-dependent manner. The dose of 10 µM and 20 µM icariin had a similar beneficial effect on chondrocytes. (**Figure 1**) The dose of 10 µM was selected for the subsequent experiments.

# Effects of Icariin on TNF-α-Induced Inflammatory Cytokines

There are various inflammatory cytokines involved in the pathologic process of OA. These cytokines, such as IL-1β, induce the production of the other inflammatory cytokines (Hong et al., 2014), leading to inflammatory milieu in chondrocytes. In the present study, we found that TNF-α induced the production of IL-1, IL-6, and IL-12 in chondrocytes. Such effects were partially blocked with the addition of icariin to TNF-α-treated cells (**Figure 2**), suggesting that icariin had an anti-inflammatory effect on chondrocytes.

### Icariin Activates Autophagy in TNF-α-Treated Chondrocytes

To investigate the effect of icariin on the autophagy of chondrocytes, the autophagy markers Atg 5 and Atg 7 were detected. The western blot and PCR results showed that autophagy markers were significantly reduced in TNF-α-treated cells when compared with those cells co-treated with icariin and TNF-α, indicating that icariin protected against TNFα-induced inhibition of autophagy (**Figures 3A–C**). In addition to Atg 5 and Atg 7, western blotting result of LC3 (a classic marker of autophagy) was also used to establish the effect of icariin on autophagy. Icariin significantly enhanced LC3- II level when adding to the TNF-α-treated cells (**Figure 3D**). During the late phase of autophagy, the fusion of lysosomes with autophagosomes may lead to a lower accumulation of LC3- II in the cytoplasm. To demonstrate this effect, bafilomycin A1 was added 24 h prior to cells co-treated with TNF-α and icariin to inhibit the fusion of lysosomes with autophagosomes. The level of LC3-II increased when bafilomycin A1 was added to the TNF-α-icariin-treated cells (**Figure 3D**), suggesting that autophagy may fail to detect when autophagosomes fusion with lysosomes. Furthermore, TEM images illustrated that the number

FIGURE 1 | (A–C) The cytotoxicity of icariin on chondrocytes was examined using the concentration of 0, 3, 5, 7, 10, and 20 µM after 24, 48, and 72 h of culture. Values were expressed as means ± SD. <sup>∗</sup>P < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

means ± SD. <sup>∗</sup>P < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

∗∗∗∗p < 0.0001. Baf: Bafilomycin A1.

of autophagosomes in cells co-treated with icariin and TNF-α was higher than that cells treated solely with TNF-α (**Figure 3E**). These results indicated that icariin activated autophagy in TNFα-treated chondrocytes.

#### Effects of Icariin on TNF-α-Induced Proliferation Inhibition and Apoptosis Activation

It is well known that inhibition of the proliferation rate correlates with decreased cell cycle progression in chondrocytes. We analyzed the cell cycle division of chondrocytes treated with PBS, TNF-α and icariin with TNF-α. According to the PI staining of flow cytometry analysis, TNF-α arrested chondrocytes in the G1 phase. However, the percentage of the cell population in the S phase increased from 22.27% to 26.11% when addition of icariin to the TNF-α-treated chondrocytes (**Figures 4A,B**). These results suggested that cell cycle progression might be delayed by TNF-α, a delay that is partially reversed by icariin treatment.

We also analyzed the effect of TNF-α on chondrocyte apoptosis by calcein-AM/PI double-staining and found that TNF-α significantly induced chondrocyte apoptosis. However, when icariin was added to TNF-α-treated cells, the percentage of cell death decreased from 46.33% to 24.33% (**Figures 4C,D**), suggesting that icariin protected against TNF-α-induced cell apoptosis. To further determine whether icariin affects the mitochondrial apoptosis pathway of chondrocytes, the antiapoptotic protein Bcl-2 level and the pro-apoptotic Bax level were detected using western blot and PCR. TNF-α significantly inhibited Bcl-2 expression but induced Bax expression. These effects were partially blocked by icariin treatment (**Figures 5A– E**). Furthermore, the effect of that TNF-α-induced Caspase-3/9 activation in chondrocytes was blocked by addition of icariin. (**Figures 5F,G**).

∗∗∗∗p < 0.0001.

# Effect of Icariin on TNF-α-Induced NO and ROS Production and Catabolism in Chondrocytes

The level of NO and ROS is known to increase when chondrocytes are exposed to inflammatory cytokines. In the present study, we found that TNF-α significantly induced NO and ROS production, an effect which could be blocked by addition of icariin (**Figure 6**).

Disruption of homeostasis in cartilage metabolism impairs the extracellular matrix and chondrocytes, resulting in cartilage degradation. The excessive matrix catabolism was usually caused by excessive mechanical joint loading or inflammatory cytokines, which was reflected by the increased levels of matrix-degrading enzymes matrix metalloproteinase (MMP) family. In the present study, TNF-α induced the production of MMP 3 and MMP 9, an effect which could be blocked by addition of icariin (**Figure 7**). These results indicated that icariin inhibited NO and ROS production and reconstituted homeostasis of metabolism in chondrocytes.

# Icariin Inhibits NF-κB Pathway Activated by TNF-α in Chondrocytes

In the present study, western blot results showed that TNFα significantly increased the level of p-p65 and decreased the level of IκBα in chondrocytes when compared with the control group, suggesting that TNF-α stimulation of the NF-κB pathway (**Figures 8A,B**). This finding also proven by the result of immunocytochemical staining that the number of p65 translocate

to nucleus in TNF-α-treated cells is higher than the control group (**Figures 8C,D**). In contrast, icariin caused the TNF-α-induced p-p65 increase and IκBα degradation reversal. Furthermore, the number of p65 translated to the nucleus also decreased by addition of icariin to TNF-α-treated cells. (**Figure 8**).

#### Involvement of NF-κB Pathway in the Autophagy Activation and Apoptosis Inhibition Induced by DHA

The NF-κB pathway plays a crucial role in various pathogenesis process of OA. To investigate whether this pathway was involved in the TNF-α-induced negative effect on chondrocytes, we used an NF-κB inhibitor, PDTC. In the present study, we found that PDTC partially blocked the TNF-α-induced inflammatory cytokines production, proliferation and autophagy inhibition, apoptosis activation, and hypercatabolism. (**Figures 2**–**8**).

# DISCUSSION

Inflammatory cytokines including TNF-α and IL-6 contribute to cartilage catabolism and degeneration in OA (Lai et al., 2014; Nasi et al., 2016). Excessive inflammatory cytokines inhibit autophagy activation, further increasing production of ROS and leading to cell death (Liu-Bryan and Terkeltaub, 2015). In the current study, we demonstrated that TNF-α suppressed autophagy in chondrocytes, which was reflected by the lower level of autophagy marker. Previous studies found that TNF-α inhibited the expression of Atg 5 and suppressed the conversion of LC3 I to II (Jiang et al., 2016). With Atg 5 knockdown, mice were more likely to develop OA with aging (Bouderlique et al., 2016), suggesting that activation of autophagy has a beneficial effect in preventing OA. When inflammatory cytokines release, the level of ROS and NO production increases, ultimately leading to chondrocyte apoptosis (Akuri et al., 2017). The present study

results consistent with previous studies which showed that TNFα induced chondrocyte apoptosis via upregulated expression level of ROS and NO. In addition, previous studies, as well as the current study, have demonstrated that TNF-α induced the pro-apoptotic protein Bax production and reduced the antiapoptotic protein Bcl-2 production, leading to hypercatabolism in chondrocytes (Ye et al., 2015). It should be noted that there are other proinflammatory cytokines contribute to the autophagy suppression and apoptosis induction in OA (Goldring and Otero, 2011). These cytokines may result in a synergistic effect in inhibiting autophagy and activating apoptosis during the pathological process of OA. Unfortunately, the present study reported that TNF-α further induced the production of other inflammatory cytokines. Therefore, it is necessary to regulate the inflammatory processes of OA. Indeed, treatments of OA with non-steroidal anti-inflammatory drugs that inhibit the release of proinflammatory cytokines is common practice (Urech et al., 2010).

The cytotoxic effects of icariin on other cell types have been investigated (Zhang et al., 2015). In the present study, icariin exhibited a positive effect on cell viability, in a dosedependent manner. Both doses of 10 µM and 20 µM icariin showed similar effect on cell viability. Previous studies reported that icariin alleviated the inflammatory response in most cell types (Pan et al., 2017; Sun et al., 2018). In the present study, we demonstrated that icariin attenuated the production of inflammatory cytokines in chondrocytes. As overproduction of inflammatory cytokines inhibit autophagy activation (Qi et al., 2014), it is logical to suppose that icariin reverses the inflammatory cytokines-induced autophagy inhibition. Some studies reported that icariin plays a beneficial role on cell survival by inhibiting autophagy (Tang et al., 2015; Li et al., 2017). In the present study, however, we found that icariin blocked the TNF-α-induced autophagy inhibition in chondrocytes. This result may be attributed to its anti-inflammatory effect on chondrocytes. When autophagy activation, it could further suppress inflammatory response in chondrocytes (Ansari et al., 2017). Given that icariin and autophagy have both been shown to attenuate inflammatory response, this may act as a positive feedback loop to suppress TNF-α-induced inflammation. Along with the production of inflammatory cytokines was suppressed by icariin, so was apoptosis. In addition, the decreased level of NO and ROS induced by icariin also contribute to the inhibition of apoptosis and catabolism (Shen et al., 2015; Qiao et al., 2018). The apoptosis results were consistent with previous studies that support that icariin had protective effect against apoptosis (Deng et al., 2017). Consequently, icariin plays a dual role of autophagy activation and apoptosis inhibition in chondrocytes.

Previous studies reported that activation of the NF-κB signaling pathway induced pro-inflammatory cytokines release (Dubey et al., 2018). Exposure of chondrocytes to a variety of inflammatory cytokines lead to the degradation of IκB, allowing p65 translocate to the nucleus (Guo et al., 2015). The present study showed that TNF-α induced NF-κB pathway activation. Icariin is known to inhibit NF-κB signaling, leading to anti-inflammatory effect (Hua et al., 2018). Thus, the TNFα-induced NF-κB activation was reversed with addition of icariin, suggesting that icariin had a negative effect on NF-κB activation. NF-κB plays a key role in both chondrocyte autophagy and apoptosis. When the NF-κB signaling pathway was activated by inflammatory cytokines, autophagy inhibition along with apoptosis activation appear, which play a synergetic effect on accelerating chondrocyte death (Jiang et al., 2016; Zhang et al., 2016). The results that autophagy markers significantly increased and apoptotic markers significantly decreased when addition of icariin or PDTC to the TNF-α-treated chondrocytes, suggested that icariin protected against OA by inhibiting the TNFα-induced NF-κB signaling pathway activation. Interestingly, we noticed that icariin treatment had a better anti-inflammatory effect than PDTC treatment on TNF-α-treated cells in terms of IL-1 and IL-6. In addition, it also had a better reversal effect than PDTC, i.e., on the negative effect of TNF-α-induced autophagy inhibition and apoptosis activation. Previous studies reported that in addition to the NF-κB signaling pathway, other signaling pathways such as the MAPK/JNK and ERK pathways are also involved in the pathogenesis process of OA (Wang et al., 2018). In the future experiments, we would like to verify whether icariin protects against OA via these or other signaling pathways.

Tissue engineering is an evolving interdisciplinary field integrating medicine, material science, biochemistry, and

biomedical engineering, which centered on development of biological alternates to restore and/or to improve tissue and organ function (Tao et al., 2013; Tao et al., 2016; Zhu et al., 2018). Along with the development of biomaterial, the drug delivery system has attracted more attention in the treatment of cancer, diabetes and OA (Tao et al., 2017b; Wang J. et al., 2017; Zhao et al., 2017; Maudens et al., 2018). Jiang et al. (2018) developed a poly (lactic-co-glycolic acid)-based nanoscale drug delivery system for the treatment of OA. Li et al. (2012) reported that icariin up-regulated the expressions of aggrecan, sox9, and collagen I of chondrocytes, features which make it a potential promoting compound for cartilage tissue engineering. Previous studies found that targeted microspheres loaded with icariin could exert colon-protective effects through reduction of the inflammatory response (Wang et al., 2016). Pan et al. (2016) also reported that icariin loaded biphasic-induced magnetic CS/nHA/MNP microcapsules is a useful drug delivery system for bone repair . Considering the various beneficial effects of icariin, it could be considered an excellent compound to be used in drugdelivery system. In the future studies, we plan to look for a good biocompatibility drug delivery system which is suitable to load icariin for the treatment of OA.

In conclusion, we demonstrated that icariin had no cytotoxic effects on chondrocytes up to the dose of 20 µM. TNF-α induced inhibition of autophagy and activation of apoptosis, and increased inflammatory cytokines, NO and ROS, as

#### REFERENCES


well as stimulating catabolism. These negative effects could be partially reversed by adding icariin to TNF-α-treated chondrocytes via inhibition of NF-κB signaling. Thus, the present study highlights that icariin induces autophagy activation and apoptosis inhibition of chondrocytes via suppression of the NF-κB signaling pathway. Future in vivo evaluation using macro-hydrogels to explore icariin's therapeutic potential in OA treatment is currently under way in our laboratory.

#### AUTHOR CONTRIBUTIONS

BM, WZ, and GL conceived and designed the experiments. JL, YL, and LH performed the experiments. BM and JW wrote the manuscript and made the same contribution to the manuscript. AP revised the language of the manuscript.

#### FUNDING

The study was supported by the Natural Science Foundation of Hubei Province (Grant No. WJ2017Q025), the Science and Technology Department of Hubei Province (Grant No. 2016CFB303), and the Free Innovation Research Fund of Wuhan Union Hospital.



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewers and handling Editor declared their shared affiliation.

Copyright © 2018 Mi, Wang, Liu, Liu, Hu, Panayi, Liu and Zhou. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-09-00605 June 4, 2018 Time: 14:17 # 12

# Human Albumin Fragments Nanoparticles as PTX Carrier for Improved Anti-cancer Efficacy

Liang Ge1,2† , Xinru You<sup>3</sup>† , Jun Huang<sup>3</sup>† , Yuejian Chen<sup>4</sup> , Li Chen<sup>2</sup> , Ying Zhu<sup>1</sup> , Yuan Zhang<sup>5</sup> , Xiqiang Liu<sup>6</sup> \*, Jun Wu<sup>3</sup> \* and Qian Hai<sup>1</sup> \*

<sup>1</sup> State Key Laboratory of Natural Medicines, China Pharmaceutical University, Nanjing, China, <sup>2</sup> School of Pharmacy, Xinjiang Medical University, Ürümqi, China, <sup>3</sup> Key Laboratory of Sensing Technology and Biomedical Instruments of Guangdong Province, School of Biomedical Engineering, Sun Yat-sen University, Guangzhou, China, <sup>4</sup> Nanjing iPharma Technology, Co., Ltd., Nanjing, China, <sup>5</sup> Department of Orthopedics, Xinqiao Hospital, Third Military Medical University, Chongqing, China, <sup>6</sup> Guangdong Provincial Key Laboratory of Stomatology, Guanghua School of Stomatology, Hospital of Stomatology, Sun Yat-sen University, Guangzhou, China

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Teresa Musumeci, Università degli Studi di Catania, Italy Huiyun Wen, Northwest University, China Romina Julieta Glisoni, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina Yue Pan, Soochow University, China

#### \*Correspondence:

Xiqiang Liu xiqiangliu@aliyun.com Jun Wu wujun29@mail.sysu.edu.cn Qian Hai qianhai24@163.com †These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 14 March 2018 Accepted: 15 May 2018 Published: 12 June 2018

#### Citation:

Ge L, You X, Huang J, Chen Y, Chen L, Zhu Y, Zhang Y, Liu X, Wu J and Hai Q (2018) Human Albumin Fragments Nanoparticles as PTX Carrier for Improved Anti-cancer Efficacy. Front. Pharmacol. 9:582. doi: 10.3389/fphar.2018.00582 For enhanced anti-cancer performance, human serum albumin fragments (HSAFs) nanoparticles (NPs) were developed as paclitaxel (PTX) carrier in this paper. Human albumins were broken into fragments via degradation and crosslinked by genipin to form HSAF NPs for better biocompatibility, improved PTX drug loading and sustained drug release. Compared with crosslinked human serum albumin NPs, the HSAF-NPs showed relative smaller particle size, higher drug loading, and improved sustained release. Cellular and animal results both indicated that the PTX encapsulated HSAF-NPs have shown good anti-cancer performance. And the anticancer results confirmed that NPs with fast cellular internalization showed better tumor inhibition. These findings will not only provide a safe and robust drug delivery NP platform for cancer therapy, but also offer fundamental information for the optimal design of albumin based NPs.

Keywords: human serum albumin fragments, nanoparticle, paclitaxel, drug delivery, anticancer

#### INTRODUCTION

With the fast growing of material chemistry and nanomedicine, biodegradable nanoscale drug delivery platforms, including nanoparticles, micelles (Wang et al., 2014a; Li W. et al., 2016; Qu et al., 2017) and liposomes, have been widely utilized for biomedical diagnosis (Park et al., 2009; Morral-Ruiz et al., 2013; Wang et al., 2014b; Hu et al., 2018) and therapy (Boussif et al., 1995; Ding et al., 2013; Xing et al., 2013; Bertrand et al., 2014; You et al., 2016; Ge et al., 2017; Li et al., 2017; Pan et al., 2018; Yang et al., 2018). Recently, a great number of functional delivery systems (Liu and Lu, 2006; Cho et al., 2011; Shrestha et al., 2012; Wu et al., 2012, 2014; Li et al., 2013; Wu and Chu, 2013; Yu et al., 2013; Hai et al., 2014; Li H. et al., 2016; Hao et al., 2017; Xu et al., 2017) have been studied. But even the nanoparticles (NPs) based on FDA approved materials, such as the poly-ε-caprolactone (PCL), poly(DL-lactic acid), poly(lactide-cocaprolactone), and poly(lactideco-glycolide) (PLGA), are still toxic for high dosage treatment (Singh and Ramarao, 2013). Then, NP systems with improved biocompatibility are highly desired (Maiti, 2011; Wang et al., 2016, 2018).

Albumin, as a biodegradable, non-toxic and non-immunogenic protein, has been used to prepare NPs (Elzoghby et al., 2012). Albumin based nanocarriers (Shimanovich et al., 2011; Altintas et al., 2013; Bakare et al., 2014; Rosenberger et al., 2014; Watcharin et al., 2014) have been reported and the albumin-bound paclitaxel (Abraxane <sup>R</sup> ) had been proved to be safe and efficient (Saif, 2013; Cecco et al., 2014). The crosslinked albumin NPs were able to increase their physical stabilities, but

**100**

the drug encapsulation efficiencies and sustained release still need to be improved (Kratz, 2008; Li et al., 2010, 2014; Elzoghby et al., 2012; Kou et al., 2018). For this goal, we hypothesized that albumin fragments based NPs could have better drug loading/release performance, which would result improved anti-cancer performance. Non-toxicity, biodegradability and preferential uptake in tumor and inflamed tissues make human serum albumin fragments (HSAFs) an ideal drug delivery system. Due to these advantages, it's motivated to develop a novel and safety nanoplatform based on HASF.

Therefore, in this report, as a model platform, a HSAF NP platform was developed as drug carriers with different crosslinking degrees and diameters using genipin as a very biocompatible crosslinker (**Figure 1A**). HSAFs were obtained via the degradation and the natural genipin crosslinker is expected to significantly reduce the toxicity while keep the similar crosslinking capability, comparing to the widely used glutaraldehyde. HSAF NPs were screened by a quantitative method based on FRET theory (**Figure 1B**) following previous report to obtain faster cellular uptake for further evaluations (Jiang et al., 2015).

# MATERIALS AND METHODS

#### HSAF NPs Preparation

Human serum albumin (HSA) was dissolved deionized water before 70% formic acid was added. CNBr was then added to degrade the HSA into fragment products (McMenamy et al., 1971). HSA degradation products were separated and purified by Superdex75 (Lapresle and Doyen, 1975; Yuan et al., 2014), and the three main peptide fragments are Fraction A299−585, Fraction B1−123, and Fraction C124−<sup>298</sup> (46−48) (w : w = 3.5:1:2.2), then the purified HSAF was used to prepare HSAF NPs. HSAF NPs were prepared and characterized according to the published protocols (Jiang et al., 2013). Then HSAF NPs with different crosslinking degrees and diameters were developed using a disulfide bond reducing method established in the previous work (Jiang et al., 2013). Briefly, a predeterminded amount of cysteine was added into the HSAF PBS solution under pH 8.0 at 37◦C. The final concentration is 5 mg/mL. After dialysis, the remaining cysteine was removed and genipin was used to do second crosslinking (1 h). The genipin residue was cleaned by same method as above and the HSAF NPs were collected by lyophilization. The HSAF NP library with formulation parameters is summarized in **Table 1**.

# Preparation and Characterization of Paclitaxel (PTX)-Loaded HSAF NPs

The PTX-loaded HSAF NPs with serious of crosslinking densities and diameters (PLC40S70, PLC70S70, PLC90S70, PLC40S160, and PLC40S260) were prepared following the same method preparing the above NPs. Briefly, the certain amount of HSAF was dissolved in PBS 8.0 at 37◦C, and then PTX were dissolved in ethanol (PTX/HSAF50 mg/g) and genipin were added. Then system was incubated for 30 min to complete the crosslinking. The NP solution was purified and concentrated using Amicon Ultra Centrifugal Filters (MWCO 100,000). The size and zeta potential of PTX-loaded HSAF NPs were evaluated by a zeta potential and particle size analyser (ZetaPlus, Brookhaven, NY, United States). The morphology of nanoparticle was verified by transmission electron microscope (H-7650, HITACHI, Japan).

#### PTX Drug Release Profiles of HSAF NPs

Dialysis was used to determine the release behavior of paclitaxel from nanoparticles. 3 mL of the nanoparticle suspension


TABLE 1 | Physical and chemical properties of HSAF and HSA NPs.

(containing 10 mg PTX) was placed in a dialysis bag (molecular weight cut-off: 13 kDa). The dialysis bags were placed in 80 mL of 1 M salicylic acid solution. Shaking was performed at a shaking speed of 100 rpm and a temperature of 37◦C. 0.5 mL of dialysate was collected at 0.5, 1, 2, 4, 6, 8, 12, 24, 36, and 48 h, respectively, and an equal volume of fresh dialyzing media was added.

#### Pharmacokinetic Studies

The PTX NPs (PLC40S70, PLC70S70, and PLC90S70) had similar diameters of about 70 nm but different crosslinking densities (41, 65, and 90%). And the NPs (PLC40S70, PLC40S160, and PLC40S260) prepared with albumin had similar crosslinking degrees around 42% but different diameters (65.9, 157.3, and 255.9 nm). The NPs (PLC40S70, PLC70S70, PLC90S70, PLC40S160, and PLC40S260) was i.v. administrated to Sprague-Dawley rats at a dose of 1 mg/kg as PTX. The blood sample was collected (100 µL) from rats into heparinized tubes at scheduled time (0, 0.083, 0.167, 0.333,

0.5, 1, 2, 4, 8, 12, and 24 h). Plasma was separated via centrifuging (4000 rpm, 10 min) and stored under −70◦C until analysis. The drug concentrations were measured by LC–MS/MS (Wang et al., 2013) as previous study. For details, the analytes were eluted with at 5% mobile phase A methanol and 95% B water phase (containing 0.1% formic acid). The flow rate was 0.3 mL/min, and the temperature of column was 30◦C. Mass analysis was operated in the positive ionization mode. Quantification was accomplished by monitoring the transition of m/z 876.0→307.8 for paclitaxel and m/z 830.3→549.0 for docetaxel (the internal standard). The spray voltage, the temperature of capillary, sheath gas pressure and auxiliary gas pressure were set at 4000 V, 350◦C, 35 and 25 Arb, respectively. The pharmacokinetic parameters were estimated via a non-compartmental analysis (WinNonlin computer program, Version 4.0; Pharsight Corporation). All the experiments were performed in accordance with the recommendations of "guidelines of the Experimental Laboratory Animal Committee of China Pharmaceutical University and the National Institutes of Health's Guide for the Care and Use of Laboratory Animals." The protocol was approved by the "Experimental Laboratory Animal Committee of China Pharmaceutical University."

#### In Vivo Imaging Study of HSAF Nanoparticles in Tumor-Bearing Mice

The NPs (C40S70, C70S70, and C90S70) had similar diameters around 70 nm but different crosslinking densities (42, 66, and 91%). And the NPs (C40S70, C40S160, and C40S260)

TABLE 2 | The diameters and drug loading efficiencies of the PTX-loaded HSAF and/or HSA NPs (n = 3).


prepared with various protein fragment concentrations had similar degrees of crosslinking around 40% but different size (65.9, 157.3, and 255.9 nm). DLC40S70, DLC70S70, DLC90S70, DLC40S160, and DLC40S260 expressed (DL means Dir was loaded) was made as the method as the NPs above, and the crosslinking densities and diameters differs not as the NPS without Dir. The certain amount of HSAF (Dir/protein 50 mg/g) was dissolved in PBS 8.0, and then the ethanol Dir was added at 37◦C water bath. 5 mg/mL Cys was added. The system was cooled for 10 min. Genipin cross-linking was completed after dialysis. The configuration of the concentration of 50 µg/mL Dir which was dissolved in the polyoxyethylene castor oil and ethanol (50:50, V/V) solution was made as a control group. DLC40S70, DLC70S70, DLC90S70, DLC40S160, DLC40S260, and Dir solution formulation (0.5 mg/kg) was injected into the tumor-bearing mice via the tail vein. After intravenous injection, intraperitoneal injection of sodium pentobarbital solution (1%, 50 mg/kg) was given to anesthetize the mice. After the anesthetization, the whole body fluorescence images were acquired using small animal in vivo near-infrared imaging system at 0.5, 1, 2, 4, and 8 h.

# In Vivo Anticancer Evaluation in Breast Cancer Models

To evaluate in vivo anticancer activity of PTX-loaded HSAF and HSA NPs, PTX-loaded HSAF and HSA NPs were made by the method as C90S70 NPs whose particle size is the smallest and the crossing link degree is the highest. The 1 × 10<sup>7</sup> /ml MCF-7 cells were re-suspended in 9% saline, and 0.1 ml cells suspension was injected to nude mice on the right axillary subcutaneous. Tumor volumes were determined on alternate day by a vernier caliper, and the tumor volumes were calculated by an equation: V (cm<sup>3</sup> ) = a × b 2 /2 (a: largest diameter; b: smallest diameter), meanwhile, mice weights were monitored three times per week. Fourteen days after tumor implantation, the volumes of tumor size were allowed to reach no less than 0.1 cm<sup>3</sup> , and mice groups (n = 8) were designed to have paclitaxel at dose of 5 mg/kg

FIGURE 3 | The cumulative release profiles of PTX from HSA NPs and HSAF NPs with various crosslinking degrees (A) and diameters (B) (n = 3).

intravenously, (A) control group received 0.9% NaCl every 2 day (B) PTX Injection (PTX equivalent of 5 mg/kg) every 2 day; (C) HSA NPs (PTX equivalent of 5 mg/kg) every 2 day; (D) HSAF NPs (PTX equivalent of 5 mg/kg) every 2 day; After 28 days of initial treatment, mice were sacrificed and tumor tissues were collected. The tumor volume and weight were used for assessment of the therapeutic activity.

#### RESULTS AND DISCUSSION

#### Synthesis and Characterization of Human Albumin Based NPs

To figure out how physical and chemical properties of NPs may affect the cellular behavior of HSAF NPs, a HSAF NP platform was prepared using genipin as crosslinker with different crosslinking degrees and sizes, but similar surface charge (zeta potentials: −20 ∼ −30 mV). The physical and chemical properties are summarized in **Table 1**. The NPs with different crosslinking degrees from 40.9 to 90.1% (C40S70, C70S70, and C90S70) were obtained by reacting with predetermined genipin. The NPs (C40S70, C40S160, and C40S260) prepared from different amount of albumin had similar crosslinking degrees around 41% but different sizes (65.9, 157.3, and 255.9 nm). These NPs are named as CxSyz: C means crosslinking; x means the crosslinking density; S means size; y means the NP size is around that number; z means is HSA or HSAF. A HSA or HSAF library could be obtained by varying the x and y. **Figure 2** showed the one example of TEM image of C70S70HSAF. The FRET indices of these NPs were in the range of 23 ∼ 32%, indicating that the NPs formulated in this study have significant FRET effects.

#### Preparation and Characterization of PTX-Loaded HSAF NPs

The PTX-loaded HSAF and/or HSA NPs were prepared as the method as the NPs above. As shown in **Table 2**, compared to the HSA NPs, the smaller diameters and higher drug loading efficiencies were obtained by the HSAF NPs when using the same formulations.

#### PTX Drug Release Profiles of HSAF NPs

In vitro PTX release profiles from HSA NPs and HSAF NPs were shown in **Figure 3**. Palitaxel release from NPs was detected by dialysis (Cho et al., 2004), the drug released was calculated at scheduled time (0.5, 1, 2, 4, 6, 8, 12, 24, 36, and 48 h). Compared with HSA NPs, PTX was released more slowly from HSAF NPs: within 48 h less than 50% of PTX was released from HSA NPs, but for HSAF NPs less than 25% PTX was released within the same time, indicating HSAF NPs could provide more possibility to delivery of PTX to specific organs and tissues than HAS NPs. For all HSA and HSAF NPs with comparable diameters, the increase of crosslinking degree decreased PTX release, perhaps because the compact structure of NPs, which was brought in by the chemical crosslink, hindered diffusion of PTX from NPs. F or NPs with comparable crosslinking degrees, the small NPs (C40S70 and C40S160) possessed faster PTX release behaviors, perhaps because small NPs possessed short drug diffusion distances.

### In Vitro Anti-cancer Evaluation of HSAF NPs

The cytotoxicity profiles of PTX-loaded HSAF NPs were compared in MCF-7 cells by MTT assay over a range of concentrations (0.25, 0.5, 1, 2, 4, 8, and 16 µg/mL). As shown in the **Figure 4**, the viability of the cells was dose-dependently decreased by the PTX-loaded HSAF NPs. Furthermore, the inhibitory effects of the NPs were increased with the NPs crosslinking density or diameter increasing at 24, 48, and 72 h, and this may result from the more efficient endocytosis brought in by the increased crosslinking degree and diameter. In comparison to the commercial product Taxol <sup>R</sup> , the weaker inhibitory effect of the PTX-loaded HSAF NPs was believed to associate with the sustained release and/or the relatively slower endocytosis of the NPs (Jia et al., 2014).

#### Pharmacokinetic Studies of HSAF NPs

The major pharmacokinetic parameters of i.v. administration of HSAF PTX NPs have been summarized in **Table 3**. PTX NPs with lower particle size (C40S70, C70S70, C90S70) showed an obvious increasing in the AUC, MRT, t1/<sup>2</sup> (P < 0.05), which correlated with an obvious decreasing in the Cl (P < 0.05). The results could be due to the lower particle size decrease the uptake of NPs by the mononuclear phagocyte system (Dobrovolskaia et al., 2008). The increasing of AUC, MRT, t1/2b and Cl could be achieved from C40S70, C70S70 and C90S70 as compared with C40S160 and C40S260 (P < 0.05), but the increment for C40S160 was not

#### TABLE 3 | Pharmacokinetic parameters of PTX after i.v. administration of NPs in mice (1 mg/kg).


Each value is the mean ± SD, n = 6. <sup>∗</sup>P < 0.05, compared with C90S260; #P < 0.05, compared with C70S160.

higher than C40S260. This result indicate that the NP size which is lower than 100 nm may be due to the inhibiting the fast uptake of NPs via the reticulo-endothelial system (RES) (Dobrovolskaia et al., 2008). The HAS PTX NPs which was made by the same method as C90S70 (C90S70HSA) was also studied, and the AUC, MRT, t1/<sup>2</sup> of HAS PTX NPs is lower than that of HSAF PTX NPs (C90S70). There are significant difference between C90S70 and C90S70HSA (P < 0.05). The drug loading rate of C90S70 is higher than that of C90S70HSA, and the drug circulation time is longer than that of C90S70HSA too.

# In Vivo Imaging Study of HSAF NPs in Tumor-Bearing Mice

As shown in **Figure 5**, most of the HSAF NPs containing the nearinfrared fluorescent probe Dir clearly enriched in the liver after

FIGURE 5 | Paclitaxel plasma concentration profiles after intravenous administration of 1 mg/kg drug to Sprague-Dawley rats (A) and in vivo distribution of Dir labeled HSAF nanoparticles DLC90S70 (A) , DLC70S70 (B) , DLC 40S70(C) , DLC40S160 (D) , DLC40S260 (E) , and Dir solution (F) in tumor-bearing mice (B). Results are expressed with the mean ± SD (n = 6). DL, Dir loading.

three kinds of HSAF NPs were delivered to tumor-bearing mice in vivo, indicating that the liver is still the barrier for NP system to achieve maximum efficient drug delivery to tumor, and how to avoid the NPs to be taken up by the liver is still priority problem for the drug delivery systems. The NPs with higher crosslinking densities or particle size distribute more to the liver. In addition to the liver, the right forelimb solid tumors in mice are the main distribution area of the HSAF NPs. From the graph, it could be observed that DLC40S70, DLC70S70, DLC90S70, DLC40S160, and DLC40S260 significantly concentrated in the tumor site at 4, 1, 0.5, 1, and 0.5 h, respectively. This difference suggests that NPs with higher size and the degree of crosslinking have the higher biodistribution to the tumor site. In this experiment, polyoxyethylene castor oil and ethanol (50:50, V/V) were used for preparing solutions. Dir was used as control, and we find Dir solutions distribute quickly to the mouse head, limbs, and solid tumors. The Dir solutions disappeared in the mice quickly and much faster than the nanoparticles.

#### In Vivo Anticancer Evaluation of HSAF NPs for Breast Cancer Models

After 14 days of tumor inoculation, the average tumor volume was around 101 ± 23.19 mm<sup>3</sup> . Then administration was continued for a total of 28 days after tumor implantation. The results (**Figure 6**) showed that tumors were significantly ( <sup>∗</sup>P < 0.01) inhibited after being treated with HSA NPs and HSAF NPs compared to PTX injection. Tumor inhibitory rate (tab.) in mice treated with HSA NPs and HSAF NPs were 63.3 and 71.4 respectively, which showed a significant (∗P < 0.01) compared to PTX injection and control groups. There was observed a nonsignificant (P > 0.05) change in tumor inhibitory rate for HSA NPs and HSAF NPs, and the tumor Inhibitory rate of HSAF NPs is higher than that of HSA NPs. There is no observable weight loss or other cytotoxicity in HSA NPs and HSAF NPs mice groups. Also, the tumor volume showed the same trend as the tumor weight, and the tumor volume of HSAF NPs group is the smallest. The dose of HSAF used in the NPS is lower than that of HSA, but the effect is better.

#### CONCLUSION

In this report, a HSAF NP system with controllable crosslinking density and size were developed for better biosafety and

#### REFERENCES


anticancer efficacy. The HSAF NP library with a series of crosslinking degrees and particle sizes were developed, and the results showed that the similar particle size of HSAF NPs had different crosslinking densities, and the highest crosslinking density combined with the smallest particle size. This may lead to a higher drug loading rate and longer drug circulation time and further higher biodistribution in tumor site. Drug loading and release tests confirmed that the HSAF NPs have better drug loading and release performance than HSA NPs. In vivo anticancer evaluations confirmed that the NPs with fast cellular uptake showed better tumor accumulation and tumor inhibition. The results provide basic information not only for the biochemical effects and biosafety of albumin based NPs, but also for regulating the physicochemical properties which are important for the in vivo delivering of drugs.

#### AUTHOR CONTRIBUTIONS

LG, JW, and QH conceived and directed the study. YC and LC prepared NPs and obtained spectroscopic results. LG and XY co-wrote the paper. LG, XY, and JH contributed to the results analysis and discussion. YiZ and YuZ provided technical support and corrections of manuscript. XL, JW, and QH oversaw the project. All authors reviewed and approved the final paper.

#### FUNDING

This work was supported by the National Natural Science Foundation of China (Grant Nos. 81760638 and 81572978), the Natural Science Foundation of Xinjiang Province (Grant No. 2017D01C200), the Thousand Talents Plan for Young Professionals and the Science and Technology Planning Project of Guangdong Province (Grant No. 2016A010103015), and the Science and Technology Program of Guangzhou (Grant No. 201707010094). Guangdong Innovative and Entrepreneurial Research Team Program (Grant Nos. 2016ZT06S029 and 2013S086). The National Natural Science Foundation of China (Grant Nos. 81372885 and 81772889) and the Major Special Research Collaborative Innovation of Guangzhou (Grant No. 201604020160).

cancer biology. Adv. Drug Deliv. Rev. 66, 2–25. doi: 10.1016/j.addr.2013. 11.009



**Conflict of Interest Statement:** YC was employed by company Nanjing iPharma Technology, Co., Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Ge, You, Huang, Chen, Chen, Zhu, Zhang, Liu, Wu and Hai. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Dual-Responsive Core Crosslinking Glycopolymer-Drug Conjugates Nanoparticles for Precise Hepatocarcinoma Therapy

Jing Wu, Jiayi Yuan, Baotong Ye, Yaling Wu, Zheng Xu, Jinghua Chen\* and Jingxiao Chen\*

Key Laboratory of Carbohydrate Chemistry and Biotechnology, Ministry of Education, School of Pharmaceutical Sciences, Jiangnan University, Wuxi, China

Nanoparticles (NPs) have demonstrated a potential for hepatocarcinoma therapy. However, the effective and safe NP-mediated drug transportation is still challenging due to premature leakage and inaccurate release of the drug. Herein, we designed a series of core cross-linking galactose-based glycopolymer-drug conjugates (GPDs) NPs with both redox-responsive and pH-sensitive characteristics to target and program drug release. Glycopolymer is comprised of galactose-containing units, which gather on the surface of GPD NPs and exhibit specific recognition to hepatocarcinoma cells, which over-express the asialoglycoprotein receptor. GPD NPs are stable in a normal physiological environment and can rapidly release the drug in hepatocarcinoma cells, which are reductive and acidic, by combining disulfide bond cross-linked core, as well as boronate ester-linked hydrophilic glycopolymer chain and the hydrophobic drug.

Keywords: glycopolymer, redox-responsive, pH-sensitive, hepatocarcinoma therapy, ASGPR

# INTRODUCTION

Human hepatocellular carcinoma (HCC) is a common type of primary liver cancer and is diagnosed in more than half a million people worldwide (de Souza et al., 2015). At present, HCC has received increasing attention because of its important effect on the physiological functions of the liver, high lethality, and the growing incidences in many regions (Torre et al., 2015). Chemotherapy is a common HCC treatment strategy that is limited because it is always accompanied by doselimiting toxicity, high rate of tumor recurrence, and drug resistance (Dutta and Mahato, 2017; Kuruvilla et al., 2017). Moreover, the phagocytosis of Kupffer cells hinders the accuracy and efficacy of chemotherapeutics (Lahmar et al., 2016). To date, nanotechnology is applied to address these problems and has obtained effective molecular-level diagnosis (Mura et al., 2013; Lim et al., 2015). It also serves as vectors, sensors, and targeting agents to achieve optimal efficacy in precise drug transportation because of its potential to alter the biodistribution and pharmacokinetics of drugs (Tao et al., 2013, 2015, 2016; Li et al., 2016; Zhang et al., 2017; Dai et al., 2018). In addition, some nanomaterials, which are known as theranostic nanosystem, provide a novel approach to obtain ideal efficacy by combining diagnosis and treatment together (Tao et al., 2017a,b; Zhu et al., 2018). However, the clinical application of the nanoparticle (NP)-mediated treatment is still challenging due to the premature drug leakage and inaccurate drug release in dilute bloodstream, thereby resulting in serious systemic side-effects to normal tissues and cells (Tao et al., 2014; Tibbitt et al., 2016; Chen et al., 2017; Rosenblum et al., 2018).

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Yaoxin Lin, Sun Yat-sen University, China Yuling Xiao, Wuhan University, China

#### \*Correspondence:

Jinghua Chen chenjinghua@jiangnan.edu.cn Jingxiao Chen tomchenjx@jiangnan.edu.cn

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 01 May 2018 Accepted: 04 June 2018 Published: 17 July 2018

#### Citation:

Wu J, Yuan J, Ye B, Wu Y, Xu Z, Chen J and Chen J (2018) Dual-Responsive Core Crosslinking Glycopolymer-Drug Conjugates Nanoparticles for Precise Hepatocarcinoma Therapy. Front. Pharmacol. 9:663. doi: 10.3389/fphar.2018.00663

**109**

Over the past decades, neoplasm pathophysiology gradually reveals distinctive hallmarks of the tumor from normal tissues (Hanahan and Weinberg, 2000; Hanahan and Weinberg, 2011; Flavahan et al., 2017). To the best of our knowledge, hepatic galactose/N-acetylglucosamine receptor, also known as asialoglycoprotein receptor (ASGPR), is specifically exposed on the surface of hepatoma cells with a considerably high amount (Rigopoulou et al., 2012; Lepenies et al., 2013). Thus, ASGPR is used as an autoantigen to achieve targeted hepatopathy therapy through specific recognition of galactose (Fu et al., 2015). Moreover, tumor microenvironment, such as acidity (Dong et al., 2013; Du et al., 2015), hypoxia (Wilson and Hay, 2011; Semenza, 2017), high level of glutathione (GSH) (Dutta et al., 2017), and overexpressed enzymes (Zhu et al., 2012; Chen et al., 2015), inspires rational design of smart NPs to respond to various biochemical and physicochemical stimuli (Fleige et al., 2012). To achieve this goal, researchers introduced cleavable linkages (Ma and Tian, 2014), such as pH-sensitive bonds [e.g., boronate ester (Liu et al., 2013; Chen et al., 2016) and "Schiff " base (Chen et al., 2014)] and redox-responsive linkages [e.g., disulfide (Yu et al., 2014; Xu et al., 2017) and Se–Se bonds (Xu et al., 2013)] to construct NPs, which not only distinguishes carcinoma cells from normal cells but also regulates the drug release procedure and precisely meets the mechanisms of various agents, such as extracellular and intracellular release (Ding et al., 2017; Shi et al., 2017). Therefore, to realize a reliable and efficient NP-mediated HCC treatment, stable loading of hydrophobic agents in the blood circulation without leakage, selective transport, and release of the drug into hepatoma cells are necessary.

Herein, we developed a core cross-linking glycopolymerdrug conjugates (GPDs) NPs with unique dual-responsive characteristics to achieve selective transportation and program release of anticancer drug for HCC treatment (**Figure 1**). Inspired by the specific recognition between ASGPR and galactose (D'Souza and Devarajan, 2015), and the cluster glycoside effect (Dimick et al., 1999; Lundquist and Toone, 2002), which can effectively improve the affinity of carbohydrate ligands for their protein receptors, we employed galactose to build glycopolymer to obtain enhanced ASGPR-mediated hepatoma cellular binding and internalization. A disulfide bond was introduced to the side-chain of glycopolymer via a dynamically covalent boronate ester between galactose moieties and phenylboronic acid, which exhibits pH-regulated characteristics. Subsequently, hydrophobic model anticancer drug doxorubicin (DOX) was conjugated with the glycopolymer to form a series of amphiphilic conjugates through a self-eliminating disulfide bond (Satyam, 2008; Roy et al., 2015). This process ensures the traceless release of DOX, thereby keeping the original chemical structure and pharmacological action of DOX. Moreover, the hydrophobic core of the self-assembly GPD NPs was interiorly crosslinked through disulfide bond to stabilize the architecture and avoid drug leakage in the physiological environment. The GSH level in the cytosol of cancer cells is much higher than that in normal cells or extracellular fluid (Wang et al., 2013). Thus, by incorporating both redox-responsive and pHsensitive characteristics into the GPD NPs, DOX can be accurately and programmatically released from the NPs in the cytoplasm of hepatoma cells, which are reductive and acidic.

### MATERIALS AND METHODS

#### Materials

β-D-Galactose pentaacetate (98%) was purchased from Alfa Aesar (China) Chemical Co. Ltd. 2-Hydroxyethyl methacrylate (HEMA, 97%) was obtained from J&K Scientific Ltd. (China). Boron trifluoride diethyletherate (BF3·Et2O, 98%) was purchased from Aladdin Reagent (Shanghai) Co., Ltd. 4-Cyano-4- (phenylcarbonothioylthio)pentanoic acid (CPA, >97%), GSH, and 2,2<sup>0</sup> -Azobis(2-methylpropionitrile) (AIBN) were purchased from Sigma-Aldrich (China) Inc. 4-Mercaptophenylboronic acid (MPBA, >95%) was obtained from Energy Chemical Co. (China). DOX hydrochloride (DOX·HCl, 99%) was provided by Beijing HvsF United Chemical Materials Co., Ltd. (China) and used as received. AIBN was recrystallized from ethanol before use. All other chemical regents and solvents were purchased from Shanghai Chemical Reagent Co. (China) and of analytical reagent grade, used directly. Dulbecco's modified Eagle's medium (DMEM), Roswell Park Memorial Institute (RPMI) 1640 medium, fetal bovine serum (FBS), penicillin-streptomycin, and Lysotracker green DND-26 were purchased from Thermo Fisher Scientific (United States). 3-(4,5 dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), trypsin, and 4,6-diamidino-2-phenylindole (DAPI) staining solution were obtained from Beyotime Biotechnology Co., Ltd. (China).

#### Synthesis of 2-[(2,3,4,6-tetra-O-acetylβ-D-galactopyranosyl)oxy]ethyl Methacrylate (GalAc-EMA)

β-D-galactose pentaacetate (5.0 g, 13 mmol) and HEMA (1.3 ml, 10 mmol) were dissolved in 50 ml dichloromethane (DCM) in an ice bath under N<sup>2</sup> atmosphere, and BF3·Et2O (4.4 ml, 34 mmol) was added dropwise into the solution. After stirring at 0◦C for 2 h, the mixture was moved to room temperature continued to be stirred overnight. Afterwards, the suspension was filtrated and the filtrate was washed with DI water and saturated NaCl solution for three times, then, dried over anhydrous magnesium sulfate. The product was obtained and purified by column chromatography (silica gel, DCM/ethyl acetate, 4/1). Yield: 73% (light yellow viscous liquid). The chemical structure of GalAc-EMA was identified with <sup>1</sup>H NMR spectrum on an AVANCE III (Bruker, Germany) equipment (Supplementary Figures 1, 2).

#### Synthesis of polyGal-EMA (PGal)

GalAc-EMA (1.00 g, 2.17 mmol), CPA (24.3 mg, 0.09 mmol), and AIBN (2.8 mg, 0.02 mmol) were dissolved in 2 ml anhydrous dimethyl sulfoxide (DMSO). The mixture was degassed with a freeze-vacuum-thaw cycle for three times under N<sup>2</sup> atmosphere. Then, the reaction was stirred at 75◦C for 24 h. Afterwards,


<sup>a</sup>The ratio was calculated form the characteristic peak area integration of <sup>1</sup>H NMR spectra.

the mixture was cooled down and the product (PGalAc) was precipitated in cold diethyl ether. The product was collected by centrifugation and dried under vacuum. Yield: 92% (pink solid, Supplementary Figure 3).

PolyGal (PGal) was obtained by deprotecting the acetyl group of PGalAc. Briefly, 0.5 g of PGalAc was dissolved in 50 ml DCM, and 1 ml CH3OH solution of CH3ONa (30%) was added dropwise. After stirring at room temperature for 20 min, the solution was adjusted to neutral pH using dilute HCl (1 M). The solution was then dialyzed with DI water for 48 h, and the PGal was collected by freeze-drying. Yield: 63% (white solid, Supplementary Figure 4).

# Synthesis of Glycopolymer-DOX Conjugates (GPDs)

PolyGal (45 mg, 0.1 mmol) was dissolved in 5 ml anhydrous DMSO and the solution was degassed with N<sup>2</sup> for 30 min. Then, 5 ml degassed DMSO solution of MPBA and disulfide-activated DOX (DOX-ss-Py), the synthesis route and <sup>1</sup>H NMR spectrum (shown in Supplementary Figures 5, 6) was added dropwise. The molar ratio of PGal, MPBA, and DOX-ss-Py was adjusted to obtain a series of GPDs with different hydrophilic/hydrophobic balances (**Table 1**). The mixture was stirred at room temperature for 12 h and poured into an ammonium bicarbonate solution (pH 8.5). After dialyzing in the dialysis tube (molecular weight cutoff, MWCO: 3.5 kDa) with ammonium bicarbonate solution for 24 h, the product was collected by freeze-drying, and the degrees of substitution (DS) of the pendant group were evaluated from <sup>1</sup>H NMR spectra (Supplementary Figures 7, 8). In addition, DOX was directly connected with PGal through succinic anhydride to form a covalent linked PGal-DOX conjugate (GDC) as the control. The drug loading (DL) amount of GPDs was determined using UV–vis spectrophotometer on a UV-2550 (Shimadzu, Japan) equipment at 480 nm in DMF, and the DL values were

calculated as follows: DL (%) = (mass of DOX) × 100/(mass of GPDs).

# Preparation and Characterizations of GPD NPs

Glycopolymer-drug conjugates were dissolved in phosphate buffer solution (PBS, pH 7.4) at different concentrations and incubated at 37◦C for 1 h to allow the self-assembly. The concentration of GPDs was optimized to be 0.5 mg/ml under the evaluation of average size and size distribution of the formed GPD NPs by dynamic light scattering (DLS) technique using a Zetasizer Nano ZS (Malvern, United Kingdom) apparatus at 37◦C. The stability of GPD NPs was estimated by recording the variation of size distribution at different times in the PBS (pH 7.4 or 5.5) with or without GSH (10 mM) at 37◦C. The morphology of the GPD NPs was observed by transmission electron microscopy (TEM) on a JEM-2100 (JEOL, Japan) instrument with an acceleration voltage of 200 kV. The samples were prepared by dripping a drop of solution onto a copper grid, dried naturally and then followed by negatively staining with phosphotungstic acid solution (0.2%, w/v), and dried in the air.

#### In Vitro Drug Release Assay

The solution of GPD3 NPs (1 mg/ml, 2 ml) was put into dialysis tubes (MWCO: 3.5 kDa) and immersed into 10 ml of PBS (pH 7.4 or 5.5, with or without 10 mM of GSH) at 37◦C, respectively, to simulate the drug release behavior in different physiological conditions. After each sampling at the assigned time intervals, the buffer was replaced with the corresponding fresh medium. The amount of released drug in the medium was determined by UV–vis spectrophotometer at 480 nm. The cumulative release ratio of drug (%) = (mass of released drug) × 100/(mass of total drug). Each value was averaged from three independent trials. The GDC NPs and DOX non-covalently loaded GDC (DOX@GDC) NPs were used as the control.

# Cell Culture

Human hepatocyte carcinoma cell line (HepG2 cells) and transformed African green monkey SV40-transformed kidney fibroblast cell line (COS7 cells) were incubated in DMEM complete medium, and human gastric adenocarcinoma cell line (MGC-803 cells) was incubated in RPMI 1640 complete medium at 37◦C in a humidified atmosphere containing 5% CO2. The medium contains 10% FBS and 1% (penicillin–streptomycin, 100 U/ml).

# Cytotoxicity Assay in Vitro

The cytotoxicity of GPD3 NPs was performed against HepG2, MGC-803, and COS7 cells by MTT assay. Briefly, cells were seeded in 96-well plates at a density of 5000 cells per well with 100 µl of complete culture medium. After cells were cultured to the logarithmic phase, the solutions of GPD3 NPs with various concentrations were added into the well. The cells were cultivated for 48 h, and then, the culture medium was replaced with 100 µl of MTT solution (0.5 mg/ml in PBS) and incubated at 37◦C for 4 h. The medium was removed and 150 µl of DMSO was added to each well for dissolving the formazan. The optical density (OD) was measured at 570 nm using Multiskan MK3 microplate reader (Thermo, United States). The relative cell viability was calculated as follows: Cell viability (%) = ODsample × 100/ODcontrol, each date was obtained from the average value of three independent trials. The cytotoxicity induced by DOX·HCl was measured as the positive control, using 5% DMSO as the co-solvent.

#### Confocal Laser Scanning Microscope (CLSM) Observation and Flow Cytometry Analysis

HepG2, MGC-803, and COS7 cells were seeded in confocal dishes at 1 × 10<sup>5</sup> cells per well, respectively. Then, each type of cells was incubated with GPD3 NPs (equivalent to 10 mg/l free DOX) for 2 and 4 h. The cells were carefully washed with PBS three times and fixed with 4% paraformaldehyde for 15 min. Subsequently, the cells were stained with DAPI (0.2 µg/ml) for 30 min and washed with PBS three times. Afterwards, the cells were viewed under a TCS SP8 confocal laser scanning microscope (CLSM, Leica, Germany).

For the quantitative analysis, three types of cells were seeded in six-well plates and incubated for 24 h, and the medium was replaced with the medium solution of GPD3 NPs (equivalent to 5 mg/l free DOX). After incubating for 1, 2, and 4 h, the cells were washed with PBS (pH 7.4) carefully for three times and collected by trypsinization. Then, the cells were resuspended in 0.5 ml of PBS (pH 7.4) and quantitatively analyzed by flow cytometry, using FACSCalibur flow cytometer (BD Biosciences, United States). The number of cells collected was 20,000, and each experiment was performed with three independent trials. To identify the ASGPRmediated internalization of GPD3 NPs, HepG2 cells were preincubated with 0.5 ml of galactose solution (2 mg/ml) or PGal (the monomer units amount was equivalent to 2 mg/ml of galactose). After 2 h incubation, the GPD3 NPs solution (equivalent to 5 mg/l free DOX) was directly added into the plate. The cells were continuously incubated for 2 or 4 h. Then, the plate was carefully washed with PBS (pH 7.4) and the mean fluorescent intensity (MFI) was measured by flow cytometer.

#### Intracellular Distribution Assay

HepG2 cells (4 × 10<sup>4</sup> cell/well) were seeded in dishes and incubated in RPMI 1640 (1 ml) containing 10% FBS for 24 h. Then, GPD3 NPs (equivalent to 5 mg/l DOX) dispersed in the culture medium were added and the cells were incubated at 37◦C for 2, 4, and 8 h. After removing the medium and washing with PBS (pH 7.4) three times, the cells were fixed with 4% paraformaldehyde. Afterwards, the cells were successively stained with 0.5 ml of LysoTracker-Green DND-26 (50 nM) for 30 min and 0.5 ml of DAPI Solution (0.2 µg/ml) for 15 min. Then, the cells were carefully washed with PBS and observed by CLSM. Free DOX was employed as the positive control and incubated for 4 h.

# RESULTS AND DISCUSSION

fphar-09-00663 July 13, 2018 Time: 16:20 # 5

### Preparation and Characterizations of GPDs

First, GalAc-EMA and DOX-ss-Py were successfully synthesized (Kularatne et al., 2010; Kumar et al., 2015), and their chemical structures were identified with <sup>1</sup>H NMR spectra (see details in the Supporting Information). Then, we employed a classical reversible addition-fragmentation chain transfer (RAFT) polymerization method and used GalAc-EMA as the monomer to synthesize a galactose-based glycopolymer, PGal. Subsequently, thiol group was introduced into the side-chain of PGal through a dynamically covalent interaction between a diol group of galactose units and a phenylboronic acid group of MPBA in a mildly alkaline condition (Deshayes et al., 2013). Afterward, DOX was connected with PGal through a designed disulfide bond, which is a redox-responsive linkage with self-eliminating characteristic (Santra et al., 2011; Li et al., 2015). A series of GPDs were received by adjusting the molar ratio of three units, that is, Gal-EMA, MPBA, and DOX. Moreover, the amount of MPBA was higher than that of DOX moiety in our design to set aside a portion of thiol group for disulfide cross-linking. As shown in **Table 1**, the molar ratios of three units of GPDs calculated from their characteristic peak area integrations of <sup>1</sup>H NMR spectra (Supplementary Figure 8) exhibited close values in comparison with their feeding ratios. These results indicated that the components of GPDs were easy to be tuned by varying the feeding ratio of the three units. In addition, GPC measurements indicated that the molecular weight of four GPDs were close to 10 kDa with narrow distributions, where their polydispersity values were lower than 1.3. These results showed a controllable nature of RAFT polymerization. The DL values of four GPDs measured using the UV–vis spectroscopy were 8.9, 10.2, 12.4, and 16.5%. These values were consistent with the results calculated from the <sup>1</sup>H NMR. Among the four GPD samples, GPD3 and GPD4 possessed higher content of DOX than the other two samples. Thus, their molecular weights were relatively higher than that of GPD1 and GPD2. Theoretically, these GPDs exhibited amphiphilic characteristic because galactose is hydrophilic, while DOX moieties show hydrophobic nature. These results demonstrated the hydrophilic/hydrophobic balance of the designed GPDs can be adjusted by varying the molar ratio of three units, which regulated the self-assembly behavior of GPDs to some extent.

# Morphology of GPD NPs

After synthesis, we separately dissolved four GPDs in PBS (pH 7.4) to allow the self-assembly of GPDs to form various NPs. As shown in **Figure 2**, all four GPDs formed homogeneous spherical NPs in an aqueous solution. The mean diameter of four GPDs estimated from the TEM images were 176.3 ± 10.5, 107.3 ± 6.6, 98.6 ± 7.9, and 21.4 ± 3.9 nm (**Table 1**). As discussed above, the self-assembly of GPDs was regulated by varying their hydrophilic/hydrophobic balance depending on the proportions of MPBA and DOX moieties in the GPDs. The average size of GPD NPs showed decreasing trend with increasing amount of

MPBA and DOX. Among the four samples, GPD3 and GPD4 have higher DOX contents and stronger hydrophobicity, thereby leading these two GPDs to form more uniform NPs than the other samples. This phenomenon is attributed to the fact that hydrophobic moieties, that is, MPBA and DOX, gathered at the core region and drive self-assembly to form NPs. Thus, the increasing hydrophobicity of GPDs increased the kernel density, reduced the critical aggregation concentration, and stabilized the architecture of NPs, thereby resulting in the decrease in particle size. Additionally, the average sizes of four GPD NPs measured from DLS were 279.7, 134.1, 113.8, and 59.2 nm (Supplementary Figure 9), thereby showing comparable results with those of TEM observation. DLS measured average size was a litter higher than the estimated average size from the TEM images. However, the difference was reduced with the increasing amount of MPBA. These results were attributed to the thiol group on MPBA charge of the core cross-linking. Thus, the increasing amount of MPBA improved the stability of core region of GPD NPs, thereby resulting in the decreasing difference of the size between hydration state for DLS and dry state in TEM observation. In addition, the ζ-potential value of four GPD NPs exhibited that their surface charges were nearly neutral, thereby indicating that hydrophilic galactose moieties coated on the surface of these NPs. Considering the DOX loading amount and the average size in aqueous medium, we chose GPD3 as the sample for the next study.

#### Stability and Environmental Sensitivity of GPD NPs

The stability and sensitivity of the self-assembly GPD NPs, which are relative to the long-term circulation and programmed drug transportation, are important for the clinical trial of polymerdrug conjugates. Thus, we evaluated the stability of GPD3 NPs in different milieus to simulate the process in various physiological environments. As shown in **Figure 3**, the TEM images illustrated that GPD3 NPs kept their morphologies GSH (10 mM) for 6 h.

fphar-09-00663 July 13, 2018 Time: 16:20 # 6

in PBS (pH 7.4, **Figure 3a**) but gradually disaggregated in reductive environment (**Figure 3b**), acidic milieu (**Figure 3c**), or the condition with both reductive and acidic characteristics (**Figure 3d**). NPs were increased in size from approximately 100 nm in PBS (pH 7.4) to approximately 200 nm in a reductive solution and continuously increased to 500 nm or more in acidic mediums and acidic buffers containing 10 mM GSH. These results demonstrated that GPD3 NPs exhibited different disaggregation rates and responsiveness levels with a variation of environments. We further employed DLS to detect the disaggregation process of GPD3 NPs in these aforementioned environments, as shown in **Figure 4**. GPD3 NPs exhibited a stable narrow size distribution in PBS (pH 7.4) after 24 h (**Figure 4A**), and this situation lasted for 2 weeks during the measurement (Supplementary Figure 10), thereby showing an outstanding stability of GPD3 NPs. However, the size of GPD3 NPs varied with the alternative environments, showing comparable regular feature with the results illustrated in TEM images. In the reductive and alkalescent solutions, GPD3 NPs loosened, and their average size increased with increasing time (**Figure 4B**) because of the breakage of disulfide bond, which is responsible for core cross-linking of GPD NPs. This process was similar to that in the acidic medium (**Figure 4C**), which leads to the fracture of boronate ester. Moreover, the size was increased intensively in the environment possessing both reducibility and acidity (**Figure 4D**) than in the above-mentioned two conditions. This variation was derived from our designed dual responsive strategy, which obtained precise drug release. In the inner hydrophobic core of GPD NPs, DOX was connected to PGal through disulfide bond and boronate ester. Between the hydrophobic core and hydrophilic shell, excess disulfide bond was employed to cross-link the core region. Thus, GPD NPs showed favorable stability in PBS (pH 7.4), which simulates normal physiological environment. Given that disulfide bond shows responsivity to reductive condition (Yu et al., 2014; Xu et al., 2017), and boronate ester is sensitive to low pH (Chen et al., 2016), GPD NPs became far more unstable in the environment with both reductive and acidic conditions, corresponding to the cytoplasmic and endosomal environment of cancer cells. Therefore, the designed GPD NPs may exhibit a rapid disaggregation in the simulated environment in the cancer cells, which are reductive and acidic, thereby showing programmed drug release feature.

#### In Vitro Drug Release Assay

To evaluate the stimuli-responsive drug release of GPD3 NPs, we investigated the DOX release in four different conditions to simulate physiologically biochemical milieus. As shown in **Figure 5A**, GPD3 NPs exhibited different release characteristics in four different conditions. In normal physiological and lowpH conditions, the DOX release rate was lower than 20% for over 200 h. However, the DOX release rate was rapidly increased to more than 70% in the medium containing GSH to induce reductive environment. In particular, the release rate exhibited nearly an idea zero-order release pattern for the first 100 h. Meanwhile, the total release rate was further increased beyond 80% in the acidic solution containing GSH. However, the GDC NP control group showed insignificant difference among the four different conditions (**Figure 5B**). The total release amount of DOX was lower than 20% because the DOX was connected with PGal through covalent bonds that is, ester and amide bonds, and it showed insensitivity to GSH and low pH. Nevertheless, the DOX in the GDC NP group showed not only insensitive release feature but also fast release in all four conditions (**Figure 5C**). Considering that the DOX was non-covalently loaded at the hydrophobic core region of GDC NPs, the release of DOX was dependent on the free diffusion of DOX molecules from the inner core to the outer region, thereby leading to undesired release at pH 7.4. Moreover, the release rate of DOX was higher in acidic conditions than that at pH 7.4. This result is due to the fact that the hydrophobic DOX loaded at the core region becomes hydrophilic in low pH, thereby improving the release rate. These results demonstrated that GPD3 NPs possessed favorable responsive drug release characteristics, which was highly relative to the presence of GSH because the core region is cross-linked by disulfide bond. Even in an acidic medium, the release of DOX was limited by the shield of disulfide bond cross-linked network. Given that the endosomal environment is acidic, and the cytoplasm of cancer cells has a higher level of GSH (approximately 20-fold than that in normal cells), thereby exhibiting a reductive environment. These results indicated that GPD3 NPs may exhibit a specifically programmed release during the transportation in cancer cells.

#### In Vitro Cytotoxicity Assay

To evaluate the cytotoxicity of GPD3 NPs, we employed COS7, HepG2, and MGC-803 cells in the investigation, as shown in **Figure 6**. The IC50 values of free DOX against three types of cells were all lower than 0.25 mg/l (**Figure 6A**). We found that free DOX exhibited insignificant difference against three types of cells. In addition, the viability of HepG2 cells was higher than that of COS7 and MGC-803 cells. The chemotherapeutic

without, or (D) with GSH (10 mM) for different periods.

agent can barely distinguish between the target and non-target cells. This result is the main reason why chemotherapy is always accompanied with serious systemic toxicity. However, our designed GPD3 NPs showed selective toxicity against the three aforementioned cells (**Figure 6B**). The IC50 value was higher than 2 mg/l against COS7 cells, but only 0.32 mg/l against HepG2

cells, and 0.89 mg/l against MGC-803 cells. The cell viability of COS7 cells were higher than those of HepG2 and MGC-803 cancer cells. In addition, HepG2 cells exhibited the highest sensitivity to GPD3 NPs among the cells employed. Considering that PGal was non-toxic (**Figure 6C**), the cell inhibition was highly derived from the conjugated DOX. These phenomena occurred because HepG2 cell is a typical type of HCC, on which a higher amount of ASPGR is exposed. However, MGC-803 cell is a kind of gastric carcinoma cell and COS7 cell is a fibroblast-like cell. These mean ASPGR is non-expressed on both MGC-803 and COS7 cells, thereby resulting in non-recognition of the designed GPD NPs. Therefore, GPD3 NPs can efficiently and selectively deliver anticancer drug DOX to HCC cells and protect normal and non-target cells, showing potential in HCC therapy.

# Cellular Uptake

To investigate the reason for the selective inhibition of GPD3 NPs against different cells, we used CLSM to observe the internalization of GPD3 NPs to different cells, as shown in **Figure 7**. Among these cells, we found brighter red fluorescence localized in HepG2 cells than those of COS7 and MGC-803 cells. Since red fluorescent DOX moieties were connected with PGal, these results showed that a higher amount of GPD3 NPs was internalized into HepG2 cells. Additionally, the red fluorescence in COS7 and MGC-803 cells showed insignificant variation with increasing incubation time. However, the fluorescent intensity was continuously increased in HepG2 cells. This phenomenon was due to the increasing amount of internalized GPD3 NPs and aggregation-caused quenching (ACQ) of DOX, which gathered at the hydrophobic core region of GPD3 NPs. As the DOX was gradually released into the cytoplasm of cells over the incubation time, the fluorescent intensity was increased with decreasing ACQ effect. These results were also consistent with the results in the drug release assay. We further used flow cytometry to quantitatively estimate the fluorescent intensity in different cells, as shown in **Figure 8A**. For the first 1 h, the MFI in three cells showed insignificant differences. However, the MFI value in HepG2 cells was increased with time and became nearly twofold when compared with that in COS7 and MGC-803 cells after 4 h. These results indicated that GPD3 NPs can selectively transport DOX to HepG2 cells and cause efficient cell inhibition. We hypothesized that this difference was due to the specific recognition between galactose, which coated on the surface of GPD NPs, and ASGPR exposed on HepG2 cells, thereby resulting in ASGPR-mediated internalization. Generally, ASGPR facilitates binding and uptake of circulating asialoglycoproteins through the recognition of the exposed terminal galactose (Lepenies et al., 2013). Nevertheless, autoimmune hepatitis, including HCC-induced inflammation, may lead to the suppression of ASGPR binding to asialoglycoproteins by the stimulatory effects of cytokines, such as interferon-γ, interleukin-2, and tumor necrosis factor, resulting in the exposure and increasing amount of ASGPR on the cell surface (Geijtenbeek and Gringhuis, 2009; Rigopoulou et al., 2012; Kuruvilla et al., 2017). Therefore, the designed GPD3 NPs, which were coated with galactose moieties on their surface, can be specifically internalized by HepG2 cells. However, the internalization of GPD3 NPs by COS7 cells and MGC-803 cancer cells was limited because ASGPR

cells after incubation of 2 and 4 h observed by CLSM. The cell nuclei were stained with blue probe DAPI.

is not expressed on normal and gastric cancer cells, thereby showing a selective characteristic. To verify this hypothesis, we employed galactose and PGal as inhibition agents to decrease the internalization of GPD3 NPs in HepG2 cells (**Figure 8B**). We found that GPD3 NPs uptake was inhibited by both galactose and PGal. However, the internalization of free DOX was not

HepG2 cells after incubation for different periods. Cell nuclei were stained with DAPI (blue), lysosomes, and endosomes were labeled with Lysotracker Green DND-26 (green). HepG2 cells incubated with free DOX (5 mg/l) was used as the control.

influenced by these two inhibitors because they can bind with ASGPR and block the ASGPR-mediated pathway of GPD3 NPs. However, the inhibitor was unable to inhibit the free diffusion of the small molecular agent DOX. Given that the conjugated DOX was connected with GPD and encapsulated in the core of NPs, the above-mentioned results indicated that the drug was transported into the cells by the ASGPR-mediated pathway. We also found that PGal exhibited higher inhibition efficiency than that of galactose. This result was attributed to the documented cluster glycoside effect, which improved the binding ability of PGal to ASGPR (Dimick et al., 1999; Lundquist and Toone, 2002). These results not only indicated that our designed GPD NPs can be specifically internalized by HCC cells, but also revealed that glycopolymer-based GPD NPs can exhibit selective and enhanced internalization to HepG2 cells.

#### Distribution Assay of GPD NPs

fphar-09-00663 July 13, 2018 Time: 16:20 # 10

To observe the distribution of GPD3 NPs in HepG2 cells, we employed lysotracker as the fluorescent probe to label endosomes and lysosomes in the cells. As shown in **Figure 9**, we found that the intensity of red fluorescence was increased in cells for the first 2 and 4 h. However, the colocalization of red and green fluorescence indicated that GPD3 NPs were mainly distributed in endosomes and lysosomes for the first 2 h. These subcellular units are acidic; thus, NPs will enlarge and escape from the endosomes and lysosomes. Afterward, we found that the red fluorescence was distributed in cytoplasm after 4 h, which verified our hypothesis mentioned above. To the best of our knowledge, GSH is overexpressed in cytoplasm of cancer cells with both reducibility and acidity. Therefore, DOX would be rapidly released out from GPD3 NPs. Subsequently, red fluorescence was colocalized with blue fluorescence after 8 h, indicating that the drug was internalized into the cell nuclei stained with DAPI. These results demonstrated our designed GPD3 NPs exhibited a programmed drug transportation characteristic. For free DOX, the red fluorescence directly distributed in the cells and concentrated upon cell nuclei after coincubation for 4 h. This result was attributed to the fact that internalization of small molecular drug primarily depends on free diffusion, thereby showing rapid but non-selective characteristic.

#### CONCLUSION

In summary, we have designed and prepared a series of dual-responsive GPD NPs for precise HCC therapy. GPD NPs possessed an adjustable size corresponding to the DL amount, thereby varying the hydrophilic/hydrophobic balance of amphiphiles. The model drug DOX was conjugated on the galactose-functionalized glycopolymer through the use of selfeliminating disulfide bond and boronate ester as linkages, thereby showing both redox-responsive and pH-sensitive characteristics.

#### REFERENCES


Moreover, the core cross-linking strategy stabilized GPD NPs in a normal physiological environment. However, performing rapid drug release feature in the milieu resulted in both reductivity and acidity. In addition, the transportation of the drug showed a programmed drug characteristic. Simultaneously, GPD NPs can be specifically internalized into HepG2 cells through an ASPGR-mediated pathway by the recognition of galactose, which coated on the surface of GPD NPs. Thus, these GPD NPs have potential uses in precise HCC therapy.

#### AUTHOR CONTRIBUTIONS

JW developed the main study. JY and BY helped in completing the synthesis. YW and ZX took part in the cell experiments and analysis. JhC and JxC drafted the manuscript and developed the study design. All authors have given final approval for this paper to be published.

# FUNDING

This work was supported by National Natural Science Foundation of China (21574059 and 51303068), Fundamental Research Funds for the Central Universities (JUSRP51709A), National First-Class Discipline Program of Light Industry Technology and Engineering (LITE2018-20), and Top-notch Academic Programs Project of Jiangsu Higher Education Institutions (PPZY2015B146).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2018.00663/full#supplementary-material



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Wu, Yuan, Ye, Wu, Xu, Chen and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Topical Use of Quercetin-Loaded Chitosan Nanoparticles Against Ultraviolet B Radiation

Wenhao Nan1,2† , Li Ding1,2† , Houjie Chen1,2, Fahim U. Khan1,2, Lu Yu<sup>3</sup> , Xinbing Sui4,5 \* and Xiaojun Shi1,2 \*

<sup>1</sup> School of Life Sciences, Tsinghua University, Beijing, China, <sup>2</sup> Graduate School at Shenzhen, Tsinghua University, Shenzhen, China, <sup>3</sup> Shenzhen Modo Biotech Co., Ltd., Shenzhen, China, <sup>4</sup> Department of Medical Oncology, Holistic Integrative Oncology Institutes and Holistic Integrative Cancer Center of Traditional Chinese and Western Medicine, The Affiliated Hospital of Hangzhou Normal University, College of Medicine, Hangzhou Normal University, Hangzhou, China, <sup>5</sup> Department of Cancer Pharmacology, Holistic Integrative Pharmacy Institutes, College of Medicine, Hangzhou Normal University, Hangzhou, China

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Xiaolong Li, Dana–Farber Cancer Institute, United States Tianjiao Ji, Boston Children's Hospital and Harvard University, United States

#### \*Correspondence:

Xinbing Sui hzzju@zju.edu.cn Xiaojun Shi shixj@sz.tsinghua.edu.cn

†These authors have contributed equally to this work.

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 15 June 2018 Accepted: 09 July 2018 Published: 26 July 2018

#### Citation:

Nan W, Ding L, Chen H, Khan FU, Yu L, Sui X and Shi X (2018) Topical Use of Quercetin-Loaded Chitosan Nanoparticles Against Ultraviolet B Radiation. Front. Pharmacol. 9:826. doi: 10.3389/fphar.2018.00826 Ultraviolet radiation is a major risk factor for human skin damage, especially solar ultraviolet-B (UVB) which can induce inflammation, photoaging, and skin cancer. Quercetin (Qu), one of flavonoid family members, has showed protective effects against UVB radiation. However, its application for topical use is limited by low hydrophilicity and poor percutaneous absorption. Herein, we found that Qu, if entrapped into TPP-Chitosan nanoparticles (TCs), can be efficiently uptake by HaCaT cells and easily permeate through the epidermis layer, meanwhile display better stability and low cytotoxicity. We also found that Qu-loaded TCs (QTCs) could notably enhance the effect of Qu on inhibiting the NF-kB/COX-2 signaling pathway as well as ameliorating the skin edema caused by UVB radiation. Therefore, this study provided a method to get rid of Qu's low hydrophilicity, enhance its percutaneous absorption and retention in the skin, and further improve its anti-UVB effect, and demonstrated that Qu-loaded chitosan nanoparticles can be used as the therapeutic agent for topical use against UVB radiation.

Keywords: Quercetin, chitosan, nanoparticles, topical use, anti-UVB radiation, skin damaging

#### INTRODUCTION

Previous studies have reported that the ultraviolet radiation (UVR) harms the human skin, particularly in fair skin populations (Madronich and Gruijl, 1993; Gallagher and Lee, 2006; Morganroth et al., 2013). Currently, due to gradual depletion of the ozone layer caused by chemical compounds containing gaseous chlorine or bromine from industries or human in atmosphere, the solar ultraviolet-B (UVB; 280–315 nm) radiation significantly penetrates to Earth's surface resulting in skin diseases including erythema, sunburn, inflammation, photoaging, oxidative stress, DNA damage, immunosuppression, and even skin cancer (Kerr and Mcelroy, 1993; Pal et al., 2015).

One of the most important mechanisms involved in UVB radiation caused skin damage is the activation of NF-κB/cyclooxygenase-2 (COX-2) signaling pathway. COX-2 one of the enzyme plays a key role in converting arachidonic acid (AA) into prostaglandins (PGE2) which increases the vascular permeability and promotes edema (Kim et al., 2005; Liu et al., 2009; Hur et al., 2010;

Pal et al., 2015; Tang et al., 2017). In addition, high COX-2 level are frequently correlated with skin inflammation and cancer development (Pentland et al., 1999; Tang et al., 2008; Mccarty, 2012; Jiao et al., 2014). Once our skin suffers from UVB radiation, IkB-α is phosphorylated through a family of serine/threonine kinases known as IKK, and subsequently falls off from NF-κB protein complex, which drives NF-κB translocating from the cytoplasm to the nucleus. Furthermore, this kind of NF-κB translocation enhances the expression of COX-2 (Kim et al., 2005; Liu et al., 2009; Hur et al., 2010; Pal et al., 2015; Tang et al., 2017).

Many flavonoids including Quercetin (Qu) play a vital role in regulating NF-κB/COX-2 signaling pathway (Serafini et al., 2010; de Alencar Filho et al., 2016). Qu is one representative flavonoid with anti-oxidant, anti-inflammation, and anti-tumor properties (Formica and Regelson, 1995). Previous research has revealed that Qu is very effective against UVB radiation through downregulating COX-2 level in vitro (Steerenberg et al., 1997). However, in vivo application of Qu is limited by its low hydrophilicity and poor percutaneous absorption (Hung et al., 2012). Therefore, many researchers managed to seek for new dosage forms of Qu to solve this problem, including Qu-loaded liposome or nanoparticles (NPs) (Hatahet et al., 2016). Our previous study has also reported using PLGA-TPGS NPs to load Qu (Zhu et al., 2016). However, although the application of PLGA-TPGS NPs significantly enhanced the protective effect of Qu, we think it is still not ideal preparation of Qu considering the natural disadvantages of PLGA-TPGS NPs. That is, PLGA-TPGS NPs, owing negative surface charge, will no doubt impede the nano-bio interaction with skin cell cytomembrane which also negatively charged (Yang et al., 2009).

Chitosan (CS), an abundant linear cationic biopolymer, has several favorable biological characteristics such as biodegradability, non-toxicity, biocompatibility, and anti-pathogen (Enríquez et al., 2006). Most importantly, CS NPs (CSs) can increase the cellular uptake due to their positive surface charge which can enhance the interaction with negatively surface charged cytomembrane (Schipper et al., 1997; Huang et al., 2002). Another advantage of CSs is its enhancement of the skin permeation of hydrophobic drugs and promotion of the retention of drugs in the epidermis, due to its interaction with the skin surface which will change the morphology of the stratum corneum and break the close conjugation of the corneocyte layers (Tan et al., 2011). Thus, theoretically Qu, if loaded into CSs, may efficiently overcome the shortcomings of Qu for topical use to protect skin from UVB radiation damage.

In this work, we compared the protective effects of Qu and Qu-loaded TPP-Chitosan NPs (QTCs) against UVB radiation in vitro and in vivo, using HaCaT cells and C57BL/6 mice models. We found that QTCs which displayed better stability and low cytotoxicity could be uptake by HaCaT cells efficiently and easily permeate through stratum corneum and epidermis. Besides, the application of CSs carrier enhanced the effect of Qu on inhibiting the NF-κB/COX-2 signaling pathway, further ameliorating the skin edema induced by UVB radiation. In this study, we firstly showed the advantages of Qu-loaded CS NPs for topical use against UVB radiation-induced skin damage.

# MATERIALS AND METHODS

### Chemicals

Quercetin and coumarin-6 were purchased from Sigma-Aldrich (St. Louis, MO, United States). Chitosan (deacetylation degree ≥ 95%, viscosity 100–200 mpa.s, biotechnology level) and sodium tripolyphosphate (TPP) were purchased from Macklin (Shanghai, China). 4',6-Diamidino-2-phenylindole dihydrochloride (DAPI) and dimethyl sulfoxide (DMSO) were purchased from Sangon Biotech (Shanghai, China). Wheat Germ Agglutinin Alexa Fluor 594 conjugate (WGA-594) were purchased from Thermo Fisher Scientific (Eugene, OR, United States). All other chemicals and reagents of the highest quality were commercially available and used as received. Antibodies against p-IkB-α, NF-κB, COX-2, GAPDH, β-actin and lamin A/C were purchased from Cell Signaling Technology.

#### Formulation and Characterization of NPs

The preparation of TPP-Chitosan NPs (TCs) was based on ionic gelation of chitosan with TPP anions reported before (Zhang et al., 2008), but with some modifications. Under magnetic stirring at room temperature, 20 mg CS was dissolved into 10 ml 1% (V/V) acetum and adjusted the pH to 5.5, then 3.5 ml TPP with the concentration of 2 mg/ml was dropwise added into CS solution. Opalescent and transparent solution was formed after transient sonication. Qu loaded TPP-Chitosan NPs (QTCs) were formed by dropwise adding 1 ml Qu ethanol solution with the concentration of 3 mg/ml into CS solution before adding TPP. The NPs were centrifuged at 8,000 rpm, 4◦C for 30 min, the supernatant was collected for HPLC and the precipitation was suspended after washed three times by ddH2O, and finally lyophilized.

The hydrodynamic diameter and zeta potential of NPs were measured by Dynamic Light Scattering Zetasizer (Zetasizer Nano ZS90, Malvern Instruments Ltd., Malvern, United Kingdom). The morphology of NPs was showed by transmission electron microscopy (TEM, Tecnai G20, FEI Company, Hillsboro, OR, United States).

#### Measurement of Drug Encapsulation Efficiency and Drug Loading Content

The drug encapsulation efficiency (EE) and drug loading content (LC) of the QTCs were measured by HPLC (LC 1200, Agilent Technologies, Santa Clara, CA, United States). The mobile phase of HPLC was 0.2% phosphoric acid solution and MeOH with the volume ratio of 55:45. The flow rate of mobile phase was 1 ml/min. The UV detection wavelength was 360 nm. A C-18 column (150 mm × 4.6 mm, GL Science Inc., Tokyo, Japan) was used. The supernatant of QTCs after centrifugation was used for HPLC to detect the concentration of Qu in the supernatant; therefore, we can detect the amount of Qu in supernatant (W1). We named the weight of lyophilized precipitation as

W<sup>2</sup> and the amount of Qu added in the preparation of QTC as W3.

$$\text{EE} = (\text{W}\_3 - \text{W}\_1) / \text{W}\_3 \times 100\%$$

$$\text{LC} = (\text{W}\_3 - \text{W}\_1) / \text{W}\_2 \times 100\%$$

#### Cell Culture

HaCaT cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco BRL, United States) supplemented with 10% fetal bovine serum (FBS, Gibco) at 37◦C with 5% CO2.

#### Cellular Uptake of Coumarin-6-Loaded TCs and Its Efficiency

Coumarin-6 was used as the fluorescent probe which was entrapped into TCs to show the cellular uptake of QTCs. HaCaT cells were seeded in 12-well plate containing cover glasses in the bottom. After the cells adhered to the glasses, 100 µg/ml coumarin-6-loaded TCs was added to the culture medium and incubated at 37◦C for 3 h. Then removed the culture medium and washed with PBS three times. Next added WGA-594 diluted by PBS to the wells and incubated at 37◦C for 20 min to stain the cytomembrane, then washed with PBS three times and fixed with 4% paraformaldehyde for 20 min, finally, cells were stained with DAPI for 15 min and washed with PBS three times. Thereafter, cells were investigated by confocal laser scanning microscope (CLSM, Olympus Fluoview FV-1000, Tokyo, Japan).

To quantitatively measure the cellular uptake efficiency of coumarin-6-loaded TCs, 100 µg/ml coumarin-6-loaded TCs were incubated with HaCaT cells in 12-well plate at 3, 6, 12, 24 (each group, n = 3), and the quantitative ratio of coumarin-6 fluorescence in cytoplasm were measured by detecting the absorbance value at 320 nm.

#### Cytotoxicity and UVB Protective Experiment in Vitro

Cytotoxicity was evaluated by MTT assay. Cells were seeded in 96-well plate, after the cell attachment rate reached 60∼70%, replaced the culture medium to new culture medium mixed with or without different concentrations of Qu (mixed with DMSO, at the concentration of no influence on cells), QTCs, TCs and incubated for another 24 h. Then removed the culture medium, added new culture medium mixed with MTT (0.5 mg/ml) and incubated 4 h. Finally, the absorbance value of 490 nm was detected.

The method to investigate the UVB protective effects of Qu, QTCs, and TCs in vitro were almost the same as cytotoxicity experiment. The difference was the time of incubation with different concentrations of Qu, QTCs, and TCs been cut to 8 h, then removed the culture medium and irradiated using a microprocessor-controlled UV Crosslinker (XL-1000, SPECTROLINKERTM, United States) of 12 mJ/cm<sup>2</sup> . After the radiation, cells were added new drug free culture medium and returned to the incubator immediately.

#### Immunoblotting

Cell and tissue lysates were separated by 12% SDS-PAGE and analyzed by immunoblotting using P-IkB-α, NF-κB, COX-2 antibodies, followed by enhanced chemiluminescence (ECL) detection. Isolation of nuclear protein was conducted using the manufacturer's protocol (Abcam, Cambridge, MA, United States).

#### Immunocytochemical Staining of NF-κB

Cell culture and UVB radiation were the same as the UVB radiation treatment mentioned before. 10 h after UVB radiation, the cells were fixed and immunofluorescence stained with NF-κB antibody. After stained with DAPI, the cells were detected by CLSM.

#### Percutaneous Absorption and Retention Study of QTCs and Qu Aqueous Solution

All animal experiment protocols were approved by the Administrative Committee on Animal Research in the Graduate School at Shenzhen, Tsinghua University. Female C57BL/6 mice (aged 6 weeks, weighing about 20 g) were purchased from Guangdong Medical Laboratory Animal Center. After execution, the dorsal skin was carefully shaved with electric clippers and then carefully cut down the dorsal skin. Scrape the subcutaneous tissue completely and carefully by using a knife, and then fixed the skin on the V-C diffusion cell to investigate the percutaneous absorption of QTCs aqueous solution and Qu aqueous solution (each group, n = 3). Take samples regularly on 2, 4, 6, 8, 12, and 24 h for HPLC. The preparation of QTCs aqueous solution was mentioned before, and the preparation of Qu aqueous solution was by dropwise adding 1 ml Qu ethanol solution (5 mg/ml) to 9 ml ddH2O under magnetic stirring.

# Skin Permeation of QTCs and Its Mechanism

Coumarin-6-loaded TCs was applied to represent skin permeation of QTCs visually. After the mice dorsal skin was carefully shaved with electric clippers, Coumarin-6-loaded TCs aqueous solution (80 mg/20 ml) was applied every 3 h, and after 12 h the treated skin was excised for frozen section to observe the skin permeation of Coumarin-6-loaded TCs. Divided the dorsal skin shaved mice into three groups (one group was for control with no treatment) and used the same method to treat shaved dorsal skin by H2O and QTCs aqueous solution, then the treated skin was excised and the paraffin section was processed for hematoxylin and eosin (HE) staining to observe the difference of stratum corneum between different treated skins.

#### UVB Radiation on Animal, Staining, and Histopathological Analysis

After anesthesia with chloral hydrate, the dorsal skin of C57BL/6 mice was carefully shaved with electric clippers, depilated with depilatory paste (Veet <sup>R</sup> ) and divided into five groups (one group was for control), then ddH2O, 10 mg/20 ml Qu aqueous solution and 80 mg/20 ml QTCs and TCs aqueous solution were applied

to the dorsal skin every 8 h. Three days later, UVB radiation was applied at the intensity of 400 mJ/cm<sup>2</sup> . After radiation, the four kinds of aqueous solutions were applied every 8 h. After 48 h, the mice were executed. The aqueous solution treated areas were excised, cut into small pieces, part of which were prepared paraffin sections and protein extraction. The protein extraction was used for immunoblotting, and the paraffin sections were used for HE and Masson staining which were further used for histopathological analysis. The other part of treated skin pieces was used for detecting the content of PGE2. Respectively weigh 100 mg different treated skin pieces, added into 1 ml PBS (pH 7.2) after cut into small pieces with intermittent ultrasonic for 1 h and then detected the absorbance value at 278 nm, Using the absorbance value per gram of skin (A/g) to represent the content of PGE2.

# Statistical Analysis

All data are presented as the mean ± SEM of no less than three independent experiments. Comparisons were performed using a two-tailed paired Student's test (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001).

# RESULTS

### Preparation and Characterization of NPs

We prepared the QTCs using the published protocol, but with minor modification (Zhang et al., 2008). As we presented in **Figures 1A–C**, Qu can firstly adhere to CS based on similarityintermiscibility theory, and then formed nanospheres through TPP-CS crosslinking under continuously stirring condition.

Next, the morphologies of QTCs and TCs were observed. From the TEM image, QTCs and TCs were all nearly spherical in shape, but the size between each was quite different (**Figures 1D,E**). The dynamic light scattering (DLS) assay showed that the average size of TCs was about 89.48 ± 2.03 nm while QTCs was about 183.63 ± 1.52 nm in **Figure 1F**, consistent with the results of TEM. Moreover, zeta potential, an essential index in the stability of NPs in suspension through electrostatic repulsion between the NPs, had been tested. As the result presented in **Figure 1G**, both QTCs and TCs showed almost the same zeta potential, about 37 mV in average. This data implied that both QTCs and TCs were highly stable in suspension. Then the drug EE and LC of QTCs were measured by HPLC. EE and LC of QTCs were 90.98 ± 1.66% and 13.15 ± 0.77%, respectively, indicating the excellent drug carrier ability of QTCs.

# Cellular Uptake, Cytotoxicity, and UVB Protective Effect of QTCs in Vitro

Many studies have been reported that CSs could enhance cellular uptake because their positive surface charge can promote the bio-nano interaction with the cytoplasmic-membrane which has negative surface charge (Schipper et al., 1997; Huang et al., 2002), which is also the reason why CSs have been often used as gene delivery carrier (Özbas˛-Turan and Akbuga, ˘ 2011). In our study, the enhancement UVB protective effect of QTCs depends on the successful internalization and sustained retention by skin cells. Therefore, we used coumarin-6 which is a kind of fluorescent dye, instead of Qu in order to visually investigate the cellular uptake of QTCs and its efficiency. HaCaT cells, the most commonly used human skin cells, were chosen as representative cells. **Figure 2A** showed the confocal images of HaCaT cells after incubation with coumarin-6-loaded TCs for 3 h. The data suggested coumarin-6-loaded TCs can be ingested by HaCaT cells and internalized into the cytoplasm. Moreover, to investigate the cellular uptake efficiency of QTCs, we measured the quantitative ratio of coumarin-6 fluorescence in cytoplasm at indicated times after incubation of coumarin-6-loaded TCs with HaCaT cells. The result showed that HaCaT cells could almost consistently uptake coumarin-6-loaded TCs in 24 h. The cellular uptake efficiency was gradually increased and reached at about 50% at 24 h (**Figure 2B**).

Next, we evaluated the cytotoxicity of Qu and QTCs in HaCaT cells. **Figure 2C** showed the in vitro cell viability of the drug formulated in QTCs and Qu at equivalent concentrations of 5, 10, 20, 40 µg/ml, respectively. High dose free Qu showed obvious cytotoxicity compared with QTCs at the same equivalent concentration at 24 h, which indicated that QTCs can significantly reduce cytotoxicity of high dose Qu. Additionally, the TCs did not show any obvious toxicity against HaCaT cells. After verifying the safety of QTCs, we furthermore tried to investigate the protective effect of QTCs compared to Qu on HaCaT cells under UVB radiation.

We found that QTCs had same protective effects on HaCaT cells as Qu after 12 mJ/cm<sup>2</sup> UVB radiation (**Figure 2D**). At the equivalent concentration of 10 µg/ml of Qu, both QTCs and Qu showed the highest protective efficiency on UVB radiation. Meanwhile, TCs showed no significant protective effect on UVB radiation.

# QTCs Enhance the Inhibition Efficacy of Qu on the NF-κB/COX-2 Signaling Pathway in Vitro

Many pharmacological researches have revealed that Qu is very effective against UVB radiation via downregulating COX-2 level in vitro through NF-κB/COX-2 signaling pathway (13, 14). In non-stimulated cells, the NF-κB dimer binds to one of the three inhibitors (IkBα, IkBβ, and IkBε) and exists in an inactive state. Various signals activate NF-κB by degrading one or more of these three inhibitors (DiDonato et al., 1997). Therefore, the hallmark of NF-κB activation is that the activated NF-κB protein will enter the nucleus, and then binds to DNA in the nucleus to upregulate COX-2 level. In order to further evaluate the protective effect of QTCs, we observed immunofluorescent staining of NF-κB protein in HaCaT cells with UVB radiation in different treatment groups. As can be seen in **Figure 3A**, after 12 mJ/cm<sup>2</sup> UVB radiation, NF-κB protein was translocated from the cytoplasm to the nucleus. But when pre-treatment with 10 µg/ml Qu, the NF-κB protein in the nucleus reduced. And the pre-treatment 80 µg/ml QTCs (consistent drug dose with 10 µg/ml Qu) showed more reduction compared with free Qu. TCs showed no obvious influence on this process. What must be mentioned was that the nucleus size of UVB irradiated HaCaT cells was smaller compared to normal HaCaT cells, and the morphology of UVB irradiated cells also changed to be abnormal, whereas the HaCaT cells pretreated with QTCs looked nearly normal condition after UVB irradiation. Representative western blotting image of nucleus NF-κB protein level also showed the same conclusion (**Figure 3B**). Furthermore, pre-treatment with 10 µg/ml Qu or 80 µg/ml QTCs (consistent drug dose with 10 µg/ml Qu) can also inhibit the over-phosphorylation of IkB-α caused by UVB radiation (**Figure 3C**). As a result, COX-2 protein expression level stayed relatively low in Qu or QTCs pre-treatment group (**Figure 3D**). Similarly, the effect of QTCs on these proteins was much stronger than free Qu. All of these data has proved that QTCs not only just remains the Qu-like inhibition effect on the

NF-κB/COX-2 signaling pathway in vitro, but also enhances its therapeutic efficacy.

#### QTCs Enhance the Percutaneous Absorption and Retention of Qu

The efficiency of an external dosage form mostly depends on the percutaneous absorption. As we mentioned before, CSs is reported as nanomaterials with the advantages of enhancement of the skin permeation of hydrophobic drugs and promotion of the retention of drugs in the epidermis, due to its interaction with the skin surface will change the morphology of the stratum corneum and break the close conjugation of the corneocyte layers (Tan et al., 2011). Therefore, we next checked the contribution of CSs on percutaneous absorption and retention of Qu in vitro. Firstly, we conducted the percutaneous absorption experiment of QTCs and Qu and compared the efficiency of percutaneous absorption in different groups. The results shown in **Figure 4A** indicated that Qu aqueous solution was very unstable and would precipitate within 30 min because of the low hydrophilicity of Qu. And the cumulative penetration amount of Qu in aqueous solution was very small even after 24 h (**Figure 4B**). On the contrary, QTCs showed better stability in aqueous solution (**Figure 4A**) and the cumulative penetration amount of Qu in QTCs shown higher level after 24 h (**Figure 4B**). Moreover, the cumulative penetration amount of Qu in QTCs presented an upward tendency in a linear line within 12 h, as showed in **Figure 4C** the r-value was 0.9732 which approaches to 1, demonstrating that QTCs can significantly improve the percutaneous absorption efficiency of Qu in vitro. Further result

shown QTCs helped the retention of Qu in skin after 24 h (**Figure 4D**).

Then, to visually show the skin permeation of QTCs in vivo, we also used coumarin-6-loaded TCs to represent QTCs. As showed in **Figure 4E**, after application of coumarin-6-loaded TCs aqueous solution on the shaved mice skin after 12 h, we found the coumarin-6-loaded TCs permeated and stored in the skin. Furthermore, we compared the HE staining of normal, ddH2O treated and QTCs aqueous solution treated mice skin, and the results showed in **Figure 4F**. The stratum corneum of QTCs aqueous solution treated mice skin was more swelling and had larger pore space than normal and ddH2O treated skin. This result was consistent with the enlarged vision in the white box in **Figure 4E**, which also showed that the stratum corneum of coumarin-6-loaded TCs aqueous solution treated skin was swelling. These results proved that QTCs can also significantly improve the percutaneous absorption efficiency of Qu in vivo.

# In Vivo Study of the Protective Effects of QTCs on UVB-Induced Skin Damage

We further investigated the protective effects of QTCs in vivo after UVB radiation. We compared the effects of four different aqueous solution containing ddH2O, 10 mg/20 ml Qu, 80 mg/20 ml QTCs, and 80 mg/20 ml TCs in mice after UVB radiation. The immunoblotting results were shown in **Figures 5A,B**. From the results, it was obvious that QTCs had a significant enhancement effect on inhibiting NF-κB/COX-2 signaling pathway by downregulating the phosphorylation of IkB-α and the expression of COX-2 when compared with Qu. However, TCs had no notably influence on NF-κB/COX-2 signaling pathway after UVB radiation.

Moreover, PGE2 can be significantly increased after UVB radiation due to the activation of NF-kB/COX-2 signaling pathway, which can induce the increase of the vascular permeability and promotes skin edema (Kim et al., 2005; Liu et al., 2009; Hur et al., 2010; Pal et al., 2015; Tang et al., 2017). Herein, we quantificationally compared the content of PGE2 in different treated skins. Mice skin treated with QTCs showed less amount of PGE2 (**Figure 5C**). And the HE and Masson staining results in **Figure 5D**, which could help to intuitively see the alteration of mice skin after UVB radiation, also suggested the slighter disorder of collagen fiber and edema of the epidermis and dermis in mice skin treated with QTCs. The thickness of epidermis and dermis of mice also varied from different treatment. Consistent with the previous results, the status of QTCs-treated mice was most close to the normal mice skin (**Figures 5E,F**). However, TCs also had no significant difference compared with non-external preparation treatment group. With all these results, we can conclude that QTCs significantly enhanced the effect of Qu on the inhibition of the increase of PGE2 and inhibition of the edema of skin, which was closer to the normal skin, compared to Qu and TCs.

# DISCUSSION

The pH of skin surface ranges from 5.4 to 5.9 (Eberleinkönig et al., 2000; Ali and Yosipovitch, 2013). Interestingly, the surface charge of CSs is positive in this pH, which could increase the interaction between CSs and negatively charged stratum corneum.

The interaction between CSs and stratum corneum changes the morphology of the stratum corneum and breaks the close conjugation of the corneocyte layers, which means CSs could notably enhance the skin permeation of hydrophobic drugs (Tan et al., 2011). What's more, due to the large quantities of amino groups on its chains, CSs exhibits a pH-sensitive behavior, so CSs can undergo volume phase transitions from swollen to collapsed states, which can significantly change the drug release capacity (Seda Tıglı Aydın and Pulat, 2012 ˘ ). Drug-loaded CSs demonstrate sustained release profiles, which may explain the promotion of drug's retention in skin (Tan et al., 2011; Hu et al., 2014). These published data are consistent with our results. In this study, we indeed verified that Qu-loaded TPP-Chitosan NPs (QTCs) could enhance the skin permeation of Qu by changing the morphology of the stratum corneum and enhance the skin retention of Qu.

It is worth mentioning that CS has inherent antibacterial and antimycotic properties due to its polycationic nature, which can inhibit the growth of a variety of pathogenic microorganisms – Gram-positive and Gram-negative bacteria, yeasts and other fungi (Rabea et al., 2003; Ignatova et al., 2013). Furthermore, CSs could also inhibit the growth of various bacteria, including E. coli, S. choleraesuis, S. typhimurium, and S. aureus (Qi et al., 2004).

Owing to these advantages, CSs can be an ideal carrier for topical drug and gene delivery (Özbas˛-Turan and Akbuga, 2011 ˘ ; Yang et al., 2013). Our study firstly elucidates the advantage of using Qu-loaded CSs for topical use to prevent UVB radiationinduced skin damage. As displayed in **Figure 6**, our finding showed that the UVB radiation stimulates the phosphorylation of IkB-α and subsequent degradation of IkB-α-NF-κB complex, which makes the activated NF-κB translocate to the nucleus from the cytoplasm. Nucleus NF-κB then activates the expression of COX-2, the key enzyme converting AA into prostaglandins (PGE2). Excess PGE2 increases the vascular permeability and promotes edema and further damage. Qu is well-known for inhibiting NF-κB/COX-2 pathway, but its topical use is largely restricted by its natural disadvantages of low hydrophilicity

#### REFERENCES


and poor percutaneous absorption. However, in QTCs drug delivery system, TCs contributes to overcome the natural disadvantages of Qu, resulting in better hydrophilicity and low cytotoxicity of QTCs drug delivery system, and also help to significantly increase the percutaneous absorption, retention and cellular uptake of Qu in the skin, which further enhances the inhibition of the NF-κB/COX-2 signaling pathway and prevents skin from UVB radiation-induced damage.

#### CONCLUSION

Our present study firstly demonstrates that the Qu-loaded chitosan NPs have the ability to prevent the UVB radiationinduce skin damage and can be used as the promising therapeutic agent.

#### AUTHOR CONTRIBUTIONS

XjS and WN conceived the idea and designed the study. WN and LD performed all the experiments, analyzed the data, and co-wrote the paper. HC and LY provided technical support. FK helped correcting the manuscript. XjS and XbS provided reagents and conceptual advice.

# FUNDING

This work was supported by grants from National Natural Science Foundation of China (Grant No. 81730108), Key Project of Zhejiang province Ministry of Science and Technology (Grant No. 2015C03055), Key Project of Hangzhou Ministry of Science and Technology (Grant Nos. 20162013A07 and 20142013A63), and Shenzhen Science and Technology Program (Grant Nos. JCYJ20170412153453623 and JCYJ20130402145002433).


**Conflict of Interest Statement:** LY was employed by company Shenzhen Modo Biotech Co., Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Nan, Ding, Chen, Khan, Yu, Sui and Shi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-09-00826 July 24, 2018 Time: 19:0 # 11

# Controlled Drug Delivery by Polylactide Stereocomplex Micelle for Cervical Cancer Chemotherapy

Kai Niu<sup>1</sup>† , Yunming Yao<sup>2</sup>† , Ming Xiu<sup>3</sup>† , Chunjie Guo<sup>4</sup> , Yuanyuan Ge<sup>5</sup> and Jianmeng Wang<sup>5</sup> \*

<sup>1</sup> Department of Otorhinolaryngology Head and Neck Surgery, The First Hospital of Jilin University, Changchun, China, <sup>2</sup> Department of Abdominal Ultrasound, The First Hospital of Jilin University, Changchun, China, <sup>3</sup> Department of Intensive Care Unit, The First Hospital of Jilin University, Changchun, China, <sup>4</sup> Department of Radiology, The First Hospital of Jilin University, Changchun, China, <sup>5</sup> Department of Geriatrics, The First Hospital of Jilin University, Changchun, China

A stable doxorubicin (DOX)-loaded stereocomplex micelle drug delivery system was developed via the stereocomplex interaction between enantiomeric 4-armed poly(ethylene glycol)–poly(D-lactide) and poly(ethylene glycol)–poly(L-lactide) to realize control drug release and improve tumor cell uptake for efficient cervical carcinoma therapy. All these DOX-loaded micelles including poly(D-lactide)-based micelle (PDM/DOX), poly(L-lactide)-based micelle (PLM/DOX), and stereocomplex micelle (SCM/DOX) exhibited appropriate sizes of ∼100 nm for the enhanced permeability and retention (EPR) effect. In addition, compared to PDM/DOX and PLM/DOX, SCM/DOX exhibited the slowest DOX releaser, highest tumor cell uptake and the most efficient tumor cell suppression in vitro. Moreover, the excellent tumor inhibiting rates of the DOX-loaded micelles, especially SCM/DOX, were verified in the U14 cervical carcinoma mouse model. Increased tumorous apoptosis and necrosis areas were observed in the DOX-loaded micelles treatment groups, especially the SCM/DOX group. In addition, all these DOX-loaded micelles obviously alleviated the systemic toxicity of DOX. As a result, SCM can be a promising drug delivery system for the future therapy of cervical carcinoma.

Keywords: controlled drug delivery, chemotherapy, cervical carcinoma, enhanced stability, stereocomplex polylactide micelle

# INTRODUCTION

Cervical carcinoma is still one of the main causes of cancer-related deaths of female patients' worldwide (Ferlay et al., 2015). In clinic, even though the use of neoadjuvant radio and chemotherapy have decreased the incidence and mortality rates of cervical carcinoma, lots of patients have suffered from intrinsic and acquired resistance to the therapy (Walch-Rückheim et al., 2016). In addition, the substantial severe side effect, low bioavailability, and poor delivery efficiency of the chemotherapeutic agent are still the major clinical challenges for the therapy. To solve these deficiencies, lots of nanodrug delivery systems including micelles (Tan et al., 2017; Wang et al., 2018), liposomes (de Jong et al., 2007; Chuang et al., 2017), nanogels (Ding et al., 2013; Li S. et al., 2018; Zhang et al., 2018), quantum dots (Tao et al., 2017a), nanosheet (Tao et al., 2017b; Zhu et al., 2018), modified nanoparticles (Tao et al., 2014, 2015, 2016; Ding et al., 2017), and nanospheres (Wang et al., 2016; El-Boubbou et al., 2017) have been developed to achieve the spatiotemporally

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Qinghua Xu, Cornell University, United States Jinshan Guo, Harvard University, United States

#### \*Correspondence:

Jianmeng Wang jmwang1981@126.com †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 05 June 2018 Accepted: 30 July 2018 Published: 14 August 2018

#### Citation:

Niu K, Yao Y, Xiu M, Guo C, Ge Y and Wang J (2018) Controlled Drug Delivery by Polylactide Stereocomplex Micelle for Cervical Cancer Chemotherapy. Front. Pharmacol. 9:930. doi: 10.3389/fphar.2018.00930

controlled drug release in the tumor sites, increase the drug accumulation in the tumor cells and alleviate the systemic toxicity (Tao et al., 2013; Rosenblum et al., 2018).

Among these nanodrug delivery systems, micelles are generally the perfect choice due to the special core–shell structures, which can make them load a wide variety of drugs with a high drug loading capacity (Ma et al., 2015; Sun et al., 2017; Zhao et al., 2017; Li D. et al., 2018). In addition, due to the appropriate volumes, micelles are very suitable as nanodrug carriers to selectively accumulate at the tumor sites through the enhanced permeation and retention effect (EPR) (Wang et al., 2015c; Chen et al., 2017a,b; Khan et al., 2017; Xu et al., 2017). However, one remaining challenge for micelles-based delivery systems is the instability of them, which often leads to premature release of payloads during the circulation in the body. Chemical cross-linking of either the core or shell is one of the traditional strategies to improve the stability of micelles. However, the application of chemical cross-linkers may unfavorably affect the bioactivity of the loaded agents and the biodegradability of the delivery system (Chen et al., 2011). As an alternative, numerous non-covalent interactions including electrostatic interactions, host-guest, hydrogen bonding and stereocomplexation have been adopted as efficient strategies to improve the stability of micelles (Kim et al., 2008; Pounder et al., 2011; Shen et al., 2017).

Stereocomplexes can be considered as physical crosslinking, which are formed by the interaction between stereoregular chains of enantiomeric polymers (Feng et al., 2017). Poly(lactide) (PLA), as a biodegradable and biocompatible aliphatic polyester, has been verified to be a typical example used for stereocomplexation. PLA has a multitude of primary structures, such as isotactic poly(L-lactic acid) (PLLA) and poly(D-lactic acid) (PDLA) and syndiotactic and atactic/heterotactic PDLLA (Li et al., 2016). It has been reported that the equimolar mixture of PDLA/PLLA could form stereocomplexes with distinctive physical and chemical stability, such as improved mechanical properties, enhanced thermal resistance, and hydrolytic stability, due to the interactions between the L-lactyl and D-lactyl unit sequences (Ikada et al., 1987; Fukushima and Kimura, 2006). Concerning the advantages of PLA stereocomplexed materials in drug delivery, various PLA-based stereocomplexed formations have been developed for the transportation and delivery of different treatment agents. For instance, Ma et al. used sequential ring-opening polymerization to successfully fabricate poly(ethylene glycol)-b-poly(L-lactic acid)-bpoly(Dlactic acid) (PEG-b-PLLA-b-PDLA) stereoblock copolymers (Ma et al., 2015). The stereoblock copolymer micelles showed higher drug loading content (DLC), slower degradation, and drug release rate. In another report, Zhao and co-workers fabricated a biodegradable stereocomplex micelles (SCMs) based on amphiphilic dextran-block-polylactide (Dex-b-PLA) for efficient intracellular drug deliveries (Zhao et al., 2013). This doxorubicin (DOX) loaded SCMs exhibited high stability and sustained release profiles in vitro. Wang et al. previously fabricated a DOX-loaded stereocomplex micelle (SCM/DOX) via the equimolar enantiomeric 4-armed poly(ethylene glycol)– polylactide copolymers (Wang et al., 2015a). All the DOX-loaded micelles, especially the SCM/DOX displayed proper sizes for EPR, controlled DOX release, and enhanced antitumor efficacy in vitro. However, the in vivo antitumor efficacy of these DOXloaded micelles and whether they can alleviate the systemic toxicity of DOX are not verified.

In order to further confirm the in vivo antitumor efficacy and systemic toxicity of DOX-loaded micelles, especially the SCM/DOX toward cervical carcinoma, in the present study, an in vivo tumor inhibition test was evaluated on the U14 cellsbearing BALB/c mouse models (**Scheme 1**). Our results indicated that all of these DOX-loaded micelles, especially the SCM/DOX, showed satisfactory tumor suppression efficacy and a higher level of safety in comparison to free DOX·HCl. These DOX-loaded micelles, especially the SCM/DOX, can serve as an excellent nanoplatform for the chemotherapy of cervical carcinoma.

#### MATERIALS AND METHODS

#### Materials

4-Armed PEG (number-average molecular weight (Mn) = 10,000 Da) was purchased from Shanghai Seebio Biotech, Inc. (Shanghai, China). DLA and LLA were obtained from Changchun SinoBiomaterials Co., Ltd. (Changchun, China) and recrystallized from ethyl acetate under argon atmosphere before use. The stereocomplex micelle (SCM) was fabricated by the equimolar mixture of the enantiomeric 4-armed poly(ethylene glycol)–polylactide copolymers. Doxorubicin hydrochloride (DOX·HCl) was purchased from Beijing Huafeng United Technology Co., Ltd. (Beijing, China). 4<sup>0</sup> ,6-Diamidino-2-phenylindole (DAPI), Alexa Fluor 488 phalloidin (Alexa 488), and 3-(4,5-dimethylthiazol-2-y1)-2,5-diphenyltetrazolium bromide (MTT) were purchased from Sigma-Aldrich (Shanghai, China). Human cervical cancer HeLa cells and mouse cervical cancer U14 cells were purchased from the American Type Culture Collection (ATCC). Clear 6-well and 96-well tissue culture polystyrene (TCP) plates were purchased from Corning Costar Co. (Cambridge, MA, United States). The deionized water used in this study was prepared through a Milli-Q water purification equipment (Millipore Co., MA, United States).

#### Preparation of DOX-Loaded Micelles

DOX-loaded micelles were prepared through a nanoprecipitation method.(Benival and Devarajan, 2012). In briefly, DOX·HCl (20.9 mg) were dissolved in 6.0 mL of Milli-Q water, and then were slowly added into 10.0 mL of PEG–PLLA copolymer solution in N,N-dimethylformamide (DMF) (10.0 mg mL−<sup>1</sup> ). After that, 2.0 mL of phosphate-buffered saline (PBS) was dropwise added into the above solution. The final solution was continuous stirring at room temperature for 12 h and subsequently dialyzed against Milli-Q water for 12 h (molecular weight cut-off (MWCO) = 3500 Da). At last, the PLM/DOX was obtained by lyophilisation. PDM/DOX and SCM/DOX were also fabricated by the same protocol.

#### In vitro Cellular Uptake

Confocal laser scanning microscopy (CLSM) and flow cytometry (FCM) were used for quantitative analysis of cell uptake in vitro.

CLSM. 2.0 × 10<sup>5</sup> of HeLa cells were seeded on glass coverslips per well in 2.0 mL of complete high glucose Dulbecco's modified Eagle's medium (HG-DMEM) in 6-well plates for 24 h. PDM/DOX, PLM/DOX, SCM/DOX, or free DOX·HCl at a final DOX concentration of 10.0 µg mL−<sup>1</sup> was added to each well. After coincubation for 2 h, the medium was removed, and the cells on glass coverslips were washed with PBS five times and fixed with 4% (W/V) PBS-buffered paraformaldehyde for 20 min at room temperature. And then, the cells were washed for five times by PBS and reacted with 0.1% (V/V) Triton X-100 in PBS for 12 min at room temperature. And then, the nuclei were then stained with DAPI for 3 min at 37 ◦C, after which the cells were washed with PBS five times. At last, the filamentous actin was stained with Alexa 488 for 30 min at 37◦C, and washed with PBS five times. The CLSM micro-images were taken by a CLSM (LSM 780, Carl Zeiss, Jena, Germany). FCM. 2.0 × 10<sup>5</sup> of HeLa cells were seeded in each well of 6-well plates and cultured for 24 h. And then, PDM/DOX, PLM/DOX, SCM/DOX, or free DOX·HCl at a final DOX concentration of 10.0 µg mL−<sup>1</sup> was added to each well. The cells without any treatment were set as control. After a 2 h co-incubation, the medium was removed, and the cells were washed with PBS five times and then digested by trypsin. Subsequently, 1.0 mL of PBS was added and collected in centrifuge tubes. The harvested cells were centrifuged at 3000 rpm for 5 min. After removing the supernatants, the bottom cells were resuspended in 0.3 mL of PBS and examined by a flow cytometer (λex = 488 nm; Beckman, CA, United States).

# Cytotoxicity Assays

The cytotoxicities of DOX-loaded micelles and free DOX·HCl with a DOX·HCl concentration of 0.16–10.0 µg mL−<sup>1</sup> were conducted toward HeLa cells and U14 cells by an MTT assay. In brief, 8.0 × 10<sup>3</sup> cells in 180.0 µL complete HG-DMEM was planted into 96-well plates and incubated at 37◦C for 24 h. And then, 20.0 µL of PBS containing various DOX formulations were added to each well and cultured for another 48 h. Subsequently,

(D) U14 cell viability after treatment with DOX·HCl and DOX-loaded micelles for 48 h. Data are presented as mean ± SD (n = 8; <sup>∗</sup>P < 0.001).

20.0 µL of MTT at a concentration of 5.0 mg mL−<sup>1</sup> was added and incubated for another 4 h. After that, the medium was carefully removed, and 150.0 µL of dimethyl sulfoxide (DMSO) was added to each well to dissolve the MTT formazan generated by the live cells. The plates were shocked for 5 min before detection. The absorbance of medium was measured at 490 nm using a Bio-Rad 680 microplate reader. The cell viability was calculated as Equation (1).

$$\text{Cell viability (\%) } = \frac{\text{A sample}}{\text{A control}} \times 100\tag{1}$$

In Equation (1), the Asample and Acontrol represented the absorbances of sample and control wells, respectively.

#### In vivo Antitumor Efficacy Assay

Female BALB/c mice (∼4 weeks) were obtained from Vital River Laboratory Animal Center (Beijing, China). All animals were carefully treated under the guidelines approved by the Institutional Animal Care and Use Committee of Jilin University. The antitumor efficacy of DOX-loaded micelles and free DOX·HCl was evaluated using subcutaneous U14 cells bearing female BALB/c mouse models. Mice treated with NS (normal saline) were used as control. When the tumor volume reached about 100 mm<sup>3</sup> , the mice were randomly divided into 6 groups (n = 8 for each group) and treated with NS, free DOX·HCl at a DOX·HCl concentration of 3.0 mg [kg Body Weight (BW)]−<sup>1</sup> or 6.0 mg (kg BW)−<sup>1</sup> , or DOX-loaded micelles at a DOX·HCl concentration of 3.0 mg (kg BW)−<sup>1</sup> via tail vein injection every 3 days for a total of 6 doses. The groups were noted as DOX-3 and DOX-6, PDM/DOX, PLM/DOX, and SCM/DOX, respectively. The tumor sizes and body weights were monitored onceevery-other-day. Tumor volume was calculated according to the following formula: tumor volume (mm<sup>3</sup> ) = 0.5 × a × b 2 , where a and b are the largest and smallest diameter of tumor, respectively. The tumor inhibition ratio was calculated using the following formula: Tumor inhibition rate (%) = (Vcontrol−Vsample)/Vcontrol

X100, where Vcontrol and Vsample represented the tumor volumes of control and sample groups, respectively. In addition, the weights of the major organs were recorded. The organ indices of all the organs of mice were calculated according to the following formula: organ index (%) = (wcontrol−wsample) × 100.

# In vivo DOX Biodistribution of Dox-Loaded Micelles at Tumor Sites

To investigate the in vivo DOX biodistribution of DOXloaded micelles, female BALB/c mice bearing U14 cells (tumor volumes were about 200 mm<sup>3</sup> ) were randomly assigned to five groups and intravenously injected with free DOX·HCl at a DOX·HCl concentration of 3.0 mg (kg BW)−<sup>1</sup> or 6.0 mg (kg BW)−<sup>1</sup> ), or DOX-loaded micelles at a DOX·HCl concentration of 3.0 mg (kg BW)−<sup>1</sup> . At 4, 12, 24, 48, and 72 h after administration, mice were sacrificed and tumors were collected, rinsed with cold PBS. Methanol was added to each tumor to extract the content of DOX. The mixture was homogenized, centrifuged, and the supernatants were collected. The amount of DOX in each tumor was determined by HPLC. The data were normalized to the tissue weight.

# Histopathological and Biochemical Analyses of Organs

The mice were sacrificed 4 days after the last treatment. After that, tumors and major organs (i.e., heart, liver, spleen, lung, kidney, and sternum) were isolated, collected, were fixed in 4% (W/V) PBS-buffered paraformaldehyde overnight except sternum, and then embedded in paraffin. The organs from healthy mice were also isolated and treated as a normal control. The paraffin-embedded tumor and organ tissues were cut at a thickness of 5 µm, and prepared for hematoxylin and eosin (H&E) staining. The histopathological sections were assessed by a microscope (Nikon Eclipse Ti, Optical Apparatus Co., Ardmore, PA, United States). For the tumor tissues, three observation fields were evaluated to get an average value of relative necrosis area with the total area of observation field as "100%." These data were analyzed by ImageJ software (National Institutes of Health, Bethesda, MD, United States). The relative necrotic area (%) of tumor tissues was calculated by Equation (2):

Relative necrotic area (%) =

$$\frac{\text{Necrotic area in turnor section}}{\text{Total area of observed turnover section}} \times 100\tag{2}$$

For the histopathological assays of sternums, the isolated sternums were handled as described previously. Four paraffin sections of each sternum were performed for H&E staining.

The damages of tissues and organs were confirmed by testing the corresponding functional enzymes in blood and organs, which were detected with commercial enzymelinked immunosorbent assay (ELISA) kits (Shanghai Lichen Biotechnology Co., Ltd., Shanghai, China). The heart indices contained creatine kinase-MB (CK-MB), creatine kinase (CK), and lactate dehydrogenase (LDH), liver-related aspartate aminotransferase (AST) and alanine aminotransferase (ALT), and kidney-associated blood urea nitrogen (BUN) and creatinine (Cr), both in serum and organs were detected by the corresponding ELISA kits according to the standard protocols provided by the suppliers.

#### Detections of Marrow Micronucleus Cell Rates and White Blood Cell Count

The marrow micronucleus cell rate (MMCR) of each group was evaluated from H&E section. For white blood cell count, 20.0 µL of anticoagulated blood from each mouse was employed to count white blood cells (WBCs).

#### Immunohistochemical Analysis

Immunohistochemistry was carried out to detect the expression of Bax, Bcl-2, caspase-3, and survivin in ex vivo tumor tissues, which were performed as described previously (Dai et al., 2011; Dolka et al., 2016).

#### Statistical Analysis

All experiments were carried out at least three times. All data are presented as mean ± standard deviation (SD) and analyzed for statistical significance using SPSS (Version 13.0, SPSS Inc., Chicago, IL, United States). &P < 0.05 was considered statistically significant, #P < 0.01 and <sup>∗</sup>P < 0.001 were considered highly significant, respectively.

# RESULTS AND DISCUSSION

#### Preparation and Characterization of DOX-Loaded Micelles

The unsatisfactory stability of micelles is still an obvious challenge for their extensive application in controlled drug delivery. Herein,

pairs of enantiomeric copolymers of PEG–PDLA and PEG–PLLA were employed as matrices to enhance the stability of micelles via stereocomplex interaction. In aqueous solution, PEG–PDLA, PEG–PLLA and an equimolar mixture of them could selfassemble into micelles due to their amphiphilic nature. They were noted as PDM, PLM, and SCM, respectively. As depicted in **Scheme 1**, DOX was encapsulated by PDM, PLM, or SCM by nanoprecipitation. In the previous work, all of these DOXloaded micelles showed clear spherical morphologies and the average diameters of the micelles were about 100, 90, and 80 nm, respectively. The appropriate sizes of these micelles made them suitable candidates to selectively accumulate at tumor sites by the EPR effect.(Kobayashi et al., 2014; Liu et al., 2014) Moreover, all the loaded micelles exhibited excellent stability both in PBS and bovine serum albumin (BSA) solutions.(Wang et al., 2015a). In addition, critical micellization concentration (CMC) is one of the key parameters to describe the physical properties of micelles. The CMCs of PDM, PLM, and SCM were 0.063, 0.052, and 0.045 g/L, respectively (Liu et al., 2014). The CMC value of SCM was lower than that of PDM and PLM, induced by the stereocomplex interaction, which played an important role in the stability of the micelle.

### Intracellular DOX Release and Cell Viability Assays

The cell uptake and intracellular release behaviors of DOXloaded micelles were employed through CLSM and FCM. As shown in **Figure 1A**, after 2 h of incubation, the intracellular DOX fluorescences of the groups treated with DOX-loaded micelles were weaker than those cultured with free DOX·HCl. The different intracellular cell uptake behavior of DOX-loaded micelles and free DOX·HCl. was most likely due to the manner of free DOX·HCl enters cells was diffusion, which was faster than the endocytosis of DOX-loaded micelles (Li et al., 2014). Furthermore, the SCM/DOX group showed higher DOX fluorescences than PDM/DOX and PLM/DOX groups. This might be related to the slower extracellular DOX release and more efficient intracellular DOX release of SCM/DOX induced by stereocomplex interaction (Liu et al., 2014). These results were further confirmed by FCM analyses. In **Figure 1B**, the intracellular DOX fluorescence intensity of SCM/DOX group was higher than that in PDM/DOX and PLM/DOX group, while lower than free DOX·HCl. Both the CLSM and FCM verified the effective internalization of DOX-loaded micelles by HeLa cells, especially the SCM/DOX.

MTT assays in HeLa cells and U14 cells revealed that the SCM/DOX displayed significant cell killing activity after 48 h incubation (**Figures 1C,D**). It should be noted that the half maximal inhibitory concentrations (IC50) values for SCM/DOX against HeLa, and U14 cells were 0.47 and 0.50 µg mL−<sup>1</sup> , respectively, which were lower than those obtained with the PDM/DOX, PLM/DOX, and free DOX·HCl [(IC<sup>50</sup> = 0.60, 0.52, and 0.49 µg mL−<sup>1</sup> , HeLa cells) and (IC<sup>50</sup> = 0.65, 0.73, and 0.68 µg mL−<sup>1</sup> , U14 cells)]. This higher cell killing activity of SCM/DOX was attributed to the enhanced cellular uptake via stereocomplex interaction and the improved intracellular DOX release.

# In vivo Biodistribution of DOX-Loaded Micelles at Tumor Sites

The in vivo DOX biodistribution at tumor sites was explored by treating tumor-bearing mice with free DOX·HCl and DOXloaded micelles, respectively. As shown in **Figure 2**, free DOX·HCl was rapidly distributed in tumors at 4 h and rapidly eliminated at 12 h. The signal of DOX in tumors was negligible after 24 h. In contrast, DOX in the DOX-loaded micelles groups showed a much slower elimination. The concentrations of DOX in the primary tumors gradually reached to the maximum contents at 12 h post injection and was significantly increased compared to the free DOX·HCl groups (both the D0X-3 and DOX-6 groups). Until 72 h later, the DOX-loaded micelles groups still showed high DOX concentrations in tumor than free DOX·HCl groups (P < 0.01). Furthermore, compared to PDM/DOX and PLM/DOX, SCM/DOX exhibited the highest DOX concentration in tumor sites, which further verified the enhanced cellular uptake via stereocomplex interaction. This long-lasting delivery of DOX-loaded micelles was beneficial to the treatment of the tumor. These results could be due to the improved EPR effect and the decreased reticulo-endothelia system (RES) elimination by the PEG coating.

#### In vivo Antitumor Efficacy Measurement

The in vivo antitumor efficacy of DOX-loaded micelles was evaluated using a U14 subcutaneous cervical carcinoma model in BALB/c mice. As shown in **Figure 3A**, in comparison to free DOX·HCl, DOX-loaded micelles showed a stronger tumor inhibition effect, probably due to the enhanced accumulation at tumor site through EPR effect.(Wang et al., 2015a,b) Importantly, SCM/DOX group exhibited better antitumor efficacy than PDM/DOX and PLM/DOX, which might be related to the improved stability of SCM/DOX, suggesting less extracellular DOX release and greater internalized in the tumor cells. Moreover, the tumor inhibition rates of PDM/DOX, PLM/DOX, and SCM/DOX were 97.8 ± 0.40, 99.0 ± 0.35, and 99.8% ± 0.02%, which were higher than that of DOX-3 and DOX-6 groups (i.e., 83.7 ± 7.01 and 89.2 ± 3.21%; P < 0.001) (**Figure 3C**). Although free DOX·HCl also showed some tumor inhibition effect, severe body weight loss was observed during therapy (**Figure 3B**), indicating its serious systemic toxicity to mice. In contrast, the DOX-loaded micelles-treated groups did not exhibit significant body weight loss, indicating satisfactory tolerance of mice to DOX-loaded micelles. In addition, organ indices were calculated by the weight ratios between organs (mg) and the whole body (g) to provide a general impression of toxicity. As shown in **Figure 3D**, no obvious difference was observed in the heart, liver, spleen, lung, kidney, and thymus indices among all the groups, indicating the DOX-loaded micelles would not lead to severe systemic toxicities when they were used in vivo. Conversely, the tumor indices of the DOX-loaded micelles groups, especially the SCM/DOX, were much lower than that of free DOX·HCl groups. These indices were consistent with

the tumor inhibition rates. H&E staining of tumor sections was performed to investigate the fate of tumor cells after experiencing treatment in the four groups (**Figure 4A**). As shown, tumor tissues from mice in control group showed no obvious necrosis or apoptosis and the tumor cells retained their normal morphology and nuclear structure, indicating that the tumor cells in control group proliferated quickly. In contrast, tumor tissues showed various degrees of necrosis in the free DOX·HCl and DOX-loaded micelles-treated groups. Furthermore, DOX-loaded micelles-treated groups showed larger necrosis area than free DOX·HCl-treated groups. In detail, the quantitative necrosis area of PDM/DOX, PLM/DOX, and SCM/DOX-treated groups were 58.6, 70.3, and 86.2% (**Figure 4B**), respectively, which were higher than that in DOX-3-treated group (30.1%) and DOX-6 treated group (55.7%) (P < 0.05). This observation indicated the effective antitumor efficacy of DOX-loaded micelles, especially the SCM/DOX.

Immunohistochemical staining analysis was performed to further verify the antitumor efficacy of DOX-loaded micelles. As shown in **Figure 5A**, the pro-apoptotic protein Bax (brown) and caspase-3 (brown) signals of from the tumor cells that received the treatment of DOX-loaded micelles, especially the

SCM/DOX, were much higher than free DOX·HCl groups. In detail, the SCM/DOX-treated group showed 2.3, 1.7, 1.4, and 1.4 times Bax signal (**Figure 5B**) (P < 0.05) and 2.5, 2.0, 1.7, and 1.4 times caspase-3 signal than DOX-3, DOX-6, PDM/DOX, and PLM/DOX (**Figure 5D**) (P < 0.05), respectively. In contrast, the expression of antiapoptotic protein Bcl-2 (brown) decreased significantly in the DOX-loaded micellestreated groups, especially the SCM/DOX group. The Bcl-2 signal of SCM/DOX group was 0.7, 0.6, 0.6, and 0.5-fold decrease compared with DOX-3, DOX-6, PDM/DOX, and PLM/DOX (P < 0.05), respectively (**Figure 5C**). In addition, survivin, which can improve the survival of tumor cells primarily, was also used to evaluate cell survival. As shown in **Figure 5E**, the expressions of survivin (brown) decreased in the DOX-loaded micelles-treated groups, especially the SCM/DOX group. The SCM/DOX group showed a 0.7, 0.6, 0.6, and 0.4-fold decrease of survival signal compared with DOX-3, DOX-6, PDM/DOX, and PLM/DOX (P < 0.05), respectively. These data clearly demonstrated that our DOX-loaded micelles, especially SCM/DOX, could serve as highly effective nano therapeutic agents.

#### In vivo Security Evaluation

The in vivo toxicity of DOX was detected by the histopathological analysis. As shown in **Figure 6A**, the obvious accumulation of neutrophils and myocardial fiber breakage were detected in the heart of free DOX·HC-treated groups, especially the DOX-6 group, indicating the evident cardiotoxicity of free DOX·HCl. All of these damages were pointed out by black arrows in **Figure 6A**. In contrast, DOX-loaded micelles-treated groups did not show neutrophils accumulation and the myocardial cells lined in order and their sarcolemma-maintained integrity, which could be due to the decreased distribution of DOX in heart.

FIGURE 7 | Biochemical parameter assays for safety evaluation. Examinations of CK, CK-MB, LDH, ALT, AST, BUN, and Cr in serum and corresponding internal organs of normal mice or U14 cervical cancer-allografted BALB/c mice after treatment with NS, DOX·HCl, or DOX-loaded micelles. Each set of data is represented as mean ± SD (n = 3; &P < 0.05, #P < 0.01, <sup>∗</sup>P < 0.001).

Besides, free DOX·HCl-treated group showed hepatotoxicity, which was revealed in the microregional necrosis of hepatocytes. In contrast, less structural disturbance was showed in the DOXloaded micelles-treated groups. In addition, free DOX·HCltreated group also showed nephrotoxicity, as judged through the shriveled glomerular and unclear cell morphology. On the contrary, the structure of the kidney in the DOX-loaded micelles-treated groups was intact. All the data demonstrated the decreased systematic toxicity of DOX-loaded micelles, which could be due to the decreased DOX release from the micelles during blood circulation and less DOX accumulated at normal tissues and organs. No obvious pathological change was found in the spleens and lungs of the DOX-loaded micelles groups, indicating the good biocompatibility of them.

The damages of small molecular chemotherapeutic drugs to chromosomes can be reflected in the increase of MMCR.(Chen et al., 1994) As shown in **Figure 6A**, various quantities of bone marrow mononuclear cells were observed in different groups. The specific quantitative proportions of them showed that the MMCRs in DOX-3 and DOX-6 groups were ascendant, while

the MMCRs of DOX-loaded micelles-treated groups were slightly ascendant (P < 0.01) (**Figure 6B**). The results were in accordance with the histopathological analysis of tissues and organs, which further demonstrated the detoxification of DOXloaded micelles.

It is reported that the count of WBC can reflect the influence of chemotherapy on the immune status.(Homma et al., 2014) In **Figure 6C**, the WBC count of the control group significantly raised in comparison to other groups, indicating that the treatments with free DOX·HCl and DOX-loaded micelles could effectively eliminate the inflammation caused by tumor (P < 0.01). In addition, the DOX-loaded micelles showed a more efficient anti-inflammatory efficacy than free DOX·HCl (P < 0.05). All the data further verified the advantages of DOXloaded micelles in the application of anti-cervical carcinoma than free DOX·HCl.

Clinical chemical parameters, including CK, CK-MB, LDH, ALT, AST, BUN, and Cr were tested to further demonstrate the security of DOX-loaded micelles in vivo. As shown in **Figure 7**, the relevant parameters for both serum and organs were obviously raised in free DOX·HCl-treated groups, especially the DOX-6 group, which indicated that free DOX·HCl caused obvious damage to heart, liver, and kidney (P < 0.05). In contrast, all the DOX-loaded micelles-treated groups, especially the SCM/DOX group, showed negligible changes of relevant parameters, which were almost the same with normal group. The data verified that all the DOX-loaded micelles, especially the SCM/DOX could minimize the damage of DOX to the body. All of the up-mentioned results were consistent with the body weight changes and immunohistochemistry results of

#### REFERENCES


tissues and organs. All these results demonstrated that all of the DOX-loaded micelles, especially the SCM/DOX, were relatively safe and could be potentially applied in clinical studies in the future.

### CONCLUSION

In this study, DOX-loaded polylactide based micelles (PDM/DOX and PLM/DOX) and stereocomplex micelle (SCM/DOX) were fabricated. In vitro studies showed that SCM/DOX increased the cellular uptake compared to PDM/DOX and PLM/DOX and exhibited the strongest cytotoxicity against HeLa cells, which was even stronger than free DOX·HCl. Furthermore, in a mouse U14 cervical carcinoma model, SCM/DOX also exhibited the most efficient antitumor efficacy compared to either PDM/DOX, PLM/DOX or free DOX·HCl. Importantly, all the DOX-loaded micelles, especially the SCM/DOX could obviously alleviate the systemic toxicity of DOX. Therefore, the stereocomplex micelle could serve as a promising nanodrug delivery system with high systemic safety for the future cervical carcinoma therapy.

#### AUTHOR CONTRIBUTIONS

These studies were conceived of and designed by all authors. Experiments were performed by KN and YY. Data analysis, data interpretation, manuscript preparations were done by MX, CG, YG, and JW.

Cancer Immunol. Immunother. 56, 1251–1264. doi: 10.1007/s00262-006-0 276-x


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Niu, Yao, Xiu, Guo, Ge and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-09-00930 August 13, 2018 Time: 9:51 # 12

# Optimization of the Linker Length of Mannose-Cholesterol Conjugates for Enhanced mRNA Delivery to Dendritic Cells by Liposomes

Fazhan Wang<sup>1</sup>† , Wen Xiao<sup>1</sup>† , Mostafa A. Elbahnasawy<sup>2</sup>† , Xingting Bao<sup>1</sup> , Qian Zheng<sup>1</sup> , Linhui Gong<sup>1</sup> , Yang Zhou<sup>1</sup> , Shuping Yang<sup>1</sup> , Aiping Fang<sup>1</sup> , Mohamed M. S. Farag<sup>2</sup> \*, Jinhui Wu<sup>1</sup> \* and Xiangrong Song<sup>1</sup> \*

<sup>1</sup> State Key Laboratory of Biotherapy, Geriatrics and Cancer Center, West China Hospital, and Collaborative Innovation Center for Biotherapy, Sichuan University, Chengdu, China, <sup>2</sup> Department of Botany and Microbiology, Faculty of Science, Al-Azhar University, Cairo, Egypt

#### Edited by:

Chao Wang, The University of North Carolina at Chapel Hill, United States

#### Reviewed by:

Xiaoqi Sun, University of Michigan, United States Rahul K. Keswani, Exelead, Inc., United States Wenmin Yuan, University of Michigan, United States

#### \*Correspondence:

Xiangrong Song songxr@scu.edu.cn Jinhui Wu wujinhui@scu.edu.cn Mohamed M. S. Farag mohamed.farag@azhar.edu.eg †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 21 May 2018 Accepted: 09 August 2018 Published: 05 September 2018

#### Citation:

Wang F, Xiao W, Elbahnasawy MA, Bao X, Zheng Q, Gong L, Zhou Y, Yang S, Fang A, Farag MMS, Wu J and Song X (2018) Optimization of the Linker Length of Mannose-Cholesterol Conjugates for Enhanced mRNA Delivery to Dendritic Cells by Liposomes. Front. Pharmacol. 9:980. doi: 10.3389/fphar.2018.00980 Liposomes (LPs) as commonly used mRNA delivery systems remain to be rationally designed and optimized to ameliorate the antigen expression of mRNA vaccine in dendritic cells (DCs). In this study, we synthesized mannose-cholesterol conjugates (MPn-CHs) by click reaction using different PEG units (PEG100, PEG1000, and PEG2000) as linker molecules. MPn-CHs were fully characterized and subsequently used to prepare DC-targeting liposomes (MPn-LPs) by a thin-film dispersion method. MPn-LPs loaded with mRNA (MPn-LPX) were finally prepared by a simple self-assembly method. MPn-LPX displayed bigger diameter (about 135 nm) and lower zeta potential (about 40 mV) compared to MPn-LPs. The in vitro transfection experiment on DC2.4 cells demonstrated that the PEG length of mannose derivatives had significant effect on the expression of GFP-encoding mRNA. MP1000-LPX containing MP1000-CH can achieve the highest transfection efficiency (52.09 ± 4.85%), which was significantly superior to the commercial transfection reagent Lipo 3K (11.47 ± 2.31%). The optimal DC-targeting MP1000-LPX showed an average size of 132.93 ± 4.93 nm and zeta potential of 37.93 ± 2.95 mV with nearly spherical shape. Moreover, MP1000-LPX can protect mRNA against degradation in serum with high efficacy. The uptake study indicated that MP1000-LPX enhanced mRNA expression mainly through the over-expressing mannose receptor (CD206) on the surface of DCs. In conclusion, mannose modified LPs might be a potential DC-targeting delivery system for mRNA vaccine after rational design and deserve further study on the in vivo delivery profile and anti-tumor efficacy.

Keywords: mRNA vaccine, dendritic cell targeting liposomes, mannose conjugates, linker length, click reaction

# INTRODUCTION

Messenger RNA (mRNA) has recently generated great attention as one of promising therapeutics with the potential for cancer immunotherapy and vaccines because the in vitrotranscribed (IVT) mRNA does not need to enter the nucleus and induces only transient protein expression without the risk of genomic integration compared with the widely investigated DNA (Sahin et al., 2014; Pardi et al., 2018). Some mRNA vaccines have been demonstrated to be effective in the preclinical mouse models of cancer (Kreiter et al., 2015;

Oberli et al., 2017; Sayour et al., 2017). Nevertheless, the anionic character of mRNA does not facilitate its penetration into cells, resulting in low antigen expression and curative effect. It has been shown that the cellular uptake rate of naked mRNA is less than 1 in 10,000 molecules (Sahin et al., 2014). Moreover, mRNA is prone to degradation by RNases present everywhere (Tsui et al., 2002). Thus, sufficiently efficacious delivery system is urgently required to target antigen presentation cells (APCs) and protect mRNA from nuclease degradation, which will be beneficial for the clinical application of more mRNA vaccines (Pardi et al., 2018).

Non-viral vectors such as lipids, lipid-like materials, polymer or hybrid systems are widely studied for delivery of mRNA vaccines, which have low unwanted immune responses in contrast to the viral systems including adeno- associated viruses, lentiviruses and the Sendai virus (Giacca and Zacchigna, 2012; Midoux and Pichon, 2015; Hajj and Whitehead, 2017). Liposomes (LPs) are the most appealing and commonly used non-viral carriers of mRNA vaccines (Markov et al., 2015; Kranz et al., 2016; Persano et al., 2017; Verbeke et al., 2017). The mRNA loaded LPs namely RNA-LPX for cancer immunotherapy have been in phase I dose-escalation trial (Kranz et al., 2016). RNA-LPX protected mRNA from RNases and the encoded antigen can be efficiently expressed in the specialized APCs, like DCs (Kranz et al., 2016). Furthermore, the antigen-specific T-cell responses were also induced in melanoma patients. However, only 1 in 3 patients showed regression of a suspected metastatic thoracic lymph node lesion. The limited antitumor efficacy of RNA-LPX indicated that the LPs were worthy of being further reformed by functionalization of particles with ligands targeting DCs.

Dendritic cells express several mannose residue-recognizing membrane lectins like CD206 (mannose receptor, MR), CD209 (DC-SIGN) and CD207 (langerin) (Caminschi et al., 2012; Le Moignic et al., 2018). Macrophages as also expressed CD206 receptor (Chen et al., 2016; Kim et al., 2017) with the ability to present antigens (Malissen et al., 2014). They can mediate endocytosis of cargos encapsulated in mannose-modified nano-preparations (Li et al., 2013; Chen et al., 2014; Wang C. et al., 2014). Of note, enhanced in vivo anticancer efficacy via mannose modification on the nano-preparations has been widely reported in the literatures (Lai et al., 2018; Le Moignic et al., 2018; Yang et al., 2018). LPs can be easily modified because phospholipids and cholesterol are typically included (Hua and Wu, 2013). These lipophilic molecules can conjugate with various moieties binding to surface receptors of the target cells with high selectivity. Our previous studies demonstrated that folic acid-conjugating LPs can specifically deliver DNA into folate receptor-overexpressing tumor (He Z.Y. et al., 2013; Yang et al., 2016). The targeting molecule folic acid was linked to cholesterol, which efficiently kept its binding specificity to folate receptor (He Z. et al., 2013). Taking into account of these, mannosylated cholesterol derivatives were designed and synthesized to prepare mannosylated LPs to help delivery to DCs in this study.

According to literatures, the length and flexibility of the space between ligand molecules and the surface of particles might be important parameters for efficient recognition of receptors (Engel et al., 2003; Stefanick et al., 2013; Jeong et al., 2014). A short linker may restrict the translational freedom of ligand, while the longer one might bury a large fraction of the conjugated ligand (Stefanick et al., 2013). The optimal linker provides a more effective ligand-receptor interaction (Stefanick et al., 2013; Jeong et al., 2014). Thus, a rational design of the targeting LPs is crucial to enhanced mRNA delivery to DCs.

In our study, MPn-CHs containing different PEG units were firstly synthesized and then used to prepare the MPn-LPs by a typical thin-film dispersion method. The mRNA encapsulating liposomes (MPn-LPX) were constructed by complexing the obtained MPn-LPs and mRNA. The preferable MPn-LPX were picked out according to the in vitro transfection efficiency of GFP-encoding mRNA on DCs. The pharmaceutical properties and preliminary cytotoxicity of the optimal delivery system were also assessed to favor its potential application for mRNA delivery.

# MATERIALS AND METHODS

#### Materials

1,2,3,4,6-Penta-O-acetyl-alpha-D-mannopyranoside was obtained from Jinan Samuel Pharmaceutical Co., Ltd. (Shandong, China). Cholesterol was supplied from Shanghai Yuanju Biology Technology Company (Shanghai, China). Cholesterol-PEG2000-N<sup>3</sup> was purchased by Shanghai Ponsure Biotech, Inc (Shanghai, China). 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP) and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) was provided by Shanghai A.V.T. Pharmaceutical Co., Ltd. (Shanghai, China). GFP-mRNA was obtained from TriLink (San Diego, CA, United States). DMEM and fetal bovine serum (FBS) were purchased from Gibco. All the other chemical reagents were of analytical grade or better without further purification unless otherwise stated.

# Cell Culture

DC2.4 cells were cultured in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin with 5% CO<sup>2</sup> at 37◦C in a humidified atmosphere.

# Synthesis and Characterization of MPn-CH

#### Synthesis of MP100-CH

As shown in **Figure 1A**, MP100-CH was obtained with similar procedures as described previously (Kim et al., 2012; Nguyen et al., 2016). In brief, Diethylene glycol (5.0 eq), paratoluensulfonyl chloride (TosCl, 1.0 eq) and triethylamine (TEA, 1.1 eq) were dissolved in anhydrous dichloromethane (DCM) and stirred for 24 hours (h) at room temperature (rt). The crude product was purified by silica gel with a mixed solvent system of DCM and methanol to harvest compound **1**. 1,2,3,4,6-Penta-O-acetyl-alpha-D-mannopyranoside (1.5 eq), compound 1 (1.0 eq) and BF3·Et2O (1.5 eq) were dissolved in anhydrous DCM. Compound **2** was acquired and purified by column chromatography. Compound 2 (1.0 eq) and sodium

azide (5.0 eq) were added into anhydrous N, N-dimethyl-Formamide (DMF) and stirred suitably for 24 h at 60◦C to prepare compound **3**. Subsequently, deacetylation of compound 3 was performed in methanol solution (HPLC grade) of sodium methoxide (NaOMe) (10:1, v/v). Then the reaction system was treated with (H+) resin to get product named compound **4**. Compound **5** was prepared as described previously (Rull-Barrull et al., 2016). In brief, Cholesterol (1.0 eq) and 3 bromopropyne (2.0 eq) were dissolved in component solvent containing anhydrous ether and anhydrous DMF (1:1, v/v). Sodium hydride (NaH, 5.0 eq) was added mildly and the solution was stirred at rt for 24 h. Finally, compound 4, compound 5 and copper iodide were mixed in equal molar ratio and dissolved in anhydrous DMF. The mixture was reacted for 24 h at rt and concentrated to obtained compound **6** (MP100-CH) via column chromatography.

#### Synthesis of MP1000-CH

The synthetic scheme was shown in **Figure 1B** and specific experimental steps were as follows. Compound **7** was prepared and purified according to our previously reported method (He Z.Y. et al., 2010). PEG<sup>1000</sup> (5.0 eq), propargyl bromide (2.0 eq) and hydrogenated sodium (NaH, 3.0 eq) were dissolved in anhydrous tetrahydrofuran (THF). The mixture was stirred at rt overnight. Compound **8** was purified by column chromatography. Compound 7 (1.2 eq), 8 (1.0 eq), DMAP (0.5 eq), and EDCI (2.0 eq) were dissolved in DCM and stirred at rt for 24 h. Compound **9** was obtained after the crude product was purified by column chromatography. Similar to the synthetic method of compound 6, compound 4 was connected to compound 9, and compound **10** (MP1000-CH) was obtained.

#### Synthesis of MP2000-CH

Compound 14 were prepared according to the scheme showed in **Figure 1C**. Briefly, compound **11** was prepared according to the synthesis method of compound 2 but replace the compound 1 with propargyl alcohol. Subsequently, compound **12** was acquired using the similar synthesis method of compound 4. Compound **14** (MP2000-CH) was obtained with the similar synthesis method of compound 10 after column chromatography.

#### General Characterization of Prepared Compounds

<sup>1</sup>H-NMR spectra of MPn-CH and other prepared compounds dissolved in CDCl3, D2O or Dimethyl Sulfoxide-D<sup>6</sup> containing TMS were recorded on a Unity Inova-400 (400 MHz) (Varian Inc., Palo Alto, CA, United States). Chemical shifts were analyzed in ppm relative to the residual solvent peaks of TMS. The

mass spectra of various compounds were obtained using a Waters Q-TOF Premier (Milford, MA, United States) equipped with the ion spray source and using N<sup>2</sup> as nebulization gas. In addition, the identity of the conjugate was also verified by Fourier Transform infrared spectroscopy (FTIR) using a Vector 22 spectrometer (Bruker, Ettlingen, Switzerland). The purity and retention time of MPn-CHs and other cholesterol derivatives were evaluated by high performance liquid chromatography (HPLC, Waters, Milford, MA, United States) at 201 nm. The mobile phase, at 1 mL/min flow rate, was composed of 100% chromatographic methanol. The retention time and purity of cholesterol derivatives were summarized in **Table 1**.

# Preparation and Characterization of MPn-LPX

#### Preparation of MPn-LPs The cationic LPs were prepared using a thin-film dispersion method with some modifications (Wang F. et al., 2018). Briefly, cationic lipid DOTAP, the helper lipid DOPE, CH and MPn-CHs at a molar ratio of 50:10:35:5 or 50:10:40:0 (**Table 2**) were dissolved in a mixture solvent of chloroform/ethanol (1:1, v/v) to prepare MPn-LPs and LPs, respectively. The organic solvents were evaporated using a rotary evaporator at 37◦C for 2 h. The lipid film was rehydrated with 2 mL RNase-free water at 60◦C for 40 min to obtain a suspension with the final lipid concentration of 6 mM. Subsequently, the above suspension was sonicated at 80 W for 3 min and filtered with a 0.22 µm sterilized filter for the following experiment. The coumarin-6 (Cou-6) loaded LPs were acquired using the similar procedure with the addition of Cou-6 into the chloroform/ethanol (1:1 v/v) solvent mixture. The fluorescent intensity of Cou-6 loaded particles was measured using CyrationTM3 (BioTek Instruments, Inc, United States).

#### Preparation of MPn-LPX

In our study, MPn-LPX was composed of DOTAP contained LPs and mRNA at N/P ratio of 3, 5 or 7, named MPn-LPX NP 3, MPn-LPX NP 5 or MPn-LPX NP 7, respectively. MPn-LPX were prepared according to previous reported methods with some modifications (Kranz et al., 2016). Briefly, mRNA was diluted by water and 1.5 M NaCl followed by adding corresponding MPn-LPs diluted with water to reach the desired ratio of N/P with the final concentration of NaCl of 150 mM. After incubated at rt for 15 min, MPn-LPX were finally obtained.

#### Size and Zeta Potential Measurements

The average particle size, size distribution (polydispersity index, PDI) and zeta potential of different formulations were recorded by Zetasizer Nano ZS90 (Malvern Instruments, Malvern, United Kingdom). All measurements were carried out using diluted samples at 25◦C and were conducted in triplicate.

# Cellular Transfection of MPn-LPX

To optimize the appropriate ratio of N/P, DC2.4 cells in the logarithmic growth period were seeded in 24 well plates at 4 × 10<sup>5</sup> cells/well and incubated for 24 h at 37◦C, followed by incubation with different N/P of LPX (0.5 µg GFP-mRNA per well) in triplicate from 3 to 7. Before transfection, the culture medium was replaced with 500 µL FBS-free DMEM. Subsequently, LPX were added. After 4 h of incubation, 500 µL complete medium was added, and the cells were incubated for another 20 h. Expression of GFP by DC2.4 cells was visualized using a fluorescence microscope (Olympus Corp., Tokyo, Japan) and transfection efficiency was obtained based on the percentage of GFP positive cells from the live cell population by flow cytometry. Additionally, mean GFP fluorescence intensity of individual cells from GFP positive cells population after transfection was measured using FlowJo software (Li et al., 2017b).

To further investigate the transfection efficiency of MPn-LPX, DC2.4 cells in 24 well plates were incubated with MPn-LPX with the N/P of 5 following the same procedure described above. Transfection efficiency and mean fluorescence intensity (MFI) of GFP positive DC2.4 cells were evaluated by flow cytometry. In brief, DC2.4 cells were captured via forward scatter (FSC) and side scatter (SSC). Live DC2.4 cells were gated as shown in Region 1 (R1), of which GFP positive cells were selected (R2). Transfection efficiency (% GFP<sup>+</sup> cells) was auto displayed with R2. MFI of GFP expression in GFP positive cells was acquired using FlowJo software. MFI was calculated after subtraction of background values of untreated DC2.4 cells. To further elaborate the kinetics of mRNA transfection in vitro, transfection efficiency of MP1000-LPX NP 5 on DC2.4 cells from 12 to 72 h has also been studied.

TABLE 1 | The retention time and purity of cholesterol derivatives was evaluated by HPLC.


Bold values represents the retention time and purity of each compound.

TABLE 2 | Formulation component and relative molar content used in the manuscript.


#### Characterization of Optimal MP1000-LPX Microscopy Investigation

The appearance and Tyndall effect of MP1000-LX were recorded by a digital camera. The morphology of MP1000-LPX NP 5 was examined by transmission electron microscopy (TEM, H-600, Hitachi, Japan). Briefly, 100 µL of MP1000-LPX suspension was added onto copper electron microscopy grids. Subsequently, they were negatively stained with 2% phosphotungstic acid for observation.

#### Gel Electrophoresis Retardation Assay

To evaluate the complexation of mRNA and MP1000-LPs, 1 µg free mRNA and MP1000-LPX (containing 1 µg mRNA) were diluted with RNase-free water. Then NorthernMax <sup>R</sup> formaldehyde load dye containing ethidium bromide (50 µg/mL) was added and mixed. After incubating the samples for 10 min in 65◦C, the samples were loaded into a 1% denaturing formaldehyde agarose gel in precooled MOPS buffer. The gel was run for 20 min at 180 V and analyzed using a molecular imager, ChemiDocTM 219 XRS system (Bio-Rad, United States). RNA MillenniumTM markers (Ambion) with bands at a range of 0.5–9 kb was included to provide size determination of the mRNA.

#### Stability Assay

For storage stability experiments, prepared MPn-LPX NP 5 were stored at 4◦C for 1 and 3 days and another 1 h at rt before particle size and transfection efficiency measurement following the similar procedure described previously (Kranz et al., 2016).

To evaluate the serum stability, 1 µg free mRNA and MPn-LPX NP 5 (containing 1 µg mRNA) were incubated in parallel with 150 mM NaCl supplemented with or without FBS at 50% final concentration at 37◦C for 2 h, respectively. To release mRNA from LPX, 1 µL of 10% Triton X-100 was added to 10 µL of MP1000-LPX samples and incubated at rt for 10 min. After mixed with NorthernMax <sup>R</sup> formaldehyde load dye, samples were treated and visualized using the similar process as described in Section "Gel Electrophoresis Retardation Assay."

#### Cytotoxicity Assay

To test potential cytotoxicity, DC2.4 cells were treated with MP1000-LPX according to the transfection procedure. Cell viability was investigated using an Apoptosis Detection Kit according to the manufacturer's protocol by flow cytometry. Three independent cytotoxicity assays were performed in duplicate.

# Cellular Uptake of MP-LPX

Cellular uptake study was performed using Cou-6 as previously reported (Xu et al., 2016). DC2.4 cells in the logarithmic growth period were collected and seeded at a density of 8 × 10<sup>5</sup> cells/well in a 24-well plate and incubated for 24 h at 37◦C. To screen the appropriate incubation time, DC2.4 cells were treated with LPX (Cou-6, 5 ng/mL) in triplicate by different time from 0.5 h to 6 h. At the end of the study, the cells were collected and washed three times with cold phosphate buffer saline. The MFI of cells was quantified by BD FACS. To further screen the uptake concentration of Cou-6, DC2.4 cells were treated in triplicate by different concentrations of Cou-6 from 2.5 µg/mL to 20 µg/mL for 2 h.

To investigate the MR mediate uptake of MP1000-LPX (Cou-6, 10 ng/mL), DC2.4 cells were pre-treated with or without 0.16 mol/L of mannose solution for 30 min followed by incubation with LPX and MP1000-LPX at 37◦C, respectively. For binding assays, DC2.4 cells were incubated at 4◦C for 30 min. Subsequently, LPX and MP1000-LPX were added and incubated for 2 h at 4◦C.

### Statistical Analysis

The data were presented as mean ± SEM unless otherwise noted. Statistical analysis was performed using Graphpad Prism 5.0. Data of two or multiple groups were analyzed using Student's t-test or non-parametric one-way ANOVA, respectively. The p-values < 0.05 were considered statistically significant.

# RESULTS

# Characterization of MPn-CH

#### Characterization of MP100-CH

We construct MP100-CH (compound **6**) according to reasonable design as shown in **Figure 1A**. The structure of compounds **1**, **2**, **3**, **4,** and **5** were validated in **Supplementary Figures S1**–**S5**, respectively. As shown in **Figure 2A**, the <sup>1</sup>H NMR spectra of compounds **4, 5** and **6** were recorded. The single peaks at δ5.33 (s) were attributed to the protons of olefinic bond (-CH2-CH = C-) in cholesterol. The single peak at δ8.01 (s) came from the protons of olefinic bond (N-CH = C-N) in coupled places. The peaks at δ3.15-3.88 (m) were attributed to the protons from the glycol unit (-O-CH2-CH2-O–CH2−) in PEG chain. These results indicated that MP100-CH has been successfully synthesized. As seen in **Figure 2B**, the mass spectrum of MP100-CH showed a peak at 740.49 (product + Na+), which was consistent with the expected molecular weight of MP100-CH. In addition, compound **6** was further confirmed by FTIR spectroscopy with of the following principal peaks: ν-OH (3700–3400 cm−<sup>1</sup> ), ν-CH<sup>3</sup> and ν-CH<sup>2</sup> (2960–2850 cm−<sup>1</sup> ), ν-CH2-O-CH2- (1210–1050 cm−<sup>1</sup> ) presence but ν-N = N-N- (2100–2270 cm−<sup>1</sup> ) attributed to compound 4, ν-CH (about 3300 cm−<sup>1</sup> ) attributed to compound 5 absence in compounds **6** (**Figure 2C**). Finally, as shown in **Figure 2D**, the spectrum of HPLC exhibited a characteristics absorption peak of MP100-CH at 9.743 min (201 nm). According to the method of area normalization, the purity of MP100-CH is 97.948%.

(D) High purity (97.984%) of MP100-CH was confirmed using HPLC. (E) Comparison of the <sup>1</sup>H NMR spectra of compounds 4, 9, and 10 in CDCl3. (F) Mass spectrum of compounds 10 was shown as (M+Na)+, m/z: 1484.8359 to 1830.8468. (G) FTIR spectrum of compounds 4, 9, and 10. (H) HPLC spectra was used to evaluate the purity of MP1000-CH (95.621%). <sup>1</sup>H NMR spectra (I) and FTIR spectrum (K) of compounds 12, 13, and 14. (J) Mass spectrum of MP2000-CH was shown as (M/2+Na)+, m/z. (L) The purity of MP2000-CH was evaluated (91.532%).

#### Characterization of MP1000-CH

To construct MP1000-CH (compound **10**), compound 4 was jointed to compound 9 via click reaction as shown in **Figure 1B**. The structure of compounds **8** and **9** were firstly confirmed in **Supplementary Figures S6**, **S7**, respectively. The <sup>1</sup>H NMR spectra of compounds **4, 9** and **10** were recorded in CDCl3. The principal proton peaks at δ2.42-2.45 (t) attributed to the protons of alkynyl group (-CH) in compounds **9** were disappeared in compounds **10** as in shown **Figure 2E**. Similar to MP100-Chol (presence of single peak at δ8.01), the successful synthesis of MP1000-Chol has also been validated. The mass spectrum of MP1000-Chol showed broad peaks from 1484.8359 to 1830.8468 (**Figure 2F**) while that of compound 9 were from 1185.8085 to 1480.0358 (**Supplementary Figure S8**). The increased molecular

weight coincided with the molecular weight of compounds **4**, which also confirmed the structure of compound 10 referring to previously report (Li et al., 2014). The FIIR spectroscopy of compounds 10 (**Figure 2G**) was similar to compounds 4 but the presence of unique ν-CH2-CO-O- of compounds 10 at around 1740 cm−<sup>1</sup> . Additionally, the characteristics absorption peak (201 nm) of compounds **10** and compounds **9** were at 13.122 min with the purity of 96.595% (**Figure 2H**) and 16.692 min with the purity of 94.881%, respectively. Consistent with expectation, the hydrophilicity of compound 10 was increased and the retention time was decreased when compared with compound 9. All of the results of <sup>1</sup>H NMR spectra, mass spectrum, FTIR spectroscopy and HPLC confirmed the successful synthesis of compounds **10**.

#### Characterization of MP2000-CH

Similar to compound **10**, compound **14** was acquired according to our designed strategies (**Figure 1C**) and authenticated via <sup>1</sup>H-NMR (**Figure 2I**), electrospray ionization mass spectrometry (ESI-MS) (**Figure 2J**), FTIR (**Figure 2K**), and HPLC (**Figure 2L**). The structure of compounds **11** and **12** were firstly confirmed in **Supplementary Figures S9**, **S10**, respectively. Of note, the measured molecular weight of compound **12** was 241.07 (product + Na+) (**Supplementary Figure S11**), which was consistent with the expected molecular weight. When compared mass spectrum of compound 14 (**Figure 2J**) with compound 13 (**Supplementary Figure S12**), nearly 200 molecular weight were increased. The retention time at 201 nm of compound 14 and compound 13 were at 6.907 min with the purity of 91.532% and 13.662 min with the purity of 91.251% (**Figure 2L**), respectively.

# Particle Size and Zeta Potential Measurement

The size and zeta potential of all formulations in this study were evaluated. There was no statistical difference between the particle size and zeta potential among different MPn-LPs or MP<sup>n</sup> LPX formulations. The particle size and zeta potential of all different MPn-LPs was about 60 nm (**Figure 3A**) and 50 mV (**Figure 3B**), respectively. In addition, the particle size and zeta potential of all different LPX was about 135 nm (**Figure 3C**) and 40 mV (**Figure 3D**), respectively. The PDI were all less than 0.3. As was shown, the size of MPn-LPX was larger and the zeta potential was lower than the corresponding MPn-LPs formulations.

# In vitro Transfection of MPn-LPX

To investigate the in vitro transfection efficacy of MPn-LPX in DC2.4 cells, the appropriate ratio of N/P of LPX was optimized firstly. GFP expression on DC2.4 cells with different treatment were observed and recorded by fluorescence microscope (**Supplementary Figures S13A–H**). As shown in **Supplementary Figure S13I**, both LPX NP 7 and LPX NP 5 achieved significant increment in transfection efficiency compared with LPX NP 3. What's more, LPX NP 5 with similar transfection efficiency to LPX NP 7 exhibited dramatically enhanced GFP fluorescence intensity (mean GFP expression level per cell) (**Supplementary Figure S13J**). By the way, calculation of MFI and GFP positive cells was shown in **Supplementary Figure S14**. Thus, MPn-LPX were prepared by setting the N/P ratio at 5:1. The GFP expression was subsequently observed by fluorescence microscope (**Figure 4A**). As shown in **Figure 4B**, MP1000-LPX induced the most GFP positive cells and the

microscope. Transfection efficiency (%GFP<sup>+</sup> cells) (B) and MFI of GFP<sup>+</sup> cells (C) were quantified. Kinetics of MP1000-LPX NP 5 transfection efficiency (D) and MFI of GFP positive cells (E) on DC2.4 cells from 12 h to 72 h. Scale bars, 200 µm. <sup>∗</sup>p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

percentage was up to 52%, which was significantly higher than any other groups (p < 0.001). However, the MFI of the GFP positive cells of MP1000-LPX NP 5 was not the best among these groups (**Figure 4C**). Taking into account of transfection efficiency and MFI, MP1000-LPX with the highest transfection efficiency and moderate MFI were selected for further study in this manuscript.

The kinetics of MP1000-LPX NP 5 transfection on DC2.4 cells was studied. As shown in **Figures 4D,E**, transfection efficiency first increased and then reached a plateau with the increase of the incubation time (from 12 to 72 h) while the MFI of GFP positive cells first increased and then decreased. In summary, the transfection efficiency achieved the maximum at 24 h and MFI was also the strongest at 24 h.

# Further Study on the Optimal MP1000-LPX

#### Morphology Examination

The appearance and morphological studies of MP1000-LPX were conducted. The colloidal solution was colorless and transparent (**Figure 5A**). Overt Tyndall effect of MP1000-LPs and MP1000-LPX colloidal solution were observed compared with water as was shown in **Figure 5B**. Representative images of size and zeta potential of MP1000-LP(X) were shown in **Figures 5C–F**, respectively. As shown in **Figure 5G**, the morphological of MP1000-LPX was observed distinct lipid membrane structure with nearly spherical in shape. Moreover, complete complexation of the mRNA with MP1000-LPs was confirmed when the N/P ratio of 3 and 5 (**Figure 5H**).

#### Stability Assessment

fphar-09-00980 September 4, 2018 Time: 9:45 # 9

The preliminary storage stability of the MP1000-LPX was determined by the size, zeta potential and transfection efficiency. The particle size and zeta potential of MP1000-LPX were determined at predetermined time of storage at 4◦C. MP1000-LPX displayed a little decrease in particle size but not zeta potential (**Figure 6A**) As shown in **Figure 6B**, the transfection efficiency of MP1000-LPX remained about 50% when stored at 4◦C for 3 days. Additionally, MP100-LPX (**Supplementary Figure S15A**) and MP2000-LPX (**Supplementary Figure S15B**) performed excellent storage stability in the preliminary test.

For serum stability, 5 µL of fresh MP1000-LPX were diluted in FBS (1:1, v/v) and incubated for 2 h at 37◦C. As shown in **Figure 6C**, the signal of naked mRNA band in serum was completely disappeared (lane 3) compared to naked mRNA alone (lane 1). MP1000-LPX did not dissociate after incubation in 50% serum (lane 5) similar to that incubation in NaCl of 150 mM (lane 4). When the free mRNA or MP1000-LPX were treated with Triton X-100, the free mRNA (lane 2) and mRNA dissociated from MP1000-LPX (lane 4) were visible in line with mRNA treated without Triton X-100 (lane 1) as shown in **Figure 6D**. Similarly, MP100-LPX and MP2000-LPX exerted good stability in the presence of serum (**Supplementary Figure S15C**). These results confirmed the adequate protection of the mRNA against degradation.

#### Cytotoxicity Assay

After incubation with indicated formulations for 24 h, cytotoxicity analysis was performed by flow cytometry. Representative figure of each condition was showed in **Figure 7A**. The percentage of living DC2.4 cells were found to be 86.7 ± 3.6 %, 86.4 ± 1.7%, and 90.1 ± 1.2% (**Figure 7B**) for control (treated with equal volume of medium), MP1000-LPX and Lipo 3K group, respectively. None significant percentage difference of living cells, early apoptosis, late apoptosis or necrosis was found among these three groups indicating that MP1000-LPX might do no harm to DC2.4 cells. Overall, MP1000-LPX NP 5 performed good safety in vitro. Considering the excellent transfection efficacy, MP1000-LPX might be one of good candidates for DC-targeting mRNA nanovaccine for in vivo application.

#### In vitro Uptake

In the preliminary uptake experiment, incubation time and concentration of Cou-6 were optimized. According to the results in **Supplementary Figures S16A,B**, 2 h and 10 ng/mL were selected for future cellular uptake experiment, respectively. To evaluate the potency of MP1000-LPX NP 5 on targeted delivery into DC2.4 cells, the fluorescent of Cou-6 internalized by cells was assayed by BD FACS. The results in **Figure 7C** indicated that intracellular uptake of MP1000-LPX was significantly higher than that of LPX at 37 and 4◦C, respectively. When pretreated the cells with mannose, no significant difference of Cou-6 in cellular uptake between MP1000-LPX and LPX was observed. Moreover, intracellular uptake of LPX and MP1000-LPX was significantly lower at 4◦C than that at 37◦C, indicating that uptake of LPs loaded with mRNA was energy dependent.

### DISCUSSION

In the presented study, we designed facile and inexpensive approach to prepare mannose-cholesterol conjugates with various linker length as synthetic ligands applied to mRNA nanovaccine. MPn-LPs were prepared by a modified thin-film dispersion method. Subsequently, the DCs targeting MPn-LPX were prepared by complexing MPn-LPs with mRNA. No significant difference in size and zeta potential was observed among MPn-LPX comprised MPn-CH with different PEG units. The effect of linker length of mannose derivatives in MPn-LPX on transfection by DC2.4 cells was investigated. Our results might provide a rational design element of mRNA vaccine.

Linker length of ligand exerted significantly effect on targeting cellular uptake (Engel et al., 2003; Stefanick et al., 2013; Jeong et al., 2014). Thus, a proper linker length of the ligands was essential for effective receptor recognize and binding. For recognition and binding to MR, linker length of mannose should consist of at least two PEG units according to previously report (Jeong et al., 2014). In our study, mannose-cholesterol conjugates with different linker length were designed and constructed by facile strategies utilizing the click reaction. In detail, mannose derivatives have been conjugated to cholesterol derivatives modified with PEG of different lengths (PEG100, PEG1000, or PEG2000). Each target product was fully characterized by <sup>1</sup>H-NMR, ESI-MS, FTIR to confirm the successful of synthesis. HPLC was used to evaluate the purity and the successful synthesis of the products as previously reported (Shariat et al., 2014). The synthetic strategies designed here offered some overt advantages over the previously reported methods for Man-C4-chol (Kawakami et al., 2000) and Man-C6-chol (Li et al., 2013), because our target products were easily to synthesis, purify and characterize. What's more, the length of PEG-linkers could be varied with desired length, resulting in many other analogous compounds with MR targeting function.

We constructed MPn-LPs with various linker lengths between cholesterol and mannose using different length of PEG linker. All MPn-LPs were constructed with the same molar ratio of mannose modified cholesterol. The MPn-LPs had similar average size and surface charge despite of the introduction of MPn-CHs with different linker lengths of PEG. We then prepared MPn-LPX and performed cellular transfection studies in DC2.4 cells with the evaluated parameters of transfection efficiency and MFI of GFP positive cells, respectively. Transfection efficiency and MFI of LPX was preliminary optimized to find optimal ratio of N/P. Our results showed no significant difference in transfection efficiency of LPX NP 5 and LPX NP 7, while both exhibited significantly higher transfection efficiency than LPX NP 3. However, MFI of the GFP positive cells of LPX NP 5

FIGURE 5 | Characterization of MP1000-LPX. Appearance (A) and Tyndall effect (B) of MP1000-LPX. Size (C) and zeta potential (D) of MP1000-LP were recorded by Zetasizer Nano ZS90, and representative images were showed. Representative images of size (E) and zeta potential (F) of MP1000-LPX. (G) TEM images of the MP1000-LPX NP 5. (H) Agarose gel electrophoresis image of mRNA maker, Free mRNA, MP1000-LPX (N/P 5) and MP1000-LPX (N/P 3).

was significantly higher compared to that of LPX NP 7. Taken the transfection efficiency and MFI of the GFP positive cells into consideration, the ratio of N/P of 5 was selected for future studies. The difference between the percentage of transfected cells and the MFI of transfected cells was consistent with previously reported results (McLenachan et al., 2013; Avci-Adali et al., 2014;

Lee et al., 2015; Li et al., 2017a). Transfection using the same transfection reagent led to similar transfection efficiency but not the MFI (**Figures 4D,E**) with the increase of the incubation time, which was consistent with previously report (Avci-Adali et al., 2014). In addition, increasing the amount of mRNA can significantly increase the average fluorescence intensity without affecting the transfection efficiency within a certain range (Avci-Adali et al., 2014). Moreover, MFI values showed the strength of the fluorescence intensity. Higher MFI values reflected a higher production of GFP by individual cell but not higher percentage of GFP positive cells.

According to literatures, upon interaction with serum, nanocarriers rapidly absorbed protein and formed a corona (Bertrand et al., 2017; Pan et al., 2017). It was the nanocarrier–corona complex, rather than the nanocarrier, that interacted with biological systems, here with a cell membrane receptor (CD206), which might partially obscure the role of target ligands (Monopoli et al., 2012). To reduce this effect, culture medium without serum was used at initial and complete culture medium with serum were added 4 h after the adding of LPX. As was known, the fate of nano-preparations in serum (mimic the in vivo environment) was very important to predict its potential therapeutic efficacy. Our preliminary test showed that the presence serum significantly affected the transfection of MP1000-LPX. We were still working on this. Hopefully, we would show the data of mRNA delivery in the presence of serum in vitro and expression of the nano-preparations in vivo in our future work.

MP1000-LPX exhibited a higher level of transfection efficiency than LPX, MP100-LPX, MP2000-LPX and positive control Lipo 3K. No size and charge variation were found in the vectors then no inference can be made on the transfection efficiency correlation with size and charge. The excellent transfection efficiency of MP1000LPX was most likely attributed to the appropriate linker length used to conjugate mannose and cholesterol. However, MP1000LPX exerted the highest transfection efficiency but moderate MFI. The inconsistency of cellular transfection efficiency and MFI observed in the field of targeted mannose modified LPX partially resulted from differences in the types of cell, incubation time, the amount of mRNA, ability of lysosome escape and types of lipid mannose modified (Kim et al., 2012; Li et al., 2013; Avci-Adali et al., 2014; Chen et al., 2015; Wang C. et al., 2015). The best effect of MP1000-LPs conjugates may be related to the above factors. However, most likely it was attributed to the linkers used to conjugate mannose and the other different component of targeting formulations (Kim et al., 2012; Wang N. et al., 2014; Wang C. et al., 2015). There are many other known and unknown factors for ligand receptor affinity beside linker length of ligand. Moreover, the intracellular metabolism of mRNA nanovaccine might also affect the expression of protein encoded by mRNA. Accurately, we cannot declare that the linker length

of MP1000-LPX is optimal for transfection by DC2.4 cells but selected MP1000-LPX as a representative formulation from our result for further investigation.

The pharmaceutical properties including particle size, zeta potential, storage stability and the ability to protect mRNA against serum degradation of optimal MP1000-LPX were then characterized systemically. MP1000-LPX displayed bigger diameter and lower zeta potential compared to MP1000-LPs, indicating the complexation of MP1000-LPs with mRNA. We were surprised to find that particle size measurement results of MP1000-LPX by Zetasizer Nano ZS90 was much larger than that by TEM although some other researchers also observed the similar phenomenon (Wang K. et al., 2018; Yang et al., 2018). The larger size distribution by Zetasizer than TEM observed in the field of size measurement partially resulted from the interference of the dispersant into the hydrodynamic diameter. Complete complexation of the mRNA with MP1000-LPs was also validated according to the results of gel electrophoresis retardation assay. The preliminary storage stability experiment revealed that MP1000-LPX could maintain its excellent transfection efficiency at least for 3 days at 4◦C, which might benefit from the protection of the mRNA against degradation. It has been reported that triMN-LPR with high zeta potential (about 35 mV) could better target human and murine dendritic cells, result in higher recruitments of DCs to draining lymph nodes, and induced significant antitumor responses (Le Moignic et al., 2018). Additionally, Folate modified cationic LPs loaded with DNMT1 gene with positive charge (>30 mV) exerted excellent in vitro targeted genome editing and in vivo antitumor effects (He et al., 2018). However, there seemed to be much cationic charge in the MP1000- LPX. Further formulation optimization will be done to balance the transfection efficiency and the high positive zeta potential in our future work.

The cellular cytotoxicity and uptake mechanism of MP1000-LPX were also evaluated. MP1000-LPX presented good safety in vitro according to the data of cell apoptosis and might be a safe formulation for in vivo application. It has been reported that the presence of free mannose could decrease the uptake of mannose modified preparations (Li et al., 2013; Wang C. et al., 2014). The amount of free mannose in this study was used according to a previously reported literature (Wang C. et al., 2014). Moreover, the experiment design was similar to

previous reports (Li et al., 2013; Wang C. et al., 2014). When DC2.4 cells were pretreated with free mannose as an inhibitor, no significant effect on uptake by LPX was observed, while uptake by MP1000-LPX was significantly decreased. This difference in uptake indicated that the enhanced uptake and transfection were mainly through the MR on DC2.4 cells in line with previous reports (Li et al., 2013; Wang C. et al., 2014). Taken together, our results of uptake in vitro confirmed that the enhanced transfection of MP1000-LPX occurs mainly via a MR-mediated mechanism and the linker length of mannose exerts a crucial role. Although MP1000-LPX exhibited higher level of transfection efficiency through the MR than MP100-LPX and MP2000-LPX, we could not exclude that other linker length of mannose modified cholesterol would exhibit more effective transfection in DC2.4 cells through MR. Nevertheless, a rational design element was proposed and more detailed future studies will be indispensable to facilitate the progression of mRNA nanovaccine.

#### CONCLUSION

In summary, MPn-CHs with different linker molecules (PEG100, PEG1000, and PEG2000) were successfully synthesized by a simple and cost-efficient method. The DC-targeting LPs complexed with mRNA were self-assembled using MPn-CHs as the targeting lipids. The linker molecules had no effect on the particle size and zeta potential of LPs and mRNA-complexed LPs but significantly affected the transfection efficiency of GFP-encoding mRNA. Unexpectedly, PEG<sup>1000</sup> rather than PEG<sup>100</sup> or the commonly used PEG<sup>2000</sup> as the linker achieved the maximal level of GFP expression. MP1000-LPX containing MP1000-CH displayed good

#### REFERENCES


profiles including small size with nearly spherical shape, good stability in serum and little cytotoxicity, indicating a hopeful DCs-targeting delivery system for mRNA vaccine.

#### AUTHOR CONTRIBUTIONS

XS conceived the project. MF and JW designed the experiments. FW, WX, and XB conducted most of the experiments. WX and QZ further performed and analyzed the transfection and characterization of formulations. FW, ME, and WX drafted the manuscript. LG and SY performed the <sup>1</sup>H-NMR and HPLC analysis. QZ and YZ performed the some preliminary experiments. WX, QZ, and AF participated in literature searching. XS and MF finished the manuscript editing. All authors reviewed and approved the manuscript.

#### FUNDING

This work was financially supported by National Key S&T Special Projects (2018ZX09201018-024) and Sichuan Province Science and Technology Support Program (2015SZ 0234).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2018.00980/full#supplementary-material


enhanced anti-inflammatory effect. Biomater. Sci. 5, 1670–1677. doi: 10.1039/ c7bm00345e


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Wang, Xiao, Elbahnasawy, Bao, Zheng, Gong, Zhou, Yang, Fang, Farag, Wu and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Tumor-Associated Fibronectin Targeted Liposomal Nanoplatform for Cyclophilin A siRNA Delivery and Targeted Malignant Glioblastoma Therapy

Phei Er Saw1,2† , Ao Zhang1,2† , Yan Nie1,2, Lei Zhang1,3 \*, Yingjie Xu<sup>4</sup> \* and Xiaoding Xu1,2 \*

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Jinzhi Du, South China University of Technology, China Dalong Ni, University of Wisconsin–Madison, United States Yue Pan, Soochow University, China

#### \*Correspondence:

Lei Zhang zhanglei646@126.com Yingjie Xu xuyingjie@shsmu.edu.cn Xiaoding Xu xuxiaod5@mail.sysu.edu.cn

†These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

> Received: 28 August 2018 Accepted: 28 September 2018 Published: 17 October 2018

#### Citation:

Saw PE, Zhang A, Nie Y, Zhang L, Xu Y and Xu X (2018) Tumor-Associated Fibronectin Targeted Liposomal Nanoplatform for Cyclophilin A siRNA Delivery and Targeted Malignant Glioblastoma Therapy. Front. Pharmacol. 9:1194. doi: 10.3389/fphar.2018.01194 <sup>1</sup> Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Medical Research Center, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China, <sup>2</sup> RNA Biomedical Institute, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China, <sup>3</sup> Department of Hepatobiliary Surgery, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China, <sup>4</sup> Department of Biochemistry and Molecular Cell Biology, Shanghai Key Laboratory for Tumor Microenvironment and Inflammation, Shanghai Jiao Tong University School of Medicine, Shanghai, China

Malignant glioblastoma (GBM) is the most aggressive brain cancer that has a very low survival rate. With the rapid development of nanotechnology in the past few decades, the use of nanoparticles (NPs) for nucleic acid delivery is expected to have a revolutionary impact on GBM therapy. However, clinical success in GBM therapy remains a formidable challenge, mainly due to suboptimal in vivo delivery of therapeutics to glioma cells. Herein, we developed an aptamer-like peptide (aptide)-decorated liposomal nanoplatform for systemic small interfering RNA (siRNA) delivery and targeted GBM therapy. This nanoplatform is mainly composed of the following key components: (i) classic liposome structure with an aqueous core that can encapsulate therapeutic siRNA; (ii) hydrophilic polyethylene glycol (PEG) chains on the outer shell to prolong blood circulation; and (iii) surface-encoded aptide to specifically target the extra-domain B (EDB) of fibronectin that over-expressed on glioma cells. After systemic administration of these new siRNA delivery NPs, they can target the glioma cells and efficiently inhibit the GBM tumor growth by silencing the expression of cyclophilin A (CypA), which is up-regulated in brain cancer and plays an important role in malignant transformation of brain cancer and maintaining glioma cell stemness. These results suggest that the reported RNA interference (RNAi) NP platform herein could become an effective tool for targeted GBM therapy.

Keywords: nanoparticle, aptide, siRNA, targeted delivery, glioblastoma

# INTRODUCTION

Glioblastoma (GBM) is the most common and aggressive form of brain cancer with poor diagnosis, difficult management, and low survival rate. Even when detected early, less than 5% of GBM patients are alive 5 years after diagnosis and the median survival rate is around 14.6 months (Behin et al., 2003; Brandsma and van den Bent, 2007; Johnson and O'Neill, 2012). Currently, multimodality treatment approach including surgical resection, radiotherapy, and chemotherapy

**158**

is the standard care for GBM patients (Meyers et al., 2013; Cheng et al., 2014; Mahmoudi and Hadjipanayis, 2014; Krizbai et al., 2016). However, the development of resistance to therapies (e.g., radiotherapy and chemotherapy) has emerged as a persistent clinical problem and ultimately induces the failure of GBM treatment. Multiple reasons have been demonstrated to contribute to this disappointed outcome, including (i) the complex structure of the brain, (ii) the heterogeneous and invasive nature of GBM, and (iii) difficulty in delivering therapeutics specifically to glioma cells (Lesniak and Brem, 2004; Pardridge, 2005; Ozdemir-Kaynak et al., 2018; Shergalis et al., 2018). Therefore, there is a critical need to develop alternative strategies for more effective GBM treatment.

Since the discovery of RNA interference (RNAi) by Fire et al. (1998), RNAi technology has demonstrated significant potential for disease treatment by silencing the expression of target gene(s), especially those encoding "undruggable" proteins (Whitehead et al., 2009; Xu et al., 2016). However, the safe and effective delivery of RNAi agents such as siRNA to target cells remains a major hurdle for the widespread clinical application of RNAi technology. RNAi agents are biomacromolecules with polyanionic characteristics, which are easily attacked by serum nucleases and cannot readily cross cell membrane. Therefore, specific delivery vehicles are required to facilitate the intracellular uptake and cytosolic delivery of RNAi agents (Yang et al., 2012; Kanasty et al., 2013; Liu et al., 2017; Xu et al., 2017a,b). Over the past decades, nanoparticles (NPs) have been demonstrated as a powerful tool toward this end, especially showing hepatocyte-specificity in non-human primates and clinical trials (Zimmermann et al., 2006; Tseng et al., 2009; Jhaveri and Torchilin, 2014; Shi et al., 2017; Durymanov and Reineke, 2018). Nevertheless, it is still challenged to accomplish the systemic delivery of RNAi agents to a particular non-liver diseased tissue (e.g., solid tumor) and cell type, followed by sufficient intra-cytosolic transport. While several RNAi NP platforms have entered into early phase clinical trials for cancer treatment (Zuckerman and Davis, 2015), substantial obstacles still remain, including long blood circulation, selective accumulation at tumor site, and efficient tumor cell internalization. Specifically, these challenges are amplified by the structural complex of the brain (Pardridge, 2005; Mahmoudi and Hadjipanayis, 2014; Shergalis et al., 2018).

To address these issues, we herein developed a liposome-based extra-domain B (EDB)-targeting nanoplatform for systemic siRNA delivery and GBM therapy. As shown in **Figure 1**, this nanoplatform is composed of a classic liposome structure (i.e., one phospholipid bilayer surrounding an aqueous core), hydrophilic polyethylene glycol (PEG) chains on the outer shell, and surface-encoded aptamer-like peptide (aptide) to specifically target the EDB of fibronectin that over-expressed on glioma cells (Borsi et al., 1992; Castellani et al., 2002; Han and Lu, 2017; Saw et al., 2017). After loading siRNA and then systemic administration, the resulting nanoplatform shows the following unique functions: (i) the hydrophilic PEG chains allow the NPs to escape immunological recognition, thus improving blood circulation; (ii) the surface-encoded aptide moieties can enhance the GBM targeting ability and intracellular siRNA delivery; and (iii) commercial available of the NP compositions and robust NP formulation enables the scale-up of this NP platform. As a proof of concept, we chose cyclophilin A (CypA) as a therapeutic target and systemically evaluated the EDB-targeting NPs for CypA siRNA (siCypA) delivery and its anticancer efficacy. CypA is a ubiquitously distributed protein belonging to the immunophilin family, which shows an activity of peptidylprolyl cis trans isomerase and plays an important role in regulation of protein folding (Wang and Heitman, 2005), trafficking (Shieh et al., 1989; Luban, 1996), assembly (Pan et al., 2008; Tanaka et al., 2011), immune-modulator and cell signaling (Jin et al., 2000; Satoh et al., 2008). It has been demonstrated that CypA is up-regulated in many cancers (e.g., liver, brain, and lung cancers) and is a key determinant for malignant transformation, epithelial to mesenchymal transition (EMT) and cancer metastasis (Yang et al., 2007; Qi et al., 2008). Recent research demonstrated that over-expressed CypA in GBM involves in maintaining glioma cell stemness via Wnt/β-catenin signaling pathway (Wang et al., 2017). Our in vivo results show that the systemic delivery of siCypA with the EDB-targeting NP platform can efficiently inhibit CypA expression in the tumor tissue and significantly inhibit GBM tumor growth.

#### MATERIALS AND METHODS

#### Materials

CypA siRNA (siCypA) and Cy5.5-labeled CypA siRNA (Cy5.5-siCypA) were acquired from Dharmacon (United States). The siRNA sequences are as follows: 5<sup>0</sup> -UGA CUU CAC ACG CCA UAA UdTdT-3<sup>0</sup> (sense); 5<sup>0</sup> -AUU AUG GCG UGU GAA GUC AdTdT-3<sup>0</sup> (antisense). Protamine sulfate and sepharose CL-4B column were purchased from Sigma Aldrich (United States). 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(10-rac-glycerol) (POPG), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE), and PEG (2000)-DSPE (ammonium salt) (PEG2000- DSPE) were purchased from Avanti Polar Lipids (United States) and used as received. N-Maleimide-PEG2000-DSPE (ammonium salt) (Mal-PEG2000-DSPE) and plant cholesterol (Chol) were purchased from A.V.T. (Shanghai) Pharmaceuticals (China). EDB-targeting aptide (APTEDB) with an additional cysteine in the β-hairpin constant loop region (sequence from N to C terminal, CSSPIQGSWTWENGK(C)WTWGIIRLEQ) was synthesized by Guangzhou IGE biotechnology Co., Ltd (China). Lipofectamine 2000 (Lipo2000) was provided by Thermofisher Scientific (United States). Real-time PCR assay kit was procured from Promega (United States). All antibodies were purchased from Abcam (United States) and used according to the manufacturer's protocol. All other chemicals were of reagent grade and used directly.

#### Methods

#### Synthesis of APTEDB Conjugated PEG2000-DSPE (APTEDB-PEG2000-DSPE)

The APTEDB-PEG2000-DSPE was synthesized via the reaction between the thiol group of APTEDB and maleimide terminal

group of Mal- PEG2000-DSPE. In brief, APTEDB and Mal- PEG2000-DSPE were dissolved in dimethyl sulfoxide (DMSO) and chloroform, respectively. Subsequently, these two solutions were mixed in a molar ratio (APTEDB: Mal-PEG2000-DSPE) of 1:2. Under nitrogen atmosphere, the mixture was stirred at room temperature for 12 h. Thereafter, the mixture was transferred to dialysis membrane (MWCO 3500) and dialyzed against deionized water for 3 days. After freeze-drying under vacuum, the APTEDB-PEG2000-DSPE was collected as a white powder.

#### Preparation and Characterizations of EDB-Targeting siRNA-Loaded NPs

The classic rehydration method was employed to prepare the EDB-targeting siRNA-loaded NPs (Saw et al., 2015). POPC, Chol, and POPG were dissolved in chloroform in a molar ratio of 4:3:3 and APTEDB-PEG2000-DSPE (2.5 wt% of the total lipid) was then added. The mixture was stirred at room temperature for 10 min to form a homogenous solution. Subsequently, the solvent was removed by using rotary evaporator and a thin lipid film was thus generated. Then, HEPES-buffered 5% glucose (HBG) containing siCypA/protamine complexes was added and the resulting mixture (2 mg/mL) was briefly sonicated to accelerate the formation of siRNA-loaded liposomes. Thereafter, extrusions were performed by using a 100 nm polycarbonate membrane to ensure the formation of uniform liposomes, which were then passed over sepharose CL-4B column to obtain the purified siRNA-loaded NPs. To prepare the siRNA-loaded NPs without EDB-targeting ability, the POPC, Chol, and POPG were dissolved in chloroform in a molar ratio of 4:3:3 and then the siRNA-loaded NPs were prepared and purified according to the method described above.

The siRNA-loaded NPs were characterized in terms of size, zeta potential, and morphology. The particle size and zeta potential were examined by dynamic light scattering (DLS, Malvern Instruments Corporation). The morphology of siRNA loaded NPs was visualized by transmission electron microscopy (TEM, Tecnai G<sup>2</sup> Spirit BioTWIN). To determine the siRNA encapsulation efficiency (EE), Cy5.5-siCypA was used to prepare the siRNA-loaded NPs according to the method aforementioned and the obtained NPs were dispersed in 1 mL of PBS. Subsequently, a small volume (5 µL) of the NP solution was withdrawn and mixed with 20-fold DMSO. The standard was prepared by mixing 5 µL of naked Cy5.5-siCypA solution with 20-fold DMSO. The fluorescence intensity of Cy5.5-siCypA was measured using a Synergy HT multimode microplate reader (BioTek Instruments), and the siRNA EE% is calculated as EE% = (FINPs / FIStandard) × 100.

#### In vitro siRNA Release

fphar-09-01194 October 15, 2018 Time: 19:26 # 4

Cy5.5-siCypA-loaded NPs were dispersed in 1 mL of PBS (pH 7.4) and then transferred to a Float-a-lyzer G2 dialysis device (MWCO 100 kDa, Spectrum) that was immersed in PBS (pH 7.4) at 37◦C. At a predetermined interval, 5 µL of the NP solution was withdrawn and mixed with 20-fold DMSO. The fluorescence intensity of Cy5.5-siCypA was determined by Synergy HT multi-mode microplate reader.

#### Cell Culture

Human glioma cells (U87MG and U251MG), skin melanoma cells (A375), prostate cancer cells (PC3), breast cancer cells (MCF-7), and Jurkat leukemic T cells were incubated in DMEM medium with 10% FBS at 37◦C in a humidified atmosphere containing 5% CO2.

#### Evaluation of the Expression of EDB and CypA in Glioma Cells

Real-time qPCR was used to evaluate the mRNA level of EDB and CypA in the glioma cells. The intracellular RNA was isolated with RiboEx using an RNA isolation kit (Geneall, South Korea). cDNA was then synthesized using reverse transcription technique using 1 µg of total RNA from each sample. The primers used for detecting the EDB domain of fibronectin were 5<sup>0</sup> -AAC TCA CTG ACC TAA GCT TT-3<sup>0</sup> (forward) and 5<sup>0</sup> -CGT TTG TTG TGT CAG TGT AG-3<sup>0</sup> (reverse). The primers for detecting the CypA were 5<sup>0</sup> -TAT CTG CAC TGC CAA GAC TGA GTG-3<sup>0</sup> (forward) and 5<sup>0</sup> -CTT CTT GCT GGT CTT GCC ATT CC-3<sup>0</sup> (reverse).

#### Targeted Cellular Uptake

Glioma cells (20,000 cells) were seeded in disks and incubated in 2 mL of DMEM medium containing 10% FBS for 24 h. Subsequently, the Cy5.5-siCypA-loaded NPs were added at a siRNA concentration of 5 nM, and the cells were allowed to incubate at 37◦ C for 4 h. After removing the medium and subsequently washing with PBS buffer thrice, the nuclei were stained with Hoechst 33342 and the cells were then viewed under a FV1000 confocal laser scanning microscope (CLSM, Olympus).

#### In vitro Gene Silencing

Glioma cells were seeded in 6-well plates (50,000 cells per well) and incubated in 2 mL of DMEM medium containing 10% FBS for 24 h. Subsequently, the cells were incubated with the siCypA-loaded NPs for 24 h. After washing the cells with PBS buffer thrice, the cells were further incubated in fresh medium for another 48 h. Thereafter, the cells were digested by trypsin and the intracellular RNA was isolated for real-time qPCR to examine the CypA expression. As a positive control, Lipo2000/siCypA complexes were prepared according to manufacturer's protocol and then incubated with the glioma cells for 4 h. After washing the cells thrice with PBS, the cells were further incubated for another 48 h and then collected for real-time qPCR analysis.

#### In vitro Cell Viability Assay

Glioma cells were seeded in 96-well plates (5,000 cells per well) and incubated in 0.1 mL of DMEM medium with 10% FBS for 24 h. Thereafter, the siCypA-loaded NPs were added at predetermined concentration and the cells were allowed to incubate for 48 h. After removing the medium and washing the cells with PBS thrice, the cell viability was measured using the AlamarBlue assay according to the manufacturer's protocol.

#### Animals

Healthy female BALB/c normal mice and nude mice (4–5 weeks old) were purchased from Sun Yat-sen University experimental animal center. All in vivo studies were performed in accordance with a protocol approved by the Institutional Animal Care and Use Committee at Sun Yat-sen University.

#### Pharmacokinetics

Healthy female BALB/c normal mice were randomly divided into three groups (n = 3) and given an intravenous injection of either (i) naked Cy5.5-siCypA, (ii) Cy5.5-siCypA-loaded NPs, or (iii) EDB-targeting NPs at a 1 nmol siRNA dose per mouse. At predetermined time intervals, 20 µL of blood was withdrawn and the wound was pressed for several seconds to stop bleeding. The fluorescence intensity of Cy5.5-siCypA in the blood was determined by Synergy HT multi-mode microplate reader. The blood circulation half-life (t1/2) was calculated according to previous report (Winter et al., 2013; Xu et al., 2017a).

#### GBM Xenograft Tumor Model

The tumor model was constructed by subcutaneous injection with 200 µL of glioma cell suspension (1:1 mixture of medium and Matrigel) with a density of 2 × 10<sup>6</sup> cells/mL into the back region of healthy female nude mice. When the volume of the tumor xenograft reached ∼100 mm<sup>3</sup> , the mice were used for the in vivo experiments.

#### Evaluation of EDB Expression in GBM Xenograft Tumor Model

The tumor tissues were excised from the GBM xenograft tumor-bearing mice and then sectioned using Leica CM 1950 Research Cyrostat (Leica Biosystems, IL, United States) at a 5 µm thickness. Subsequently, the tumor sections were washed in PBS and fixed with 4% (w/v) paraformaldehyde. After blocking with 2% bovine serum albumin (BSA) for 1 h, the tumor sections were incubated with anti-BC-1 antibody (EDB specific antibody) overnight at 4◦ C. Then, the tumor sections were washed in PBS and Alexa Fluor 594-conjugated goat anti-mouse IgG was added. After 1 h incubation at room temperature, the tumor sections were washed with PBS thrice and mounted with DAPI containing mounting medium. The tumor sections were finally viewed under CLSM to examine the EDB expression.

#### Inhibition of Tumor Growth

fphar-09-01194 October 15, 2018 Time: 19:26 # 5

GBM xenograft tumor-bearing nude mice were randomly divided into three groups (n = 4) and intravenously injected with (i) PBS, (ii) siCypA-loaded NPs or (iii) EDB-targeting NPs at a 1 nmol siRNA dose per mouse once every two days. All the mice were administrated four consecutive injections and the tumor growth was monitored every two days by measuring perpendicular diameters using a caliper and tumor volume was calculated as follows:

$$\mathcal{V} = \mathcal{W}^2 \times \mathcal{L}/2$$

where W and L are the shortest and longest diameters, respectively.

#### Histology

After the aforementioned treatment, the mice in each group were sacrificed at end of the evaluation period, and the tumor tissues and main organs (heart, liver, spleen, lung, and kidney) were collected. After fixing with 4% paraformaldehyde and then embedding in paraffin, the tissue was sectioned and stained with hematoxylin-eosin (H&E) and then viewed under an optical microscope. In addition, the apoptosis in the tumor tissues was examined by terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay according to manufacturer's protocol.

#### RESULTS AND DISCUSSION

#### Preparation and Characterizations of the NPs

EDB of fibronectin has been demonstrated to be highly expressed in the aggressive and malignant GBM (Borsi et al., 1992; Castellani et al., 2002). Therefore, EDB-targeting nanoplatform could be used a robust vehicle to deliver various therapeutics for GBM treatment. Arising from the excellent targeting ability of aptamer, we previously developed aptamer-like peptide (aptide) and demonstrated its strong ability to specifically bind EDB with high affinity (K<sup>d</sup> ∼16 nM) (Saw et al., 2013, 2015, 2017). Based on the high EDB expression in GBM and strong targeting ability of aptide, we herein conjugated the EDB-targeting aptide (APT-EDB) to PEG2000-DSPE, which was then formulated with other lipids (POPC, Chol, and POPG) to obtain the APT-EDB-decorated liposomes (denoted APT-EDB NPs) for the systemic siRNA delivery and targeted GBM treatment (**Figure 1**). In this work, we employed this nanoplatform to deliver siCypA because it can specifically silence the expression of cancer-associated CypA that over-expressed in GBM (Yang et al., 2007; Qi et al., 2008).

glioma cells (U87 and U251). (B) The mRNA level of CypA in the normal brain tissues and GBM xenograft tumor tissues of nude mice. (C) Western blot analysis of the EDB expression in the glioma cells (U87 and U251). (D) Immunofluorescence analysis of the EDB expression in the GBM xenograft tumor tissues of nude mice. The nuclei and EDB were stained with blue and green fluorescence, respectively. ∗∗P < 0.01; ∗∗∗P < 0.001.

**Figure 2** shows the morphology of the siCypA-loaded APT-EDB NPs. Similar as other reported liposomes (Saw et al., 2013, 2015, 2017), the APT-EDB NPs show a spherical morphology with an average size of ∼112 nm determined by DLS (**Figure 2D**). Compared to the NPs without APT-EDB decoration (∼104 nm, **Figure 2C**), there is around 10 nm increase after APT-EDB decoration, suggesting the success in the APTEDB decoration because APT-EDB shows a hydrodynamic size of ∼5 nm (Kim et al., 2012). Moreover, due to the presence of PEG chains on the outer layer, these siRNA-loaded NPs

show a high stability. As shown in **Figure 2E**, either with or without APT-EDB decoration, possibly due to the presence of weak interaction between the particles and protein, there is a slight change in the particle size and polydispersity density (PDI) when incubating the NPs in 10% FBS-containing PBS solution for 8 h. However, this interaction will become stable as the incubation time increases and therefore there is no obvious size change after 8 h incubation. After obtaining these stable NPs, we next labeled the siCypA with fluorescent dye of Cy5.5 to examine the siRNA encapsulation efficiency (EE%) and release behavior. Through analyzing the fluorescence intensity, the EE was determined as ∼90%. The siRNA release profile is shown in **Figure 2F**. As can be seen, the NPs show a sustained siRNA release behavior. Around 35% of loaded siRNA can be released within 12 h and the cumulative release reaches ∼50% 48 h later.

### Determination of EDB and CypA Expression

The main purpose of this work is to develop EDB-targeting siRNA delivery nanoplatform for GBM treatment by silencing the CypA expression. Prior to evaluating the gene silencing efficacy of the APT-EDB NPs, we first examined the CypA and EDB expression in the glioma cells. We chose different cancer cell lines and examined the corresponding CypA expression using realtime qPCR. As shown in **Figure 3A**, compared to skin melanoma cells (A375), Jurkat leukemic T cells, prostate cancer cells (PC3), and breast cancer cells (MCF-7), glioma cells (U87 and U251) show a much higher CypA expression. This encouraging result suggests that CypA is organ-specific over-expressed. To further verify this result, we compared the CypA expression in U87 and U251 xenograft tumors with normal brain tissues. As can be seen, the U87 or U251 xenograft tumors show around twofold or fourfold higher CypA expression that that of normal brain tissues (**Figure 3B**), implying that silencing CypA expression can be used as a potential strategy for GBM treatment.

After validation of the high CypA expression in the glioma cells, we next examined the EDB expression on these cells. As shown in **Figure 3C**, western blot analysis demonstrates that both U87 and U251 cells have a high EDB expression. Moreover, if using these two cell lines to construct xenograft tumor model, EDB is also highly expressed in the tumor tissues (green fluorescence, **Figure 3D**). All these results are consistent with previous reports (Yang et al., 2007; Qi et al., 2008) and indicate that EDB is indeed a suitable biomarker for targeted GBM therapy. Combining with results of PCR analysis (**Figures 3A,B**), because U251 cells show a relatively higher level of CypA and EDB expression compared to U87 cells, we chose U251 cells to evaluate the targeting and gene silencing ability of the APT-EDB NPs.

#### Evaluation of GBM-Targeting Ability and in vitro Gene Silencing

The GBM-targeting ability was evaluated by incubating U251 cells with the siCypA-loaded NPs. From the fluorescent images shown in **Figure 4A**, compared to naked siRNA, higher amount of siRNA can be observed and distributed in the cytoplasm of the

cells incubated with the siRNA-loaded NPs. More importantly, due to the presence of specific recognition between aptide and EDB, the cellular uptake of APT-EDB NPs is much higher than that of the NPs without aptide decoration. This result is further proven by the flow cytometry analysis. As shown in **Figure 4B**, U251 cells show stronger ability to internalize the APT-EDB NPs and intracellular mean fluorescence intensity (MFI) is more than fivefold stronger than the cells incubated with the NPs without aptide decoration (**Figure 4C**).

After validation of the GBM-targeting ability of the APT-EDB NPs, we next examined their in vitro gene silencing efficacy. **Figure 5A** shows the mRNA level of CypA in U251 cells treated with the siCypA-loaded NPs. With the GBM-targeting ability to improve the cellular uptake, the APT-EDB NPs shows an effective gene silencing and there is around 80% decrease in the mRNA level of CypA in the cells at a siRNA concentration of 10 nM. This high gene silencing efficacy is similar as the commercial available Lipo2000 and is more than twofold higher than that of the NPs without aptide decoration. With this suppressed CypA expression, the cells have a very low viability. As shown in **Figure 5B**, lower than 50% of the cells are alive when treated with

FIGURE 6 | (A) Blood circulation profile of the naked Cy5.5-siCypA and the Cy5.5-siCypA-loaded NPs with (APT-EDB NPs) or without aptide decoration (NPs). (B) Overlaid fluorescent image of the main organs of the GBM xenograft tumor-bearing mice treated with naked siRNA, and the siCypA-loaded NPs with (APT-EDB NPs) or without aptide decoration (NPs). (C) Biodistribution of the siRNA obatined from (B). (D) Tumor growth profile of the GBM xenograft tumor-bearing mice treated with PBS, and the siCypA-loaded NPs with (APT-EDB NPs) or without aptide decoration (NPs). Intravenous injections are indicated by the arrows. (E,F) Weight (E) and representative photograph (F) of the tumor tissues from the mice in each group after 28 day evaluation period. (G–I) TUNEL staining of the collected tumor tissues in (F). TUNEL-positive apoptotic cells were stained with green fluorescence. <sup>∗</sup>P < 0.05; ∗∗P < 0.01.

FIGURE 7 | Histological section of the main organs of the GBM xenograft tumor-bearing mice after systemic treatment with PBS, and the siCypA-loaded NPs with (APT-EDB NPs) or without aptide decoration (NPs). H&E; magnification 100×.

the APT-EDB NPs at a siRNA concentration of 10 nM. In contrast, at the same siRNA concentration, around 90% of the cells are still alive when treated with the NPs without aptide decoration. This result highlights the importance of the GBM-targeting ability of the siRNA-loaded NPs to their gene silencing in glioma cells.

#### Evaluation of in vivo Anti-tumor Efficacy

Encouraged by the strong EDB-targeting ability and effective gene silencing of the APT-EDB NPs, we finally evaluated their in vivo anti-tumor efficacy. The pharmacokinetics was first examined by intravenously injecting the Cy5.5-siCypA-loaded NPs to normal adult mice. As shown in **Figure 6A**, the naked siRNA is rapidly cleared from the blood and its half-life (t1/2) is less than 10 min. In contrast, the APT-EDB NPs show much longer blood circulation with blood t1/<sup>2</sup> of around 4.68 h, which is comparable to the t1/<sup>2</sup> of the NPs without aptide decoration (∼5.12 h). This long circulation feature is mainly attributed to protection by the PEG chains on the outer layer (Knop et al., 2010) and will ensure the accumulation of the APT-EDB NPs in the tumor tissues via enhance and permeable retention (EPR) effect (Bertrand et al., 2014). The biodistribution results in **Figures 6B,C** also demonstrate our statement. As shown in **Figure 6B**, the siRNA loaded NPs show higher tumor accumulation than that of naked siRNA. Moreover, due to the presence of APT-EDB targeting ligand, the APT-EDB NPs show more than twofold higher tumor accumulation than that of the NPs without aptide decoration (**Figure 6C**).

The in vivo anti-tumor efficacy was examined by intravenous injection of the siCypA-loaded APT-EDB NPs to GBM xenograft tumor-bearing mice once every two days at a 1 nmol siRNA dose per mouse. As shown in **Figures 6D–F**, after four consecutive injections, the tumor growth is obviously inhibited compared to the mice treated with PBS or NPs without aptide decoration. There is around sevenfold increase in the tumor size (from ∼50 to ∼350 mm<sup>3</sup> ) (**Figure 6B**). However, for the mice treated with PBS or the NPs without aptide decoration, there is around 14-fold (from ∼50 to ∼690 mm<sup>3</sup> ) or 11-fold (from ∼50 to ∼550 mm<sup>3</sup> ) increase in the tumor size. This tendency is further supported by the result of TUNEL assay. From the images shown in

#### REFERENCES


**Figures 6G–I**, more apoptotic cells can be observed in the tumor section of the mice treated with the siCypA-loaded APT-EDB NPs (**Figure 6I**). Noting that, the administration of the NPs does not induce apparent in vivo toxicity. As shown in **Figure 7**, no noticeable histological changes can be found in the tissues from heart, liver, spleen, lung, or kidney of the mice treated with PBS or siRNA-loaded NPs. All these results indicate that the NP platform developed in this work could be potentially used as a safe and efficient gene delivery system for GBM treatment.

#### CONCLUSION

In summary, we have developed a robust liposome-based NP platform for systemic siRNA delivery and targeted GBM treatment. This NP platform can target the glioma cells through the specific recognition between their surface-encoded aptide and over-expressed EDB on glioma cells, thereby leading to improved intracellular siRNA delivery and more effective gene silencing in glioma cells. In vivo results show that this long-circulating NP platform can target the GBM tumor tissues and obviously inhibit the tumor growth by silencing the CypA expression. Taken together, the NP platform developed herein could potentially serve as an effective delivery tool for non-invasive GBM treatment.

#### AUTHOR CONTRIBUTIONS

PS and XX conceived and designed the experiments. PS, AZ, YN, and YX performed the experiments. PS, YN, LZ, YX, and XX analyzed the data and co-wrote the paper.

# FUNDING

This work was supported by the National Natural Science Foundation of China (81874226 and 81803020) and the Thousand Talents Program for Distinguished Young Scholars.



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling Editor declared a past co-authorship with the authors PS and XX.

Copyright © 2018 Saw, Zhang, Nie, Zhang, Xu and Xu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Effects of Chitosan Oligosaccharides on Human Blood Components

Xi Guo1,2, Tong Sun1,2, Rui Zhong<sup>3</sup> , Lu Ma<sup>2</sup> , Chao You1,2,4, Meng Tian1,2,4 \*, Hao Li <sup>2</sup> \* and Chengwei Wang<sup>5</sup> \*

*<sup>1</sup> Department of Neurosurgery, West China Hospital, Sichuan University, Chengdu, China, <sup>2</sup> Neurosurgery Research Laboratory, West China Hospital, Sichuan University, Chengdu, China, <sup>3</sup> Institute of Blood Transfusion, Chinese Academy of Medical Sciences & Peking Union Medical College, Chengdu, China, <sup>4</sup> West China Brain Research Centre, West China Hospital, Sichuan University, Chengdu, China, <sup>5</sup> Department of Integrated Traditional and Western Medicine, West China Hospital, Sichuan University, Chengdu, China*

Chitosan oligosaccharide (COS) is known for its unique biological activities such as anti-tumor, anti-inflammatory, anti-oxidant, anti-bacterial activity, biological recognition, and immune enhancing effects, and thus continuous attracting many research interests in drug, food, cosmetics, biomaterials and tissue engineering fields. In comparison to its corresponding polymer, COS has much higher absorption profiles at the intestinal level, which results in permitting its quick access to the blood flow and potential contacting with blood components. However, the effects of COS on blood components remain unclear to date. Herein, two COS with different molecular weight (MW) were characterized by FTIR and <sup>1</sup>H NMR, and then their effects on human blood components, including red blood cells (RBCs) (hemolysis, deformability, and aggregation), coagulation system [activated partial thromboplastin time (APTT), prothrombin time (PT), thrombin time (TT), and the concentration of fibrinogen (Fib)], complement (C3a and C5a activation), and platelet (activation and aggregation), were comprehensively studied. In the case of RBCs, COS exhibited a low risk of hemolysis in a dose and molecular weight dependent manner and the irreversible aggregation was observed in their high concentration. For coagulation system, COS has a mild anticoagulation activity through blocking the intrinsic coagulation pathway. In addition, COS showed no effect on complement activation in C3a and C5a and on platelet activation while inhibition of platelet aggregation was evident. Finally, the mechanism that effects of COS on blood components was discussed and proposed.

#### Edited by:

*Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China*

#### Reviewed by:

*Shiguo Chen, Shenzhen University, China Jian Zhong, Shanghai Ocean University, China*

#### \*Correspondence:

*Meng Tian 6744710@qq.com; tianmong007@gmail.com Hao Li coscolh@126.com Chengwei Wang 21270526@qq.com*

#### Specialty section:

*This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology*

> Received: *19 October 2018* Accepted: *16 November 2018* Published: *03 December 2018*

#### Citation:

*Guo X, Sun T, Zhong R, Ma L, You C, Tian M, Li H and Wang C (2018) Effects of Chitosan Oligosaccharides on Human Blood Components. Front. Pharmacol. 9:1412. doi: 10.3389/fphar.2018.01412* Keywords: chitosan oligosaccharide (COS), red blood cells (RBC), coagulation, complement, platelet

# INTRODUCTION

Chitosan oligosaccharide (COS) is an oligomer of chitosan with an average molecular weight (MW) < 5,000 Da, and its chemical structure, like chitosan, is a linear binary copolymer consisting of β-1, 4-linked 2-acetamido-2-deoxy-β-D-glucopyranose (GlcNAc) and 2-amino-2-deoxy-β-D-glucopyranose (GlcN) (Kumar et al., 2004; Muanprasat and Chatsudthipong, 2017). The great research interest on this oligomer results not only from its physical-chemical properties such as better water solubility and cationic nature at neutral pH, but also from its biological activities, e.g., anti-tumor, anti-inflammatory, anti-oxidant, anti-bacterial activity, biological recognition, and immune enhancing effects (Liu et al., 2009; Lu et al., 2014, 2015; Zhang et al., 2014, 2018; Huang et al., 2016; Kunanusornchai et al., 2016; Mattaveewong et al., 2016; Ding et al., 2017; Kalagatur et al., 2018). These unique properties and activities continuously attract many research interests in drug, food, cosmetics, biomaterials and tissue engineering fields (Swiatkiewicz et al., 2015; Lee et al., 2017; Bai et al., 2018; El-Sayed et al., 2018; Nan et al., 2018).

Although the use of COS has been broadened in many research and applied areas, there are some dilemmas that need to be elucidated. For example, it is well-known that oral administration of COS has much higher absorption profiles at the intestinal level than that of its corresponding polymer, chitosan, which results in permitting its quick access to the blood flow and potential contacting with blood components (Chae et al., 2005). This is also applicable to chitosan based tissue-engineering scaffolds, because the degraded fragments that mainly composed of COS would finally enter into the blood circulation. For these reasons, the effect of COS on the blood such as its main components, including red blood cells (RBCs), coagulation, protein adsorption, complement, and platelets, has to be taken into account, since the hemocompatibility of the COS has not been reported to date. In general, chitosan formed a coagulum in contact with whole blood where platelets showed distinct adhesion to its surface within a short time, and the clotting time was reported to reduce by 40% compared to whole blood alone (Rao and Sharma, 1997; Chen et al., 2017). As a result, chitosan is normally used as a hemostatic dressing for wound healing in clinic instead of a blood-contact medical device (Muzzarelli et al., 2005). The hemostatic property was attributed to the polycationic characteristic of chitosan and its non-specific binding to cell membranes resulting from the positively charged amino groups along the molecular chains (Benesch and Tengvall, 2002). In the case of COS, however, some different phenomena were observed. Lin et al. compared the whole blood clotting time of COS and chitosan and found that the former showed no hemostatic effect (Lin and Lin, 2003). Similarly, Fernandes et al. studied the interactions of COS with human RBCs and the results showed that no significant hemolysis was evident and the damage of COS on the RBCs was dependent on the concentration and MW of the used samples (Fernandes et al., 2008). The above reports indicated that COS appears to be better hemocompatiblility than chitosan. Besides the positively charged amino groups, there are numerous hydroxyl groups along the molecular chains of the COS. Previous studies on the effect of functional groups such as hydroxyl, carboxyl, and amino groups on the blood components have suggested that hydroxyl groups resulted in a mild anticoagulation as well as complement activation in spite of no report on COS (Sperling et al., 2005). In this regard, it can be expected that the hydroxyl groups along the molecular chains of COS probably also have some specific effects on the blood components.

Herein, we hypothesize that COS would affect the blood components in a different manner comparable to chitosan, which might be related with its MW and functional groups along the molecular chains. To address this hypothesis, the effects of COS with two different MW on the human blood components were comprehensively investigated in this work (**Figure 1**). Specifically, the effect of COS on RBCs included hemolysis, deformability and aggregation; the blood coagulation was evaluated by coagulation time, including prothrombin time (PT), activated partial thromboplastin time (APTT) and thrombin time (TT), and the concentration of fibrinogen (Fib); for complement system, complement 3a (C3a) and 5a (C5a) using ELISA method were considered; and in terms of platelet, platelet activation, and aggregation were determined. Based on above results, the mechanism of action was discussed and proposed.

# MATERIALS AND METHODS

# Materials Preparation

Two COS samples with distinct MW, 3 (COS-3K) and 5 (COS-5K) kDa, were purchased from Zhejiang Aoxing Biotechnology Co., Ltd. (Zhejiang, China). Both samples were obtained by enzymatic hydrolysis of chitosan derived from crab shells. PEI and Heparin sodium were purchased from Aladdin Co., Ltd. (Chengdu, China). Coagulation-related kits were purchased from Union Bio-tech Co., Ltd. (Chengdu, China). The ELISA Kit II was purchased from Becton-Dickinson Co., Ltd, (USA). Flow cytometry-related agents, anti-CD61-fluorecein isothiocyanate (FITC) and anti-CD62p-phycoerythrin (PE) were purchased from BD Pharmingen, BD Bioscience Co., Ltd. Adenosinediphosphate and epinephrine, which were platelet aggregationinducers, were purchased from Kelong Co., Ltd. (Chengdu, China). COS-3K/5K were mixed with normal saline to get the COS stock solution.

# Structure Characterization

The structure of the two COS was characterized by Fourier transform infrared spectra (FTIR) and <sup>1</sup>H nuclear magnetic resonance spectroscopy (NMR). FTIR analyses of the COSs were recorded with KBr compressed pellets on a Nicolet 670 FT-IR Spectrometer. <sup>1</sup>H NMR analyses were recorded on a Bruker AV II-400MHz spectrometer. The DD was calculated by <sup>1</sup>H NMR according to equation (1), where "ACH3" and "AGlcNH−2," respectively correspond to the integral of the N-acetyl proton signal and H-2 proton signal of GlcN units.

$$DD(\%) = 1 - \frac{\frac{1}{3}A\_{CH\_3}}{\frac{1}{3}A\_{CH\_3} + A\_{GlcNH-2}} \times 100\tag{1}$$

# Blood Collection

The study was approved by Ethical Committee of Institute of Blood Transfusion, Chinese Academy of Medical Sciences & Peking Union Medical College. Anticoagulated whole blood samples (3.8% sodium citrate/blood 1:9) and fresh frozen plasma (FFP) anticoagulated with citrate-phosphate-adenine used in this study were obtained from Chengdu Blood Center. For RBC tests, RBC suspension was gained by washing anticoagulated whole blood with normal saline for three times and then adjusting hematocrit (HCT) to 10%. The whole blood samples were centrifuged at 1,200 g for 20 min under 4◦C to obtain plateletpoor plasma (PPP) and at 150 g for 10 min under 4◦C to acquire platelet-rich plasm (PRP).

#### Red Blood Cells Hemolysis

Washed RBC suspension with 10% HCT (270 µl) was mixed with COS-3K/5K solution (30 µl) to obtain a final COS concentration

of 0.01, 0.1, 0.5, 1 mg/ml. The mixtures then were gently centrifuged at 1,500 g for 5 min after 1 h incubation under 37◦C. Supernatant (0.02 ml) obtained after centrifuge was added to ortho-tolidine solution (1 ml; 0.2 g in 60 ml acetic acid) and hydrogen peroxide (1 ml; 1 g/L). After incubation for 10 min, the reaction was stopped by mixed with acetic acid (10 ml; 10%). The spectrophotometer was used to measure absorbance at a wavelength of 435 nm. A standard hemoglobin solution with concentration of 100 mg/L was used as a standard control and diluted water (DW) was used as a hemolysis control. To figure out the extent of hemolysis, the following equation was used (Lewis, 1965).

$$Henny(\%) = \left(\frac{A\_1}{A\_S} \times 100\right) \times \frac{100 - Hct(\%)}{C\_{Hb}(\lg/L) \times 1000} \tag{2}$$

A<sup>1</sup> means the absorbance (435 nm) of sample; A<sup>S</sup> means the absorbance of standard sample (100 mg/L); Hct (%) means hematocrit; CHb means concentration of hemoglobin.

#### Deformability

To measure RBC deformability, the washed RBC suspension (270 µl) were incubated with COS stock solution (30 µl) to attain the final concentration of 0.01, 0.1, 0.5, and 1 mg/ml. After 1 h incubation, the suspension was centrifuged at 1,500 g for 4 min and then the precipitate was gently mixed with 1 ml of polyvinylpyrrolidone solution (15% in PBS). A laser-diffraction Ektacytometer system (LBY-BX, Beijing Pencil Instrument Co., Ltd, China) was used. About 0.5 ml of the mixture was added to the system and the sheared between two concentric cylinders of the machine, in which size of gap was 0.5 mm, under four different shear stresses of 0.39, 0.77, 1.54, and 7.7 Pa (corresponding to shear rates of 50, 100, 200, and 1,000) at 37◦C. Then the parameter was presented as elongation index (EI) by the system through measuring the diffraction of passing laser. Normal saline and diluted water mixed with RBC suspension were used as standard and positive control, respectively.

#### Aggregation

To determine whether COS can lead to RBC aggregation, anticoagulated whole blood samples (270 µl) was mixed with COS stock solution (30 µl) to attain the final concentration of 1 and 5 mg/ml for 1 h incubation. The mixtures then were centrifuged at 1,000 g for 3 min. The precipitate (3 µl) and supernatant (40 µl) were gently mixed and mixtures (4 µl) were made into slides. Images of slides were captured by a digital microscope camera. Polyetherimide (PEI), which can lead to obvious erythrocytes aggregation, was used as positive control and normal saline was used as standard control.

#### Coagulation

Activated partial thrombin time (APTT), prothrombin time (PT), thrombin time (TT) and the concentration of fibrinogen (Fib) were measured to determine the effect of COS on coagulation system by a modified strategy according to previous study (Nikitina et al., 2018). Briefly, COS stock solution (60 µl) were mixed with FFP (540 µl) to attain a final concentration of 0.01, 0.1, and 0.5 mg/ml and then was incubated at 37◦C for 3 min. A coagulation analyzer (Instrumentation Laboratory ACL ELITE, USA) was used to test the incubated FFP. The parameter range available for the machine is 6 s to 245 s for APTT, 5 s to 165 s for PT, and 3 s to 169 s for TT. NS was used as standard control and heparin (HP, final concentration of 0.75 IU/ml) was used as positive control. Values of above coagulation tests were mean from three measurements and results were expressed as mean ± standard deviation (SD).

#### Complement

Complement activation reflects the effect of COS on complement system. We focused on the activation of human complement C3 and C5 in circulation and measured their cleft fragment C3a and C5a. The method of measurement was followed by the standard protocol of ELISA Kit II (Becton-Dickinson Co., Ltd, USA). Briefly, serum (90 µl) was incubated with COS (final concentration of 0.1 and 1 mg/ml) at 37◦C for 1 h. Then incubated serum (100 µl) and standard solution (100 µl) of the kit were, respectively, mixed with ELISA dilution (50 µl) in antibody coated well-sealed under room temperature for 2 h. After that, wells were washed and added detection antibody and enzyme concentrate of the kit for 1 h incubation. Finally, the wells were read and absorbance of 450 nm was measured by spectrophotometer (EON, Bio-Tech CO., Ltd, USA). The concentration of C3a and C5a was calculated according to the standard curve of standard samples of kit.

#### Platelet

#### Activation

CD61, a specific marker of platelet surface, and CD62p, the marker of activated platelet, were involved in this test. Plateletrich plasm (PRP) (90 µl) was incubated with COS stock solution (10 µl) at 37◦C for 1 h to obtain a final concentration of 0.1 and 0.5 mg/ml. Incubated PRP (5 µl) was mixed with anti-CD61 fluorecein isothiocyanate (5 µl), anti-CD62p-phycoerythrin (5 µl), and PBS buffer (10 mM, 40 µl) in dark for 15 min and then added 400 µl of PBS buffer. The flow cytometry (Becton-Dickinson, San Jose', CA, USA) were used to measure the number of CD62p-expressing platelets. Normal saline and thrombin (10 U/ml) were used as normal control and positive control, respectively (Zhong et al., 2013).

#### Aggregation

To investigate whether platelet aggregation was affected by COS, PRP (270 µl) was incubated with COS stock solution (30 µl) at 37◦C for 1 h. After incubation, we added adenosine diphosphate (0.1 mM, 12.5 µl) and epinephrine (0.15 mM, 12.5 µl) to the incubated PRP to induce platelet aggregation. And the aggregometry (MODEL700, CHRONO-LOG CO. LTD, USA) was used to measure extent of aggregation. PRP incubated with normal saline and thrombin (10 U/ml) were used as control groups.

#### Statistical Analysis

All results were described as means ± standard deviation (SD). Statistical analysis of all data was performed using SPSS 19 (IBM, Statistic Package for Social Science). A value of p < 0.05 was considered as being statistically significant.

#### RESULTS AND DISCUSSIONS

#### Structure Characterization of COS

The FTIR spectra of the two COS are shown in **Figure 2**, both of which displayed characteristic absorptions that are similar to that of chitosans with high DD as reported previously. The absorption band at around 3,430 cm−<sup>1</sup> is attributed to the N-H and O-H stretching vibrations (Chen et al., 2016; Wu et al., 2017). The peaks at 2,880 and 2,930 cm−<sup>1</sup> are assigned to C-H stretching vibrations, and the peaks at 1,650, 1,630, and 1,320 cm−<sup>1</sup> are, respectively, assigned to amide I, II, and III. The band at 1,380 is corresponds to C-H bending and C-H stretching vibrations, and the band in the range 1,150– 890 are assigned to the characteristics of its polysaccharide structure. As previous literature reported, the COS prepared by chemical degradation approaches such as oxidative degradation with peroxide hydrogen shown a weakened amide Iband due to the H-abstraction at C-1 and C-2 during degradation that both led to the partly deamination. However, the two COS samples in this study obtained by enzymatic hydrolysis remain exhibited a strong amide Iband, which indicated that the two COS samples are both degradation products with high DD.

The chemical structure of the two COS samples was further studied by <sup>1</sup>H NMR. As shown in **Figure 3**, the two samples exhibited similar <sup>1</sup>H NMR spectrum. According to previous literatures, signals at 2.0 and 3.1 ppm are assigned to protons of CH3 and H-2 of GlcN, respectively. In the low field, the signals at 3.5–4.0 ppm are assigned to H-3, 4, 5, 6 of GlcN and H-2, 3, 4, 5, 6 of GlcNAc, and the signals at 4.5 ppm are attributed to H-1 of GlcN (Trombotto et al., 2008). Compared to <sup>1</sup>H NMR spectrum of the chitosan as previous reported, there is no significant

difference for the two samples. In addition, the DD values for the two samples were calculated by <sup>1</sup>H NMR spectrum according to equation (1). The calculated DD were 88.4 and 87.4% for COS-5K and COS-3K, respectively, which is consistent with the results of the FTIR.

#### Red Blood Cells

In order to investigate the effect of two COS on RBCs, three main parameters of RBCs including hemolysis, deformability and aggregation were studied. Hemolysis is a crucial and direct parameter evaluating the safety of biomaterials and defined as the percentage of RBCs lysis when the biomaterial interacting with RBCs suspension (Fernandes et al., 2008). According to ISO10993-5, percentage of hemolysis induced by exogenous materials under 5% was considered as a low risk (Weber et al., 2018). As shown in **Figure 4**, percentages of hemolysis presented in a dose-dependent manner. As the concentration increasing from 0.1 to 1 mg/ml, the hemolysis increased from 0.30 ± 0.10 to 1.15 ± 0.14% (p < 0.01) in COS-3K group and from 0.30 ± 0.06 to 1.42 ± 0.30% (p <0.01) in COS-5K group. When the concentration below 0.1 mg/ml, the hemolysis in both COS groups had no significant difference with that in NS group. We also observed that hemolysis in COS-5K group was significantly higher than that in COS-3K group when the concentration was above 0.1 mg/ml, indicating that the hemolysis of COS also presented as a MW-dependent manner.

As for deformability of RBCs, an index representing the ability of RBCs to change their biconcave disk shape against the exogenous force from vascular walls, we used four different sheer forces in the tests (Zhao et al., 2010). As shown in **Figures 5A–D**, as sheer force increasing from 0.39 to 7.7 Pa (corresponding sheer rate 50 to 1,000), elongation index was generally increased except in DW group with a stable and lowest EI due to containing no intact RBCs. Under sheer rate of 50 and 100, EI presented a dose dependent manner where EI was significant higher at 0.5 and 1 mg/ml than 0.01 and 0.1 mg/ml. Though under sheer rate of 200, there was no statistical difference on EI between 0.5, 1, and 0.01, 0.1 mg/ml in COS-3K group, EI in COS-5K group had the same trend as under 50 and 100 sheer rate.

Results of erythrocytes aggregation was in accordance with hemolysis. As shown in **Figure 6**, at the concentration of 1 mg/ml, RBCs displayed biconcave shape and some erythrocytes rouleaux defined as reversible aggregation were observed compared to normal distribution of RBCs in NS group. However, with the concentration increasing to 5 mg/ml, RBCs all aggregated forming erythrocytes clusters which are irreversible aggregation and the normal shape of erythrocytes was absent. As previous report, amidocyanogen of chitosan could neutralize the negatively charged neuraminic acid residues located on the surface of RBCs (Fernandes et al., 2008). Likely, the positive charge of COS can change the distribution of charge and break the electrokinetic balance on erythrocyte surface leading to aggregation in a dose-dependent manner.

#### Coagulation

Effect on coagulation is an essential property of biomaterial due to both anti-coagulation and pro-coagulation can cause fetal

FIGURE 5 | Deformability of RBC detected by a laser-diffraction Ektacytometer system under shear rate of 50 (A), 100 (B), 200 (C), 1,000 (D) with the presence of COS with different concentration. NS, normal saline; DW, diluted water. \**p* < 0.05; \*\**p* < 0.01.

complications or side effects on human such as bleeding and thrombosis formation (Zhang et al., 2013). In order to study effect on coagulation, APTT, PT, TT, and Fib were involved in the tests. It is defined as disordered hemostatic property of human blood when APTT changes more than 10 s or PT/TT changes more than 3 s (Yang et al., 2008). In our study, as shown in **Figure 7A**, high concentration (0.5 and 0.1 mg/ml) COS prolonged APTT compared to NS group (p < 0.01) in a dose-dependent manner and the maximum APTT even exceeded the machine available value (>245 s). In contrast, TT was normal in all COS group (**Figure 7B**). For PT showing in **Figure 7C**, though it was slightly longer at 0.1 and 0.5 mg/ml group, all the PT values were in normal range when considering 3 s mobility scale. Similar to TT, concentration of Fib in COS groups was equal to that in NS group (**Figure 7D**). Within our expectation, heparin (0.75 IU/ml) group (TT positive control) showed a prolonged TT value which exceeded 169 s.

APTT and PT, respectively, reflect the status of intrinsic and extrinsic blood coagulation, while TT is used to check the effect on the conversion from fibrinogen to insoluble fibrin induced by thrombin or whether there exist an anticoagulant when APTT and PT are prolonged. Our study indicated COS influenced intrinsic coagulation pathway due to prolonging APTT while PT, TT, and Fib were in normal range. To the best of our knowledge, it is the first time to report that COS has anticoagulation activity through blocking the intrinsic coagulation pathway, which is different from chitosan with pro-coagulant activity. On the other hand, it should be noted that the anticoagulation activity of the COS is mild compared to other anticoagulants such as alginate sulfates, whose APTT was prolonged by 180 s in the

FIGURE 7 | Analysis of influence on coagulation caused by COSs. (A) APTT of COSs and heparin. (B) TT of COSs and heparin. (C) PT of COSs and heparin. (D) Concentration of Fib. \*\**p* < 0.01 vs. NS group; # means the parameter exceeded machine available. NS, normal saline; HP, heparin.

concentration of 80µg/ml, while the APTT of the COS was around 80 s in the concentration of 0.1 mg/ml.

that determined whether the chitosan induced complement activation.

#### Complement

Complement activation is an important indicator of hemocompatibility of biomaterials. Complement system is an innate part of immune system and can be activated via three pathways, classical pathway, alternative pathway and lectin pathway (Coulthard and Woodruff, 2015). Once complement system activated, C3 cleaves to fragments C3a and C3b, and leads to its downstream generation of C5a. As shown in **Figure 8**, at low concentration, COS seemed to induce more C3a than NS group than that at high concentration. However, the difference has no statistical significance.

Chitosan is suggested to activated complement system by deplete or absorb C3 and C5 proteins from serum (Benesch and Tengvall, 2002). Interestingly, in our study, COS demonstrated relative complement compatibility that C3a and C5a in all groups was in a normal range. Similar to our findings, Marchand et al. reported that chitosan with 80% and 95% DD do not activate complement system for that C3 and C5 bind to chitosan as intact proteins (Marchand et al., 2010). Hence, DD might be a significant factor

#### Platelet

CD62P, a platelet activation marker, was used to investigate the effect of COS on platelet. As shown in **Figures 9A,B**, with presence of COS at 0.1 and 0.5 mg/ml, expression of CD62P was in normal range (p > 0.05) indicating platelet was not activated by COS. Heparin demonstrated a strong ability to stimulate platelet that CD62P was remarkable high in heparin group. In terms of platelet aggregation, reduced aggregation was observed in all COS groups (p < 0.05 for all COS) compared to normal saline (**Figure 9C**). Moreover, there was no statistical significant difference between two concentrations or two COSs.

Platelet was a vital component in blood participating in coagulation cascade and provided large scale of coagulation factors, such as platelet factor 3 and factor Va, when activated leading to generate platelet micro particles (Gorbet and Sefton, 2004). In coagulation process, induced platelet micro particles promote platelet adherence to fibrin resulting in congregation of platelet and anti-bleeding function. In accordance to coagulation results, COS did not activate platelet and inhibited platelet aggregation indicating it played as an anti-coagulant. Under such

concentrations of 0.5 and 0.1 mg/ml, COS with two MW has no difference in ability of inhibiting platelet.

# DISCUSSION OF MECHANISM

Chitosan derived from chitin is a cationic polysaccharide that is widely used as a hemostasis, wound dressing, or antibacterial material in clinic (Kumar et al., 2004; Muzzarelli et al., 2005). However, application in blood-contacting is occluded by its inferior compatibility with blood components such as procoagulant function and complement activation. Interesting, here it was found that COS exhibited quite different effects on human blood components comparable to chitosan (**Figure 10**). To explain this phenomenon, a mechanism was proposed as follows based on literatures and our results.

The procoagulant function of chitosan is due to its cationic nature that induced by positively charged amino groups, in which these groups induction of fibrinogen adsorption, activation of platelets, releasing of coagulation factors, and then trigger extrinsic coagulation pathway (Benesch and Tengvall, 2002; Zhang et al., 2013). However, when chitosan was degraded into COS, the decreased MW resulted in decreasing of local positive charge density and thus weakened the electrostatistic interactions between the positively charged amino groups and the blood components in comparison to that of chitosan. In the case of RBCs, the electrostatic interactions results in membrane damage, hemolysis as well as aggregation, while low risk of hemolysis, deformability and aggragation were observed for COS-3k and 5k resulting from the decreased MW and positive charge density. Similarly, the decreased positive charge density may reduce the procoagulant function of chitosan as well. However, our results shown that the both two COS exhibited a mild anticoagulation activity in a dose-dependent manner. The anticoagulation activity probably related with the hydroxyl groups along the molecular chains, since it has been demonstrated that hydroxyl groups are capable of anticoagulation as previously reported (Sperling et al., 2005).

# CONCLUSION

In summary, the molecular structure of the two COS with different MW was characterized by FTIR and <sup>1</sup>H NMR, and then the effects of the COS on the human blood components, including RBCs, coagulation system, complement, and platelet, were studied in this work. The results indicated that: (i) COS exhibited a low risk of hemolysis in a dose and MW dependent manner and the irreversible aggregation was observed in their high concentration; (ii) COS has a mild anticoagulation activity through blocking the intrinsic coagulation pathway; (iii) COS has no effect on complement activation in C3a and C5a; (iv) COS has no effect on platelet activation while inhibition of platelet aggregation was evident. Finally, the mechanism of action was discussed and proposed.

#### AUTHOR CONTRIBUTIONS

XG: results discussion, writing the article, and hemocompatibility evaluation. TS: results discussion, hemocompatibility evaluation. RZ: hemocompatibility evaluation, statistic analysis. LM: results discussion. CY: evaluating the results of study. MT: article design, NMR spectroscopy, and FTIR characterization. HL and CW: results discussion, statistic analysis.

#### REFERENCES


#### ACKNOWLEDGMENTS

This work is supported by National Natural Science Foundation of China (No. 81401528, 51403238), The National Key Research and Development Program of China (No. YS2018YFA010082), Sichuan Province Science and Technology Key R&D Project (No. 2018SZ0029, 2018SZ0100, 2015SZ0051), and 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (No. ZY2016102, and ZY2016203).

cell morphology and membrane protein structure. Biomacromolecules 9, 3346–3352. doi: 10.1021/bm800622f


in human umbilical vein endothelial cells. Pharmacol. Res. 59, 167–175. doi: 10.1016/j.phrs.2008.12.001


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Guo, Sun, Zhong, Ma, You, Tian, Li and Wang. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# A Robust Nanoparticle Platform for RNA Interference in Macrophages to Suppress Tumor Cell Migration

Shi Liang1,2,3, Junmeng Zheng1,2, Wei Wu1,2,3, Quan Li1,2,3, Phei Er Saw1,2 , Jianing Chen1,2,3, Xiaoding Xu1,2 \*, Herui Yao1,2,3 \* and Yandan Yao1,2,3 \*

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China Bingyang Shi, Macquarie University, Australia Guoqing Pan, Jiangsu University, China

#### \*Correspondence:

Xiaoding Xu xuxiaod5@mail.sysu.edu.cn Herui Yao yaoherui@126.com Yandan Yao yaoyand@mail.sysu.edu.cn

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 28 October 2018 Accepted: 30 November 2018 Published: 14 December 2018

#### Citation:

Liang S, Zheng J, Wu W, Li Q, Saw PE, Chen J, Xu X, Yao H and Yao Y (2018) A Robust Nanoparticle Platform for RNA Interference in Macrophages to Suppress Tumor Cell Migration. Front. Pharmacol. 9:1465. doi: 10.3389/fphar.2018.01465 <sup>1</sup> Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Medical Research Center, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China, <sup>2</sup> RNA Biomedical Institute, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China, <sup>3</sup> Breast Tumor Center, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China

Macrophages are one of the most abundant immune cells in the solid tumor and their increased density is associated with the specific pathological features of cancers, including invasiveness, metastasis, immunosuppression, neovascularization, and poor response to therapy. Therefore, reprogramming macrophage behavior is emerging as a promising therapeutic modality for cancer treatment. RNA interference (RNAi) technology is one of the powerful strategies for the regulation of macrophage activities by silencing specific genes. However, as polyanionic biomacromolecules, RNAi therapeutics such as small interfering RNA (siRNA) cannot readily cross cell membrane and thus specific delivery vehicles are required to facilitate the cytosolic siRNA delivery. Herein, we developed a robust nanoparticle (NP) platform for efficient siRNA delivery and gene silencing in macrophages. This NP platform is composed of biodegradable poly (ethylene glycol)-b-poly (ε-caprolactone) (PEG-b-PCL), poly (ε-caprolactone)-b-poly (2-aminoethyl ethylene phosphate) (PCL-b-PPEEA), and PCL homopolymer. We chose CC-chemokine ligand 18 (CCL-18) as a proof of concept therapeutic target and our results demonstrate that the CCL-18 silencing in macrophages can significantly inhibit the migration of breast cancer cells. The successful regulation of the macrophage behavior demonstrated herein shows great potential as an effective strategy for cancer therapy.

Keywords: macrophages, RNAi, nanoparticle, cancer therapy, siRNA delivery

# INTRODUCTION

Macrophages are important cells of immune system with two major phenotypes, i.e., proinflammatory phenotype (M1) and anti-inflammatory phenotype (M2) (Mantovani et al., 2008; Noy and Pollard, 2014; Ostuni et al., 2015). In solid tumors, tumor-associated macrophages (TAMs) are one of the most abundant cell types (up to 50% of the tumor mass) and are present at all

stages of tumor progression (Caillou et al., 2011; Ginhoux et al., 2015; Parayath et al., 2018). Numerous clinical and epidemiological studies have demonstrated that TAMs are primary M2-like macrophages (Gordon and Martinez, 2010; Sica and Mantovani, 2012; Bronte and Murray, 2015) and their increased density is associated with the specific pathological features of cancers, including invasiveness, metastasis, immunosuppression, neovascularization, and poor response to therapy (Qian and Pollard, 2010; McAllister and Weinberg, 2014; Bronte and Murray, 2015). Therefore, macrophages represent an important therapeutic target and strategies that can effectively regulate undesirable macrophage activities are always pursued for future cancer therapy.

One of the promising strategies for the regulation of macrophage activities is using RNA interference (RNAi) technology to silence specific genes (Aouadi et al., 2009; Kortylewski et al., 2009; Yu et al., 2013). Since its discovery in 1998, RNAi technology has demonstrated significant potential for disease treatment by silencing the expression of target gene(s), especially those encoding "undruggable" proteins (Melnikova, 2007; Burnett and Rossi, 2012; Kanasty et al., 2013; Zuckerman and Davis, 2015). The key challenge is the safe and effective delivery of RNAi therapeutics such as small interfering RNA (siRNA) to aberrant macrophages (e.g., TAMs). Due to its polyanionic and macromolecular characteristics, naked siRNA cannot readily cross the cell membrane and thus requires specific delivery vehicles to facilitate its intracellular uptake and cytosolic delivery for bioactivity (Whitehead et al., 2009; Zhang et al., 2012; Shi et al., 2017; Xu et al., 2017b). Over the past decade, nanoparticles (NPs), which present the advantage of preferential and selective accumulation at tumor sites via the enhanced permeation and retention (EPR) effect, have been widely used for cancer treatment (Chen et al., 2017; Zeng et al., 2017; Liu et al., 2018; Xiao et al., 2018). Up to now, numerous innovative NPs have been developed to enhance the siRNA delivery efficacy (Tseng et al., 2009; Xu et al., 2016, 2017a; Cheng et al., 2017; Saw et al., 2018). However, a substantial number of these NPs are designed to directly target tumor cells. At present, modest effort has been made to develop RNAi NPs for the modulation of undesirable macrophage activities.

Herein, we developed a robust RNAi NP platform for the efficient regulation of macrophage activities. This NP platform is composed of biodegradable poly (ε-caprolactone)-bpoly (2-aminoethyl ethylene phosphate) (PCL-b-PPEEA), poly (ethylene glycol)-b-poly (ε-caprolactone) (PEG-b-PCL), and PCL homopolymer (**Figure 1**). Through optimizing the NP size by adjusting the formulation, we demonstrated that larger size NPs can deliver siRNA and silence target gene in macrophages with higher efficacy. As a proof-of-concept, we chose CC-chemokine ligand 18 (CCL-18) as a therapeutic target and evaluated the influence of CCL-18 silencing on the macrophage activities. CCL-18 is a key factor secreted by TAMs to induce cancer cell epithelial-mesenchymal transition (EMT), enhance breast cancer metastasis, and reduce patient survival (Chen et al., 2011; Su et al., 2014; Nie et al., 2017). Our results show that the optimal NP platform can efficiently silence the CCL-18 expression in macrophages, leading to significant inhibition of breast cancer cell migration (**Figure 1**).

# MATERIALS AND METHODS

#### Materials

Methoxyl-poly (ethylene glycol) (Meo-PEG114-OH, M<sup>n</sup> = 5000), phorbol myristate acetate (PMA), N-2-hydroxyethylpiperazine-N 0 -2-ethanesulfonic acid buffered saline (HEPES), stannous octoate [Sn (Oct)2], RNase A, and heparin sulfate were acquired from Sigma-Aldrich and used directly. ε-Caprolactone (CL) was provided by Sigma-Aldrich and distilled before use. The poly (ε-caprolactone) homopolymer with 34 repeating units (PCL34) and polydispersity of 1.21 was synthesized according to the previous report (Wang et al., 2006). The block copolymers, methoxyl-poly (ethylene glycol)-block-poly (ε-caprolactone) (mPEG114-b-PCL41) and poly (ε-caprolactone)-block-poly (2-aminoethyl ethylene phosphate) (PCL25-b-PPEEA17), were synthesized according to our previous reports (Wang et al., 2013; Liang et al., 2015). The degree of polymerization of each repeating unit was calculated based on proton nuclear magnetic resonance (1HNMR) analysis. Dulbecco's modified Eagle's medium (DMEM), 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazoliumbromide (MTT), fetal bovine serum (FBS), and trypsin were purchased from Gibco BRL. Lipofectamine 2000 (Lipo) RNAi MAX transfection kit and DAPI were provided by Invitrogen Corp. Fluorescent dye (Cy5) labeled CCL-18 siRNA (Cy5-siCCL-18) and negative control siRNA (siNC) were acquired from Suzhou Ribo Life Science Co., The siRNA sequences are as follows: siNC, 5 0 -TTG GGA AAA TGA GTG GTT dTdT-3<sup>0</sup> (sense) and 5 0 -AAC CAC TCA ACT TTT TCC CAA dTdT-3<sup>0</sup> (antisense); siCCL-18, 5<sup>0</sup> -ACA AGU UGG UAC CAA CAA ATT-3<sup>0</sup> (sense) and 5<sup>0</sup> -UUU GUU GGU ACC AAC UUG UGC -3<sup>0</sup> (antisense). The fluorescent dye was labeled at the 5<sup>0</sup> -end of the sense strand of siCCL-18. All other organic solvents or reagents were analytical grade and used without further purification.

#### Methods

#### Preparation and Characterizations of Nanoparticles (NPs)

The NPs with different sizes were prepared by using the classic nanoprecipitation method. The polymers, mPEG-b-PCL (10 mg), PCL-b-PPEEA (10 mg), and PCL (50 mg) were, respectively, dissolved in 1 mL of acetonitrile and methyl alcohol (v/v, 1:1). Subsequently, mPEG-b-PCL and PCL-b-PPEEA were mixed in a molar ratio of 1.5:1 and then added dropwise to 10-fold volume of deionized water which was under vigorously stirring. After stirring for another 20 min, the NP suspension was transferred into a rotary evaporator to remove the organic solvent. The final NP suspension was dispersed in deionized water at a concentration of 1 mg/mL. To adjust the NP size, different amount of PCL was mixed with the mixture of mPEG-b-PCL and PCLb-PPEEA (molar ratio, 1.5:1) and the resulting NPs were

prepared according to the same method described above. The size distribution and zeta potential of the NPs were examined by dynamic light scattering (DLS, Malvern Instruments Corporation). The morphology of the NPs was observed by

#### Gel Retardation Assay

BioTWIN).

The NPs prepared above were mixed with the siCCL-18 aqueous solution (20 mM) at different N/P ratios. After incubating at room temperature for 20 min, the formed NP/siCCL-18 complexes were electrophoresed on a 1% agarose gel at 120 mV for 10 min in pH 8.3

transmission electron microscopy (TEM, Tecnai G<sup>2</sup>

Tris/borate/EDTA buffer (89 mM Tris, 89 mM boric acid, 2 mMEDTA). The siRNA bands were visualized with ethidium bromide staining under a UV transilluminator at a wavelength of 365 nm. Naked siCCL-18 was used as control.

#### In vitro siRNA Release

The NP/Cy5-siCCL-18 complexes at an N/P ratio of 10:1 were prepared according the same method described above and then suspended in pH 7.4 PBS solution at a siRNA concentration of 200 nM. Subsequently, the siRNA loaded NP suspension was transferred to a dialysis device (MWCO 100 kDa) that was immersed in pH 7.4 PBS solution at 37◦C. At a predetermined

can knock down CCL-18 expression (e) and thus CCL-18 secretion from the macrophages would be blocked (f), leading to the inhibition of tumor migration (g).

Spirit

interval, 5 µL of the NP suspension was withdrawn and mixed with 20-fold DMSO. The fluorescence intensity of Cy5-siCCL-18 was determined by a microplate reader.

#### Evaluation of the Stability of NP/siRNA Complexes

The NP/siCCL-18 complexes were prepared at an N/P ratio of 10:1 and then dispersed in DMEM containing 10% FBS at 37◦C. At a predetermined interval, the size distribution of the NP/siCCL-18 complexes was examined by DLS.

#### Cell Culture

THP-1 cells (human monocytic leukemia cell line) and human breast cancer cells (MDA-MB-231 cell line) were obtained from the American Type Culture Collection (ATCC) and incubated in DMEM medium with 10% FBS at 37◦C in a humidified atmosphere containing 5% CO2.

#### Construction of M2-Like Macrophages

THP-1 cells were incubated in DMEM medium containing 10% FBS and PMA (10 ng/mL) at for 8 h. Subsequently, the medium was replaced by fresh DMEM medium containing interleukin-4 (IL-4, 20 ng/mL). After 48 h incubation, the cells will differentiate into M2-like macrophages and the mRNA level of CCL-18 was examined using quantitative reverse transcription real-time PCR (qRT-PCR).

#### Confocal Laser-Scanning Microscope (CLSM)

After successful construction of the M2-like macrophages, the cells (50,000 cells) were seeded in round disks and incubated in 2 mL of DMEM medium containing 10% FBS. After 24 h incubation, the medium was replaced and the NP/Cy5 siCCL-18 complexes were added to the disks at a siRNA concentration of 200 nM. After 4 h incubation, the medium was removed and the cells were washed with pH 7.4 PBS solution thrice. Finally, the nuclei were stained with DAPI and the cells were viewed under a Carl Zeiss LSM 710 CLSM.

#### Flow Cytometry

The M2-like macrophages were seeded in six-well plate (50,000 cells per well) and incubated in 2 mL of DMEM medium containing 10% FBS. After 24 h incubation, the medium was replaced and the NP/Cy5-siCCL-18 complexes were added at a siRNA concentration of 200 nM. After 4 h incubation, the medium was removed and the cells were washed with pH 7.4 PBS solution thrice. Finally, the cells were collected for flow cytometry analysis using a BD FACSCalibur flow cytometer.

#### Cytotoxicity

The cytotoxicity of the NPs was evaluated by using MTT assay. The M2-like macrophages were seeded in 96-well plates at 5,000 cells per well and incubated in 100 µL of DMEM medium containing 10% FBS. After 24 h incubation, the medium was removed and different amounts of the NPs suspended in the culture medium were added. After 48 h incubation, the medium was removed and the cell viability was examined using the MTT assay according to the manufacturer's protocol.

#### CCL-18 Silencing

The M2-like macrophages were seeded in six-well plate (50,000 cells per well) and incubated in 2 mL of DMEM medium containing 10% FBS. After 24 h incubation, the medium was replaced and the NP/siCCL-18 complexes were added at different siRNA concentrations. After 48 h incubation, the medium was removed and the cells were washed with pH 7.4 PBS solution thrice. The intracellular mRNA was isolated and the mRNA level of CCL-18 was examined using qRT-PCR.

#### Inhibition of Migration

MDA-MB-231 cells (50,000 cells) were seeded in round disks and incubated in 2 mL of DMEM medium containing 10% FBS. After 24 h incubation, the cells in the predesigned area of the disks were removed using tips. After washing the cells with PBS thrice,

the cells were incubated with the conditioned medium from the macrophages, in which the CCL-18 has been silenced by the NP-180/siCCL-18 at a siRNA concentration of 400 nM. After 48 h incubation, the medium was removed and the cells were viewed under optical microscope after washing with PBS thrice.

# RESULTS AND DISCUSSION

#### Preparation and Characterizations of Nanoplatform

Starting from the commercial available mPEG-OH and ε-CL, we employed ring-opening polymerization (ROP) to synthesize the mPEG-b-PCL and PCL homopolymer (Sun et al., 2008). The amphiphilic polymer PCL-b-PPEEA was also prepared by ROP method (Sun et al., 2008), in which the cationic PPEEA segment was used to complex siRNA via electrostatic interaction. By mixing these three polymers in acetonitrile and methyl alcohol (v/v, 50:50) followed by the addition to deionized water, well-defined NPs can be formed with spherical morphology (**Figure 2A**). In this self-assembly system, the amphiphilic PEG-b-PCL and PCL-b-PPEEA can spontaneously self-assemble into NPs with hydrophobic PCL chains embedded in the cores and hydrophilic PEG and PPEEA chains positioned on the surface that can, respectively, stabilize the NPs and complex

negatively charged siRNA. Moreover, with the increasing feed amount of PCL homopolymer, the average size of the resulting NPs increases from ∼40 to 180 nm (**Figure 2B**). The possible reason is that the increased percentage of PCL homopolymer in the NPs induces the size increase of their hydrophobic inner cores (Mao et al., 2014; Liang et al., 2015). In this work, we prepared four types of NPs with the size of 40, 90, 130, and 180 nm (respectively, denoted NP-40, NP-90, NP-130, and NP-180) to evaluate their ability to deliver CCL-18 siRNA (siCCL-18) for gene silencing in macrophages.

We next used gel retardation assay to evaluate the siRNA loading capacity of the NPs. As shown in **Figure 3**, due to the presence of cationic PPEEA segment on the surface, all the NPs can effectively load siCCL18 at an N/P ratio of 10:1, without apparent size change after siCCL18 loading. Furthermore, all the siCCL18 loaded NPs show good stability in 10% FBScontaining cell culture medium (**Figure 4**). More importantly, in comparison with the naked siCCL18, the use of the NPs can protect the siCCL18 from degradation by RNase. As shown in **Figure 5A**, with the protection by the NPs, the addition of RNase does not induce the siRNA degradation and the loaded siRNA can still bind with the NPs even under electric field. In contrast, without the NP protection, the naked siRNA has been degraded by the RNase and thus no siRNA band can be observed in the gel retardation assay experiment. All these results demonstrate the NP platform developed in this work shows a strong ability to load and protect the siCCL-18, which would thus ensure its bioactivity when used for regulation of macrophage activities. Notably, since we only varied the feed amount of hydrophobic PCL homopolymer to adjust the NP size, while the other two polymers (Meo-PEG-b-PCL and PCLb-PPEEA) remain constant in the NP formulations, all the NPs showed the similar ability to load and release the siRNA (**Figure 5B**).

#### Evaluation of CCL-18 Silencing

After validation of the siRNA loading ability of the NPs, we next examined whether these NPs can deliver siRNA to macrophages for gene silencing. THP-1, a human monocytic leukemic cell line, was used to construct M2 type macrophage-like cells through treatment with PMA and interleukin-4 (IL-4) (Liang et al., 2015). As shown in **Figure 6A**, the high expression of CCL-18, a well-known chemokine generated by M2-type macrophages (Chen et al., 2011), demonstrated the success of THP-1 cells differentiation from monocytes to M2-type macrophages (denoted THP-1 originated macrophages). With this encouraging result, we subsequently encapsulated dye-labeled siCCL-18 (Cy5-siCCL-18) into the NPs and investigated their cellular uptake by the differentiated THP-1 macrophages obtained above. From the flow cytometry (FACS) analysis shown in **Figure 6B**, the THP-1-originated macrophages show higher uptake of the Cy5-siCCL18 loaded NPs compared to the naked siRNA. The intracellular mean fluorescence intensity (MFI) is at least 1.3-fold stronger than that of the macrophages treated with the naked siRNA. Among these NP formulations, the uptake of NP-180 is highest and the possible reason is that macrophages are apt to internalize foreign materials with large size (Tabata and Ikada, 1988; Champion et al., 2008). **Figure 6C** shows the fluorescent images of macrophages incubated with the Cy5-siCCL-18 loaded NP-180. Similar as the results of FACS analysis, the THP-1-originated macrophages show strong ability to internalize the siRNA loaded NP-180 as seen with bright red fluorescence and these NPs are mainly dispersed in the cytoplasm where siRNA functions (Whitehead et al., 2009). Although the THP-1 originated macrophages show higher cellular uptake of the NP-180, these NPs do not induce apparent cytotoxicity (**Figure 7A**).

Based on the results of FACS analysis and toxicity assay, the NP-180 shows higher uptake by the THP-1-originated macrophages with negligible toxicity. Therefore, we chose this NP platform as siCCL-18 delivery tool to examine its gene silencing efficacy in macrophages. As shown in **Figure 7B**, the NP-180 can indeed transport siCCL-18 into the THP-1 originated macrophages and thereby down-regulate CCL-18 expression. Compared to the macrophages without any treatment (Control), the mRNA level of CCL-18 is down-regulated by 20% at a siCCL-18 concentration of 100 nM and more than 70% of CCL-18 is suppressed at a siCCL-18 concentration of 400 nM.

#### Evaluation of the Inhibition of Migration

It is known that CCL-18 is an important factor secreted by TAMs that can enhance breast cancer metastasis and therefore reduce patient survival (Chen et al., 2011; Su et al., 2014). Previous reports demonstrate that CCL-18 released by TAMs in breast cancer promotes the invasiveness of cancer cells by triggering integrin clustering and enhancing their adherence to extracellular matrix and silencing the CCL-18 expression can inhibit the consistent activation

FIGURE 5 | (A) Gel retardation assay of the siRNA loading capacity of the NP-40, NP-90, NP-130, and NP-180; (B) Cumulative siRNA release profile of the complexes formed between Cy5-siCCL-18 and the NP-40, NP-90, NP-130, or NP-180.

of downstream signal pathways to prevent the migration of breast cancer (Chen et al., 2011). Therefore, we finally simulated tumor microenvironment in vitro by incubating breast cancer cell line (MDA-MB-231) with the medium that was obtained from THP-1-originated macrophage (denoted conditioned medium), and examined the influence of CCL-18 silencing on the behavior of MDA-MB-231 cells. As shown **Figure 7C**, due to presence of secreted CCL-18 by macrophages, the addition of conditioned medium to MDA-MB-231 cells can enhance their migration compared to the cells incubated in normal medium. In contrast, after using siCCL-18 loaded NP-180 to suppress CCL-18 expression in macrophages, the migration of MDA-MB-231 cells is significantly inhibited when incubated with the conditioned medium. This result is consistent with our previous reports (Chen et al., 2011; Su et al., 2014; Nie et al., 2017), and highlights the importance of CCL-18 to the breast cancer cell migration.

# CONCLUSION

We have developed a robust RNAi NP platform for efficient gene silencing in M2-type macrophages. Through varying the percentage of hydrophobic PCL homopolymer in the NP formulation, we successfully constructed four types of NPs with different sizes and systemically evaluated their siRNA loading ability and gene silencing efficacy. Experimental results demonstrate that the NP platform with larger size (NP-180) shows higher cellular uptake and efficient CCL-18 silencing in macrophages, leading to efficient inhibition of the breast cancer cell migration. Notably, this NP platform may passively target the tumor tissues via the EPR effect. However, the NP size is very large (∼180 nm) and will affect the therapeutic effect if used for in vivo regulation of macrophage behaviors. We are currently optimizing the NP formulation and small size NPs will be developed in the future for in vivo regulation of macrophage behaviors and cancer treatment.

#### AUTHOR CONTRIBUTIONS

fphar-09-01465 December 12, 2018 Time: 14:19 # 9

SL, HY, YY, and XX conceived and designed the experiments. SL, JZ, WW, QL, and JC performed the experiments. SL, PS, XX, HY, and YY analyzed the data and co-wrote the paper.

#### REFERENCES


#### FUNDING

This work was supported by grants from the Natural Science Foundation of China (81874226, 81272897, 81772837, 81372819, and 81702618), the Thousand Talents Program for Distinguished Young Scholars, the National Key Research and Development Program of China (2017YFA0106300 and 2016YFC1302300), the Guangdong Science and Technology Department (2014A050503029, 2015B050501004, and 2016B030229004), the Natural Science Foundation of Guangdong Province (2017A030313828), and the funding from Guangzhou Science and Technology Bureau (201704020131).

nanoparticle-mediated CDK4 siRNA delivery. Mol. Ther. 22, 964–973. doi: 10.1038/mt.2014.18



mannosylated polymeric micelles. Mol. Pharm. 10, 975–987. doi: 10.1021/ mp300434e


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Liang, Zheng, Wu, Li, Saw, Chen, Xu, Yao and Yao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Enhanced Anti-tumor of Pep-1 Modified Superparamagnetic Iron Oxide/PTX Loaded Polymer Nanoparticles

#### Baoyan Wang<sup>1</sup>† , Weijun Wu<sup>1</sup>† , Hongjin Lu<sup>2</sup> , Zhi Wang<sup>3</sup> and Hongliang Xin<sup>2</sup> \*

<sup>1</sup> Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, China, <sup>2</sup> Department of Pharmaceutics, School of Pharmacy, Nanjing Medical University, Nanjing, China, <sup>3</sup> Department of Pharmacy, The Second Affiliated Hospital of Nanjing Medical University, Nanjing, China

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Quanyin Hu, Massachusetts Institute of Technology, United States Veronica Lazar, University of Bucharest, Romania

#### \*Correspondence:

Hongliang Xin xhl@njmu.edu.cn †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 29 October 2018 Accepted: 21 December 2018 Published: 22 January 2019

#### Citation:

Wang B, Wu W, Lu H, Wang Z and Xin H (2019) Enhanced Anti-tumor of Pep-1 Modified Superparamagnetic Iron Oxide/PTX Loaded Polymer Nanoparticles. Front. Pharmacol. 9:1556. doi: 10.3389/fphar.2018.01556 Superparamagnetic iron-oxide nanoparticle (SPION) has gained tremendous attention for drug delivery applications due to their unique properties. In this study, we developed a dual targeted delivery system with paclitaxel (PTX) and SPION co-loaded PLGA nanoparticles, modified with Pep-1 peptide (Pep-NP-SPION/PTX), to achieve magnetic targeting and active targeting for tumor treatment. SPION was synthesized by a coprecipitation method and was then encapsulated with PTX simultaneously into PLGA nanoparticles. After that, the non-complex was conjugated with Pep-1 through chemical modification. The resulting Pep-NP-SPION/PTX showed a spherical morphology and an average size of 100 nm. The enhancement cellular uptake of Pep-NP-SPION was demonstrated in in vitro through cell experiments. The IC<sup>50</sup> value of Pep-NP-SPION/PTX and NP-SPION/PTX was determined to be 10.2 and 19.4 µg/mL, respectively. A biodistribution study showed that obvious higher accumulations of Pep-NP-SPION was observed in tumors, compared with that of non-targeting nanocomposites. Moreover, under the condition of a magnetic field, both NP-SPION and Pep-NP-SPION exhibited much higher tumor distribution. Furthermore, Pep-NP-SPION/PTX presented desirable in vivo anti-tumor effects based on active targeting and magnetic targeting characteristics. Altogether, Pep-NP-SPION/PTX can offer magnetic targeting and receptor mediated targeting to enhance the anti-tumor outcome.

Keywords: tumor, drug delivery system, superparamagnetic iron oxide nanoparticle, interleukin-13 receptor α2, dual targeting nanocarrier

# INTRODUCTION

Current tumor chemotherapy still faces a major problem with the of lack of selectivity of drugs on tumor cells, which leads to a narrow therapeutic index of most anti-tumor drugs (Wong et al., 2016). As a result, the achievement of an adequate therapeutic effect requires a high concentration of antitumor drugs, which in turn enhances the systematic toxicity. Thus, in order to reduce the toxicity of a chemotherapy drug at its minimum dose, specific nanoparticulate drug delivery systems, such as nanoparticle, liposome and polymeric micelle, could provide a non-invasive treatment strategy due to its passive targeting properties (Huang et al., 2016; Muralidharan et al., 2017;

Roberts et al., 2017). However, these conventional drug delivery systems lack the capability of active targeting for tumor section.

Recently, superparamagnetic iron-oxide nanoparticles (SPION) consisting of cores made of iron oxides, have been considered to be attractive in cancer theranostic applications, since they can be delivered to the required tissue through an external magnetic field (Hachani et al., 2016). Due to the superparamagnetism, a high saturation field and a high field irreversibility, SPION has been widely used for various biological applications (Li et al., 2016). Moreover, SPIONs could lose their magnetization and become highly dispersed even after the removal of the magnetic field (Nagesh et al., 2016). However, biological application of SPION was limited because of the high surface hydrophobicity, making them susceptible to being ingested and eliminated by mononuclear phagocyte systems (MPS). In order to prolong the circulation time, it is necessary to modify the surface of SPION by amphiphilic copolymer coating to convert hydrophobic SPION into hydrophilic ones (Mirsadeghi et al., 2016; Silva et al., 2016).

Under an external magnetic field, SPION-encapsulated polymer nanoparticles could easily reach around the tumor section. In order to increase the active targeting of nanoparticles and uptake by tumor cells, receptor-mediated endocytosis could serve as a versatile targeting strategy via linking nanoparticles with multifunctional ligands to construct a potentially multiple targeting drug nanocarrier (Zhao et al., 2015; Rabiej et al., 2016; Shen et al., 2018). It is reported that various receptors are over-expressed on tumor cells, including folate receptor, neuropilin-1 receptor and transferrin receptor (Wang et al., 2011; Du et al., 2015; Song et al., 2015). Based on these receptors, drug delivery systems have been modified with corresponding targeting ligands and explored to deliver drugs through receptormediated endocytosis. The interleukin 13 receptor α2 (IL-13Rα2) is a subtype of the interleukin-13 receptor family, which is overexpressed on tumor cells (Mintz et al., 2002; Balyasnikova et al., 2012). It has been reported that IL-13Rα2, acting as a decoy receptor, has an intimate relationship with the progression of a tumor and can undergo internalization after binding to ligands (Kawakami et al., 2001). This property indicates that IL-13Rα2 could serve as a promising targeted moiety for anti-tumor drug delivery.

Pep-1 peptide (CGEMGWVRC) that was screened by the phage display library, could bind to IL-13Rα2 with high affinity and specificity and could be exploited to target ligand to tumor cells (Pandya et al., 2012). In our previous study, we demonstrated that Pep-1 conjugated paclitaxel (PTX) loaded nanoparticles, could be internalized into tumor cells via IL-13Rα2 mediated endocytosis (Wang et al., 2014). However, the accumulation of the targeted drug delivery system in the tumor tissue was still rather low and could only be by enhanced penetration and retention (EPR) effects (Wilhelm et al., 2016). Therefore, we aimed at developing PTX and SPION co-loaded polymer nanoparticles with Pep-1 peptide modification as a dual targeting nanocarrier (designated as Pep-NP-SPION/PTX) for tumor treatment in this study. SPION was prepared using a co-precipitation method and loaded into PEG-PLGA polymer nanoparticles that were modified with Pep-1 peptide to form Pep-NP-SPION/PTX (**Figure 1A**). As showed in **Figure 1B**, after intravenous (i.v.) injection, Pep-NP-SPION/PTX was expected to accumulate at the tumor tissue in the presence of an external magnetic field and then be internalized into tumor cells through IL-13Rα2 mediated endocytosis, which would reduce the uptake of Pep-NP-SPION/PTX by the MPS and enhance the anti-tumor efficiency of PTX. These physical-chemical properties of the dual targeted nanocarrier were also systematically characterized. Furthermore, the in vitro biological targeted capability of Pep-NP-SPION/PTX was investigated. Finally, the in vitro and in vivo anti-tumor effect of Pep-NP-SPION/PTX was studied using a cell and subcutaneous xenograft tumor mice model, respectively.

# MATERIALS AND METHODS

#### Materials

Qleic acid, Iron(II) chloride, iron(III) chloride and ammonium hydroxide were purchased from Sinopharm Chemical Reagent Co., Ltd. (China). Methoxyl poly(ethylene glycol)-co-poly(D,Llactic-co-glycolic acid) copolymer(MePEG-PLGA, 40 KDa) and Maleimidyl-poly(ethylene glycol)-co-poly(D,L-lactic- coglycolic acid) copolymer(Male-PEG-PLGA, 41.5 KDa) were purchased from Daigang Biomaterial Co., Ltd. (Jinan, China). PTX was purchased from Zelang Medical Technology Co., Ltd. (Nanjing, China). 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) was purchased from Beyotime Biotechnology Co., Ltd. (Nantong, China). Penicillin-streptomycin, RPMI 1640 medium, fetal bovine serum (FBS) and 0.25% (w/v) trypsin solution were obtained from Gibco BRL (Gaithersburg, MD, United States).

# Cell Line

The C6 cell line was obtained from the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). The cell line was cultured in a RPMI 1640 medium, supplemented with 10% FBS, 1% penicillin and 100 mg/mL streptomycin sulfate. Cells were cultured in incubators maintained at 37◦C with 5% CO2. All experiments were performed in the logarithmic phase of cell growth.

#### Animals

Mice (male, 4–5 weeks, 20–25 g) were supplied by the Department of Experimental Animals, Nanjing Medical University (Nanjing, China) and maintained under standard a housing environment. All animal experiments were performed in accordance with protocols and evaluated and approved by the ethics committee of Nanjing Medical University.

#### Synthesis of SPION Coated With Oleic Acid (OA-SPION)

Superparamagnetic iron-oxide nanoparticle was prepared by coprecipitating ferrous and ferric salts in an alkaline medium (Schleich et al., 2013). Briefly, 0.10 g of FeCl2·4H2O and 0.27 g of FeCl3·6H2O were added into 50 mL PEG600, which was placed

under a vacuum overnight at 40◦C to degas. The mixture was then heated to 125◦C under nitrogen protection. Then, 2 mL ammonium hydroxide was added, and the mixture was reheated to 125◦C. 5 min later, 20 mL of oleic acid was added into the solution at 125◦C and the mixture was kept at this temperature for another 30 min. Finally, the black precipitate was separated by an external magnet and washed three times with absolute ethanol and ultrapure water to remove the unreacted reagent, respectively. Dry SPION coated with oleic acid was prepared through freeze-drying for further use.

# Preparation of Pep-NP-SPION/PTX

Paclitaxel and SPION loaded PLGA-based nanoparticles were prepared via an emulsion/solvent evaporation method (Wilhelm et al., 2016). Firstly, the conjugation of Pep-1 peptide to PEG-PLGA copolymer via the reaction of maleimide with thiol group was carried out as previously reported (Wilhelm et al., 2016). Then, 19 mg MePEG-PLGA and 2 mg Pep-PEG-PLGA polymer were dissolved in 1 mL dichloromethane (DCM) containing PTX (1 mg/mL) and OA-SPION (Fe concentration: 4 mg/mL). Then the mixture was added to a 2 mL 1% (w/v) sodium cholate aqueous solution, which was finally sonicated using a probe sonicator (Xin Zhi Biotechnology Co., Ltd., China) for 5 min at 190 W output. The suspension was added drop-wisely into 10 mL 1% (w/v) sodium cholate solution under moderate stirring. A rotary evaporator was used to evaporate the redundant DCM at 40◦C. The solutions were centrifuged at 13500 g for 40 min to remove the excess excipient. After being washed three times by water, the Pep-NP-SPION/PTX was collected and re-dispersed in PBS for further use.

#### Characterization of Pep-NP-SPION/PTX Morphology, Particle Size and Zeta Potential of Pep-NP-SPION/PTX

The morphology of Pep-NP-SPION/PTX was characterized by the transmission electron microscope (TEM, Philips CM 100) operating at a voltage of 100 kV. The particle size and zeta potential of nanoparticles were assessed by dynamic light scattering (DLS, Zs90, Malvern, United Kingdom).

#### Determination of SPION Loading Content

The Fe content was measured using the phenanthroline spectrophotometric method. Lyophilized nanoparticles were dissolved in DCM, and then dried under nitrogen. The residue was dissolved with 10% (v/v) HCl solution. The yellow solution was combined with aqueous solutions of hydroxylamine

hydrochloride (10%, w/v) and sodium acetate buffer solution (pH = 5, 1 mol/L). Then 0.1% phenanthroline solution was added. After 0.5 h, the absorbance at 510 nm was measured and the Fe content was determined based on a comparison of a standard curve.

#### Encapsulation Efficiency and Loading Capacity of PTX

The amount of PTX was measured by high performance liquid chromatography (HPLC) with UV detection at 227 nm (LC-10AT, SHIMADZU, Japan). The mobile phase consisted of acetonitrile and water (47:53, v/v) with a gradient elution pumped at a flow rate of 1.0 mL/min and the column was maintained at 30◦C. The calibration curve was linear in the range of 0.1– 100 µg/mL with a correlation coefficient of R <sup>2</sup> = 0.9998.

To determine the encapsulation efficiency (EE%) and loading capacity (LC%) of Pep-NP-SPION/PTX, the lyophilized samples were dissolved in DCM, and then dried under nitrogen. The residue was dissolved in mobile phase solution and the concentration of PTX was analyzed by HPLC as described. The EE% and LC% were calculated as indicated below (n = 3).

$$\text{EE\%} = \frac{\text{Amount of PTX in the nanopparticles}}{\text{Total amount of PTX added}} \times 100\%$$

$$\text{LC\%} = \frac{\text{Amount of PTX in the nanopparticles}}{\text{Nanopparticles weight}} \times 100\%$$

$$\text{Namoparticles we'}$$

#### In vitro Release

The in vitro release kinetics of PTX from Pep-NP-SPION/PTX was measured by a standard dialysis method in PBS (0.04 M, pH 7.4) or HAc-NaAc buffer (0.04 M, pH 5.0) containing 0.1% (w/v) Tween-80 at 37◦C. Briefly, nanoparticles were suspended in 1 mL of mediator solution and placed in a dialysis bag (MWCO 3000). The dialysis bag was immersed in 30 mL of the release medium which was shaken at 37◦C. At predetermined time intervals, a portion of 0.3 mL dialysate was harvested, and the same volume of fresh mediator solution was added. The PTX concentration of samples was determined by HPLC as described above.

# In vitro Evaluation

#### In vitro Cellular Uptake

C6 cells were seeded into a 24-well plate at the density of 1 × 10<sup>5</sup> cells per well. After 24 h incubation, the cells were incubated with NP-SPION and Pep-NP-SPION at different concentrations (Fe concentration: 25 and 50 µg/mL) at 37◦C for 1 h, respectively. After that, the cells were washed three times with cold PBS and fixed with 4% formaldehyde for 10 min. Then, these cells were stained with Pearls' reagent (2% potassium ferrocyanide/6% HCl: 1/1) for 30 min. Finally, after washing by PBS for three times, the uptake of nanoparticles in the C6 cells was observed using a microscope.

#### In vitro Cytotoxicity

An MTT assay was used to evaluate the cell cytotoxicity of Pep-NP-SPION/PTX. C6 cells were seeded into 96-well plates at the density of 5000 cells/well and incubated at 37◦C in a 5% CO<sup>2</sup> atmosphere, and then incubated with Taxol <sup>R</sup> , NP-SPION/PTX and Pep-NP-SPION/PTX at different PTX concentration (1, 5, 10, and 20 µg/mL) after 24 h, respectively. All the concentrations of Fe used were based on the corresponding PTX concentrations (Fe concentrations: 3.9, 19.6, 39.2, and

78.4 µg/mL). After 48 h, the culture medium was discarded and 20 µL MTT solution was added into each well and incubated for additional 4 h. Then the unreacted dye was removed and 200 µL of DMSO was added to each well. Finally, the optical density was measured by a microplate reader at wavelength of 490 nm.

#### In vivo Evaluation

fphar-09-01556 January 18, 2019 Time: 16:27 # 5

#### Biodistribution of Pep-NP-SPION/PTX in Tumor Tissue

C6 cells (3 × 10<sup>6</sup> cells suspended in 100 µL PBS) were injected into the armpit of the right anterior limbs of mice subcutaneously. When the tumor volume reached about 300 cm<sup>3</sup> , the mice were divided into four groups randomly, then intravenously administrated with NP-SPION/PTX and Pep-NP-SPION/PTX (Fe dose of 15 mg/kg), respectively. The influence of the external magnetic field was also evaluated. 2 h after injection, the mice were sacrificed, and the tumor tissues were harvested and fixed in 4% paraformaldehyde. After paraffin embedding, the tumors were cut into 5 µm and stained with Prussian blue staining.

#### In vivo Anti-tumor Efficacy

In vivo tumor growth inhibition was carried out to evaluate the anti-tumor efficacy of Pep-NP-SPION/PTX. Subcutaneous xenograft tumor mice model was established as described above. With the tumor volume reached about 100 mm<sup>3</sup> , the mice were divided into six groups randomly, and then intravenously administrated with saline, Taxol <sup>R</sup> , NP-SPION/PTX and Pep-NP-SPION/PTX (PTX dose: 4.5 mg/kg, Fe dose: 15.6 mg/kg), respectively. The influence of external magnetic field was also evaluated. The formulation was given every other day for four injections. 15 days later, the mice were sacrificed with tumor collection and tumor volume was calculated (the formula: π/6 × larger diameter × smaller diameter<sup>2</sup> ).

# RESULTS

# Synthesis of SPION Coated With Oleic Acid

In order to improve the hydrophobicity and dispersibility of SPION, their surfaces were conjugated by oleic acid through physical adsorption. The FTIR spectra of SPION and OA-SPION are shown in **Figure 2**. The absorption peak at 569 cm−<sup>1</sup> was the characteristic absorption peak of Fe-O bond. The spectrum of OA-SPION showed a sharp absorption peak at 1644 cm−<sup>1</sup> , which was attributed to the C = O stretching vibrational absorption of oleic acid. However, the characteristic vibrational absorptions were not present in the spectrum of SPION. These results

demonstrated the successful modification of oleic acid on the surface of SPION.

#### Characterization of Pep-NP-SPION/PTX

Pep-NP-SPION/PTX was prepared via an emulsion/solvent evaporation method. As shown in **Figure 3**, Pep-NP-SPION/PTX was well-dispersed in water and showed a brown color before magnetic separation. When exposed to a magnetic field, the solution become more and more colorless and the nanoparticles were collected near the magnet. Therefore, the magnetic property of SPION was not affected by the encapsulation of PEG-PLGA.

The morphology and size of Pep-NP-SIPION/PTX was characterized by TEM and DLS, respectively. A TEM photograph showed that Pep-NP-SPION/PTX was uniformly spherical and of regular shape with a narrow size distribution. As shown in **Table 1**, the particle size of NP-SPION/PTX and Pep-NP-SPION/PTX were 129.8 ± 1.2 nm and 101.6 ± 1.0 nm, respectively. The particle size difference between NP-SPION/PTX and Pep-NP-SPION/PTX might be due to the modification of hydrophilic Pep-1 peptides, which enhanced the hydrophilicity of polymer nanoparticles. The zeta potential of NP-SPION/PTX and Pep-NP-SPION/PTX were below −20 mV.

The PTX LC% and EE% of Pep-NP-SPION/PTX were 3.54 ± 0.14% and 81.30 ± 3.50%, respectively, compared with 3.22 ± 0.86% and 66.85 ± 1.84% for NP-SPION/PTX. The change of encapsulation efficiency and loading capacity might be attributed to the modification of polymer materials with Pep-1 peptide. The iron loading of NP-SPION/PTX and Pep-NP-SPION/PTX was 12.41 ± 2.75 mg and 13.53 ± 3.38 mg/100 mg polymer, respectively.

#### In vitro Drug Release

The release profiles of PTX-loaded magnetic nanoparticles under different pH conditions are presented in **Figure 4**. Tween-80 was used to increase the solubility of PTX in a buffer solution and to avoid the binding of PTX to the nanoparticle surface. Herein, the drug release behavior was studied under a simulated physiological environment (pH 7.4) and an acidic condition (pH 5.0) because of the lower pH value in tumor tissues, due to excess lactic acid produced by hypoxia and acidic intracellular organelles (Xu et al., 2012).

Paclitaxel-loaded magnetic nanoparticles all exhibited an initial burst release profile with total PTX releases of about 60% before 12 h. After 72 h, NP-SPION/PTX and Pep-NP-SPION/PTX showed a total drug release of 63.4 and 65.8% of PTX at pH 7.4, respectively, while 82.3 and 87.1% of the PTX in PBS


(pH 5.0), respectively. These results suggested that solution pH could affect the release pattern of PTX from nanoparticles. The pH responsive release of the dual targeted nanoparticles may be beneficial for enhancing the anti-tumor efficacy of the PTX in the acidic tumor microenvironment.

#### In vitro Cell Experiment In vitro Cellular Uptake

It has been reported that IL-13Rα2 was upregulated in C6 cell line (Daines et al., 2006), so the C6 cell was used as the cell model to investigate the in vitro uptake of Pep-NP-SPION in this study. The cellular uptake of Pep-NP-SPION was studied quantitatively by Prussian blue staining. The number of blue granules within C6 cells was related to the nanoparticle concentration. As shown in **Figure 5**, the blue granules intensity was increased with the increase of iron concentrations ranging from 25 to 50 µg/mL after 1 h incubation, which indicated that the cellular uptake of Pep-NP-SPION exhibited a concentrationdependent mode. Moreover, the cellular uptake of Pep-NP-SPION was obviously higher than that of NP-SPION. These results implied that the modification of Pep-1 peptide could enhance the cellular uptake of nanoparticles through receptormediated endocytosis.

#### In vitro Cytotoxicity

The in vitro cell viability of NP-SPION/PTX and Pep-NP-SPION/PTX was evaluated by MTT assay. The cytotoxicity of various PTX formulations exhibited a concentration-dependent pattern with the increase of PTX concentration ranging from 1 to 20 µg/mL. Different degrees of cytotoxicity were found in all the PTX formulations. As shown in **Figure 6**, the cell viability of Pep-NP-SPION/PTX was even lower than 50% at PTX concentration of 10 µg/ml. The IC<sup>50</sup> value of Pep-NP-SPION/PTX and NP-SPION/PTX is 10.2 and 19.4 µg/mL, respectively.

# Biodistribution of Pep-NP-SPION/PTX in Tumor Tissue

The in vivo tumor targeting capability of magnetic nanoparticles was studied qualitatively by Prussian blue staining. Iron oxide was indicated as blue and cell nucleus as red. As shown in **Figure 7**, in the presence of an external magnetic field, both NP-SPION/PTX and Pep-NP-SPION/PTX exhibited much higher tumor distribution than those without magnetic field intervention, which indicated that the constructed nanoparticles had an attractive magnetic targeting character for improving tumor distribution. Moreover, compared with the NP-SPION/PTX group, Pep-NP-SPION/PTX showed an obvious higher distribution no matter whether an external magnetic field was added, which revealed that the modification of Pep-1 peptide could accelerate the accumulation of nanoparticles into the tumor section via IL-13Rα2 mediated endocytosis. Altogether, based on the active targeting and magnetic targeting, Pep-NP-SPION/PTX can be used as a potential dual targeting nanocarrier for the treatment of tumors.

# In vivo Anti-tumor Efficacy

In this study, mice bearing subcutaneous tumor xenograft were used to evaluate the in vivo anti-tumor efficacy of Pep-NP-SPION/PTX. The mice were intravenously injected with Taxol, NP-SPION/PTX and Pep-NP-SPION/PTX (PTX dose: 4.5 mg/kg, Fe dose: 15.6 mg/kg) every other day for four consecutive administrations with tumor sizes recorded every 3 days, respectively. As shown in **Figure 8**, the tumor size of the saline group was obviously all larger than that of the PTX formulations. In the initial days, there was no significant difference in tumor size among these groups. At the experimental terminal, the tumor size of all PTX formulations followed the order: Pep-NP-SPION/PTX with magnetic field < Pep-NP-SPION/PTX < NP-SPION/PTX with magnetic field < NP-SPION/PTX < Taxol < Saline. These results showed that the modification of Pep-1 peptide could improve the anti-tumor efficacy of nanoparticles through IL-13Rα2-mediated

endocytosis, consistent with our previous study. Moreover, the encapsulation of SPION endowed nanoparticles with the magnetic targeting property and enhanced the anti-tumor efficacy with the magnetic field. Together, Pep-NP-SPION/PTX constructed in this study could offer a potential magnetic targeting and receptor mediated targeting to enhance the antitumor efficacy for tumor treatment.

# CONCLUSION

In this study, we successfully constructed PTX and SPION coloaded polymer nanoparticles with Pep-1 peptide modification as a dual targeted nanocarrier for tumor treatment. The cellular uptake of Pep-NP-SPION/PTX showed a concentrationdependent manner and significantly enhanced than that of unmodified NP-SPION/PTX. Pep-NP-SPION/PTX exhibited cytotoxicity comparable to Taxol at the PTX concentrations rangingfrom1 to20µg/mLin vitrocell experiments. Furthermore, Pep-NP-SPION/PTX showed a satisfactory tumor accumulation and the magnetic field significantly enhanced the bio-distribution of nanoparticles in the tumor section. More importantly, the in vivo anti-tumor efficacy showed that Pep-NP-SPION/PTX exhibited desirable anti-tumor efficacy and the magnetic field could also enhance the anti-tumor efficacy. Altogether, these results indicate that the modification of Pep-1 peptide could enhance the active targeting property through receptor-mediated endocytosis and the encapsulation of SPION which obviously improves the physical targeting property of nanoparticles under an external magnetic field. The physical magnetic targeting and IL-13Rα2 mediated active targeting characteristics could synergistically increase the targeted efficiency for a tumor. Therefore, Pep-NP-SPION/PTX could serve as a potential dual targeting nanocarrier for tumor therapy.

#### AUTHOR CONTRIBUTIONS

fphar-09-01556 January 18, 2019 Time: 16:27 # 9

HX designed the experiments. BW and WW performed the experiments. BW and HX wrote the main manuscript. HL and ZW prepared the figures and tables. All authors reviewed the manuscript.

### FUNDING

This work was supported by the grants from the National Natural Science Foundation of China (31671018), the Natural Science

#### REFERENCES


Foundation of Jiangsu Province-Excellent Young Scientist Fund (BK20160096), the 2016 Qing Lan Program of Jiangsu Province, and the 2017 Six Talent Peaks Project of Jiangsu Province.

#### ACKNOWLEDGMENTS

The authors acknowledge the support from the School of Pharmacy, Fudan University and the Open Project Program of Key Lab of Smart Drug Delivery (Fudan University), Ministry of Education, China (SDD2012-4).


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Wang, Wu, Lu, Wang and Xin. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Polypeptide Nanogels With Different Functional Cores Promote Chemotherapy of Lung Carcinoma

Kai Niu<sup>1</sup>† , Nan Li<sup>2</sup>† , Yunming Yao<sup>3</sup>† , Chunjie Guo<sup>4</sup> , Yuanyuan Ge<sup>5</sup> and Jianmeng Wang<sup>5</sup> \*

<sup>1</sup> Department of Otorhinolaryngology Head and Neck Surgery, The First Hospital of Jilin University, Changchun, China, <sup>2</sup> Department of Neonatology, The First Hospital of Jilin University, Changchun, China, <sup>3</sup> Department of Abdominal Ultrasound, The First Hospital of Jilin University, Changchun, China, <sup>4</sup> Department of Radiology, The First Hospital of Jilin University, Changchun, China, <sup>5</sup> Department of Geriatrics, The First Hospital of Jilin University, Changchun, China

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Yanqi Ye, University of North Carolina at Chapel Hill, United States Rongxin Su, Tianjin University, China Bailiang Wang, Wenzhou Medical University, China

\*Correspondence:

Jianmeng Wang jmwang1981@126.com †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 30 October 2018 Accepted: 14 January 2019 Published: 04 February 2019

#### Citation:

Niu K, Li N, Yao Y, Guo C, Ge Y and Wang J (2019) Polypeptide Nanogels With Different Functional Cores Promote Chemotherapy of Lung Carcinoma. Front. Pharmacol. 10:37. doi: 10.3389/fphar.2019.00037 Two kinds of tumor microenvironment-responsive polypeptide nanogels were developed for intracellular delivery of cytotoxics to enhance the antitumor efficacies and reduce the side effects in the chemotherapy of lung carcinoma. The sizes of both doxorubicin (DOX)-loaded nanogels methoxy poly(ethylene glycol)–poly(Lphenylalanine-co-L-cystine) [mPEG–P(LP-co-LC)] and methoxy poly(ethylene glycol)– poly(L-glutamic acid-co-L-cystine) [mPEG–P(LG-co-LC)] (NGP/DOX and NGG/DOX) were less than 100 nm, which was appropriate for the enhanced permeability and retention (EPR) effect. The bigger and smaller scale of nanoparticle could induce the elimination of reticuloendothelial system (RES) and decrease the in vivo circulating halflife, respectively. The loading nanogels were stable in the neutral environment while quickly degraded in the mimic intracellular microenvironment. Furthermore, the DOXloaded reduction-responsive nanogels showed significantly higher tumor cell uptake than free DOX·HCl as time went on from 2 to 6 h. In addition, these DOX-loaded nanogels showed efficient antitumor effects in vivo, which was verified by the obviously increased necrosis areas in the tumor tissues. Furthermore, these DOX-loaded nanogels efficiently reduced the side effects of DOX. In conclusion, these reduction-responsive polypeptides based nanogels are suitable for the efficient therapy of lung carcinoma.

#### Keywords: nanogel, reduction-responsive, controlled drug delivery, lung carcinoma, antitumor

# INTRODUCTION

Lung carcinoma has been the most common malignancy in both men and women with high mortality (Kanodra et al., 2015; Kernstine, 2017). Chemotherapy, as a traditional treatment, is important to the treatment of lung cancer. However, the main therapeutic platforms available for chemotherapy drugs currently have the disadvantages of not being able to effectively aggregate in tumor cells and causing various systemic side effects. Due to these deficiencies, various nanomedicine delivery systems have been studied such as micelles (Feng et al., 2017; Shen et al., 2017; Xu et al., 2017; Kosakowska et al., 2018; Li et al., 2018b; Sun et al., 2018), nanogels (Ding et al., 2011; Guo et al., 2017, 2018; Jiang et al., 2018b; Li et al., 2018d; Zhang et al., 2018; Zhu et al., 2018), polymer–drug conjugates (Zhao et al., 2017; Cong et al., 2018; Li et al., 2018a; Yang et al., 2018), liposomes (Allen and Cullis, 2013; Zhao et al., 2013; Piffoux et al., 2018), and so forth

(Ding et al., 2014; Tao et al., 2015, 2017; Chen et al., 2017a; Armstrong and Stevens, 2018; Ji et al., 2018; Jiang et al., 2018a; Li et al., 2018b,c; Xiao et al., 2018), which help to improve the drug accumulation in the tumor, to achieve the effective control of drug release in the tumor lesions, and to reduce the side effects (Yu et al., 2016; Zhang et al., 2017). Among them, nanogels have the strong core–shell structures by crosslinking, which enables them not only to exhibit high drug loading capacity, but also to prevent drug leakage (Ye et al., 2016; Chen et al., 2017b). In addition, the suitable sizes of the nanogels made them can efficiently accumulate at the tumor lesions by the enhanced permeability and retention (EPR) effect (Acharya and Sahoo, 2011). More importantly, additional functionality is given to crosslinkers to achieve "switch on/off " release of drugs from nanogels in tumor cells (Mura et al., 2013). In addition, the stability of nanogels is one of the major obstacles, which handers the in vivo application of them. Fortunately, the chemistry crosslinking has been widely used to enhance the stability and functionality of nanogels, which is more stable than non-covalent interactions (Ryu et al., 2010; Jiang et al., 2018a). Furthermore, cross-linking nanoparticles with different functional chemical crosslinkers could controllably release the laden drugs according to the redox, low pH, and high enzyme level of tumor microenvironments (Zha et al., 2011).

Due to the different metabolic pathways of tumor cells, the microenvironments of tumor tissues show hypoxia, low sugar, and low pH (Parks et al., 2013). It is worth noting that malignant cells show the reductive intracellular microenvironment (Ge and Liu, 2013). Therefore, the reduction-sensitive polymeric nanoparticles have attracted more and more attention in the realm of smart antitumor drug delivery (Cheng et al., 2013; Phillips and Gibson, 2014).

Herein, we reported the drug delivery potential of reductionsensitive polypeptide nanogels formulations, which could suppress lung carcinoma cell proliferation at low dose and reduce unwanted adverse effects. The reduction-responsive methoxy poly(ethylene glycol)–poly(L-phenylalanine-co-Lcystine) (mPEG–P(LP-co-LC)) and methoxy poly(ethylene glycol)–poly(L-glutamic acid-co-L-cystine) mPEG–P(LG-co-LC) nanogels were prepared to selectively deliver chemotherapy agents (**Scheme 1**). Typically, nanogels loaded with doxorubicin (DOX) were used as models for clinical antitumor drug. Results showed that both DOX-loaded nanogels exhibited satisfactory antitumor activity and higher safety than free DOX·HCl. These DOX-loaded nanogels are able to serve as satisfactory nanoplatforms for the therapy of lung carcinoma.

# MATERIALS AND METHODS

Methoxy (ethylene glycol) (mPEG) was purchased from Aladdin Industrial Co., Ltd. (Shanghai, China) and mPEG-NH<sup>2</sup> was synthesized by trimethylamine modified. L-Glutamic acid (LC), L-phenylalanine (LP), and L-cystine (LC) were obtained from GL Biochem, Ltd. (Shanghai, China). LG NCA, LP NCA, and LC NCA were prepared as reported in previous work (Huang et al., 2015; Shi et al., 2017). Doxorubicin hydrochloride (DOX·HCl) was bought from Beijing Huafeng United Technology Co., Ltd. (Beijing, China). Glutathione (GSH) (used for cell culture) was bought from Aladdin Reagent Co., Ltd. (Shanghai, China). Both 4,6-Diamidino-2-phenylindole (DAPI) and methyl thiazolyl tetrazolium (MTT) were obtained from Sigma–Aldrich (Shanghai, China). Hematoxylin and eosin were purchased from Merck Company (Darmstadt, Germany).

#### Characterizations

Proton Nuclear Magnetic Resonance (1H NMR) spectra were detected on a Bruker AV 600 NMR spectrometer (Billerica, MA, United States) using deuterated dimethyl sulfoxide (DMSO-d6) as the solvent. Fourier-transform infrared spectroscopy (FT-IR) was performed on a Bio-Rad Win-IR instrument (Bio-Rad Laboratories Inc., Cambridge, MA, United States) by using potassium bromide method. The morphology of NGP/DOX and NGG/DOX nanogels were visualized on JEM-1011 transmission electron microscope (TEM; JEOL, Tokyo, Japan). Sizes of NGP/DOX and NGG/DOX were determined by dynamic laser scattering (DLS) measurement on a WyattQELS instrument (DAWN EOS, Wyatt Technology Corporation, Santa Barbara, CA, United States), and the scattering angle was set at 90◦ .

#### Syntheses and Characterizations of mPEG–P(LP-co-LC) and mPEG–P(LG-co-LC)

The reduction-responsive mPEG–P(LP-co-LC) nanogel was synthesized through the one-step ROP of LP NCA and LC NCA with amino-terminated mPEG (mPEG–NH2) as a macroinitiator according to the reports in our previous works. Firstly, 2 g of mPEG-NH2, 2.3 g of NCA LP NCA and 0.9 g of LC NCA were dissolved in 100 mL of DMF and stirred for three days. The obtained solution was poured into 700 mL of the ethyl ether twice, and the white precipitate was collected. After vacuum drying, mPEG–P(LP-co-LC) was obtained with the yield of 75.2%. mPEG–P(BLG-co-LC) was prepared in a similar route. mPEG—P(LG-co-LC) was generated by removing the benzyl group from mPEG–P(BLG-co-LC). mPEG–P(BLG-co-LC) was dissolved in dichloroacetic acid (100 mg mL−<sup>1</sup> ) and a 33 wt. % solution of HBr in acetic acid was added subsequently (20 mL for 1 g copolymer). After stirring for 1 h at 30◦C, the mixture was precipitated into diethyl ether (10 times volume of the reaction solution). The obtained product was further dried under vacuum at room temperature for 24 h after washing twice with diethyl ether (Yield: 80.5%). The products were characterized by proton nuclear magnetic resonance (1H NMR), Fouriertransform infrared (FT-IR), inductively coupled plasma optical emission spectroscopy (ICP-OES), and elemental analysis.

#### DOX Encapsulation

Nanogel was loaded with DOX by nanoprecipitation method. Firstly, 50 mg of nanogel was dispersed in 20 mL of N,Ndimethylformamide (DMF). After adding 10 mg of DOX·HCl, the solution was further stirred for 12 h at room temperature. Then, 2 mL of phosphate buffered saline (PBS; 0.01 M) was slowly mixed into the above solution, along with 18 mL of MilliQ water.

The final mixture was dialyzed against MilliQ water for 24 h (molecular weight cut-off (MWCO) = 3500 Da) after stirring for 12 h at room temperature. The MilliQ water was changed every 2 h. Finally, NGP/DOX and NGG/DOX nanogels were obtained by filtration and lyophilization.

The drug loading content (DLC) and drug loading efficiency (DLE) were detected by standard curve method, using fluorescence spectroscopy on a Photon Technology International (PTI) Fluorescence Master System with Felix 4.1.0 software (PTI, Lawrenceville, NJ, United States; kex = 480 nm). The DLC and DLE of NGP/DOX and NG/DOX were calculated according to Eqs. (1) and (2), respectively.

$$\text{DLC}(\%) = \frac{\text{Weight of Drug in Nanogel}}{\text{Weight of Drug} - \text{Loaded Nanogel}} \times 100\% \quad \text{(1)}$$

$$\text{DLE}(\%) = \frac{\text{Weight of Drug in Nanogel}}{\text{Weight of Feeding Drug}} \times 100\% \tag{2}$$

#### In vitro DOX Release

In vitro drug release profiles of DOX from NGP/DOX and NGG/DOX nanogels were performed in PBS (pH 7.4) with or without 10 nM GSH. 10 mL of nanogel (0.1 mg mL−<sup>1</sup> ) aqueous solution was transferred into an end-sealed dialysis bag (MWCO = 3500 Da). The release experiment was carried out by putting the end-sealed dialysis bag into the corresponding release medium (100 mL) at 37◦C with continuous shaking at 75 rpm in the dark. At fixed time intervals, 2 mL of release medium was removed and an equal volume of fresh medium was added. The amount of released DOX was measured by the fluorescence spectrophotometer (kex = 480 nm).

#### Cell Culture

Under the conditions of 37◦C and 5% (V/V) carbon dioxide (CO2), the human lung Lewis cells were cultured in RPMI-1640, which was supplemented with 10% (V/V) fetal bovine serum (FBS), penicillin (100 IU mL−<sup>1</sup> ), and streptomycin (100 IU mL−<sup>1</sup> ).

#### Intracellular DOX Release

The intracellular DOX release from NGP/DOX and NGG/DOX were measured by confocal laser scanning microscopy (CLSM) toward Lewis cells. The cells (15,000 cells) were seeded in disks, incubated in 1 mL of RPMI-1640 medium containing 10% FBS for 24 h, and pretreated with 10 mM GSH or buthionine sulfoximine (BSO) for 2 h. After removing the medium and subsequently washing three times with PBS (pH 7.4) solution, 1 mL of NGP/DOX and NGG/DOX solution in RPMI-1640 was added, with a final DOX dose of 10 µg mL−<sup>1</sup> . The cells treated with equivalent free DOX without GSH pretreatment were used as control. After another 2 h of incubation, the cells were washed with PBS for five times, and fixed with 4% (W/V) PBS-buffered paraformaldehyde at room temperature for 30 min. The cellular nuclei were then stained at 37◦ C for 3 min using DAPI. A CLSM (Carl Zeiss, LSM 780, Jena, Germany) was used to view the intracellular localization of DOX.

#### Cytotoxicity Assays

fphar-10-00037 February 1, 2019 Time: 16:24 # 4

The cytotoxicities of NGP/DOX, NGG/DOX and free DOX were evaluate din Lewis cells at different conditions. The cells were planted in 96-well plates (7 × 10<sup>3</sup> cells per well) in 200 µL of RPMI-1640 medium supplemented with 1X penicillin/streptomycin and 10% fetal bovine serum. After incubation for 24 h at 37◦C, the cells were pretreated with 10 mM GSH or BSO for 2 h. Subsequently, each culture medium was replaced by 180 µL of RPMI-1640 containing NGP/DOX, NGG/DOX and free DOX at equivalent concentrations, respectively, with the DOX ranging from 0.3 to 18.4 µM. After a further 24, 48, and 72 h incubation, 20 µL MTT (5 mg mL−<sup>1</sup> ) in PBS was added to each well, followed by another 4◦h incubation at 37◦C. Then the sediment was dissolved in 150 µL DMSO after the medium was removed. The absorbance at 490 nm of the above solution was determined on an ELx808 microplate reader (Bio-Tek Instruments, Inc., Winooski, VT, United States). The percentage of cell viability was determined by comparing the absorbance of the sample cells and the control cells [Eq. (3)].

$$\text{Cell}(\%) = \frac{A\_{\text{Sample}}}{A\_{\text{Control}}} \times 100\% \tag{3}$$

#### Animal Procedures

5-Week-old male BALB/c mice weighting 18 ± 0.2 g were supplied by the Jilin University Experiment Animal Center (Changchun, China). All animal experiments were performed according to the Guidelines for Animal Care and Use of Jilin University. The tumor grafted mouse model was established by subcutaneous injection of 100 µL of cell suspension containing 2 × 10<sup>6</sup> Lewis cells in PBS into the armpit of right anterior forelimb.

#### In vivo Antitumor Assessments

The tumor volumes and mice's body weights were monitored every two days from the second day after the inoculation of Lewis cells (that was, Day 1). When tumor volume increased to about 100 mm<sup>3</sup> after 8 days of inoculation, the nude mice were randomly divided into 7 groups (n = 10), that was, free DOX, NGP/DOX or NGG/DOX at a DOX dose of 3 or 6 mg (kg BW)−<sup>1</sup> and normal saline (control group). The formulations of DOX were recorded as DOX/3, DOX/6, NGP/DOX/3, NGP/DOX/6, NGG/DOX/3 and NGG/DOX/6, respectively. At the same time, the treatments began with injecting100 µL of normal saline and various DOX preparations in normal saline into the tail vein of mice for four times every 5 days. The tumor sizes were measured every day, and the body weights were measured every subsequent day. Tumor volumes [Eq. (4)] and body weights were used to evaluate the antitumor efficacy and security in vivo.

$$V(\text{mm}^3) = \frac{\text{L} \times \text{S}^2}{2} \tag{4}$$

In Eq. (4), L (mm) was the largest diameter of tumor, and S (mm) was the smallest diameter.

The following formula was used to calculate the tumor inhibition ratio:

$$\text{Tumor inhibition rate (\%)} = (V\_{\text{control}} - V\_{\text{sample}}) / V\_{\text{control}} \tag{5}$$
 
$$\text{Eumor (\%)} \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{ } \dots \text{ } \text{ } \text{ } \text{$$

In Eq. (5), Vcontrol and Vsample represented the tumor volumes of control groups and sample groups, respectively.

The weights of the major organs were recorded. The organ indices of all the organs of mice were calculated by [Eq. (6)].

$$\text{Organ(\%)} = \frac{W\_{\text{Organ}}}{W\_{\text{Body}}} \times 100\% \tag{6}$$

#### Immunohistochemical Analyses of Tumor Tissues

On day 27, the Lewis lung carcinoma-grafted BALB/c mice were killed by cervical dislocation 5 days after the last injections. The tumors and major organs (heart, liver, spleen, lung, kidney, thymus, and marrow) were isolated at first, and then fixed with 4% (W/V) paraformaldehyde overnight, followed by dehydration, clearing, wax infiltration, and embedding. The paraffin-embedded tumors and organ tissues were cut at a thickness of 5 µm for hematoxylin and eosin (H&E) staining. Paraffin sections with a thickness of 3 µm were used for immunohistochemical staining (including caspase-3, survivin, Bax, and Bcl-2) to assess the pathological and immunological characteristics of tumor tissues. The instruments used included Leica RM 2245 paraffin machine (Leica, Germany), Leica HI1210 fishing machine (Leica, Germany), Leica EG1150H embedding machine (Leica, Germany), Leica HI1220 booth machine (Leica, Germany), Olympus BX51 microscope (Olympus, Japan), and Motic image analysis system (Motic Industrial Group Co., Ltd., Xiamen, China).

#### Histopathological and Biochemical Analyses of Organs

The major internal organs and tissues (heart, liver, spleen, lung, kidney, thymus, and marrow) were collected at the same time. The organs from healthy mice were used as controls. All the organs involved were divided into two parts as follows: (i) one part (excluding marrow) fixed with 4% (W/V) PBS-buffered paraformaldehyde was prepared for the histopathological analyses by H&E staining. (ii) The other

part was used to detect the organ function-related biochemical indicators, including blood urea nitrogen (BUN), creatinine (Cr), alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase (CK), creatine kinase-MB (CK-MB) and lactate dehydrogenase (LDH), by commercial enzyme-linked immunosorbent assay (ELISA) kits. The biochemical indicators in serum were also tested. Briefly, 300 µL of blood without anticoagulant was centrifuged at 3000 rpm for 10 min. The serum was then collected to detect the clinical biochemical parameters. The data from healthy nude mice were used as controls. The histopathological results were detected and analyzed by Olympus BX51 microscope and Motic image analysis system, respectively.

### Detections of White Blood Cell (WBC) Count and Bone Marrow Cell Micronucleus Rates (BMMRs)

On day 27, 20 µL of blood (anticoagulated through enucleation method) was taken from each nude mouse to count the WBCs. The sternums from BALB/c mice were decalcified and fixed for 10 days after being placed in 10% (V/V) formic acid-formalin solution. The data from normal nude mice were used as controls. Then, the tissues were dehydrated, cleared, wax infiltrated, and embedded. For each sternum, paraffin sections with a thickness of 5 µm were collected for H&E staining, with an interval of 50 µm. The H&E-stained section was used to evaluate the BMMR.

### Statistical Analyses

All tests were carried out at least three times, and the relevant data were expressed as mean ± standard deviation (SD). Statistical analysis was performed using SPSS 13.0 statistical software (SPSS Inc., Chicago, IL, United States), ∗∗∗P < 0.05 was considered statistically significant, and ∗∗P < 0.01 and <sup>∗</sup>P < 0.001 were as considered significant differences.

# RESULTS AND DISCUSSION

# Characterizations of NGP/DOX and NGG/DOX

These reduction-responsive nanogels consisted of mPEG (hydrophilic shell) and disulfide-cross-linked P(LP-co-LC) or P(LG-co-LC) (hydrophobic core). The disulfide bond of the LC segment endowed the nanogels with reduction-responsiveness (Huang et al., 2015; Shi et al., 2017). These DOX-loaded nanogels were prepared through a modified nanoprecipitation method

by adding DOX aqueous solution to dimethylformamide (DMF) solution containing NGP or NGG nanogel (**Scheme 1**) (Ding et al., 2013a; Huang et al., 2015; Liu et al., 2015; Shi et al., 2017). The drug loading content (DLCs) and drug loading efficiency (DLEs) of NGP/DOX and NGG/DOX were calculated to be 9.8, 54.7 and 15.2, 91.5 wt.%, respectively. As shown in the TEM images, NGP/DOX, and NGG/DOX showed a spherical morphology at a diameter of about 80 and 93 nm, respectively (**Figures 1A,B**). The R<sup>h</sup> of NGP/DOX and NGG/DOX detected by dynamic laser scattering (DLS) were 78.1 ± 3.5 and 93.3 ± 3.5 nm, respectively (**Figures 1A,B**). The hydrodynamic size of NGP/DOX and NGG/DOX was slightly larger than the diameter detected by TEM due to the swelling of nanogels in the aqueous condition (Huang et al., 2015; Shi et al., 2017). Appropriate size facilitated the efficient accumulation of nanogels in the tumors (Ding et al., 2013b,c, 2015).

### In vitro Release Performance and Tumor Cell Inhibition

The DOX release performances of NGP/DOX and NGG/DOX were carried out in PBS with 0 or 10.0 mM glutathione (GSH). As shown in **Figure 1C**, during the 72 h test, only around 30%

of DOX was released in the PBS with 0 mM GSH. In contrast, an obviously increased DOX release was observed in the PBS with 10.0 mM GSH. In detail, the proportions of cumulative released DOX from NGP/DOX and NGG/DOX in the initial 12 h were 47.8 and 39.8% in PBS with GSH (GSH+), respectively. After 72 h of incubation, the DOX released from NGP/DOX and NGG/DOX in the GSH+ medium (72.4 and 74.7%) were more than twice that in the GSH− medium (31.0 and 28.1%), respectively. The accelerated release of DOX should be due to GSH breaking the disulfide bond of nanogels (Huang et al., 2015). After 24 h, the drug release of nanogels was relatively decelerated and sustained. The release results indicated the NGP/DOX and NGG/DOX could efficiently release the DOX in tumor cells according to the different redox potential between intracellular and extracellular microenvironments, which might have more obvious antitumor activity.

Confocal laser scanning microscopy (CLSM) assays were performed toward Lewis lung carcinoma cells pretreated without (GSH−), with 10.0 mM GSH (GSH+) or buthionine sulfoximine (BSO) to verify the satisfactory intracellular DOX release of NGP/DOX and NGG/DOX. NG/DOX with a dosage of 10.0 µg mL−<sup>1</sup> DOX·HCl was used to co-cultured with the GSH− or GSH+ cells for 2 and 6 h. The unpretreated cells co-cultured with equivalent free DOX·HCl were prepared as a control. As shown, the DOX fluorescence was shown in all the Lewis lung carcinoma cells treated with DOX formulations (**Figure 2**). As

mean ± SD (n = 6).

time went on, the DOX fluorescence in the nanogels with GSH pretreatment groups were higher than that of the free DOX·HCl and BSO pretreated groups. As far as we knew, only the DOX released from nanogels in the cells could be detected by CLSM (Ding et al., 2013c). These results verified the efficient endocytosis as well as the reduction-responsive intracellular DOX release of NGG/DOX and NGP/DOX. These results might be related to that with the pretreatment of GSH, the intracellular GSH content was increased, which facilitated the rapid release of DOX from nanogels. Moreover, the results were consistent with the in vitro DOX release kinetics in PBS (**Figure 1C**). It was interesting that the tumor cells incubated with free DOX·HCl exhibited the strongest DOX fluorescence at 2 h. This result was related to the way of free DOX·HCl entered cells was diffusion, which was faster than the endocytosis of nanogels (Ding et al., 2013b,c).

The in vitro antitumor activities of NGG/DOX and NGP/DOX were estimated by MTT assay at 24, 48, and 72 h. As shown in **Figure 3**, compared with the BSO pretreated and unpretreated groups, GSH pretreated nanogels showed obviously higher growth inhibition efficiency in Lewis cells during the test time. This more satisfactory inhibition of GSH pretreated nanogels might be related to the intracellular GSH content was increased after GSH pretreatment, which facilitated the rapid release of DOX from nanogels and showed stronger tumor killing effect. As time went on, the half maximal inhibitory concentrations (IC50s) of NGP/DOX decreased, but the order was constant: BSO < GSH− < GSH+. At 72 h, the IC50s of BSO, GSH−, and GSH+ pretreated groups were calculated to be 3.78, 3.30, and 2.61 µM mL−<sup>1</sup> , respectively. The NGG/DOX groups showed similar results, and the IC50s of BSO, GSH−, and GSH+ pretreated groups were calculated to be 12.12, 7.92, and 4.65 µM mL−<sup>1</sup> at 72 h. The NGP/DOX and NGG/DOX groups pretreated by GSH had the lowest IC50, demonstrating their enhanced suppressor capability against the tumor cells and their advantage as potential antitumor drug formulations.

#### In vivo Antitumor Efficacy

The most important indicators of drug use in vivo were efficacy and safety, which must be rigorously evaluated before any newly formulations could be used clinically. Tumor-grafted animal models are the main method to evaluate the antitumor activities of drug delivery systems. As shown, compared with free DOX·HCl, DOX-loaded nanogels exhibited stronger antitumor effect (**Figure 4A**). This result was related to two factors: one was related to the EPR effect (Huang et al., 2015; Shi et al., 2017), which resulted more DOX-loaded nanogels accumulated in tumor sites, the other was due to the more controlled release of DOX from nanogels, which resulted less DOX was released in circulation and more DOX accumulated in tumor sites. The tumor inhibition rates of NGP/DOX/3, NGP/DOX/6, NGG/DOX/3, and NGG/DOX/6 groups were 98.9 ± 0.2%, 99.4 ± 0.03%, 97.15 ± 0.4%, and 99.1 ± 0.1%, which were much stronger than that of DOX/3 and DOX/6 groups (i.e., 87.4 ± 2.1% and 95.8 ± 3.0%; P < 0.001). Despite DOX also showed antitumor effect, the weight loss of mice was very obvious during the therapy (**Figure 4B**), especially in the DOX/6 group, indicating the severe systemic toxicity of DOX. In contrast, the groups treated with DOX-loaded nanogels showed tiny body weight loss, indicating the effective attenuation effect DOX-loaded nanogels. Furthermore, organ indices were provided to offer a general impression of the system toxicity of DOX. The tumor index was also calculated. As shown, the organ indices exhibited no obvious difference except the tumor indices (**Figure 5**), indicating these nanogels would efficiently suppress the tumor growth while not cause systemic toxicities in vivo.

In addition, the histopathological and immunohistochemical tests were performed to further confirm the antitumor efficacy of DOX-loaded nanogels (**Figure 6**). In this study, four immunohistochemical stainings caspase-3, survivin, Bax, and Bcl-2 were performed (**Figure 6**). It is well known that the process of cell apoptosis is regulated and controlled by various apoptotic genes. Notably, the caspase family plays a crucial role in the process of apoptosis (Huang et al., 2015; Shi et al., 2017). The activation of apoptosis-inducing factor caspase-3 is the key pathway for a variety of stimuliinduced apoptosis (Huang et al., 2015; Shi et al., 2017). As shown, the signals of pro-apoptotic protein Bax (brown) and caspase-3 (brown) of the nanogels treatment groups were much stronger than those of the free DOX·HCl treatment groups. In contrast, the anti-apoptotic protein Bcl-2 (brown)

<sup>∗</sup>P < 0.001).

showed an obvious decline in the nanogels treatment groups. Furthermore, survivin was also tested to evaluate the cell survival (Huang et al., 2015; Shi et al., 2017). As shown, the signals of survivin (brown) were obviously reduced in the DOXloaded nanogels treatment groups. These results fully verified that our DOX-loaded nanogels, especially NGP/DOX/6 and NGG/DOX/6, could be served as efficient nano-therapeutic agents.

As shown in **Figure 6**, H&E staining showed universal mitosis and mild hemorrhagic necrosis in the control

FIGURE 7 | Ex vivo histopathological analyses (i.e., H&E) of lung tumor, visceral organ, thymus, and marrow sections after all treatments with control, NS, free DOX 3 and 6 mg (DOX/3, DOX/6), NGP/DOX, and NGG/DOX at a dose of 3.0 or 6.0 mg DOX·HCl equivalent per kg body weight (NG/DOX/3, NG/DOX/6). Magnification: 200×.

group, indicating the rapid cell growth. In contrast, all the DOX formulations treatment groups showed varying degrees of tumor growth suppression. Specifically, DOX formulations caused a reduction in mitosis and extensive hemorrhage and necrosis. The treatment groups were ranked as follows according to the relative amount of necrotic tissues: NGP/DOX/6.0 > NGG/DOX/6.0 > DOX/6.0 > NGP/DOX/3.0 > NGG/DOX/3.0 > DOX/3.0. Furthermore, the results of semiquantitative in **Figures 10C,F** showed the necrotic areas of NGP/DOX/6 and NGP/DOX/6 treatment groups were 1.4 and

1.2 times larger than those of the free DOX·HCl/6 treatment group, respectively.

#### In vivo Security Evaluation

In this study, systematic safety was evaluated by monitoring the physical conditions and body weights changes, by analyzing the pathological morphology of various organs, by detecting the biological parameters from organs and serum, and by examining the BMMR and WBC levels after therapeutics. The in vivo systematic toxicity of DOX was reflected by body weight and histopathology of organs. Similar body weight gain trends were observed in each group of nude mice within the initial 1–10 days (**Figure 4B**). After that, the body weights of the nanogels treatment groups still showed similar

growth trends on day 11–26. This result might be related to that controlled release of nanogels, which resulted less DOX release in the circulation. On the contrary, the body weights of free DOX·HCl treated groups showed significant downward trends within the same time interval, and larger doses cause more weight loss (P < 0.001). Especially the dose of 6.0 mg (kg BW)−<sup>1</sup> DOX treatment group caused severe weight loss, indicating the toxicity of DOX were dosedependent.

The histopathological analyses of major organs were shown in **Figure 7**. Significant neutrophil accumulation and myocardial fiber breakage were observed in the showed in the free DOX·HCl treatment groups. In contrast, the neutrophil accumulation did not occur in the nanogels treatment groups, the myocardial cells were arranged orderly, and the sarcolemma-maintained integrity, probably relate to the reduced accumulation of free DOX·HCl in heart. Moreover, the microregional necrosis of hepatocytes in the free DOX·HCl treatment group indicated that free DOX·HCl had significant hepatotoxicity. On the contrary, less structural interferences were observed in the nanogels treatment groups. Moreover, the nephrotoxicity of free DOX·HCl was also reduced by the nanogels, which was verified by the intact structure of the kidneys in the nanogels treatment groups. These results indicated that the DOX-loaded nanogels could effectively reduce

systematic toxicity, probably due to the satisfactory stability of DOX-loaded nanogels. All the results confirmed that the DOXloaded nanogels had good biocompatibility.

The relevant clinical parameters of heart (CK, CK-MB, and LDH), liver (ALT and AST), and kidney (BUN and Cr) can reflect the function of corresponding organs. In this work, these clinical parameters were detected to verify the safety of DOXloaded nanogels in vivo. As shown, the parameters of all organs except heart in each treated group were within the normal range (**Figure 8**). Free DOX·HCl/6 could cause serious damage to the heart (**Figure 8E**). The relevant parameters in serum were also tested by the commercial ELISA kits. All the clinical parameters of mice treated with DOX-loaded nanogels were equal or lower than normal nude mice treated with NS except CK (**Figure 9**). These results indicated that DOX-loaded nanogels did not cause serious organ dysfunction.

The number of WBC was tested to reflect the influence of chemotherapy drugs on the immune system. As shown in **Figures 10A,D**, the WBC counts of the NS treatment group was increased obviously. On the contrary, the DOX formulation treatment groups did not show increased WBC counts. The results indicated that the treatments with DOX formulations could effectively reduce the inflammation induced by tumor (P < 0.01).

The genotoxicity caused by chemotherapy drugs can be quantified by BMMR (Wang et al., 2018). As shown in **Figures 10B,E**, H&E-stained marrow sections were used to observe bone marrow mononuclear cells. These histopathological sections were also used to calculate BMMRs. Compared with normal mice, the BMMRs increased obviously in Lewis lunggrafted mice. For the mice treated with DOX formulations,

REFERENCES


a dose-related increase in BMMR was observed. Moreover, the BMMRs of the groups treated with free DOX·HCl were significantly greater than those of groups treated with DOXloaded nanogels (P < 0.001). These results demonstrated that the physiological damage caused by free DOX was dose-dependent. Fortunately, DOX-loaded nanogels could effectively mitigate this injury.

#### CONCLUSION

In this work, reduction-responsive DOX-loaded nanogels (NGP/DOX and NGG/DOX) were prepared by classical nanoprecipitation method. In vitro studies showed that both NGP/DOX and NGG/DOX groups exhibited stronger cellular uptake of Lewis cells compared with free DOX treatment groups. Furthermore, all the NGP/DOX and NGG/DOX groups exhibited more efficient antitumor efficacy than the free DOX·HCl treatment groups in the Lewis lung carcinoma grafted nude mouse model. Importantly, all the DOX-loaded nanogels could significantly reduce the systemic toxicity of DOX. Therefore, these polypeptides nanogels with high systemic safety could serve as promising nanodrug delivery platforms for the future lung carcinoma chemotherapy.

#### AUTHOR CONTRIBUTIONS

All authors conceived and designed the study. KN, NL, and YY performed the experiments. KN, NL, YY, CG, YG, and JW analyzed and interpreted the data, and prepared the manuscript.


efficient intravesical chemotherapy of bladder cancer. Adv. Sci. 5:1800004. doi: 10.1002/advs.201800004


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Niu, Li, Yao, Guo, Ge and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-10-00037 February 1, 2019 Time: 16:24 # 15

# Ceramide-Graphene Oxide Nanoparticles Enhance Cytotoxicity and Decrease HCC Xenograft Development: A Novel Approach for Targeted Cancer Therapy

Shi-Bing Wang1,2† , Ying-Yu Ma<sup>2</sup>† , Xiao-Yi Chen<sup>1</sup> , Yuan-Yuan Zhao<sup>1</sup> and Xiao-Zhou Mou1,2 \*

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Xiaowei Zeng, Sun Yat-sen University, China Luying Wang, Beijing Forestry University, China

\*Correspondence:

Xiao-Zhou Mou mouxz@zju.edu.cn †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 30 October 2018 Accepted: 21 January 2019 Published: 08 February 2019

#### Citation:

Wang S-B, Ma Y-Y, Chen X-Y, Zhao Y-Y and Mou X-Z (2019) Ceramide-Graphene Oxide Nanoparticles Enhance Cytotoxicity and Decrease HCC Xenograft Development: A Novel Approach for Targeted Cancer Therapy. Front. Pharmacol. 10:69. doi: 10.3389/fphar.2019.00069 <sup>1</sup> Key Laboratory of Tumor Molecular Diagnosis and Individualized Medicine of Zhejiang Province, Zhejiang Provincial People's Hospital, People's Hospital of Hangzhou Medical College, Hangzhou, China, <sup>2</sup> Clinical Research Institute, Zhejiang Provincial People's Hospital, People's Hospital of Hangzhou Medical College, Hangzhou, China

Despite substantial efforts to develop novel therapeutic strategies for treating hepatocellular carcinoma (HCC), the effectiveness and specificity of available drugs still require further improvement. Previous work has shown that exogenous ceramide can play a key role in inducing the apoptotic death of cancer cells, however, the poor watersolubility of this compound has hampered its use for cancer treatment. In the present study, we used polyethylene glycol (PEG) and polyethylenimine (PEI) co-conjugated ultra-small nano-GO (NGO-PEG-PEI) loaded with C6-ceramide (NGO-PEG-PEI/Cer) as a strategy for HCC treatment. We assessed the biological role of NGO-PEG-PEI/Cer, and we assessed its antitumor efficacy against HCC both in vitro and in vivo in combination with the chemotherapeutic drug sorafenib. We found that NGO-PEG-PEI significantly enhanced the cellular uptake of C6-ceramide. By investigating the mechanism of cellular delivery, we determined that the internalization of NGO-PEG-PEI/Cer progressed primarily via a clathrin-mediated mechanism. The combination of NGO-PEG-PEI/Cer and sorafenib exhibited synergy between these two drugs. Further work revealed that NGO-PEG-PEI/Cer may play a role in subverting multidrug resistance (MDR) in HCC cells by inactivating MDR and Akt signaling. NGO-PEG-PEI/Cer also significantly inhibited tumor growth and improved survival times in vivo, and the synergetic effect of NGO-PEG-PEI/Cer combined with sorafenib was also observed in drug-resistant HCC xenografts. In conclusion, our NGO-PEG-PEI nanocomposite is an effective nano-platform for loading C6-ceramide for therapeutic use in treating HCC, exhibiting high cancer cell killing potency in this tumor model. The NGO-PEG-PEI/Cer/sorafenib combination additionally represents a promising potential therapeutic strategy for the treatment of drug-resistant HCC.

Keywords: hepatocellular carcinoma, graphene oxide, ceramide, apoptosis, drug-resistant

# INTRODUCTION

fphar-10-00069 February 6, 2019 Time: 21:24 # 2

Hepatocellular carcinoma (HCC) is the third most common cause of cancer-related death globally, accounting for approximately 80% of all types of primary liver cancer (Torre et al., 2015). Currently, HCC treatment is largely focused on a combination of surgery, chemotherapy, drug-targeted therapy and radiofrequency ablation as appropriate. These comprehensive treatments can significantly improve prognosis and prolong the life span of patients (Forner et al., 2012). Unfortunately, the majority of patients diagnosed with HCC are not eligible for surgery, leaving systemic therapy as the primary treatment option in those patients with advanced disease (Lencioni et al., 2016; Yu, 2016). Although conventional anticancer drugs have been used for the treatment of HCC, their high toxicity and relative non-specificity impede longterm application (Keating and Santoro, 2009). In addition, several clinical studies have demonstrated that conventional cytotoxic chemotherapy has low response rates and severe side effects (Thomas et al., 2008). Therefore, targeted drug delivery strategies and targeted therapeutics are the current key topics of research interest among those seeking to treat this deadly disease.

Ceramide, the simplest of the sphingolipids, is composed of a sphingosine base and amide-linked acyl chains varying in length from C14 to C26 (Ponnusamy et al., 2010). Ceramides have been reported to act as bioeffectors capable of mediating various cellular processes, including proliferation and apoptosis of cancer cells (Alphonse et al., 2013; Camacho et al., 2013). Work has revealed that ceramide has numerous effects on cell function, with the potential to induce cell growth arrest, senescence, apoptosis, and autophagy (Hannun and Obeid, 2008). Interestingly, ceramide can also mediate alternative pre-mRNA splicing, thereby enabling cells to express proapoptotic isoforms of bcl-x and caspase-9 (Chalfant et al., 2002; Massiello and Chalfant, 2006). Given these findings, ceramide has attracted tremendous attention in the field of cancer therapy as a potentially powerful tumor suppressor (Henry et al., 2013). C6-ceramide in particular has been widely used for the treatment of malignant tumors (Tagaram et al., 2011; Overbye et al., 2017). For example, Adiseshaiah et al. found that the synergistic combination therapy of nanoliposomal C6-ceramide and vinblastine is associated with a disruption of autophagy in HCC and colorectal cancer (Adiseshaiah et al., 2013). Tagaram et al. also demonstrated that ceramide induces p-AKT-dependent apoptosis in human HCC cells in vitro and suppresses xenograft tumor growth in vivo (Tagaram et al., 2011), exerting an inherent tumorkilling effect. However, ceramide is highly hydrophobic, which largely limits its application in vivo, necessitating the search for a suitable carrier for ceramide delivery that does not restrict its pharmacological effects.

Recently, nanoparticle therapy has been identified as a potential multi-modal approach to enhance therapeutic efficacy and reduce side effects associated with cancer treatment (Davis et al., 2008; Goncalves et al., 2013; Tao et al., 2017a, 2018). Nanoparticles are associated with a more targeted localization to tumors and an active mode cellular uptake, making it possible to achieve controlled-release drug delivery and specific gene transfection (Davis et al., 2008; Tao et al., 2013; Zhu et al., 2018). Graphene, a class of two-dimensional carbon nanomaterials with desirable physical and chemical properties, has attracted great interest in many different fields including biomedicine (Feng and Liu, 2011; Chen et al., 2012). Nano-graphene oxide (NGO) is reported to act as potential nano-platform for the delivery of anticancer chemotherapy drugs and genes (Zhang et al., 2010; Zhang L. et al., 2011). NGO was reported to be suitable for loading doxorubicin (DOX) at rates as high as 235% (the weight ratio of loaded drug to carriers) (de Melo-Diogo et al., 2018). However, the toxicology of NGO has been focused recently, which showed that its cellular toxicity in vitro is closely related to its surface functionalization (Hu et al., 2011). Zhang et al. developed DOX-loaded NGO-PEG (Polyethylene Glycol) as a strategy for chemo-photothermal synergistic therapy in one system, which significantly enhanced the therapeutic efficacy of cancer treatment in vivo and in vitro (Zhang W. et al., 2011).

NGO has great potential for use as delivery vehicles designed to enhance cancer treatment, So our collaborator developed PEG and PEI (Polyethylenimine) co-conjugated ultra-small nano-GO (NGO-PEG-PEI) as a novel gene delivery carrier, and found that it showed excellent stability against salts and serum (Feng et al., 2013). In the present study, we used these nanoparticles for loading C6-ceramide, and we found that this formulation allows C6-ceramide to travel through the bloodstream and target tumor cells via enhanced cellular permeability and retention, facilitating its potential clinical use as a novel therapeutic strategy. Additionally, through in vitro and in vivo studies we also investigated the antitumor efficacy and molecular mechanisms of NGO-PEG-PEI/Cer combined with other chemotherapy drugs in HCC.

#### MATERIALS AND METHODS

#### Synthesis and Characterization of NGO-PEG-PEI/Cer

NGO-PEG-PEI was kindly provided by Dr. Kai Yang at the School of Radiation Medicine and Protection (SRMP) of Soochow University (Suzhou, China). Briefly, GO was obtained by oxidation of graphite following the modified Hummers method. Preparation of NGO-PEG-PEI was performed according to previous description (Feng et al., 2013). A mixture of GO solution (0.5 mg/ml) with 6-armed amine-terminated PEG (0.5 mg/ml) was under sonication for 5 min. Then EDC (0.5 mg/ml) was added, after another 5 min sonication, the mixture was stirred gently for 10 min at room temperature. The mixture was stirred for 6 h at room temperature following the second time addition of EDC (1 mg/ml) after being sonicated with PEI (2.5 mg/ml) for 5 min. After that, the mixture was washed with deionized water by 100 nm Milli-Q membrane filter (Millipore, Bedford, MA, United States) 3 times, and we obtained NGO-PEG-PEI re-suspended in water.

NBD C6-ceramide (6-((N-(7-Nitrobenz-2-Oxa-1,3- Diazol-4-yl)amino)hexanoyl)Sphingosine) (N1154, Thermo Fisher Scientific, MA, United States) solution with gradient concentration was prepared and its absorbance at 536 nm was measured. The standard curve was drawn according to different concentrations. Then C6-ceramide was mixed with a certain concentration of NGO-PEG-PEI solution in equal volume and oscillated overnight. After centrifuging for 30 min at 8000 rpm, the absorbance of supernatant was determined, and the concentration of free drug in supernatant was obtained according to the standard curve. Then NGO-PEG-PEI/Cer was prepared according to the maximum loading of C6-ceramide carried by NGO-PEG-PEI.

After loadinging the C6-Ceramide with NGO-PEG-PEI, PBS was added to make the final volume of 1.0 ml. The average size and zeta potential of the NGO-PEG-PEI/Cer complex were then measured with dynamic laser scattering (DLS) and a Zetasizer 3000HS particle analyzer (Malvern Instrument Inc., Worcestershire, United Kingdom), respectively. The sizes and zeta potential values were presented as the average values of three measurements.

#### Cell Culture and Maintenance

The human HCC cell lines HepG2, HuH7, and PLC/PRF/5 were purchased from the Cell Bank of Type Culture Collection of Chinese Academy of Sciences (Shanghai, China). HuH7- SR is sorafenib-resistant HuH7 cell line, which was retained in our lab. All the cell lines were authenticated by shorttandem repeat profiling and cultured in Dulbecco's Modified Eagle's Medium (GIBCO, Carlsbad, CA, United States) supplemented with 10% heat inactivated fetal bovine serum (GIBCO). Cells were incubated in a 5% CO<sup>2</sup> humidified incubator at 37◦C.

#### Cell Uptake of NGO-PEG-PEI/Cer

HCC cells were plated into 6-well plates at a density of 1 × 10<sup>5</sup> cells/well and incubated with NGO-C6 in dulbecco's modified eagle's medium (DMEM) containing 10% fetal bovine serum. The cells were rinsed by PBS and then collected. The uptake ratio of NGO-PEG-PEI/Cer by HCC cells was measured by flow cytometry (Beckman, NJ, United States) ) using NBD labeled on NGO-PEG-PEI/Cer. The whole procedures were operated in dark place.

#### Cell Viability Assay

For cell proliferation analysis, cells were dispensed in 96 well culture plates at a density of 5 × 103 cells/well. After attachment, cells were treated with Ceramide-C6, NGO-PEG-PEI/Cer, Sorafenib or combination therapy with NGO-PEG-PEI/Cer and Sorafenib at given concentration and time. The medium added with PBS was a blank control. Cell survival rate was evaluated by a standard 3-(4,5-dimethylthiazol-2 yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma, St. Louis, MO, United States), Medium was removed and fresh medium containing MTT (5 mg/ml) was added to each well. The cells were incubated at 37◦C for 4 h, after draw off the supernatant of each well carefully and then an equal volume 150 µl of dimethyl sulfoxide (DMSO) was added to each well and mixed thoroughly on concentrating table for 10 min. The absorbance from the plates was read at 595 nm with Cytation 3 Multi-Mode Reader (BioTek, Vermont, United States).

#### Cell Apoptosis Assay

Apoptosis staining Kit containing Annexin V- fluoresceine isothiocyanate (FITC)/Propidium iodide (PI) (KeyGene Biotech, Nanjing, China) was used to detect cell apoptosis according to the manufacture's protocol. Cells were stained with 5 µl Annexin V-FITC and 5 µl PI after 48 h of treatment of NGO-PEG-PEI, C6 or NGO-PEG-PEI/Cer and then keep in dark at room temperature for 15 min. After that, these cells were analyzed by flow cytometer (Novo cyte 3130, ACEA Biosciences, Santiago, CA, United States).

#### Western Blotting Analysis

Cells were harvested in lysis buffer (Beyotime, China) involving 1% Complete Mini-Protease Inhibitor Cocktail (Roche Diagnosis, Switzerland). Protein extractions were quantified using the bicinchoninic acid (BCA) kit (Thermo scientific, MA) and heated for 10 min at 100◦C. Thirty microgram of protein was resolved in 12% SDS-PAGE and transferred to nitrocellulose membrane (Millipore, Germany). After blocked for 1 h at 37◦C, the membranes were immunobloted with different antibodies overnight at 4◦C. Antibodies against Caspase-8, Caspase-9, Caspase-3, recombinant poly ADP ribose polymerase (PARP), P-glycoprotein (P-gp), multidrug resistance 1 (MDR1), AKT serine/threonine kinase 1 (Akt), jun proto-oncogene (c-Jun), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were purchased from Abcam (Shanghai, China). Antibodies against Caspase-8 (1:1000), Caspase-9 (1:1000), Caspase-3 (1:1000), PARP (1:500), XIAP (1:1000), cIAP-1 (1:1000), cIAP-2 (1:1000), Survivin (1:1000), livin(1:1000), P-gp(1:500), and MDR1 (1:1000) were purchased from EMD Millipore Corporation (Billerica, MA, United States). Membranes were then washed with TBST and incubated with Horseradish Peroxidase (HRP) conjugated goat anti-rabbit or anti-mouse antibody (1:5000) for 1 h at room temperature. Finally, blots were detected using ChemiDocTM MP Imaging System (Bio-Rad) with a SuperEnhancedchemiluminescence detection kit (Applygen, Beijing, China).

#### Clonogenic Assay

The ability of the HCC cells to generate in vitro colonies was determined using clonogenic assay. Briefly, cells were incubated six-well plates at a concentration of 500 cells/well after treatment with NGO-PEG-PEI, C6 and NGO-PEG-PEI/Cer. The medium was regularly changed for 2 weeks until colony formation. Then the colonies were gently washed with phosphate buffer saline (PBS) after the supernatants removed, fixed with cold methanol for 20 min, and stained with crystal violet 0.1% in PBS at room temperature for 10 min followed by air-drying.

Images were captured and the total number of colonies/well was counted.

# In vivo Antitumor Effect and Systemic Toxicity

Male Balb/c mice (6 weeks old) were obtained from Laboratory Animal Center of Zhejiang Chinese Medical University (Hangzhou, China). Tumor bearing mice were established by subcutaneous injection of 5 × 10<sup>6</sup> HCC cells (HuH7 or HuH7-SR) in 200 µl PBS into the flank region of Balb/c mice. The dimension of tumors was monitored by digital calipers. Average tumor volume is about 90 mm<sup>3</sup> . Then the mice were randomized into 4 different treatment groups (7 mice per group): control group (PBS), NGO-PEG-PEI group, C6-ceramide group, NGO-PEG-PEI/Cer group. Then the mice bearing tumors in different groups were intravenous injected with PBS, NGO-PEG-PEI, C6-ceramide, NGO-PEG-PEI/Cer, respectively. After treatment, tumor volumes were tracked every 3 days by digital caliper measurements.

Control and nanocomposites-treated mice were sacrificed at 18 days after treatment. Major organs of those mice were collected, fixed in 4% formalin, conducted with paraffin embedded sections, stained with hematoxylin and eosin (H&E) and immuhistochemistry, and examined under a digital microscope.

This study was carried out in accordance with the recommendations of Laboratory Animal Center of Zhejiang Chinese Medical University, and the project was approved by the ethics committee of Zhejiang Provincial People's Hospital.

#### Statistical Analysis

The Statistical Package for the Social Sciences (version 13.0; SPSS Inc., Chicago, IL, United States) was used to perform all statistical analyses. Continuous data were analyzed using paired t-test or Wilcoxon rank test. Categorical data were analyzed using χ <sup>2</sup> or Fisher's exact test. Survival analysis was estimated by Kaplan–Meier method accompanying the log-rank test to calculate differences between the curves.

For all tests, P-values were obtained from two-tailed statistical tests and p-values less than 0.05 were considered statistically significant.

#### RESULTS

#### Synthesis and Characterization of NGO-PEG-PEI/Cer Nanoparticles

Graphene oxide (GO) has been extensively explored in nanomedicine for its excellent physiochemical, electrical, and optical properties. In this study, polyethylene glycol (PEG) and polyethylenimine (PEI) were covalently conjugated to GO via amide bonds, yielding a physiologically stable dualpolymer-functionalized nano-GO conjugate (NGO-PEG-PEI) of ultra-small size. The synthesis and characterization of the NGO-PEG-PEI has been reported elsewhere (Feng et al., 2013). The NGO-PEG-PEI/Cer nanoparticles were generated via electrostatic interaction between the positively charged cationic NGO-PEG-PEI and the negatively charged C6-ceramide surface (**Supplementary Figure 1**). To determine optimal conditions for generating the NGO-PEG-PEI/Cer complex, the absorbance of the NGO-PEG-PEI/Cer were measured with various concentrations of the C6-Ceramide by UV Spectrophotometer. The size distribution and zeta potentials of the NGO-PEG-PEI/Cer complex were measured with various concentrations of the DA3 polymer by DLS and a zeta potential analyzer, respectively (**Supplementary Figure 1**).

We incorporated trace amounts of C6-ceramide into NGO-PEG-PEI formulations to quantify the amount of NGO-PEG-PEI/Cer delivery compared with naked C6-ceramide administration. Pharmacokinetic studies revealed that NGO-PEG-PEI formulations delivered C6-ceramide more effectively and efficiently than did mock administration of C6-ceramide in the presence of 10% FBS (**Figure 1A**). NGO-PEG-PEI delivery resulted in a threefold increase in ceramide accumulation in HepG2 cells, with a maximal accumulation observed at approximately 18 h. Additionally, confocal microscopy demonstrated that NGO-PEG-PEI/Cer nanoparticles were present in the cytoplasm and enhanced transduction efficiency of C6-Ceramide in cancer cells (**Figure 1B**).

We next investigated the mechanism by which C6-ceramide is released or transferred from NGO-PEG-PEI vehicles into cellular membranes. To elucidate the mechanism of cellular delivery of NGO-PEG-PEI/Cer, we studied the effects of inhibitors of various endocytotic mechanisms, including chlorpromazine (clathrin inhibitor), amiloride (actin inhibitor), methylβ-cyclodextrin (MβCD, caveolae inhibitor), genistein (PTK inhibitor), and dynasore (dynamin inhibitor). Cells without inhibitor pretreatment were studied as controls under the same experimental conditions. Relative green fluorescence levels were measured after transduction with NGO-PEG-PEI/Cer complexes, and the experimental results suggested that the cell entry process for NGO-PEG-PEI is via clathrin-mediated endocytosis, as chlorpromazine led to a 32.36% inhibition of internalization (**Figures 1C,D**).

# In vitro Antitumor Activity of NGO-PEG-PEI/Cer

fphar-10-00069 February 6, 2019 Time: 21:24 # 6

A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed 48 h post-infection with NGO-PEG-PEI/Cer to evaluate the cytotoxicity of the nanocomposites in the HepG2, HuH7 or PLC/PRF/5 cell lines in the presence of 10% FBS. The results indicated a significantly higher inhibition of cell growth in cells treated with NGO-PEG-PEI/Cer relative to those treated with C6-ceramide in a dose-dependent fashion (**Figure 2A**).

To further confirm the inhibitory effect of NGO-PEG-PEI/Cer on the proliferation of HCC cells, we conducted a clone formation assay and found that NGO-PEG-PEI/Cer could effectively inhibit the proliferation of HCC cells (**Figure 2B**).

FIGURE 3 | NGO-PEG-PEI/CER induced apoptosis in HCC cells in vitro. (A) Apoptosis was analyzed via Annexin V-FITC/PI double staining. HepG2, HuH7, or PLC/PRF/5 cells were infected with NGO-PEG-PEI/Cer (10 µM), C6-ceramide (10 µM), or NGO-PEG-PEI (100 µg/ml) for 24 h. Florescence was then analyzed by flow cytometry. Data are presented as mean ± SD of three separate experiments. (B) HepG2 cells were infected with NGO-PEG-PEI/Cer (10 µM), C6-ceramide (10 µM), or NGO-PEG-PEI (100 µg/ml) for 24 h. Whole cell extracts were prepared and immunoblotted to detect caspase pathway activation. GAPDH was used as a loading control. <sup>∗</sup>P < 0.05.

FIGURE 4 | NGO-PEG-PEI/Cer enhances sorafenib-mediated growth inhibition in HCC cells. (A) Cells were treated with NGO-PEG-PEI/Cer and/or sorafenib for 48 h, and cell viability was then determined by MTT assay. Data are presented as mean ± SD of three separate experiments. (B) The potential synergistic effect of sorafenib combined with NGO-PEG-PEI/Cer on HCC cells was assessed by Chou–Talalay Combination Index (CI) analysis using the CalcuSyn software. The middle curve line indicates the simulated combination index values, which are expressed as the log10 (CI) ± 1.96 SD, encircled by two lines of algebraic evaluation of the 95% confidence intervals. The log10 (CI) values attained at given fractional affects represent an antagonism between the treatments when >0, an additive efficiency when equal to 0 and a synergism when <0. This was quantified by CIN analysis and expressed as CIN versus fraction affected. Where calculable, 95% confidence intervals are shown. (C) The sorafenib-resistant HuH7 cell line (HuH7-SR) was used to illuminate the mechanism by which NGO-PEG-PEI/Cer may influence sorafenib-resistant HCC. NGO-C6(10 µM), C6-ceramide (10 µM), or NGO (100 µg/ml) was used to treat HuH7-SR cells. Uninfected cells served as control. Forty eight hours later, whole cell extracts were prepared and immunoblotted. GAPDH was used as a loading control.

# NGO-PEG-PEI/Cer Treatment Induces Apoptosis in vitro

To address the underlying mechanism by which NGO-PEG-PEI/Cer induces cytotoxicity, we evaluated NGO-PEG-PEI/Cer associated apoptosis in vitro by flow cytometry. We observed significant increases in apoptosis in the HepG2, HuH7 or PLC/PRF/5 cell lines treated with NGO-PEG-PEI/Cer relative to those threated with C6-ceramide, NGO-PEG-PEI or PBS (**Figure 3A**).

We further evaluated apoptosis by assessing the expression of apoptosis-related proteins in HepG2 cells at 48 h postinfection using Western blotting analysis. The results indicated a significant increase in the activation of caspases 3, 8, and 9, and increased PARP cleavage in NGO-PEG-PEI/Cer-treated cells (**Figure 3B**). Taken together, these findings indicated that NGO-PEG-PEI/Cer effectively induced apoptosis via caspase activation.

### Combined Treatment With Sorafenib and NGO-PEG-PEI/Cer Results in Synergistic Efficacy

To determine whether NGO-PEG-PEI/Cer enhances the cytotoxic effect of sorafenib, we analyzed the viability of HCC cells after co-treatment with NGO-PEG-PEI/Cer and sorafenib. The HepG2, HuH7, and PLC/PRF/5 cells were treated with

sorafenib (0.125, 0.25, 0.5, or 1 µM) with or without NGO-PEG-PEI/Cer (1.25, 2.5, 5, or 10 µM). The combination of NGO-PEG-PEI/Cer with sorafenib significantly inhibited cell growth as compared with treatment with sorafenib or NGO-PEG-PEI/Cer alone (**Figure 4A**). Next, the synergistic effects of sorafenib combined with NGO-PEG-PEI/Cer on HCC cells were quantified by combination index (CIN) analysis and expressed as CIN versus fraction affected in **Figure 4B**. These results revealed that the combination of sorafenib and NGO-PEG-PEI/Cer has a synergistic tumor killing effect.

These results raised the question of the mechanism by which NGO-PEG-PEI/Cer influences sorafenib-resistant HCC cell lines. Therefore, we generated a sorafenib-resistant HuH7 cell line (HuH7-SR) and assessed the expression of multidrug resistancerelated proteins or Akt, phospho-Akt, c-Jun, phospho-c-Jun of these HuH7-SR cells treated with NGO-PEG-PEI/Cer, C6 ceramide, or NGO-PEG-PEI by Western blot analysis in the presence of 10% FBS. Compared with the sensitive HuH7 cells, HuH7-SR cells exhibited markedly elevated levels of multidrug resistance-related proteins (P-gp and MDR1) and Akt, phospho-Akt, c-Jun, phospho-c-Jun. Our results further demonstrate that NGO-PEG-PEI/Cer has the capacity to reduce levels of multidrug resistance-related proteins and Akt, phospho-Akt, c-Jun, phospho-c-Jun (**Figure 4C**). Taken together, these findings indicated a synergistic repressive effect of the combination of sorafenib and NGO-PEG-PEI/Cer treatment on HCC cell proliferation.

#### Enhanced Cytotoxic Effect of Co-treatment With Sorafenib and NGO-PEG-PEI/Cer in vivo

We developed two hepatoma carcinoma tumor xenograft mouse models using the HuH7 cells and HuH7-SR cells in BALB/c athymic nude mice to evaluate the effects of NGO-PEG-PEI/Cer treatment or co-treatment with sorafenib and NGO-PEG-PEI/Cer in vivo (**Figure 5A**). Anti-tumor efficacy was evaluated by plotting tumor growth curves over a 42 or 45 days observation period. The mean tumor volume was significantly decreased in mice injected with sorafenib, NGO-PEG-PEI/Cer, and the combination therapy relative with those injected with PBS (**Figures 5B,C**). Furthermore, co-treatment of sensitive HuH7 and HuH7-SR cells with sorafenib and NGO-PEG-PEI/Cer was more effective than sorafenib (P = 0.001 and 0.002, respectively) and NGO-PEG-PEI/Cer alone (P = 0.001 and 0.003, respectively). Co-treatment with sorafenib and NGO-PEG-PEI/Cer was also associated with a higher survival rate than treatment with PBS, sorafenib, or NGO-PEG-PEI/Cer (**Figures 5D,E**).

The tumor histopathological changes were further evaluated by hematoxylin and eosin (H&E) staining and immunohistochemistry (IHC). The combined treatment with sorafenib and NGO-PEG-PEI/Cer resulted in greater cytotoxicity than either single treatment as evidenced by H&E staining. Moreover, very high Caspase-3 expression was evident in the tumor tissues from mice which received the combined treatment based on IHC staining with an anti-Caspase-3 antibody (**Figure 6**). Results from this experiment showed significantly higher rates of apoptosis in the combination treatment group when compared with either individual treatment (**Figure 6**).

# DISCUSSION

Ceramide is reported to modulate cell death, cycle arrest, metastasis, stress responses, and pro-inflammatory responses in cancer cells (Saddoughi and Ogretmen, 2013; Kitatani et al., 2016), especially in HCC (Morales et al., 2007). Despite its key role in regulating tumor cell growth and death, its cell impermeability and its tendency to undergo precipitation in aqueous solutions have limited the use of ceramide as a therapeutic agent.

With the development of nanotechnology, application of nanomedicine in drug delivery becomes an area of rapid growth and advancement given its significant ability to enhance therapeutic efficacy, minimize side effects of drugs, and enhance drug bioavailability in vivo (Chen et al., 2016; Taghdisi et al., 2016; Tao et al., 2016, 2017b). Therefore, many nanomaterials – namely liposomes, micelles, dendrimers, carbon nanotubes, polymers, inorganic metallic nanolayers, and graphene oxides – have been explored in efforts to design nanocarriers for different drugs (Elgadir et al., 2015; Pattni et al., 2015). Recently, NGO has been explored in the field of biomedicine, which was found to be promising for drug delivery (Huang et al., 2012; Jayakumar et al., 2012). It is attractive in part due to its ease of synthesis, controlled particle size, and high surface area for drug loading. Although NGO without additional surface coatings appears

FIGURE 6 | Histopathological analysis. Subcutaneous HuH7 or HuH7-SR tumors were collected 18 days after injection, and sections were analyzed by the indicated methods. H&E staining showed that tumor tissues treated with the NGO-PEG-PEI/Cer or combination of sorafenib and NGO-PEG-PEI/Cer exhibited the greatest amount of cell death. An immunohistochemical analysis demonstrated there was strong expression of caspase-3 in xenografts treated with NGO-PEG-PEI/Cer as well as in those from the combined therapy group.

to exhibit dose-dependent toxicity, well functionalized NGO with biocompatible coatings such as PEG has been found to reduce obvious toxicity (Yang et al., 2013a,b). Other studies have explored the use of NGO for DNA plasmid and siRNA delivery after being functionalized by PEI (Feng et al., 2011; Zhang L. et al., 2011). Our collaborator also found that PEG and PEI functionalized NGO may enhance gene delivery (Feng et al., 2013). In the present study, we used NGO-PEG-PEI to carry C6, and we determined that the maximum drug loading capacity of 100 µg/ml NGO-PEG-PEI was 56 µM C6-ceramide. Next, in order to verify the enhanced intracellular trafficking of NGO-PEG-PEI/Cer, we assessed the cell uptake of NBD labeled NGO-PEG-PEI/Cer. Our results revealed that NGO-PEG-PEI significantly enhanced the cellular uptake of C6 and resulted in a threefold increase in ceramide accumulation with a maximal accumulation observed after approximately 18 h, which revealed that NGO-PEG-PEI formulations delivered C6-ceramide more effectively and efficiently. The mechanism of cellular delivery was also explored, established that this intracellular internalization of NGO-PEG-PEI/Cer progressed mainly via a clathrinmediated mechanism.

C6-ceramide was cytotoxic and anti-proliferative when employed against a panel of human melanoma cells (Jiang et al., 2016), as well as against cervical cancer and colon cancer cells (Overbye et al., 2017). Nanoliposomal C6-ceramide has also been reported to induce cell apoptosis of HCC cells in vitro, concomitant with an accumulation of cells in the G2 phase of the cell cycle (Tagaram et al., 2011). In the present study, cell viability and toxicity were also detected after treatment with NGO-PEG-PEI/Cer, and our results showed that NGO-PEG-PEI/Cer exhibited high tumor cell killing potency as it was capable of both reducing HCC cell proliferation and increasing apoptosis. These findings thus suggest that NGO-PEG-PEI/Cer may induce potent, specific antitumor cytotoxicity.

Sorafenib, which is the first multi-target, multi-kinase inhibitor to be developed, is systematically used in the treatment of advanced liver cancer and has been proven to be effective (Wilhelm et al., 2006; Cheng et al., 2009). However, many patients do not respond to sorafenib, or they develop drug resistance after several months of sorafenib treatment. Therefore, it is urgent that enhancers or synergistic agents be identified for combination use with sorafenib to improve the clinical treatment of HCC. In our study, NGO-PEG-PEI/Cer combined with sorafenib was utilized for treating HCC, and the CI50 was determined to establish a synergistic effect. Our results showed that combination treatment with both drugs showed clear synergism effects superior to single drug treatment. NGO-PEG-PEI/Cer combined with sorafenib displayed an overall CI50 value <1, confirming the synergy between these two drugs. Further work revealed that the NGO-PEG-PEI/Cer/sorafenib combination may exert anti-multidrug resistance (anti-MDR) activities in HCC cells by significant inactivation of MDR and Akt signaling. These results illustrated that NGO-PEG-PEI/Cer combined with sorafenib can mediate an efficient synergistic therapeutic effect for antitumor therapy in drug-resistant HCC.

To evaluate the effects of NGO-PEG-PEI-C6 on HCC tumorbearing mouse xenografts, survival time and tumor volume were evaluated, revealing that NGO-PEG-PEI/Cer significantly delayed tumor growth and improved survival times. Li et al. previously found that injection of nanoliposome-loaded C6 ceramide slowed tumor growth by reducing proliferation and increasing apoptosis in HCC (Li et al., 2018), suggesting that delivery system of ceramide via nanoparticle may be an effective strategy for the treatment of human HCC in vitro and in vivo. The synergetic effects of NGO-PEG-PEI/Cer combined with sorafenib were also evident in our mouse model, with evidence of tumor growth inhibition and improved overall survival improvement in mice receiving the combination regimen. These findings suggest that the NGO-PEG-PEI/Cer/sorafenib combination represents a potential therapeutic strategy for the treatment of drug-resistant HCC in vivo.

In summary, using nanotechnology-based advances in drug delivery, we utilized NGO-PEG-PEI nanocomposites as an effective nano-platform for loading C6-ceramide for therapeutic treatment of HCC. This formulated composite exhibited excellent anti-cancer efficacy in vitro and in vivo, facilitating its potential clinical use. Furthermore, combined treatment with NGO-PEG-PEI/Cer and sorafenib achieved a superior therapeutic effect in drug-resistant HCC, suggesting that NGO-PEG-PEI/Cer has great potential to treat drug-resistant HCC when used as a synergistic agent in combination with sorafenib.

# AUTHOR CONTRIBUTIONS

X-ZM conceived and designed the experiments. S-BW and Y-YM carried out the majority of experiments and drafted the manuscript. X-YC analyzed the results and revised the manuscript. Y-YZ collected and analyzed the data. All authors read and approved the final manuscript.

#### FUNDING

This work was supported by the Zhejiang Provincial Natural Science Foundation of China (No. LQ16H160017 to Y-YM, No. LY17H180010 to X-YC), the National Natural Science Foundation of China (No. 81672430 to X-ZM, No. 81602706 to S-BW), Funds of Science Technology Department of Zhejiang Province (No. LGF18H160026 to Y-YZ).

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar.2019. 00069/full#supplementary-material

FIGURE S1 | Synthesis and characterization of NGO-PEG-PEI/Cer. (A) The zeta potential value of the NGO-PEG-PEI were measured. (B) A schematic illustration showing the synthesis of NGO-PEG-PEI conjugate and the preparation of NGO-PEG-PEI/Cer complex. (C) The absorbance of the NGO-PEG-PEI/Cer were measured with various concentrations of the C6-Ceramide by UV Spectrophotometer. (D) The average size (nm) of the NGO-PEG-PEI/Cer complex was measured.

#### REFERENCES

fphar-10-00069 February 6, 2019 Time: 21:24 # 10



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Wang, Ma, Chen, Zhao and Mou. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Lipid Nanoparticles-Encapsulated YF4: A Potential Therapeutic Oral Peptide Delivery System for Hypertension Treatment

Shengnan Zhao<sup>1</sup>† , Jinhua Li1,2† , Yang Zhou1,2† , Lingjing Huang<sup>3</sup> , Yanfei Li<sup>3</sup> , Juanjuan Xu<sup>3</sup> , Chunmei Fu<sup>3</sup> \*, Xia Guo<sup>2</sup> \* and Jian Yang<sup>1</sup> \*

<sup>1</sup> School of Applied Chemistry and Biological Technology, Shenzhen Polytechnic, Shenzhen, China, <sup>2</sup> Key Laboratory of Birth Defect and Related Disorders of Women and Children, Department of Pediatric Hematology/Oncology, West China Second University Hospital, Sichuan University, Chengdu, China, <sup>3</sup> Key Laboratory of Drug Targeting and Drug Delivery System (Ministry of Education), West China School of Pharmacy, Sichuan University, Chengdu, China

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Yanqi Ye, University of North Carolina at Chapel Hill, United States Wenliang Li, Jilin Medical University, China

#### \*Correspondence:

Chunmei Fu fuchunmei@scu.edu.cn Xia Guo guoxkl@163.com Jian Yang jiany@szpt.edu.cn †These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 27 September 2018 Accepted: 25 January 2019 Published: 19 February 2019

#### Citation:

Zhao S, Li J, Zhou Y, Huang L, Li Y, Xu J, Fu C, Guo X and Yang J (2019) Lipid Nanoparticles-Encapsulated YF4: A Potential Therapeutic Oral Peptide Delivery System for Hypertension Treatment. Front. Pharmacol. 10:102. doi: 10.3389/fphar.2019.00102 Drugs are administered orally in the clinical treatment of hypertension. Antihypertensive peptides have excellent angiotensin converting enzyme inhibitors activity in vitro. However, the poor oral bioavailability and therapeutic effect of antihypertensive peptides were mainly caused by rapid degradation in gastrointestinal and the short circulation time in blood, which remain to be further optimized. Therefore, the novel oral peptide delivery system is urged to improve the oral absorption and efficacy of peptide drugs. In this work, Tyr-Gly-Leu-Phe (YF4)-loaded lipid nanoparticles (YF4-LNPs) combined the advantages of polymer nanoparticles and liposomes were developed, which could greatly enhance the oral bioavailability and ameliorate the sustained release of peptide drug. YF4 loaded nanoparticles (YF4-NPs) were firstly prepared by a double-emulsion internal phase/organic phase/external phase (W1/O/W2) solvent evaporation method. YF4-NPs were further coated by membrane hydration-ultrasonic dispersion method to obtain the YF4-LNPs. The optimal YF4-LNPs showed a small particle size of 227.3 ± 3.8 nm, zeta potential of −7.27 ± 0.85 mV and high entrapment efficiency of 90.28 ± 1.23%. Transmission electronic microscopy analysis showed that the core-shell lipid nanoparticles were spherical shapes with an apparent lipid bilayer on the surface. Differential scanning calorimetry further proved that YF4 was successfully entrapped into YF4-LNPs. The optimal preparation of YF4-LNPs exhibited sustained release of YF4 in vitro and a 5 days long-term antihypertensive effect in vivo. In summary, the lipid nanoparticles for oral antihypertensive peptide delivery were successfully constructed, which might have a promising future for hypertension treatment.

Keywords: lipid nanoparticles, antihypertensive peptide, hypertension, oral administration, sustained release, continuously antihypertensive effect

# INTRODUCTION

Nanotechnology has been widely used to improve the oral absorption and therapeutic efficacy of small molecule peptide drugs, which have shown tremendous potentials but great challenges (Olbrich et al., 2001; Tan et al., 2009; Li et al., 2013; Yang et al., 2013; Thi et al., 2015; Wang et al., 2018). Poly-(lactic-co-glycolic) acid (PLGA) nanoparticles have attracted much attention, owing to

**223**

the unique properties of biodegradability, biocompatibility, and sustained release (Musumeci et al., 2006). Liposomes also have been applied for drug delivery due to the superior biocompatibility, drug absorption and nontoxicity (Wang and Liu, 2013). Unfortunately, both of them have some unsatisfied disadvantages in peptide delivery, such as leakage and safety issues, which seriously limited the delivery efficiency. It is necessary to deal with these problems for improving the oral delivery efficiency of peptides (Luo et al., 2006).

Lipid coating PLGA nanoparticles system (lipid nanoparticles, LNPs) has not only combined the advantages of polymer nanoparticles and liposomes but efficiently avoided the defects of them (Muller and Keck, 2004; Alavi et al., 2017). The dual advantages of the particles and vesicle make it an excellent oral drug carrier with high biocompatibility and sustained release (Xie et al., 2018). In this system, drugs can be efficiently encapsulated in the nanoparticles core and/or the lipid bilayers, resulted in increased drug load capability. The drug diffusion rate can be delayed by the polymer core. Additionally, the stability of the LNPs can be further increased by the lipid shell (Nogués et al., 2006). However, it is still challenging to achieve high encapsulation efficiency (EE) and decent particle size when incorporating hydrophilic drugs into LNPs.

Hypertension has become a strong risk factor for cardiovascular disease and affected almost 2 billion people. In detail, it has highly correlated with the heart attack, cerebral hemorrhage, stroke, kidney failure and blindness (Kjeldsen, 2017). Angiotensin-converting enzyme (ACE), located in various tissues, is potent to affect the cascade of process that trigger the increasing of blood pressure. Hence, ACE inhibitors (ACEIs) are commonly used to decrease blood pressure (Girgih et al., 2016). However, the synthetic ACEIs were reported to produce negative side effects such as dry cough, angioedema, itch, diarrhea (Tenenbaum et al., 2000; Abassi et al., 2009), and renal impairment. Alternatively, natural antihypertensive peptide fragments have gradually entered the field of human vision due to the sound security.

YF4, a polypeptide extracted from milk protein, has been reported for the opioid activity for the first time in 1986 (Yoshikawa et al., 1986). Other research groups have conducted other pharmacological studies and found that it had hypotensive activity as well (Espejo-Carpio et al., 2013). However, the antihypertensive peptides are usually accompanied by extremely poor stability and easily degraded by gastric acid and pepsin in the gastrointestinal (GI). At present, the existing peptide drugs in clinical are given mainly by injection (Boelsma and Kloek, 2010; Majumder et al., 2013). Because of the short circulation time (Agrawal et al., 2015) and degradation of polypeptide drugs in the GI, the patients need to be injected frequently which results in the poor compliance (Cleland et al., 2012). To address these issues, novel oral peptide delivery systems are urged to improve their stability in the stomach and intestines in order to improve its therapy effect with a convenient and economic way (Niu et al., 2018).

In the present study, we chose the antihypertensive peptide YF4 as a model drug to systematically optimize the lipid nanoparticles and investigate it's potential in oral peptide delivery. The formulation parameters of YF4-LNPs were systematically investigated. The physic-chemical properties including particle size, surface charge, EE, Transmission electron microscope (TEM), Differential scanning calorimeter (DSC), and in vitro release profile were then characterized systemically. Additionally, the in vivo antihypertensive efficacy was assessed on spontaneously hypertensive rats.

# MATERIALS AND METHODS

# Materials

YF4 (purity > 99%) was gained from Phtdpeptides Co., Ltd. (Zhengzhou, China). PLGA (MW = 15 kDa; LA/GA = 75:25) was purchased from Jinan Daigang Biomaterial Co., Ltd. (Jinan, China). Poly (vinyl alcohol) (PVA, MW = 30–70 kDa, HD, 80%) and TPGS were obtained from Sigma-Aldrich (St. Louis, MO, United States). Soybean phosphatidylcholine (SP) was gained from Shanghai A.V.T. Pharmaceutical Co., Ltd (Shanghai, China). Carbomer 996 was supplied by Hang Zhou Carbokar Import & Export Co., Ltd. (Hang Zhou, China). Poloxamer 188 was acquired from BASFSE (Germany). mPEG2000-Chol (purity > 98.6%) was synthesized by State Key Laboratory of Biotherapy, Sichuan University. Cholesterol was purchased from Shanghai source poly Biotechnology Co., Ltd.). All other reagents were of analytical grade and were used unless otherwise stated.

# Animals

Twelve weeks old male spontaneously hypertensive rats (SHRs) with a circa weight between 250 and 320 g were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. After a week of adaption, animals were admitted to experiments (six per group). Animals were maintained under 12 h dark and light cycles at controlled temperature, 55% humidity, 22◦C. The rats were free access to food and water. All the experiments were approved and supervised by the State Key Laboratory of Biotherapy Animal Care and Use Committee (Sichuan University, Chengdu, Sichuan, China). In this study, groups of rats were used for the oral administration of saline group, Captopril group vs. YF4 group, YF4 vs. YF4-NPs vs YF4-LNPs.

# Preparation of YF4-LNPs

YF4-NPs were prepared by an easy and controllable method. The YF4-NPs were further coated by membrane hydration-ultrasonic dispersion method to obtain the lipid nanoparticles (YF4-LNPs). Firstly, the core of YF4-LNPs (YF4-NPs) was prepared by a double-emulsion internal phase/organic phase/external phase (W1/O/W2) solvent evaporation method (Yu et al., 2016). In brief, 50 µL YF4 (10 mg/mL) deionized water solution was used to form the inner aqueous phase. The 1 mL organic phase (mixed solution of acetone and acetonitrile, initial volume ratio was 5:1) containing PLGA was used as the organic phase, and then mixed with the inner aqueous phase by probe sonication in ice bath to form the primary W1/O emulsion. The primary emulsion was then added to the 4 mL external phase solution and further sonicated to obtain the final W1/O/W2 double emulsion. The organic phase in the ultimate emulsion was rapidly removed by evaporation under vacuum at 37◦C to get YF4-NPs. At the same time, SP (lipid), cholesterol (Chol) and mPEG2000-cholmixture (the mass ratio was 4:1:0.25) were dissolved in 2 mL chloroform at a certain proportion. The organic solvents were subsequently removed using a rotary evaporator (R-201Shanghai Shen Shun Biological Technology Co., Ltd.) to produce a thin film of lipid at 37◦C. The lipid film was hydrated with 4 mL prepared YF4- NPs for 1 h at 60◦C to obtain a suspension. YF4-LNPs were finally acquired by Ultrasonic cell crushing and isolating machine (VCX130, American Systems on ICs).

# Characterization of YF4-LNPs

#### Particle Size and Zeta Potential

fphar-10-00102 February 15, 2019 Time: 17:48 # 3

The mean particle size, size distribution and ζ potential were measured by Zetasizer (Zetasizer Nano-ZS 90; Malvern Instruments Ltd., Malvern, United Kingdom) at 25◦C. The prepared YF4-LNPs were diluted with deionized water and experiments were conducted in triplicate. All the data were presented as mean ± SD.

#### Entrapment Efficiency and Drug Loading

Briefly, the obtained supernatant by ultrafiltration centrifuging the colloidal suspension during the preparation of YF4-LNPs was stored to determine the EE and drug loading capacity (DL) of YF4-LNPs as previously described (Yu et al., 2016). The untrapped YF4 in supernatant was determined by high performance liquid chromatography (HPLC, Waters Alliance 2695) to calculate the amount of YF4 unpackaged into YF4-LNPs. Analyses were performed in triplicate and the values were expressed as mean ± SD. EE and DL of YF4 were calculated as following formulas:

$$\begin{aligned} \text{EE\%} &= \\\\ \left(\frac{\text{TotalYF4amount} - \text{Amount ofYF4insupernatant}}{\text{TotalYF4amount}}\right) & \times 100 \\\\ \text{DL\%} &= \end{aligned}$$

$$\left(\frac{\text{TotalYF4amount} - \text{Amount ofYF4insuperrantant}}{\text{Totalweight} \|\text{liquidanopparticles}}\right) \times 100$$

#### Stability

YF4-LNPs were stored at 4◦C for two weeks to investigate the preliminary stability. The changes of particle size and EE of YF4- LNPs were examined by the methods described above. Analyses were performed in triplicate and the values were expressed as mean ± S.D.

#### Appearance and Morphology

The appearance and Tyndall effect of YF4-LNPs were observed by a digital camera. The morphology of the YF4-LNPs was examined by TEM (H-600, Hitachi, Japan). In brief, the prepared samples were diluted to proper concentration by deionized water and then were placed on a copper electron microscopy grid and negatively stained with 2wt% phosphotungstic acid solution for 30 s for observation.

#### Differential Scanning Calorimetry (DSC)

The physical state of YF4 loaded in YF4-LNPs was verified by DSC (200PC, Netzsch, Karlsruhe, Germany). Freeze-dried YF4- LNPs, blank LNPs, YF4, and the physical mixture of blank LNPs and free YF4 with the same mass ratio as those in YF4-LNPs were heated from 75 to 250◦C at a heating ramp of 20◦C/min under nitrogen atmosphere at a flow rate of 50 mL/min.

#### In vitro Release Studies

The in vitro release profiles of YF4-LNPs were investigated in PBS buffer at pH 1.0 (simulated gastric fluid), pH 4.5, pH 6.8 (simulated different intestinal fluids), and pH 7.4 (physiological pH condition) by dynamic dialysis method (Boelsma and Kloek, 2010). In brief, free YF4 solution, YF4-NPs, and YF4-LNPs were first dispersed in release media in dialysis bags (MWCO 3000) and then were shaken at 37◦C with a speed of 100 rpm. One milliliter buffer was removed and replaced with equal volume fresh release medium at 0, 2, 4, 8, 12 h, respectively. The content of YF4 was measured by HPLC after centrifugation for 10 min at 13,000 rpm as described in section of "Entrapment efficiency and drug loading." Analyses were performed in triplicate and the values were expressed as mean ± S.D.

#### In vivo Antihypertensive Efficacy

Before the experiment, we first verified the antihypertensive activity of YF4 in SHRs. Simultaneously, the captopril and the saline water were used as the positive control and control group, respectively. SHRs were randomly divided into three groups (n = 6). Control group were treated with 0.9% saline water, and the other groups were treated with captopril (5 mg/kg) or the same dose of YF4 dissolved in saline water. All rats were administered at a single oral dose. Blood pressure was indirectly recorded using the tail-cuff method (BP-2010A, Softron Beijing Biotechnology, China) at each of the following times: 0, 2, 4, 8, 12, and 24 h after administration.

In YF4-LNPs pharmacodynamics study, the SHRs were randomly divided into three groups (n = 6). Control group were treated with 0.9% saline water, the other groups of SHRs were treated withYF4 and YF4-LNPs. All the groups of rats were given the same dose at 0.8 mg/kg body weight. Blood pressure was indirectly recorded as described above. Aside from free YF4 group, all the other groups were needed to detect blood pressure at more time points every day after administration until the pressure reached to normal level. All the SBP were measured for three times to get the average value at each time point.

#### Statistical Analysis

The obtained data were analyzed using the Graph Pad Prism 5. Data were analyzed by one-way analysis of variance. p < 0.05 was considered a statistically difference, and p < 0.01 was considered a statistically significant difference.

# RESULTS

# Characterization of YF4-LNPs

Several factors that could affect the features of YF4-LNPs were systematically optimized to produce the desirable YF4-LNPs,

including the PLGA concentration, the volume of acetone in the organic phase, the volume of the inner aqueous medium, the pH of the aqueous medium, the external water phase, the PVA concentration, the PVA volume, the ultrasonic time, PLGA to lipid material ratio, and the ultrasound power. Firstly, the influence of PLGA concentration on the EE and particle size was investigated in **Figure 1A**.The EE and the particle size were positively correlated with the PLGA amount within the range of 10–50 mg. **Figure 1B** showed that the EE was dually increased and then suddenly decreased when acetone volume was above 200 µL while the particle size was negatively correlated with the increasing of acetone volume in the organic phase. In **Figure 1C**, the EE firstly increased and then decreased dramatically with the increase of inner aqueous volume, whereas the particle size was negligibly affected by the volume of internal phase. As seen in **Figure 1D**, the EE occurred a sudden enlargement when the pH reached 13, whereas there were no obvious changes in particle size. As shown in **Figure 1E**, the PVA was used as the external water phase, both the EE and the particle size were desirable. **Figure 1F** illustrated the influence of the PVA concentration on EE and particle size. As PVA concentration increased, the EE firstly increased dramatically and then decreased over 1%. Simultaneously, the particle size was observed to slightly fluctuate as PVA concentration changed. As the volume of PVA increased, a slight downward trend of EE and a fluctuation of particle size were observed in the **Figure 1G**. In **Figure 1H**, as internal phase/organic phase (W1/O) ultrasonic time was prolonged, the EE increased dually firstly and then dropped, while particle size was relatively stable. Furthermore, the effect of the ultrasonic time of internal phase/organic phase/external phase (W1/O/W2) on EE and particle size was also investigated in **Figure 1I**. EE was firstly increased under short-term ultrasound condition, but was suddenly compromised when the ultrasonic time was beyond 45 s, while the particle size decreased with the ultrasonic time prolonged. Different ratios of PLGA vs. lipid bilayer materials (PLGA/LPs) were investigated in **Figure 1J**. The EE was also significantly affected by the PLGA/LPs ratio, while the particle size was relatively stable. **Figure 1J** showed that the formulations exhibited a better size and higher EE when the PLGA/LPs ratio was 1:1.5. It was shown in **Figure 1K** that with the ultrasound power increased from 100 to 500 W, both the EE and particle size exhibited a rapidly decreasing tendency. Overall, the optimal YF4-LNPs were demonstrated to significantly improve the drug incorporation with an EE of 89.88 ± 1.23% (n = 3) and DL of 2.18 ± 0.25%. The average particle size was 227.3 ± 3.8 nm with a narrow size distribution (PDI = 0.09 ± 0.02) (**Figure 2A**). The zeta potential of YF4-LNPs was slightly negative, with the value of −7.27 ± 0.85 mV (**Figure 2B**). The colloidal suspension was observed as slightly blue opalescence with strong Tyndall effect (**Figure 2C**).

As shown in the TEM image in **Figure 3A**, YF4-LNPs were generally spherical, homogeneous and the lipid bilayer

was well coated on the nanoparticles. In **Figure 3B**, the DSC analysis showed that the free YF4 displayed an obviously sharp endothermic peak at 225.6◦C. The YF4-LNPs only presented two smooth endothermic peaks at 140.1◦C, 205.2◦C, with the absent sharp endothermic peak of free YF4. However, in the physical mixture group, there were three endothermic peaks at 145.3◦C, 202.3◦C, and the peak at 226.9◦C, which represented the free YF4.All the obtained statistics suggested that YF4 was successfully encapsulated into the YF4-LNPs. The stability study displayed that YF4- LNPs remained stable without change of EE and size for at least 1 week (**Figure 3C**). When YF4-LNPs were stored for 2 weeks, the EE of YF4 got a sharp drop and the particle size became larger, which illustrated that YF4-LNPs could keep stable in 4◦C for 1 week and were suitable for developing an oral administration.

#### In vitro Release

In vitro release profiles of YF4 from YF4-LNPs were presented in **Figure 4** and **Supplementary Figure S2**, the free YF4 exhibited a burst release, and over 80% of YF4 released in 6 h, while those loaded in YF4-LNPs were released gradually without apparent burst release in all of the release media. Within a 12 h period, compared with the approximately total release of free YF4 at pH 4.5, 6.8, only less than 40% of YF4 was released from YF4- LNPs. Besides, over more than 40% of YF4 was still packed in the YF4-LNPs while the free YF4 was quickly released in the acidlike solution in the stomach (pH 1.0). All the results illustrated that YF4 was efficiently packed in YF4-LNPs and might achieve a sustained release in small intestine.

#### In vivo Antihypertensive Efficacy

To investigate the blood pressure lowing (BPL) effect of YF4, captopril was chosed as the positive control. As seen in the **Figure 5**, compared with saline group, the BPL effect of YF4 at a dose of 5 mg/kg presented at 2 h after a single oral administration was 23.0 mmHg, which was equal to captopril with the same dose. However, both of them showed the short BPL effect.

Then we prepared the lipid nanoparticles encapsulated YF4 and further investigated the BPL effect of YF4-LNPs in SHRs. As **Figure 6** showed, compared with saline group, YF4 and YF4-LNPs at a dose of 0.8 mg/kg all emerged a BPL effect. Administration of 0.8 mg/kg of YF4 signicantly decreased SBP by 15.6 mmHg at 4 h post-administration and SBP returned to the untreated level at 12 h post administration. And the decreased SBP between 2 and 120 h post administration was observed after treated by 0.8 mg/kg YF4 LNPs. YF4-LNPs at a dose of 0.8 mg/kg strikngly decreased the blood pressureby 43.5 mmHg about 2 h post administration. Amazingly, even after 120 h, obviously BPL effect was still seen as a significantly reduction of 20.5 mmHg. The SBP returned to the initial level at 144 h post administration.

# DISCUSSION

The hypertension patients are required to lifelong medication, which severely affects the living quality. Oral administration is preferred due to its convenience and safety. However, the oral antihypertensive peptides still face enormous difficulties, such as the inactivation by acid and enzyme in the GI (Richard, 2017), low effienciency in penetrateing the mucous layer of the intestinal

(Lai et al., 2009), and unable to be efficiently absorbed by small intestinal epithelial cells, which result in the depressing therpeacy effect (Ensign et al., 2012). These defects seriously limit the application of antihypertensive peptides, thus it is essential to develop a new oral antihypertensive peptide system to solve the problem of poor stability (Chen et al., 2015) and achieve the goal of long-term effect of blood pressure reduction.

PLGA nanoparticles, as a safe drug-loading system, can effectively slow the drug release and protect drugs from gastrointestinal degradation (Lozoya-Agullo et al., 2017; Mustafa et al., 2017). Then coating the lipid membrane on the particles can further protect the drugs from GI and improve the contact between the nanoparticles and the small intestinal epithelial cells, which may achieve the goal of prolonging the circulation in the blood. The formed LNPs are potential to increase the stability, permeability and bioavailability of drugs (Zhang et al., 2010; Fang et al., 2014) in the GI.

The formed YF4-LNPs were prepared by a simple and controllable method (Yu et al., 2016). It was reported that EE and partice size both have enormous influence on the quality

of nano-preparions (Johnstone et al., 2013). The utilization and absorption of drug can be improved by nano-preparions with higher EE and smaller particle size (Yu et al., 2016). Hence, factors which might influent the EE and size were optimized systemically, including the PLGA concentration, the volume of acetone in the organic phase, the pH of the aqueous medium, the volume of the inner aqueous medium, the external water phase, the PVA concentration, the PVA volume, the ultrasonic time, lipid material to nanomaterial ratio, and the ultrasound power. The an ever-growing PLGA viscosity in the organic phase caused droplets became larger and net shear stress became less, which would limit the free YF4 in inner aqueous diffuse into organic phase (**Figure 1**; Kiss et al., 2011; Turk et al., 2014). Therefore, the increasing PLGA concentration was indicated to lead to higher EE and larger particle size. The addition of acetone in the organic phase plays a great role in the nanoparticles forming (Anarjan et al., 2011). The addition of acetone in the organic phase could significantly affect the EE and particle size. The presence of acetone promoted organic phase to diffuse to inner phase (Luo et al., 2016), causing the carrier materials PLGA quickly precipitated and contained YF4 in the water phase, attributing to the EE increased. Whereas, too much acetone in the organic phase could also increase the opportunity of diffusion from inner to organic phase, which might result in the drug leakage (Yu et al., 2016). In addition, the interfacial energy at the oil/water interface was also reduced by the addition of acetone in the organic phase and thereby the stability of droplets was enhanced, which contributed to the smaller sizes of nanoparticles (Wang and Anderko, 2013). The opportunity of entrapment of the internal phase by the organic phase was enhanced as the W1/O volume ratio increased (Sang, 1999), which attribute to an excellent EE and a smaller particle size. However,when the W1/O volume ratio is too low or high, the W1/O emulsion structure was not complete, causing the poor EE and particle size. There is no doubt that the pH and ionizing status of peptide was highly related with the solubility of peptide (Ragab et al., 2004), which could affect the peptide loading state in the nanoparticles. When the pH was in isoelectric point (pI) of the YF4 peptide (Silva et al., 2013), in alkaline state, the state of charge distribution and the ionization state changed, the EE was significantly improved.

The effect of external water phase was also investigated. As an emulsifying agent, different external water phases make a great difference on the EE and particle size (Peng et al., 2007), which mainly related to viscosity and surface tension (Rayat and Feyzi, 2011). Although these surfactants can significantly reduce the surface tension, it is easy to lead to excessive viscosity of the aqueous phase, eventually leading to a larger particle size, the drug is not easy to be encapsulated, and the encapsulation efficiency is very low. Therefore, the optimal external aqueous media must possess these comprehensive properties to obtain higher EE and more appropriate particle size. According to above analysis, PVA was selected as the externalwater phase. Then we further evaluated the impacts of the concentration and volume of PVA. When the concentration of PVA was below 1%, the outer aqueous viscosity was too small to stable the droplet, the EE and the particle size were both unsatisfyed. However, the EE was declined as the increase of PVA concentration and volume (Song et al., 2008; Moralescruz et al., 2012; Sun et al., 2016), which might be attributed to the increase viscosity. In addition, the corresponding shear stress decreased as the increase viscosity, and led to the diffusion of YF4 from the inner aqueous phase into outside hampered, which might explain the lagrer particle size.

Ultrasonic duration might affect the drug leakage and the homogeneity of particle size in the preparation of LNPs. Moreover, peptides are generally denatured under high intensity ultrasonic shear systems (Gentile et al., 2012). Consequently, the proper ultrasonic duration was seemed as vital in the preparation of YF4-LNPs. The EE was increased firstly and suddenly decreased with longer ultrasonic time, which might be attributed by the broken coarse emulsion drops into nanodroplets (Fricker et al., 2010). However, too long ultrasonic time would lead to the leakage of YF4 from the internal phase to the external phase, causing a decrease in EE. Besides, prolonging W1/O/W2 ultrasonic time for a short time could significantly enhance the EE and decrease the particle size, caused by the increasing net sheer force, which is similar to the changes of ultrasonic power.

Ultimately, the ratio of PLGA/lipid materials was further optimized. As a result, a higher EE and a better particle size of YF4-LNPs were gained at 1:1.5 ratio of PLGA/lipid materials, which might result from the increased affinity between drug and the formulation (Luo et al., 2006).

To investigate the entrapment state of YF4 into lipid nanoparticles, DSC analysis were performed. The DSC showed that the YF4 was successfully entrapped in the YF4-LNPs. Additionally, the YF4-loaded nanoparticles were coated with lipid film according to TEM images, indicating the successful construction of lipid nanoparticles.

The release of YF4 from lipid nanoparticles was dramatically lower than the unincorporated YF4, which might be attributed to the surprising core-shell structure that could result in the tunable and sustained drug release profiles (Zhang et al., 2008; Fang et al., 2014). Although, our results are unsatisfactory, there is no obvious gastric protective and obviously sustained release in the intestinal tract, our study does protect the free drug from degradation in the stomach to a certain extent, providing more possibilities for its slow release in the intestinal tract and so as to prolonging

the circulation in the blood and exerting its efficacy. Our result showed that only less than 30% of YF4 diffused into the dispersed medium after 12 h in all pH conditions, indicating a prolonged release of YF4-LNPs, which might lead to a potent and prolonged therapeutic efficacy of YF4-LNPs. For YF4- LNPs, the release percentages in media with different pH can provide some evidence that YF4-LNPs could provide sustainable release of YF4 in both stomach and intestine and protect YF4 from degradation.

It has been reported that ACEI peptides have amazing hypotensive activity in vitro (Escudero et al., 2012). We then firstly verified the BPL effect of YF4 compared with ACEI positive drug captopril shown in **Figure 5** (Nurminen et al., 2000). However, the oral duration is short and the dosage of antihypertensive peptides is far higher than that in literature (Yu et al., 2016). Oral delivery of peptides has enormously continuous challenge due to its poor stability in the GI tract and low permeability through the intestinal epithelium membrane (Friedman and Amidon, 1991). The pharmacokinetics of YF4 in rats was studied and showed in the **Supplementary Figure S1** and **Supplementary Table 1**. The results showed that the half-life of YF4 was only 2.91 h, just like that of conventional polypeptide and protein drugs. Thus, the new oral peptide delivery systems need to be developed. LNPs exhibit a wide range of erosion times, tunable biodegradation and mechanical properties, such as improving the connection between the nano-preparations and the intestinal epithelium cells, protecting the drugs from the GI, which might explain the antihypertensive effect in SHRs of YF4-LNPs. The YF4-LNPs exhibited a much better antihypertensive efficiency, especially YF4-LNPs showed a ten-time prolonged BPL effect. Currently, the mechanism of enhancement in therapeutic efficacy is still under investigation. The efficiency of YF4- LNPs being absorbed into systemic circulation requires further quantification. The association between elevated therapeutic efficacy and the increase in AUC or bioavailability needs further elucidation.

Over all, we systematically optimized the factors that could affect the features of LNPs, and finally constructed YF4-LNPs with a high entrapment efficiency and small particle size. Then, we evaluated its antihypertensive efficiency in vivo, and found that YF4-LNPs displayed stronger BPL effect and prolonged the effect up to 120 h. This may be explained by the controlled release of YF4 and the protection effect of the nanoparticles core and lipid membrane avoiding enzymatic degradation. This may provide some methodological clues and insights for the therapy of hypertension.

#### REFERENCES


#### CONCLUSION

In this study, we have developed uniform-sized PLGA lipid nanoparticles for oral delivery of peptides. The LNPs was successfully prepared by thin-film membrane hydrationultrasonic dispersion method and systematically optimized. Then, the properties of YF4-LNPs were investigated, including the stability, the in vitro release profile. To our surprise, YF4- LNPs were considerable stable and showed a sustained release profile. Then the DSC curves and TEM illustrated that the successfully loading of YF4 and the coating of lipid bilayer. YF4- LNPs exhibited an enhanced antihypertensive function with a longer duration (up to 120 h in SHRs), which suggested the sustained release. In summary, all these data supported the belief that YF4-LNPs would be a promising platform for oral hypertension treatment. However, the preparation process of YF4-LNPs was relatively complex, which might be difficult to be scaled up. Hence, a simplified process is worthy of being further investigated to fabricateYF4-LNPs for the clinical translation study. Moreover, the pharmacokinetics of YF4-LNPs is also necessary to be investigated in the future.

#### AUTHOR CONTRIBUTIONS

JY conceived the project. XG, CF, and JY designed the experiments. SZ, JL, and YZ conducted most of the experiments. SZ and JL drafted the manuscript. LH performed some preliminary experiments. YL and JX participated in literature searching. JY, XG, and CF finished manuscript editing. All authors reviewed and approved the manuscript.

#### FUNDING

This work was financially supported from the Science and Technology Project of Shenzhen (JCYJ20170413155047512), Sichuan Province Science and Technology Support Program (16ZC2698) and National Natural Science Foundation of China (No. 81472162).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2019.00102/full#supplementary-material


randomized controlled trial. Nutr. J. 9:52. doi: 10.1186/1475- 2891-9-52



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Zhao, Li, Zhou, Huang, Li, Xu, Fu, Guo and Yang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Co-delivery of Cisplatin(IV) and Capecitabine as an Effective and Non-toxic Cancer Treatment

Xiao Xiao1,2† , Ting Wang<sup>3</sup>† , Leijiao Li<sup>4</sup> \*, Zhongli Zhu<sup>4</sup> , Weina Zhang<sup>5</sup> , Guihua Cui1,2 \* and Wenliang Li1,2 \*

<sup>1</sup> School of Pharmacy, Jilin Medical University, Jilin, China, <sup>2</sup> Center for Biomaterials, Jilin Medical University, Jilin, China, <sup>3</sup> Department of the Gastrointestinal Surgery, The First Hospital of Jilin University, Changchun, China, <sup>4</sup> Jilin Provincial Science and Technology Innovation Center of Optical Materials and Chemistry, School of Chemistry and Environmental Engineering, Changchun University of Science and Technology, Changchun, China, <sup>5</sup> Common Subjects Department, Shangqiu Medical College, Henan, China

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Jinshan Guo, Harvard University, United States Xiaoding Xu, Sun Yat-sen University, China

#### \*Correspondence:

Leijiao Li lileijiao@ciac.ac.cn Guihua Cui 675053025@qq.com Wenliang Li wenliangl@ciac.ac.cn

†These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 07 December 2018 Accepted: 28 January 2019 Published: 19 February 2019

#### Citation:

Xiao X, Wang T, Li L, Zhu Z, Zhang W, Cui G and Li W (2019) Co-delivery of Cisplatin(IV) and Capecitabine as an Effective and Non-toxic Cancer Treatment. Front. Pharmacol. 10:110. doi: 10.3389/fphar.2019.00110 A strategy for preparing composite micelles (CM) containing both cisplatin(IV) [CisPt(IV)] prodrug and capecitabine using a co-assembly method is described in this study. The CM are capable of an effective release of the anticancer drug cisplatin(II) [CisPt(II)] and capecitabine via acid hydrolysis once they are internalized by cancer cells. Moreover, the CM display a synergistic effect in vitro and the combination therapy in the micellar dosage form leads to reduced systemic toxicity and enhanced antitumor efficacy in vivo.

Keywords: co-delivery, cisplatin(IV), capecitabine, cancer treatment, composite micelles

# INTRODUCTION

Utilization of single small molecule anticancer agents in clinical thermotherapy is very rare owing to the rapid development of drug resistance in tumor cells (Di Francesco et al., 2002; Ahmad, 2010). Hence, drug combination is predominantly used in the clinical setting (Joensuu et al., 1998; Feldmann et al., 2007; Woodcock et al., 2011). Combination therapy presents its own set of advantages, related to the improved medication compliance and the enhanced ability to formulate combined drug profiles. From the pharmacokinetic stand point, the positive effects and adverse effects of a combination therapy may be specific to the relative dosages providing a simpler overview compared to single drug profiles. Moreover, the interaction between the individual drugs in a specific combination drug may also provide additional positive effects which may be absent in the individual drug profiles. The active ingredients used in combination drugs rarely exhibit any adverse interactions and are scrupulously reviewed by the Food and Drug Administration (FDA) in the United States (Love et al., 1989; Sitzia and Huggins, 1998). Among the combination chemotherapy regimens, co-administration of capecitabine and platinum-based drugs is the most common.

Platinum-based anticancer agents are one of the most commonly used chemotherapeutic drugs in the treatment of various solid tumors (Lebwohl and Canetta, 1998; Pasettoa et al., 2006; Xiao et al., 2017; Shen et al., 2018). However, almost all platinum drugs possess some inherent and serious side effects which influence drug resistance (Sharma et al., 2005; Windebank and Grisold, 2008). Cisplatin (cis-diamminedichloroplatinum, CDDP) is the first approved platinum drug, which has been used as a standard chemotherapeutic agent for more than 30 years (Weiss and Christian, 1993; Rosenberg and Lippert, 1999). Due to the low concentration of chloride, cisplatin is hydrolyzed inside the cell and is converted to the highly reactive species [Pt(NH3)2Cl(OH2)]+, which creates 1,2-GpG intrastrand adducts within the DNA. The adducts inhibit DNA transcription and replication, ultimately leading to cancer cell apoptosis (Jamieson and Lippard, 1999;

Boulikas et al., 2007). Nevertheless, severe side effects, including nephrotoxicity, neurotoxicity, ototoxicity and myelosuppression, limit the clinical application of cisplatin. Moreover, the intrinsic and acquired resistance developed by various cancers is another main reason for limiting application of cisplatin (Madias and Harrington, 1978; Ekborn et al., 2000).

Fluorouracil (5-FU) is one of the most widely used medications for the treatment of various cancers, such as colon, esophageal, stomach, pancreatic, breast, and cervical cancers. Unfortunately, 5-FU is distributed to both the healthy and the cancer cells following systemic administration making the drug cytotoxic and leading to death of healthy cells (Arkenau et al., 2003). Capecitabine is an antitumor fluoropyrimidine carbamate, which is very specific only toward cancer cells owing to the activation of tumor-specific enzymes (Ssif et al., 2008).

In comparison to single drug anti-cancer treatment, enhanced tumor therapy efficiency is often observed in patients administered a combination therapy (Joensuu et al., 1998; Feldmann et al., 2007; Woodcock et al., 2011). For example, Xeloda is indicated for the first-line treatment of advanced gastric cancer and is a combination of capecitabine and platinum regimen (DÍaz-Rubio et al., 2002). However, ensuring a correct dosage of a given combination drug in the context of endocytosis within the cell remains a huge challenge in small molecule-based combination therapy. In addition, unfavorable side effects likely accompany the combination of small molecule drugs. To improve the therapeutic efficacy and to reduce the side effects, nanocarriers are used to encapsulate the combination drugs (Duncan, 2003; Kwangjae et al., 2008; Maeda, 2010; Bae and Park, 2011; Tao et al., 2013, 2014, 2016, 2017, 2018; Zhu et al., 2018; Ling et al., 2019). Nanomedicine approach can assure selective accumulation of the drug at tumor sites via the enhanced permeation and retention effect, in turn protecting the drug from premature degradation and blood clearance.

In this work, two polymer-drug conjugates [MPEG-b-P(LAco-MCC)-COOH/capecitabine and MPEG-b-P(LA-co-MCC)- OH/CisPt(IV)] were synthesized (**Scheme 1**). Because of the amphiphilic properties of these polymer-drug conjugates, they were able to be co-assembled into composite micelles (CM) creating a combination therapy drug delivery system (**Scheme 2**).

The resulting polyprodrug-based combination chemotherapy regimen displays many unique features. Firstly, the relative drug doses of cisplatin(IV) and capecitabine can be easily adjusted by modulating the ratio of PEG-PLA/CisPt(IV). Next, the polymer micelles-based prodrug delivery system provides the active drugs [cisplatin(II) and 5-FU] with double protection. Specifically, both cisplatin(IV) and capecitabine are located in the core of the CM resulting in effective protection mechanism. Additionally, the other protection stems from the prodrug formulation. After a series biotransformation events, cisplatin(IV) and capecitabine get converted to active agents' cisplatin(II) and 5-FU, respectively. Therefore, it is critical for small molecule drug delivery to assure constant levels of the two drugs in the circulation. Importantly, the polymer micelle-based prodrug strategy provides an effective protection for the drugs. Compared to other small moleculebased combination therapies, our method displays significantly lower systemic toxicities. Finally, once the polymeric CM enter the tumor cells, prodrugs cisplatin(IV) and capecitabine will be released by the hydrolysis of PEG-PLA/cisplatin(IV) and PEG-PLA/capecitabine conjugates in the micelles as shown in **Scheme 2**. Moreover, cisplatin(IV) can be converted to cisplatin(II) under the intracellular reducing conditions (i.e., high concentration of glutathione (GSH) and ascorbic acid). Cisplatin(II) can also become conjugated to basic groups of DNA, in turn killing the tumor cells. Additionally, capecitabine can be converted to 5-FU under the enzymatic catalysis. Specifically, 5-FU can be converted to fluorouridine triphosphate (FUTP) and fluorodeoxyuridine triphosphate (FdUTP), which are then incorporated into RNA and DNA, respectively, ultimately leading to cell death.

#### MATERIALS AND METHODS

#### Materials

The polymers MPEG-b-P(LA-co-MCC)-COOH and MPEGb-P(LA-co-MCC)-OH were synthesized as previously described. N,N'-dicyclohexylcarbodiimide (DCC) and N-hydroxybenzotriazole (HOBt) were purchased from Sigma-Aldrich. CisPt(IV) prodrug was synthesized as reported in our earlier work and capecitabine was purchased from the Zhejiang Haizheng Pharmacy Co. Ltd. All other chemicals and solvents were used without additional purification steps.

#### General Measurements

<sup>1</sup>H-NMR spectra were measured by a Unity-300 MHz NMR spectrometer (Bruker) at room temperature. Fourier Transform infrared (FT-IR) spectroscopy was performed using a Bruker Vertex 70 spectrometer. A Quattro Premier XE system (Waters) equipped with an electrospray interface was used for the mass spectroscopy (ESI-MS) assessments. The total platinum (Pt) content in the polymer-Pt(IV) conjugates and in samples obtained from the dialysis bags in drug release experiments was determined by inductively coupled plasma optical emission spectrometer (ICP-OES, iCAP 6300, ThermoScientific, United States). The quantitative determination of trace levels of Pt was measured by ICP-MS (Xseries II, ThermoScientific, United States). Size and size distribution of micelles were determined by dynamic light scattering (DLS) with a vertically polarized He–Ne laser (DAWN EOS, Wyatt Technology, United States). The morphology of the polymer micelles was evaluated by transmission electron microscope (TEM) performed on a JEOL JEM-1011 electron microscope. Particle size and zeta potential measurements were conducted on a Malvern Zetasizer Nano ZS. (Zetasizer Nano ZS is a high performance dual angle particle size and molecular size analyzer that uses DLS combined with "NIBS" optics to enhance the detection of aggregates, as well as small or diluted samples, and poles. Low or high concentration sample).

#### Synthesis of N-Hydroxy-Succinamide (NHS) Ester of MPEG-b-P(LA-co-MCC)

MPEG-b-P(LA-co-MCC) (0.5 g, 0.0625 mmol) was dissolved in 5 mL of CH2Cl<sup>2</sup> in a flask, to which DCC (128.9 mg,

0.625 mmol), NHS (71.9 mg, 0.625 mmol), and DMAP (76.35 mg, 0.625 mmol) were added. The reaction mixture was constantly stirred in an ice bath for 24 h. Next, it was filtered to remove 1,3 dicyclohexylurea (DCU). The filtrate was then added dropwise to cold diethyl ether (50 mL), the resultant precipitate was collected by filtration and dried under vacuum to obtain the NHS ester of MPEG-b-P(LA-co-MCC).

#### Synthesis of MPEG-b-P(LA-co-MCC)-OH

The NHS ester of MPEG-b-P(LA-co-MCC) (0.5 g, 0.0625 mmol) was dissolved in dried CH2Cl<sup>2</sup> and 2-aminoethyl alcohol (4.6 mg, 0.075 mmol) was added to the polymer solution. The reaction mixture was stirred at room temperature for 24 h and then precipitated with ether to obtain the final product MPEG-b-P(LA-co-MCC)-OH.

#### Synthesis of MPEG-b-P(LA-co-MCC)-OH/Pt(IV) Conjugates

Cisplatin(IV)-COOH [abbreviated as CisPt(IV)] was synthesized as described previously. CisPt(IV) was conjugated to the polymer MPEG-b-P(LA-co-MCC)-OH with pendant hydroxyl groups using N-dicyclohexyl carbodiimide/4-dimethylaminopyridine (DCC/DMAP). Briefly, CisPt(IV) (43.3 mg, 0.081 mmol) was first dissolved in dried N,N-Dimethylformamide (DMF) (1 mL) in a flask under stirring conditions. The polymer MPEG-b-P(LAco-MCC)-OH (0.5 g, 0.0625 mmol) was then dissolved in 3 mL of C2H2Cl<sup>2</sup> and then added to the CisPt(IV) solution. Next, 128.9 mg of DCC, 76.35 mg of DMAP and 95.7 mg of HOBt were added into the solution. The reaction mixture was stirred for 2 days at room temperature. Afterward, the mixture was filtered to remove DCU. The filtrate was slowly added to diethyl ether to precipitate the crude product, which was then dried to yield a pale yellow powder. Next, the powder was dissolved in DMF and placed into a dialysis bag and dialyzed against water to remove unreacted CisPt(IV). Finally, the solution was lyophilized to obtain the MPEG-b-P(LA-co-MCC)-OH/Pt(IV) conjugates.

# Synthesis of MPEG-b-P(LA-co-MCC)/Capecitabine Conjugates

Capecitabine was conjugated to the polymer MPEG-b-P(LA-co-MCC) with pendant carboxyl groups using the DCC/DMAP/HOBt method. Briefly, capecitabine (43.3 mg, 0.081 mmol) was first dissolved in dried DMSO in a flask under stirring conditions and 0.5 g MPEG-b-P(LA-co-MCC) dissolved in 3 mL of C2H2Cl<sup>2</sup> was then added. Afterward, 128.9 mg of DCC, 76.35 mg of DMAP, and 95.7 mg of HOBt were added. The reaction mixture was stirred for 2 days at room temperature. Next, it was filtered to remove DCU, precipitated by ether and collected by filtration. The product was dried to obtain a red powder. The powder was then dissolved in DMF and it was placed into a dialysis bag and dialyzed against water to remove unreacted capecitabine. Finally, following the dialysis, the solution was lyophilized to yield the MPEG-b-P(LA-co-MCC)/capecitabine conjugates.

#### Preparation of MPEG-b-P(LA-co-MCC)- OH/Pt(IV) and MPEG-b-P(LA-co-MCC)/ Capecitabine CM

The CM were prepared by nano-precipitation method. In brief, 50 mg of MPEG-b-P(LA-co-MCC)-OH/Pt(IV) conjugates and 50 mg of MPEG-b-P(LA-co-MCC)/capecitabine conjugates were dissolved in a flask filled with 10 mL of DMF. Next, the solution was added in a dropwise fashion into the flask under stirring to form a micellar solution. The solution was then dialyzed against water to remove DMF and then freeze-dried to obtain the MPEG-b-P(LA-co-MCC)-OH/Pt(IV) and MPEG-b-P(LA-co-MCC)/capecitabine CM.

# In vitro Drug Release of CM

The CM (5 mg) were dissolved in PBS (10 mL, 0.1 M, pH 7.4, or pH 5.5) and then placed into a pre-swelled dialysis bag (3500 MWCO), which was immersed into PBS (100 mL) at 37◦C in a shaking culture incubator (Wang et al., 2012). At 1 h intervals, 1.5 mL of sample solution was withdrawn from the dialysate and measured for the Pt and capecitabine concentration using ICP-OES and HPLC, respectively. After sampling, fresh PBS (1.5 mL) was immediately added back into the incubation medium. The process was repeated for 8 h. The Pt released from the micelles was expressed as the percentage of the cumulative Pt or capecitabine outside the dialysis bag to the total Pt or capecitabine inside the micelles.

#### Cell Lines and Cell Culture Conditions

Human Colorectal Tumor Cells HCT-8 (ATCC) were cultured in RPMI 1640 media (Hyclone, Logan, UT, United States) supplemented with 10% fetal bovine serum (FBS, Life Technologies, United States), 100 U/mL penicillin and 100 µg/mL streptomycin (Ameresco, Life Technologies, United States) at 37◦C with 5% CO2.

#### Cell Viability Studies

The HCT-8 cells harvested in a logarithmic growth phase were seeded in 96-well plates at a density of 1 × 10<sup>4</sup> cells/well and incubated in 100 µL/well RPMI 1640 medium for 12 h. The medium was then replaced by CM with a Pt concentration ranging from 0.1 to 432 µM. To assess cell viability, 20 µL of MTT solution (5 mg/mL) was added and the plates were incubated for another 4 h at 37◦C. Next, the culture medium containing MTT was removed and 150 µL of DMSO was added to dissolve the formazan crystals that were formed. Finally, the plates were shaken for 10 min and the absorbance was measured at 492 nm using a microplate reader.

# Synthesis of MPEG-b-P(LA-co-MCC)-OH/CisPt(IV) Conjugates

With CisPt(IV) in hand, we hoped to create a carrier polymer with hydroxyl groups. Therefore, MPEG-b-P(LA-co-MCC)-OH was synthesized according to our previous report (Xiao et al., 2012). CisPt(IV) was conjugated to the polymer MPEG-b-P(LA-co-MCC)-OH using DCC/DMAP method and the ICP-OES measurement of MPEG-b-P(LA-co-MCC)-OH/CisPt(IV) conjugates indicated a Pt content of 9.3 w/w%.

#### Synthesis and Characterization of MPEG-b-P(LA-co-MCC)/Capecitabine Conjugates

Capecitabine was conjugated to the polymer MPEG-b-P(LA-co-MCC) with pendant carboxyl groups using the DCC/DMAP/ HOBt method. The <sup>1</sup>H-NMR spectra of capecitabine, MPEGb-P(LA-co-MCC) and MPEG-b-P(LA-co-MCC)/capecitabine in deuterated chloroform (CDCl3) solvent were collected and are summarized in **Figure 1**. Compared to the <sup>1</sup>H-NMR spectrum of MPEG-b-P(LA-co-MCC), we identified additional new peaks in the MPEG-b-P(LA-co-MCC)/capecitabine conjugates. These peaks may likely be attributed to capecitabine. Moreover, we collected the UV–vis spectra of capecitabine solutions at different concentrations (0.01–0.1 mg/mL) and summarized the data in **Figure 2A**. Using the spectra, we were able to calculate the standard curve of absorbance at 304 nm vs. capecitabine concentration (**Figure 2B**). This standard curve and the UV– vis spectra of MPEG-b-P(LA-co-MCC)/capecitabine conjugates revealed that the capecitabine content in the conjugates was 17% w/w.

### Preparation of the Mixed Polymer-Drug Micelles

Drug combination particles were prepared by the self-assembly method. The obtained particles were spherical micelles with a mean diameter of 55 nm (according to TEM) and 110 nm

FIGURE 3 | (A) TEM and (B) DLS characterization of the drug combination particles.

TABLE 1 | The IC50 values of free drug CPT/CisPt(II) combinations and mixed polymer-drug micelles M[CPT/CisPt(II)] with various CPT/CisPt(II) ratios evaluated in HCT-8 cells at 48 h.


(according to DLS) (**Figure 3**). The resulting particle size using the DLS was larger compared to the TEM assessment, likely due to the shrinkage during the TEM sample preparation process. The particles displayed a zeta potential of −4.4 mV, which is almost neutral, hence we expected the particles to have a good stability in vivo.

#### RESULTS AND DISCUSSION

#### Drug Release

Drug release experiments were performed at different pH values as stated in the literature (Choi et al., 1998). Data presented in **Figure 4** show the release kinetics (0–50 h) for Pt and capecitabine in the combination drugs in two conditions at pH 5.0 and pH 7.4. Overall, the drug release curves showed that the two drugs displayed sustained release at both pH 5.0 and

pH 7.0 and the maximum release occurred at approximately 12 h and then remained stable for 50 h. Moreover, the cumulative release percentages of Pt and capecitabine at pH 5.0 were higher compared to values at pH 7.4. Additionally, we observed that the Pt release was more sensitive to pH compared to the capecitabine release. Importantly, the intracellular compartments for the endosomes and lysosomes are acidic in cancer cells, allowing for the release of these two drugs into cancer cells and thereby improving the therapeutic effect.

#### In vitro Cytotoxicity

fphar-10-00110 February 19, 2019 Time: 12:58 # 8

We calculated the IC<sup>50</sup> values of the combination therapy of free drug capecitabine(CPT)/CisPt(II) combinations and mixed polymer-drug micelles M(CPT/CisPt(II)) with various CPT/CisPt(II) ratios in HCT-8 cells at 48 h (**Table 1**). We observed that both free drugs CPT/CisPt and mixed polymerdrug micelles M[CPT/CisPt(II)] exhibited an enhancement of combination potency accompanied by dose-response profiles being shifted toward lower drug concentrations when the CPT/Pt molar ratio increased (**Figure 5**).

At 48 h, the IC<sup>50</sup> values of free CisPt(II) alone and M[CisPt(II)] were 13.66 and 21.47 µM, respectively. The IC<sup>50</sup> values of both free drug combinations and the polymerdrug micelles were sharply reduced when the CPT/Pt molar ratio increased (**Table 1**). These data suggest that synergistic cytotoxicity of CPT/CisPt(II) formulations against HCT-8 cells was evident. More importantly, the mixed polymer-drug micelles, M[CPT/CisPt(II)], had similar synergistic effects in HCT-8 cells. In short, M(PTX/Pt) (PTX is short for paclitaxel) displayed significant time- and dose-dependent inhibitory effects allowing for the dissociation and the release of the drugs into the cells.

#### CONCLUSION

In this study, we successfully created a synthetic strategy for drug co-delivery by conjugating and co-assembling, wherein

#### REFERENCES


two different drugs, including CisPt(II) based drugs, are enveloped into one carrier polymer allowing for the delivery of the combination therapy. Our method exhibited effective synergistic effect in vitro between the free CTP/CisPt(II) and the composite M(PTX/Pt) micelles. Furthermore, the polymer-based combination of capecitabine and cisplatin(IV) prodrug displayed safer and more efficacious inhibition of HCT-8 cell growth compared to the small molecules individually. We hope that our combination strategy can be extended to other anticancer drugs. Given the significant effect of the combination therapy, we believe this strategy will likely be utilized in the clinic in the foreseeable future.

#### AUTHOR CONTRIBUTIONS

LL, GC, and WL designed the experiments. XX and TW carried out the experiments and wrote the manuscript. ZZ and WZ helped analyzing the experimental results.

#### FUNDING

This work was supported by the Keypoint Research and Invention Program (No. AWS17J016), the National Natural Science Foundation of China (Grant No. 21871246), the Grant of Jilin Province Science and Technology Committee (Nos. 20180101194JC, 20170520149JH, and 20180520009JH), Jilin Province Science and Technology project of traditional Chinese medicine (2017103), Jilin Province Education Department the Science and Technology development project (Nos. JJKH20181117KJ and JJKH 20191059KJ), the Youth Foundation project of Jilin Province Health and Family Planning Commission (No. 2016Q053), the Science and Technology Key Project of Henan Province (182102310611), and National College Student Innovation Training Program (201713743002).


PLGA-TPGS diblock copolymer for breast cancer therapy. Acta Biomater. 9, 8910–8920. doi: 10.1016/j.actbio.2013.06.034


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Xiao, Wang, Li, Zhu, Zhang, Cui and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fphar-10-00110 February 19, 2019 Time: 12:58 # 9

# Efficient Treatment of Sporothrix globosa Infection Using the Antibody Elicited by Recombinant Phage Nanofibers

Feng Chen<sup>1</sup> , Rihua Jiang<sup>1</sup> \*, Shuai Dong<sup>2</sup> and Bailing Yan<sup>3</sup> \*

<sup>1</sup> Department of Dermatology, China-Japan Union Hospital of Jilin University, Changchun, China, <sup>2</sup> Department of Gynecology and Obstetrics, The First Hospital of Jilin University, Changchun, China, <sup>3</sup> Department of Emergency, The First Hospital of Jilin University, Changchun, China

#### Edited by:

Chao Wang, The University of North Carolina at Chapel Hill, United States

#### Reviewed by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China Bailiang Wang, Wenzhou Medical University, China Everardo López-Romero, Universidad de Guanajuato, Mexico

#### \*Correspondence:

Rihua Jiang 635597795@qq.com Bailing Yan yanbailing@163.com

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 08 December 2018 Accepted: 11 February 2019 Published: 27 February 2019

#### Citation:

Chen F, Jiang R, Dong S and Yan B (2019) Efficient Treatment of Sporothrix globosa Infection Using the Antibody Elicited by Recombinant Phage Nanofibers. Front. Pharmacol. 10:160. doi: 10.3389/fphar.2019.00160 Antifungal therapy is used to treat sporotrichosis. However, there are several limitations in this therapy, such as development of drug resistance and potential health risks including liver injury. The purpose of our study was to evaluate the antifungal efficacy of antibody against the hybrid phage nanofibers displaying KPVQHALLTPLGLDR (phage-KR) in a fungal-infected mouse model. In this study, we extracte an antibody against hybrid phage nanofibers (phage-KR) from immunized mice and passively inoculate Sporothrix globosa (S. globosa) infected mice. The study shows that the antibody exhibits efficient inhibition efficacy of the S. globosa infection, including reduction of the progressive fungi colonizing, increase of animal survival rate and relief of organ inflammation in the mice. The results indicate that antibody against phage-KR may act as a potential strategy for safe and efficient treatment of S. globosa infections.

Keywords: sporotrichosis treatment, recombinant phage nanofibers, Sporothrix globosa, antibody treatment, immune response

#### INTRODUCTION

Sporotrichosis is known to be an acute or chronic subcutaneous mycosis, which can affect humans and other mammals (Barros et al., 2011). Especially, the global spread of the cases of sporotrichosis tends to increase in recent years. The disease is usually categorized into four types according to clinical symptoms, including fixed cutaneous, lymphocutaneous, multifocal or disseminated, and extracutaneous (Ramos-e-Silva et al., 2007; Barros et al., 2011). It is believed that Sporothrix globosa (S. globosa) is the only pathogenic species in northeast China, despite clinical presentations or the regions where it is isolated (Yu et al., 2013).

Currently, antifungal therapy is used to treat sporotrichosis worldwide. However, there are certain limitations in this therapy. Whether administered continuously or intermittently, antifungal treatments can lead to the development of drug resistance in Sporothrix (Donnelly et al., 2008). Potential health risks associated with long-term exposure to antifungal agents may cause liver injury in asymptomatic patients, especially in patients with liver disorders, children, and pregnant women (Tuccori et al., 2008). Hence, it is crucial to find an alternative treatment for sporotrichosis, such as antibacterial materials (Lin et al., 2017; Li et al., 2018).

Gp70, a glycoprotein of 70 KDa and a major adhesin expressed on cell surface of Sporothrix schenckii, is found to be associated with virulence of fungus (de Almeida et al., 2015;

Rodrigues et al., 2015). It plays a key role in immunization modulation and host defense. The monoclonal antibody (mAb) against Gp70 is a candidate for vaccination against sporotrichosis, which may induce strong protection (de Almeida et al., 2015). Four peptides of Gp70 may be involved in this protection. Immunization was confirmed by mass spectrometry of digested Gp70 and the epitopes on the peptides were confirmed by using an epitope-finding algorithm.

In our previous studies, it has been found that recombinant phage displaying peptide KPVQHALLTPLGLDR (KR, one of the four peptides) could enhance the immune responses of T helper (Th) 1 and Th17 cells and elicit antibody against Gp70 in BALB/c mice, leading to the inhibition of subsequent infection of the mice (Chen et al., 2017). Phage displaying other three peptides cannot elicit efficient protective immune response. So, in this study, only the antibody against hybrid phage displaying KPVQHALLTPLGLDR (phage-KR) was extracted from the immunized mice and then passively inoculated into S. globosa-infected mice. The curing efficacies of the treatment including anti-fungal effects, alleviation of inflammation of the organs and improvement of animal survival rate by the treatment were evaluated (**Scheme 1**).

# MATERIALS AND METHODS

#### Animals

BALB/c mice (6–8 weeks of age, 20–25 g body weight) were received from Beijing HuaFuKang Biological Technology Co., Ltd. (China). An animal facility with specific pathogen-free conditions was used to raise the mice. All animal procedures in this study were performed in accordance with the Guidelines for Care and Use of Laboratory Animals of Jilin University and approved by the Animal Ethics Committee of The First Hospital of Jilin University (Protocol No. 2017-096-01).

#### Strain and Culture Conditions

This study was carried out in accordance with the recommendations of Guidelines for Use of Patient Specimens, Ethics Committee of China-Japan Union Hospital of Jilin University. The protocol was approved by the Ethics Committee of China-Japan Union Hospital of Jilin University. All subjects gave written informed consent in accordance with the Declaration of Helsinki. Cultured isolates were obtained from the patients who were diagnosed with invasive sporotrichosis. Sequence searches in GenBank revealed that all isolates were S. globosa strains. The isolates were allowed to grow on Sabouraud dextrose agar slants at 28 ◦C for 7 days. Fungus was then added to brain heart infusion (BHI) broth and cultured at 37◦C for 7 days. The conidia taken from the cultures were diluted to 1 × 10<sup>8</sup> cells/mL (Lyon et al., 2013; Nunes Mario et al., 2014). The yeast cells were heat-killed for 2 h at 60◦C. The heat-killed S. globosa (HK-SP) were conserved at 4 ◦C (Tachibana et al., 1999).

#### Phages

The sequence of peptide KR was displayed on the gene III of f388-55 phage vector previously. Phage expressing peptide KR could elicit antibody against S. globosa and induce a mixed Th1/Th17 response (**Supplementary Figure S1**). Wild type phages were produced as described previously and conserved in our laboratory (Wang et al., 2014). The phage pellet was allowed to resuspend in PBS.

# SDS-PAGE

Expression of peptide KR by recombinant phage was tested by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The samples of phage were boiled for 10 min in an equal volume of 2× sample loading buffer containing 100 mM Tris– HCl (pH 8.3), 4% SDS, 20% glycerol, and 0.02% bromophenol blue. Proteins were then electrophoresed. The protein bands were shown by silver-staining according to the procedure by Schagger and von Jagow (1987).

# Production of Antibodies

The BALB/c mice were randomly divided into four groups. At a weekly interval, the BALB/c mice were injected intraperitoneally for immunization for four times with different formulations, including 100 µl of PBS containing 25 µg phage-KR nanofibers (denoted as group RP), 100 µl of PBS with 25 µg wild-type phage nanofibers (denoted as group Mock), 100 µl of PBS with 10<sup>8</sup> HK-SP (denoted as group HK-SP), or PBS only as the negative control (denoted as group PBS). One week after the last immunization, sera were collected from the immunized mice, and IgG antibody was extracted and purified from the sera based on the manufacturer's procedure by using HiTrap Protein G HP column (a product of GE General Electric, United States).

#### Western Blotting

The serum collected from the mice with disseminated sporotrichosis containing antibodies against Gp70 of S. globosa or control individuals (de Almeida et al., 2015). The protein was denatured, electrophoresed, and transblotted onto a nitrocellulose membrane in Tris/Glycine buffer. The membrane was blocked in TBS-T with 5% (w/v) non-fat milk at 4◦C overnight. Following washing with TBS-T for four times, the nitrocellulose membrane was cultured in a 1:80 dilution of serum in TBST with 5% non-fat milk at 37◦C for l h. Following washing, the membrane was further cultured at 37◦C with goat antimouse IgG conjugated with peroxidase (obtained from Vector Laboratories Inc., of United States) for 1 h, and then stained with 3-amino-9-ethylcarbozole (AEC) for acting as a chromogen.

#### Immunofluorescence

1 × 108 sporophores were cultured in 6-well tissue culture plates with RPMI 1640 medium for 4 h. Following slight washing using PBS, the cells were cultured with the sera containing antibodies against recombinant phage displaying peptide KR (RP) or wild type phage (Mock) for 2 h on ice. The adherent cells were fixed using 4% paraformaldehyde for 20 min, followed by washing with PBS three times. The cells were then permeabilized with

0.5% Triton X-100, followed by further washing with TBST three times. After blocking with 3% bovine serum albumin (BSA) at 37◦C, the cells were cultured with a 1:1000 dilution of antiphage g3p (pIII) antibody in PBS buffer (pH 7.4) at 4◦C for 1 h, followed by incubation with Cy3 labeled goat-anti-mouse IgG (1:2000) (Abbkine, United States) for another 45 min at 37◦C and staining by 4<sup>0</sup> ,6-diamidino-2-phenylindole (DAPI) for 10 min. With washing by PBS containing 0.1% Tween 20 for three times, the cells were viewed by a laser scanning confocal microscope.

# Treating S. globosa With Antibodies Against Sporotrichosis

The mouse model infected with disseminated S. globosa was established by intravenous injection of 0.2 ml (1 × 10<sup>8</sup> cells/ml) of S. globosa suspension, then the mice were randomly divided into four groups (n = 6 per group). One day after infection, mice were intravenously given the following antibodies once every 3 days for a total of three administrations: (1) purified antibody against phage-KR (100 µL); (2) purified antibody against wild-type phage (100 µL); (3) purified antibody against HK-SP (100 µL); and (4) 100 µL PBS. All dosages were 5 mg per kg body weight. Finally, the mice were sacrificed 2 weeks after the final infection.

# Assessment of Protection

To evaluate the colony forming units (CFU) of S. globosa at a unit of per gram of tissue, all the kidneys were excised from the mice under aseptic condition (n = 6 in each group). The kidneys were weighed and then homogenized in 3 ml of sterile saline by using the glass tissue homogenizers. After that, saline was used to dilute the tissue homogenate, which was then put up on Sabouraud's dextrose agar. Additionally, for the histological analysis, the kidneys of the mice were removed and fixed with formalin [10% (v/v)]. The buffered paraffin-embedded tissues were sectioned into slices with 3∼5 µm thickness for hematoxylin and eosin (H&E) staining.

To evaluate the survival time of the mice, the mouse model infected with disseminated S. globosa was established by intravenous injection of 0.2 ml (5 × 10<sup>8</sup> cells/ml) of S. globosa suspension, and the mice were separated into four groups (n = 10 per group). One day post infection the mice were given

mouse; Lane 4: RP phage with normal serum; Lane 5: WT phage with normal serum.

antibodies intravenously as described above. The survival time was monitored for 2 weeks following treatment.

#### Assessment of Liver and Renal Injury

Toxicity was assessed 1 day after passive therapy. Blood samples (50 µl in each mouse) were obtained through the lateral tail vein, followed by transferring into 0.5 mL centrifugation tubes. For each sample, clinical biochemistry parameters were measured by ELISA diagnostic kits (Rongsheng, Shanghai, China; Yutong, Jiangsu, China, respectively) (Chen et al., 2017).

#### Statistical Analysis

We examined the differences in the survival time between different groups by the Log-rank test. Analysis of Variance was used for analysis of the data. The criterion for statistical significance was p < 0.05.

# RESULTS

#### Production of Recombinant Phage

Expression of phage-KR was evaluated by using SDS-PAGE (**Figure 1A**). Sera collected from mice infected with systematic sporotrichosis reacted with the fusion protein band in hybrid phage (**Figure 1B**). The results indicated that the hybrid recombinant phage displayed the peptide KR on the surface.

#### Antibody Response Against Phage Displaying Peptide KR in Immunized Mice

Our results showed that immunization with the hybrid phage expressing peptide KPVQHALLTPLGLDR (phage-KR) produced antibodies in the sera, which is able to bind Gp70 (**Figure 2**). By immunofluorescence assay, we demonstrated that the antibodies could also recognize Gp70 expressed by S. globosa (**Figure 3**). Hence, the results showed that the recombinant phage displaying peptide KR could exhibit similar function to Gp70 for treatment of S. globosa infection collectively. Namely, the peptide KR

expressed on the surface of the phage is able to mimic Gp70, a cell surface component, and induce the mice to generate antibodies which can bind to Gp70 for treatment of the infection.

### Assessment of Using Antibody Against Recombinant Phage to Protect Systemic S. globosa Infection

To assess the feasibility of using antibody against Gp70 (anti-Gp70) IgG to treat sporotrichosis, IgG was collected and purified from the sera of the immunized mice, followed by administrating to the mice that were infected with S. globosa at a lethal dose (1 × l0<sup>8</sup> cells). The survival time of the mice was monitored over 14 days (**Figure 4**). It was found that the group injected with antibody against phage-KR exhibited the highest survival rate (80%). In contrast, the mice treated with PBS showed a much lower survival rate of only 30%. For the group treated with antibody against HK-SP, an increased survival rate (70%) was also found at 14 days following infection. Significantly enhanced survival rate was obtained in the mice treated with antibody against phage-KR compared to the PBS-treated mice. The CFUs

(excitation: 550 nm, emission: 570 nm); (f) merged picture of image (d) and image (e). Recombinant phage (RP); Wild-type phage (WT).

in the kidneys of the animals immunized with antibody against phage-KR or antibody against HK-SP were statistically decreased, compared to the mice treated with antibody against wild-type phage or PBS injected mice (**Figure 5**). The histopathological changes between these groups were consistent with the results of the survival study. Macroscopically, there were no lesions observed in internal organs. However, extensive lymphocytes and neutrophils were observed in the kidney of the group MOCK and group PBS by H&E staining, clearly indicating inflammation. In contrast, only medium levels of lymphocytes and neutrophils were observed in the kidney of the group treated with antibody against phage-KR or antibody against HK-SP (**Figure 6**).

#### Liver and Renal Function

The toxicological assessment of the treatments was carried out. Clinical biochemistry parameters, including aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, glucose, urea, creatinine, total protein, and albumin, indicated values in normal biological ranges. No obvious difference between the treated groups and control group was observed.

# DISCUSSION

Sporotrichosis is currently distributed throughout the world with a significant increase in human and animal cases over the last two decades. By far, considerable efforts have been made to develop vaccines for human infections caused by Sporothrix species. We have previously demonstrated that phage displaying KR could

P = 0.79.

elicit protective immunity against fungal diseases (Chen et al., 2017). In this study, the potential treatment efficacy of antiphage-KR IgG against Sporotrichosis was explored. We show that anti-HK-SP or anti-phage-KR antibody can reduce the extent of the damage and improve the survival rate of the mice infected with S. globosa.

Sporotrichosis is a subcutaneous mycosis resulted from the S. schenckii complex, and routine antifungal drugs are not suitable for all patients. Nascimento et al. (2008) demonstrated that the Gp70 molecule is a putative adhesin for fibronectin and laminin, as well as anti-Gp70. Additionally, the yeast cells opsonized with anti-Gp70 mAb increase the phagocytic index (de Almeida et al.,

2015). The immunity inhibits interaction between S. schenckii yeast cells and the subendothelial matrix.

The humoral immune response appears to prevent and control sporotrichosis infection in mice. Antibodies play an important role in protecting the host from fungal infections, including the agglutination of fungal cells, impeding of fungal attachment, increase of the phagocytosis by host effector cells, neutralizing of immunoregulatory molecules, as well as the complement activation (Nascimento et al., 2008). The incapacity of immune sera to mediate protection from fungi indicates insufficient amounts of protective antibody instead of a fundamental incapability of antibodies in protecting against fungal pathogens. However, phage nanofibers could display peptide KR on the surface through genetic means, allowing the peptide KR to possess a conformation analogical to the native protein. Furthermore, they have the ability to induce the antibody response. Additional benefits include the easiness of production and cost-effectiveness, as well as nontoxicity to humans.

The immunogenicity of epitopes is increased when they are expressed on the phage. Mice sera raised against hybrid phage reacted with peptide KR (**Figure 2**), suggesting that the specific response of antibody to the hybrid phage may be motivated by peptide KR instead of the components of phage. Antibodies are naturally generated products of the immune system that interact with other immune components. Additionally, protective antibodies may act through the complement-mediated lysis, promotion of phagocytosis, as well as Fc-mediated release of cytokine and direct antimicrobial efficiency (Casadevall et al., 2004). Furthermore, IgG was collocted and purified from the sera of the immunized mice, and then administered into the mice infected with S. globosa, leading to significant reduction of the

number of CFU and decreased accumulation of inflammatory cells in the kidney. Moreover, injection of IgG significantly enhanced the survival rate of mice infected with S. globosa. Our study demonstrates that the antibody against S. globosa elicited by phage-KR can efficiently protect the mice from disseminated infection of S. globosa. Further studies should examine the protection of phage-KR in other Sporothrix strains.

Itraconazole-resistant strains of S. schenckii complex have been reported (Oliveira et al., 2011; Ottonelli Stopiglia et al., 2014). The use of phage-KR without adjuvant may lead to significant improvement in the efficacy of antifungal therapy, as well as a reduced inflammation reaction. This approach may avoid the adverse side effects of the antifungal drug, especially for the patients with liver dysfunction, pregnant women, and children, who are not suitable for antifungal drug therapy (Kauffman et al., 2000). It is worth mentioning, although the mechanism for phage responses still remains unknown, no obvious adverse effects of phage have been detected in considerable preclinical studies carried out in various animal models. Although phage-KR would not replace conventional antifungal agents, we show that phage-KR may be an important alternative to existing therapies in some cases (Almeida-Paes et al., 2007).

#### CONCLUSION

In conclusion, this study shows clearly the therapeutic efficiency of the humoral immune response generated by recombinant phage for the treatment of S. globosa infection in a fungal infected mouse model. It is notable that the antibody elicited by phage-KR

#### REFERENCES


nanofibers displays efficient inhibition of the infection, such as decrease of the progressive fungi colonizing, alleviation of kidney inflammation and reduction of the mortality rate of the mice with sporotrichosis. Overall, the strategy of using antibody against phage-KR can be an efficient and safe approach for the treatment of sporotrichosis infection.

#### DATA AVAILABILITY

All datasets generated for this study are included in the manuscript and/or the **Supplementary Files**.

#### AUTHOR CONTRIBUTIONS

FC, BY, and RJ conceived and designed the experiments and analyzed the data. FC and SD performed the experiments. FC and BY wrote the manuscript.

#### FUNDING

This work was supported by the Department of Science and Technology of Jilin Province (#20180414033GH).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2019.00160/full#supplementary-material


humoral immune response in sporotrichosis: toward discovery of potential diagnostic and vaccine antigens. PLoS Negl. Trop. Dis. 9:e0004016. doi: 10.1371/ journal.pntd.0004016


challenge in BALB/c mice. Hum. Vaccine Immunother. 10, 1057–1063. doi: 10.4161/hv.27714

Yu, X., Wan, Z., Zhang, Z., Li, F., Li, R., and Liu, X. (2013). Phenotypic and molecular identification of Sporothrix isolates of clinical origin in Northeast China. Mycopathologia 176, 67–74. doi: 10.1007/s11046-013- 9668-6

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Chen, Jiang, Dong and Yan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Carbon-Based Nanomaterials for Biomedical Applications: A Recent Study

#### Debabrata Maiti<sup>1</sup> , Xiangmin Tong<sup>2</sup> , Xiaozhou Mou<sup>2</sup> \* and Kai Yang<sup>1</sup> \*

<sup>1</sup> State Key Laboratory of Radiation Medicine and Protection, School of Radiation Medicine and Protection, School for Radiological and Interdisciplinary Sciences (RAD-X), Collaborative Innovation Center of Radiation Medicine of Jiangsu Higher Education Institutions, Soochow University, Suzhou, China, <sup>2</sup> Key Laboratory of Tumor Molecular Diagnosis and Individualized Medicine of Zhejiang Province, Zhejiang Provincial People's Hospital, Hangzhou, China

The study of carbon-based nanomaterials (CBNs) for biomedical applications has attracted great attention due to their unique chemical and physical properties including thermal, mechanical, electrical, optical and structural diversity. With the help of these intrinsic properties, CBNs, including carbon nanotubes (CNT), graphene oxide (GO), and graphene quantum dots (GQDs), have been extensively investigated in biomedical applications. This review summarizes the most recent studies in developing of CBNs for various biomedical applications including bio-sensing, drug delivery and cancer therapy.

#### Edited by:

Wei Tao, Harvard Medical School, United States

#### Reviewed by:

Dalong Ni, University of Wisconsin-Madison,

> United States Gang Liu, Xiamen University, China Peng Huang, Shenzhen University, China

#### \*Correspondence:

Xiaozhou Mou mouxz@zju.edu.cn Kai Yang kyang@suda.edu.cn

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 15 October 2018 Accepted: 15 November 2018 Published: 11 March 2019

#### Citation:

Maiti D, Tong X, Mou X and Yang K (2019) Carbon-Based Nanomaterials for Biomedical Applications: A Recent Study. Front. Pharmacol. 9:1401. doi: 10.3389/fphar.2018.01401 Keywords: carbon nanomaterials, biomedical applications, biosensor, drug delivery, cancer therapy

# INTRODUCTION

In the field of science and technology, carbon-based nanomaterials (CBNs) are becoming attractive nanomaterials (Cha et al., 2013; Wang et al., 2014, 2015; Tiwari et al., 2015; Lin et al., 2016; Mukhopadhyay et al., 2016; Zhang et al., 2017). Due to the existence of diverse allotropes of carbon, from renowned allotropic phases such as amorphous carbon, graphite and diamonds to newly discovered auspicious carbon nanotubes (CNTs), graphene oxide (GO), graphene quantum dots (GQDs) and fullerene, carbon-based materials have recently become prized (Mostofizadeh et al., 2011). Each member of the carbon family exhibits inimitable features and has been widely exploited in diverse biological applications including biosensing, drug delivery, tissue engineering, imaging, diagnosis and cancer therapy (Hong et al., 2015; Bhattacharya et al., 2016). In 1991, Sumio Iijima first observed the formation of multiwall CNTs from carbon arc discharge. After some years, Prof. Sumio Iijima and Donald Bethune individually perceived single wall CNTs (Monthioux and Kuznetsov, 2006). Afterward, research on CNTs proliferated quickly. CNTs were described as hollow cylinders consisting of graphitic sheets and were classified into single walled carbon nanotube (SWCNT) and multi walled carbon nanotube (MWCNT) (**Figure 1**). SWCNTs, with a cylindrical nanostructure, are made by rolling up a single graphitic sheet with a high aspect ratio. MWCNTs contain few graphitic layers in the rolling pattern, with an interlayer spacing of 3.4Å (Odom et al., 1998; Eatemadi et al., 2014). As a consequence of its unique mechanical, electrical and structural diversity, it gives superior strength, flexibility and electrical conductivity toward various biological entities, which is useful for sensing, medical diagnosis and treating various diseases (Wu et al., 2010; Hwang et al., 2013; Roldo and Fatouros, 2013; Kumar et al., 2017). However, among the various allotropes of carbon, graphene is considered the most attractive material owing to its unique intrinsic properties. About 70 years ago, in 1947, Wallace evaluated the electronic structure of graphene and McClure deduced the corresponding wave equation in 1956. The name "graphene" was first introduced in 1987 by Mouras and co-workers as "graphitic intercalation compounds (GIC)" (Sun et al., 2013).

**250**

Over the last two decades, research on graphene has greatly increased, and various exceptional properties have been observed by investigators. Graphene is described as the planar graphitic sheet of graphite, consisting of sp<sup>2</sup> hybridized carbon network with a carbon-carbon distance of 1.42Å and an interlayer spacing of 3.4Å (**Figure 1**; Erickson et al., 2010). Graphene exhibits a number of exceptional properties that lend to its potential favorability for bio-applications. The prospect of easy functionalization causes the enrichment of functional groups on its surface, which in turn facilitates the specific and selective detection of several biological segments. Furthermore, its extremely large surface area, chemical purity and free π electrons render it an ideal candidate for drug delivery (Yang et al., 2013; Zhang et al., 2013; Pattnaik et al., 2016). Moreover, with the help of its feasible behavior toward different fluorescent dyes, therapeutic agents and other biomaterials, it is widely used for in vivo imaging, diagnosis and treatment of cancer. Another recently invented and attractive biomaterial from the carbon family is GQDs, which is defined as a zero-dimensional graphene sheet with a lateral dimension of less than 100 nm in one to a few layers (3–10) (Song et al., 2015). During the conversion of two-dimensional graphene sheets into GQDs, the GQDs endow excellent photoluminescence due to quantum confinement (Wang et al., 2016). Interestingly, as compared to other fluorescent dye or semiconductor quantum dots, the GQDs exhibit superior biocompatibility and resistance to photo-bleaching. Additionally, GQDs carry keen features of graphene, such as a large surface area and available π electrons, which make the GQDs a smart nanomaterial for a wide range of biomedical applications, including imaging, targeted drug delivery, biomolecules sensing, cancer therapy and so on (Zheng et al., 2015; Kumawat et al., 2017; Li et al., 2017; Chen et al., 2018).

Recently, by utilizing the inherent properties of different newly invented CBNs, these have been modified and extensively used in biomedicine, including applications for bio-sensing, drug delivery and cancer therapy. This encouraged us to conduct a comprehensive review on CBNs in biomedical applications. Regarding the same issue, a few more reviews and prospective articles have been conducted, and most of them have discussed synthesis, characterizations and, to a lesser extent, biomedical applications. Moreover, many of these review articles have discussed overall research that has been carried out over last two decades. In this review we thoroughly recapitulate the most recent progress of CBNs for biomedical applications in the last half decade and offer our own point of view of the field. We expect that this review article will direct researchers to design developed CBNs for superior biomedical applications.

# CARBON NANOTUBES (CNTs) FOR BIOMEDICAL APPLICATIONS

#### Carbon Nanotubes as Biosensors

Owing to their exceptional structural, mechanical, electronic and optical properties, CNTs have been regarded as a new generation nanoprobes (Tîlmaciu and Morris, 2015). Their high aspect ratio, high conductivity, high chemical stability and sensitivity (Zhao et al., 2002) and fast electron-transfer rate (Lin et al., 2004) make them exceedingly fit for biosensing applications. The basic element of CNT-based biosensors is the immobilization of biomolecules on its surface, therefore enhancing recognition and the signal transduction process. On the basis of their target recognition and transduction mechanisms, these biosensors are largely categorized into electrochemical and electronic CNT-based biosensors and optical biosensors. CNTs have been renowned as promising materials for improving electron transfer, which makes them appropriate for combining electrochemical and electronic biosensors (Jacobs et al., 2010; Holzinger et al., 2014; Kumar et al., 2015; Wang et al., 2015; Yang et al., 2015; Hou et al., 2016; Zribi et al., 2016).

Numerous CNT-glucose biosensors based on the conjugation of glucose oxidase have been designed. Zhu et al. (2014) used carbon nanotube non-woven fabrics (CNTFs) to sense glucose from a glucose oxidase-impregnated polyvinyl alcohol solution. The Gaitan Group have emphasized the effect of surface chemistry and the structure of glucose oxidase-coated MWCNT in electrochemical glucose sensing (Gaitán et al., 2017). Electrochemical biosensors built on CNTs have further been designed for detecting nitric oxide and sensing epinephrine (Ulissi et al., 2014; Mphuthi et al., 2016). Bisker et al. (2016) established 20 distinct SWCNT corona phases for detecting human blood proteins. The study revealed that the specific corona phase was capable of recognizing fibrinogen with high selectivity and resulted in a decrease of florescence intensity of SWCNT >80% at saturation (**Figure 2A**). However, absorption intensity remained unchanged with little red shift (**Figure 2A**, inset). The fluorescent response of SWCNT with a smaller diameter was more pronounced compared to the larger diameter nanotube, displayed in the excitation–emission profiles of the SWCNT sample before (**Figure 2B**) and after (**Figure 2C**) the fibrinogen adding. The fibrinogen recognition was tested in the human blood serum environment. Recently, the same group demonstrated that label-free detection of individual proteins' efflux from Escherichia coli (bacteria) and Pichia pastoris (yeast) in real time was possible by using SWCNT (Landry et al., 2017). Baldo et al. (2016) successfully developed a MWCNTbased device detecting arginase-1. The Tuan Group developed a CNT-based field effect transistor (FET) as a conducting channel with a length and width of 15 and 700 µm. The CNTbased field effect transistor (CNTFET) was used directly in a DNA solution under a high current of 1.91 A (Xuan et al., 2017). The Zhou Group has explored the DNA-mediated SERS property of SWNTs, which permitted the ultrasensitive detection of a broad range of ctDNA in human blood. The T-rich deoxy-ribonucleic acid (DNA)-mediated surface-enhanced Raman scattering (SERS) of SWNTs could sense a KRAS G12DM content as low as 0.3 fM, with a detection of 5.0 µL from the sample volume (Zhou et al., 2016). Their photophysical properties, such as fluorescence emission in the NIR region and excellent photo stability, make SWCNTs effective optical probes in biomedicine. Jena et al. (2017) designed single-stranded DNA functionalized SWCNTs, which responded to the lipid content in the endosomal lumen of live cells. From NIR photoluminescence of the

SWCNTs, the lipid content was measured via solvatochromic shift (Jena et al., 2017).

#### Carbon Nanotubes for Drug Delivery

Among the different carbon allotropes, CNTs have attracted escalating attention as a highly competent vehicle for transporting various drug molecules into the living cells because their natural morphology facilitates non-invasive penetration across the biological membranes (Chen et al., 2008; Das et al., 2013; Liu et al., 2013; Panczyk et al., 2016). Generally, drug molecules are attached to CNT sidewalls via covalent or non-covalent bonding between the drug molecules and functionalized CNT. But each of these processes has advantages or disadvantages. The covalent interaction makes the drug-loaded CNT stable in both the extra- and intracellular compartments, meaning that such a phenomenon has a lack of sustained release of the drug inside the cellular microenvironment of cancer cells, which is a shortcoming in the drug delivery system. Non-covalent interaction facilitates the controlled release of the drug in the acidic condition of tumor sites but suffers from stability in extracellular pH levels. Hence, the utilization of the inner hollow cavity of CNT for drug loading provides the ideal isolation of the drug from the physiological environment. In order to overcome the discrepancy of drug release in the tumor cell microenvironment, some external stimuli have been tested via temperature, electric field, light or a combination of these. To evaluate the temperature-responsive release of biomolecules, the Shin Group fabricated chitosan-functionalized CNT with thermosensitive polymer, poly-N-Isopropyl acryl amide (NIPAAm) and 1 butyl-3-. 21 vinyl imidazolium (NIPAAm-co-BVIm), followed by encapsulating the bovine serum albumin (BSA) at body temperature (37◦C). The release of the BSA occurred just above the lower critical solution temperature (LCST) of poly-VBIm (38–40◦C) (Kang et al., 2017). Shi et al. (2015) used an electric field to release the ibuprofen from a hybrid hydrogel composed of sodium alginate (SA), bacterial cellulose (BC), and multi-walled carbon nanotubes (MWCNTs). Estrada et al. (2013) studied the temperature and near infrared (NIR) light-responsive release of methylene blue (MB) from multi-walled carbon nanotube (MWCNT)–k-carrageenan hydrogel. However, to date, many drugs have been loaded onto the CNT including doxorubicin (Huang et al., 2011), paclitaxel (Singh et al., 2016), docetaxel (Raza et al., 2016), oxaliplatin (Lee et al., 2016), etc., to demonstrate the efficiency for in vitro and in vivo cancer treatments. The Dai Group have extensively studied functionalized CNT for the purpose of in vitro and in vivo drug delivery (Dhar et al., 2008; Liu et al., 2008, 2009a,b). Their group discovered a new strategy to make CNT highly water soluble to entrap drug molecules (Liu et al., 2007). The Jain Group evaluated and compared the in vitro and in vivo cancer targeting tendency of doxorubicin (DOX)-loaded folic acid (FA) and estrone (ES)-anchored PEG functionalized MWCNTs (DOX/ES-PEG-MWCNTs) on MCF-7 tumor-bearing Balb/c mice (Mehra and Jain, 2015). After 43 days, the mice treated with DOX/ES-PEG-MWCNTs showed a longer survival span compared to those groups treated with free DOX (18 days) or PBS (12 days). The Khandare Group reported calcium phosphate (CaP)-crowned drug loaded multiwall carbon nanotubes (CNT–GSH–G4– CaP) could be considered as a nanocapsule for intracellular delivery of an anticancer drug (Banerjee et al., 2015). The schematic diagram for the encapsulation and release of drug molecules from the nanocapsule is described in **Figure 2D**. They systematically studied pH triggered CaP dissolution and drug release in subcellular compartments such as lysosomes (pH5.0) (**Figure 2E**). Additionally, zero premature release at physiological pH supported the drug-loaded nanocapsule for effective anticancer treatment. Risi et al. (2014) steadily observed the efficient loading and releasing of a new anticancer drug on CNT. In order to improve the biocompatible nature of CNT, Xu et al. (2016) developed an amine-terminated PEG functionalized polydopamine (PDA) (shell)-CNT (core) nanosystem for drug delivery. The Picaud group investigated theoretically on the loading and releasing of cisplatin onto/from CNT (Mejri et al., 2015).

#### Carbon Nanotubes for Cancer Therapy

Carbon nanotubes are widely used in biomedical applications due to their versatile properties. These are the attractive candidates for the carrying of anticancer drugs, genes and proteins for chemotherapy (Adeli et al., 2013; Eskandari et al., 2014; Amenta and Aschberger, 2015; Hwang et al., 2017). Moreover, strong NIR light absorption capability renders CNTs as efficient photothermal agents. Su et al. (2017) developed iRGDpolyethyleneimine (PEI) functionalized MWCNT followed by conjugation with candesartan (CD). The functionalized iRGD-PEI-MWCNT-CD was assembled with plasmid AT (2) [pAT (2)]. iRGD and CD were used to target αvβ3-integrin and AT1R of tumor endothelium and lung cancer cells, respectively. The CD

treatments at different times. (H) Digital photographs of tumors and tumor-bearing mice after different treatments. Copyright Bisker et al. (2016) Nature publishing

as a chemotherapeutic exhibited synergistic downregulation of VEGF upon combining with pAT (2) and inhabited angiogenesis effectively (Su et al., 2017). The Zhou group designed a DOX-loaded MWCNT-magneto fluorescent carbon quantum dot (CQD) nanocomposite for chemo- and photothermal therapy (Zhang et al., 2017). The negative surface charge of the GdN@CQDs-MWCNTs facilitated binding with positively charged DOX molecules. The material had a high ability to absorb NIR light. On in vivo photothermal therapy, the temperature of the tumor site of the mice treated with GdN@CQDs-MWCNTs/DOX-EGFR was increased to 51.8◦C under laser irradiation at the power density at 2 W/cm<sup>2</sup> for 5 min. No significant change in temperature of the control group treated the mice's tumor site (**Figure 2F**). This heating effect favored the release of DOX and photothermal therapy, as revealed by the suppression of tumor volume (**Figures 2G,H**). Recently, Dong et al. (2017) used DOX-loaded TAT-chitosan functionalized

group, Banerjee et al. (2015) Royal Society of Chemistry, and Zhang et al. (2017) Elsevier.

(upper) CNT–G4–GSH–DOX (lower). (F) In vivo photothermal images under 5 min NIR laser (808 nm, 1 W/cm<sup>2</sup>

MWCNT nanosystem for combining chemo and photothermal therapy. In order to enhance apoptosis in cancer cells, the Dong-woo group used a PEG-coated CNT-ABT737 nanodrug to target mitochondria (Kim et al., 2017). Cytosol release of the nanondrug resulted in apoptosis of lung cancer cells through abruption of the mitochondrial membrane. Finally, the material exhibited effective in vivo therapeutic efficacy. Moreover, the localized heating effect under NIR irradiation induced therapeutic performance. The Chen Group developed a gold nanoparticle-coated carbon nanotube ring (CNTR) with superior Raman and optical signal properties, resulting in the improvement of the photoacoustic (PA) signal and photothermal conversion behavior of the CNTR@Au (Song et al., 2016). The material exhibited a significant outcome in image-guided cancer therapy. The surface plasmon resonance (SPR) absorption by gold in SWNT-Au-PEG-FA nanomaterials improved photothermal cancer killing efficacy (Wang et al., 2012;

) irradiation. (G) Tumor growth curves after different


Bao et al., 2016). Some current observations based on CNTs for different cancer therapy have listed in **Table 1**.

#### GRAPHENE OXIDE FOR BIOMEDICAL APPLICATIONS

#### Graphene Oxide as Biosensor

Graphene oxide is capable of dynamically interacting with the probe and/or for the transduction of a specific response toward the target molecules. This transduction process is achieved by fluorescence, Raman scattering and electrochemical reaction. On the basis of this, GO are broadly used as biosensors (Kim et al., 2017; Suvarnaphaet and Pechprasarn, 2017), and we discuss here the most recent works on the progress of GObased nanoarchitecture in biosensing applications. Graphene nanomaterials have been extensively used for the selective electrochemical sensing of single- and double-stranded DNA (Liu et al., 2012; Tang et al., 2015). The high sensitivity could be attributed to the excellent electrochemical properties of graphene, the strong ionic interaction between the negatively charged – COOH groups and the positively charged nucleobases, and the robust π–π stacking between the nucleobases and honeycomb carbon framework. The Rahigi group developed reduced graphene nanowire (RGNW) biosensors for electrochemical detection of the four bases of DNA (guanine, tyrosine, adenine and cytosine) by checking oxidation signals of the discrete nucleotide bases (Akhavan et al., 2012). The RGNW exhibited tremendous stability, with only 15% variation in the oxidation signals upon an increase in differential pulse voltammetry (DPV) up to 100 scans. Recently, Zhang and co-workers designed carboxyl (−COOH) functionalized GO and polyaniline (PANI) modified GO. They successfully detected DNA via DPV with ranges between 1 × 10−<sup>6</sup> and 1 × 10−<sup>14</sup> (Cheng et al., 2017). Johnson and co-workers designed a label-free DNA biosensor based on graphene field effect transistors (GEFTs) functionalized with single-stranded probe DNA. This highly sensitive biosensor offered a broad analytical range with a detection limit of 1 fM for 60-mer DNA oligonucleotides (Ping et al., 2016). By the same group, a device based on gold nanoparticledecorated GEFTs (Au NP-Gr-FETs) was fabricated by the physical vapor deposition method. Thiol-functionalized Au NP-Gr-FETs were able to detect DNA with a detection limit of 1 nM and exhibited high specificity against no complementary DNA (Gao et al., 2016). A single-layer graphene (SLG)-based FET biosensor was able to detect a very low concentration of DNA (10 fM) (Zheng et al., 2015). Kim et al. (2016) developed a graphene surface modified vertically aligned silicon nanowire

for detecting oligonucleotides with sensitivity and selectivity. They first decorated oligonucleotides on the surface of Si nanowire arrays and followed by hybridization to the probe, resulting in an increase in the biosensor (**Figure 3A**). It was observed that the current of the biosensor was increased from 19 to 120% with an increase in concentration of DNA from 0.1 to 500 nM (**Figure 3B**; Kim et al., 2016). Park et al. (2014) evaluated the adsorption and desorption mechanism of single- and double-stranded DNA on GO. They observed that ssDNAs were preferentially adsorbed on GO whereas dsDNA exhibited lower affinity. Alternatively, recently it was studied that adsorption of DNA on GO is length-dependent (Huang and Liu, 2018). Prabowoa et al. (2016) introduced a novel idea for the detection of Mycobacterium tuberculosis DNA hybridization using graphene deposited on a SPR-sensing chip. The use of GO-based nanomaterials for glucose sensing is now growing prosperously (Cheng et al., 2017; Kumar et al., 2017). A device based on graphene gated electrodes with glucose oxidase exhibited superior selectivity and enhanced glucose sensitivity with a detection limit of 0.5 mM (Zhang et al., 2015). The Jun group fabricated reduced graphene oxide (RGO) with phenyl butyric acid (PBA), which could be used as a linker to bind glucose. The well-modulated RGO-based radio frequency (RF) sensor device was capable of detecting glucose levels in the range between 1 and 4 mM (Park et al., 2016). The Chen Group prepared a highly stable and reusable graphene-bismuth composite device, which was capable of detecting glucose in a wide linear range of 1–12 mM with a high sensitivity of 2.243 µAmM−<sup>1</sup> cm−<sup>2</sup> and with a low detection limit of 0.35 mM (Mani et al., 2015). Carbon modified graphene/fullerene C60 composite was fruitfully designed to detect glucose in the range of 0.1–12.5 mM. The device showed a limit of detection (LOD) of 35 µM, with high sensitivity of 55.97 µAmM−<sup>1</sup> cm−<sup>2</sup> (Thirumalraj et al., 2015). Ponpandian's group successfully developed hydroxyapatite 1-D nanorods on a graphene nanosheet modified with glassy carbon electrode. The device exhibited an excellent sensing property in a wide range of 0.1–11.5 mM with a LOD of 0.03 mM and greater sensitivity of 16.9 µAmM−<sup>1</sup> cm−<sup>2</sup> (Bharath et al., 2015).

#### Graphene Oxide for Drug Delivery

Utilizing the extremely large surface area and available π electrons, graphene is suitable as a drug carrier. Wang et al. (2012) loaded a high amount of doxorubicin (DOX) on phospholipid monolayer coated graphene and subsequently observed the sustained release of DOX to a greater extent

at an acidic pH compared to a basic pH (Liu et al., 2012). DOX could be loaded on a graphene sheet via physisorption followed by surface modification by PEG-NH<sup>2</sup> in order to enhance stability and compatibility in a biological medium (Zhang et al., 2013). Nandi and co-workers were able to load both a hydrophilic drug (DOX) and a hydrophobic drug (indomethacin) successfully on poly-N-isopropyl acrylamide (PNIPAM) grafted GO (GPNM) via π–π interaction, H-bonding and hydrophobic interaction (Kundu et al., 2015). They grafted PNIPAM covalently with GO through the free radical polymerization process (FRPP). The controlled release of DOX was favorable in an acidic pH due to the enhancement of hydrophilicity, higher solubility to DOX and a minimization of the hydrogen bonding interaction between DOX and the GPNM surface. Xu et al. (2014) loaded paclitaxel (PTX) onto GO-PEG via π–π stacking and hydrophobic interactions and the loading capacity was calculated to be 11.2 wt%. Zhao et al. (2015) designed well-defined polymethylmethacrylic acid (PMMA)-coated polyethylene glycol (PEG) modified graphene oxide nanoparticles (GON), which were highly dispersed in PBS solution, and acted as an efficient drug delivery system (**Figure 3C**). PMMA brushes capably reduce the impulsive release of DOX in the stimulated normal tissues and accelerates DOX release in the tumor tissues in response to a reducing agent, glutathione (GSH) (10 µM) (**Figure 3D**). Furthermore, strong fluorescence of DOX (green) indicated a persistent release of DOX from DOX-loaded PEGylated alginate (ALG-PEG) grafted GON and its internalization (**Figure 3E**; Zhao et al., 2014). The Tan group designed DOX-loaded GO followed by modification with hyaluronic acid (HA), which was used as a targeting agent and to enhance the stability of the HA-GO-DOX nanohybrid (Song et al., 2014). Encouraged by the high loading of DOX on GO, recently Mahdavi et al. (2016) have fruitfully carried out a simulation study on DOX loading and releasing in GO at different pH points. In doxorubicin (DOX)-loaded p-aminobenzoic acid polyethyleneimine (PEI), biotin, b-Cyclodextrin (b-CD) conjugated graphene oxide (rGO) nanosystem, the PEI and biotin were used to enhance the stability and targeting efficacy, respectively. The b-Cyclodextrin (b-CD) acted as host molecules for accommodating guest molecules, such as water insoluble anticancer drugs (Wei et al., 2014).

#### Graphene Oxide for Cancer Therapy

Recently, GO has been considered to be an exciting nanomaterial due to its inherent size- and shape-dependent optical properties, unique physicochemical behavior, extremely large surface to

Red staining. THP-1 cells were seeded into 8-well chamber slides and incubated with f-MWCNTs at 120 µg/ml in complete RPMI 1640 for 3 h. After fixation, cells were stained with Magic Red (ImmunoChemistry Technologies), wheat germ agglutinin-Alexa 488, and Hoechst 33342 dye, followed by visualization under a confocal 1P/FCS inverted microscope. (B) ROS production in GLC-82 cells treated with 100 mg/L of GO for 48 h. The positive control was prepared by culturing the cells with RPMI-1640 containing 100 µM of H2O<sup>2</sup> for 1 h prior to the addition of DCFH-DA. The cells without DCFH-DA treatment was taken as a negative control. The control means that cells without exposure to GO were labeled by the DCFH-DA. (C) GQDs with different sizes on the membrane after 100 ns MD simulation: (I) GQD7-small size, (II) GQD61-small size, (III) GQD151-large size, and (IV) GQD275-large size. The GQDs are shown by a VDW model with VMD. N atoms (blue) and P atoms (yellow) in the membrane are also shown in the VDW model. (D) The angles between different GQDs and the x–y plane of the lipid membrane as a function of simulation time. Copyright Li et al. (2013, 2014) and Liang et al. (2016) American Chemical Society.

volume ratio and versatile surface properties, which make it ideal nanomaterial for cancer therapy (Kumar et al., 2017; Nejabat et al., 2017). Yu et al. (2017) designed αvβ6-targeting peptide (HK-peptide) functionalized and photosensitizer (HPPH) coated GO (GO (HPPH)-PEG-HK). The GO (HPPH)-PEG-HK activated dendritic cells and significantly prevented tumor growth and lung metastasis by increasing the infiltration of cytotoxic CD8<sup>+</sup> T lymphocytes within tumors as evidenced

by in vivo optical and single-photon emission computed tomography (SPECT)/CT imaging (Yu et al., 2017). The Chen Group fabricated a DOX-loaded RGO-gold nanorods vehicle for combined photothermal therapy and chemotherapy. A large release of DOX was observed due to the NIR photothermal heating effect and acidic nature of the tumor microenvironment (Song et al., 2015). The tight packing of Au NPS on GO led to an enhancement of the absorption peak from 528 to 600 nm. Under laser light (808 nm, 1.0 W/cm<sup>2</sup> ), Au (30 nm)-GO (20 nm) showed the maximum temperature increase of 23.2◦C (Kang et al., 2017). Cheon et al. (2016) claimed that a DOX-loaded BSA functionalized graphene sheet could be a powerful tool for combining chemo- and photothermal therapy for brain tumors. Regarding the clinical application, the Chen Group fabricated hyaluronic acid-chitosan-g-poly (N-isopropyl acrylamide) (HACPN) grafted DOX-folic acid-GO thermosensitive hydrogel for breast cancer therapy (Fong et al., 2017). Su et al. (2016) designed a novel material consisting of dual chemotherapeutics loaded sponge-like carbon material on graphene nanosheet (graphene nanosponge) supported lipid bilayers (lipo-GNS) modified with tumor targeting protein. The well fabricated ultrasmall lipo-GNS (40 nm) showed a significant accumulation in the tumor site and, therefore, successful suppression of the xenograft tumors in 16 days (Su et al., 2016). Shao et al. (2017) designed a mesoporous silica (MS) coated polydopamine that functionalized RGO followed by modification with hyaluronic acid (HA) and DOX loading. The pH dependent and near infrared-triggered DOX release made the RGO@MS (DOX)-HA an effective chemo-photothermal agent (Shao et al., 2017). Very recent, Dai et al. (2017) designed TiO2-MnOx conjugated graphene composite as a smart material for tumor eradication. Our group developed <sup>131</sup>I labeled PEG functionalized nano RGO for combined radio and photothermal therapy (**Figure 3F**). Effectual tumor accumulation of <sup>131</sup>I-RGO-PEG was observed after its intravenous injection as confirmed by gamma imaging (**Figure 3G**). RGO exhibited strong near-infrared (NIR) absorbance and could induce effective photothermal heating of the tumor under NIR light irradiation. <sup>131</sup>I was able to emit b rays to kill cancer cells (**Figure 3H**; Chen et al., 2015). Some more recent studies based on GO nanomaterials for different cancer therapies have been listed in **Table 1**.

#### GRAPHENE QUANTUM DOTS (GQDs) FOR BIOMEDICAL APPLICATIONS

#### Graphene Quantum Dots (GQDs) as Biosensors

Recently, GQD-based biosensors have largely been developed for practical applications in clinical analysis and disease diagnosis. On the basis of excellent photoluminescence (PL), electro chemiluminescence (ECL) and electrochemical behaviors of GQD, these have been widely used for detecting biomacromolecules including DNA, RNA, proteins or glucose molecules with better selectivity and sensitivity (Xie et al., 2016; Kumawat et al., 2017). Qian et al. (2014) developed DNA probe-functionalized reduced GQDs to detect DNA based on the Furrier Resonance Energy Transfer (FRET) fluorescence sensing method. The Qui group successfully designed a Zr4<sup>+</sup> coordinated phosphorylated peptide-GQD conjugate that was capable to detect casein kinase II (CK2) in the range between 0.1 and 1.0 ml−<sup>1</sup> with a detection limit of 0.03 ml−<sup>1</sup> (Wang et al., 2013). Zhang et al. (2017) developed pyrene-1-boronic acid (PBA) functionalized GQD for glucose sensing (**Figure 4A**). They observed that glucose sensitivity was strongly dependent on the PBA concentration as revealed from the significant shift of Dirac voltage with an increase in the concentration of PBA (**Figure 4B**). Moreover, the significant enhancement of relative capacitance with an increase in glucose concentration further suggested that the PBA functionalized GQD could be used as a perfect probe for glucose sensing (**Figure 4C**). The Wei group prepared an electrochemifluorescent polyvinyl alcohol (PVA)/GQD nanofiber for highly sensitive and selective detection of both H2O<sup>2</sup> and glucose (Zhang et al., 2015). Here, after adsorption of glucose oxidase (GOD) onto the (PVA)/GQD nanofiber, the molecular recognition between GQD and glucose triggered the production of H2O2, which was detected by fluorescent GQD. The detection of cancer cells in early stage of disease has become a perquisite paradigm. In this regard, Wang et al. (2016) designed Pd NPs decorated N-doped GQD (NGQD) for cancer detection. The NGQD@NC@Pd HNS hybrid material exhibited significant electrochemical reduction of H2O2. Hence, it was possible to detect various living cancer cells (Xi et al., 2016).

# Graphene Quantum Dots (GQDs) for Drug Delivery

Graphene quantum dots possesses some unique features, such as a single atomic layer with small lateral size and an oxygenrich surface that renders it suitable for loading drug molecules and enhancing stability in physiological media. In addition, the fluorescent property of GQD makes it an appropriate platform for the traceable delivery of the drug into the cancer cells (Cheng et al., 2015; Pistone et al., 2016; Srivastava et al., 2016). Hence, GQDs have been widely used for drug delivery in various diseases from last decade. The Zhu group loaded DOX on a GQD-embedded zeolite imidazolate framework (ZIF-8), where ZIF-8 was used as an efficient drug carrier. DOX-loaded ZIF-8/GQD nanoparticles effectively showed acidic pH responsive drug release behavior (Tian et al., 2017). Intracellular drug delivery and the real-time monitoring of drug release could be possible from DOX-loaded aptamer/GQD capped fluorescent mesoporous silica nanoparticles. In the adenosine triphosphate (ATP)-rich cytoplasm of the tumor cells, the ATP aptamer caused the release of the GQDs from nanocarriers, resulting in the release of DOX (Zhang et al., 2015). On the basis of the salient physicochemical properties of GQDs, the Wei group developed DOX loaded GQD followed by conjugation with Cy5.5 dye via a cathepsin D-responsive (P) peptide (Ding et al., 2017). The drug-loaded nanoconjugate showed improved tissue penetration and cellular uptake properties, which in turn facilitated superior therapeutic performance both in vitro and in vivo. The cellular uptake of 4T1 cells and release of

DOX were evaluated by confocal laser scanning microscopy (CLSM) (**Figure 4E**). The GQD-P-Cy treated cells exhibited blue fluorescence, implying promising internalization. The invisible fluorescence signal of Cy5.5 from GQD-P-Cy treated cells indicted its satisfactory biocompatibility. The green fluorescence signal around 565 nm from the DOX@GQD-P-Cy treated cells demonstrated the DOX releasing from GQD. The strong in vivo fluorescence signal of DOX from the tumor site signified the great accumulation of DOX inside the tumor (**Figure 4F**). Nigam et al. (2014) developed a GQD-conjugated gemcitabineloaded HSA nanoformulation for targeted drug delivery. In this nanosystem, albumin helped to deliver gemcitabine to the tumor cells via the gp60 pathway (Nigam et al., 2014). Pietro and colleagues designed biotin-conjugated DOX-loaded GQD for targeted drug delivery in cancer therapy (Iannazzo et al., 2017). Sui et al. (2016) fabricated a cisplatin-GQD nanoconjugate for enhanced anticancer activity. In this nanoconjugate, GQD helped to improve cellular uptake and then cisplatin assisted to enhance nuclear uptake by interacting with DNA (Sui et al., 2016). Wang et al. (2014) demonstrated that ligand modified DOX-loaded GQD-folic acid nanocarrier improved selective cell labeling, targeted drug delivery and the real-time monitoring of cellular uptake.

# Graphene Quantum Dots (GQDs) for Cancer Therapy

Owing to its outstanding physicochemical property, low toxicity, good hydrophilicity, stable intrinsic fluorescence property and surface functional groups, various kinds of nanomedicines, from chemotherapeutics to radioisotopes, were conceivable for loading and usage for cancer treatments (Iannazzo et al., 2017). The Lee group fabricated hydrophobic anticancer drug, curcumin loaded GQDs for synergistic chemotherapy (Some et al., 2014). Ge et al. (2014) synthesized GQD, which showed tremendous singlet oxygen (1O2) generation capability and photodynamic therapy (PDT) via in vivo therapy. The diameter of GQD was in the range between 2 and 6 nm as revealed from scanning transmission electron microscopy (STEM) (**Figure 4G**). Their group explored how GQD was able to generate singlet oxygen (1O2) under irradiation in presence of 2, 2, 6,6-tetramethylpiperidine as observed from ESR peaks (**Figure 4H**). However, absence of an ESR signal in the presence of 5-tert-butoxycarbonyl-5-methyl-1 pyrroline N-oxide under irradiation indicated that no other ROS was generated. Moreover, no significant diffusion of GQD was at the injection site (**Figure 4I**). On in vivo PDT, a tumor of female BALB/c mice treated with GQD started to diminish after 9 days and after 17 days (**Figure 4**). Yao et al. (2017) explored that GQD capped magnetic mesoporous silica nanoparticles have the ability to produce heat under an alternating magnetic field (AMF) and/or under NIR irradiation. The material exhibited efficient chemo-photothermal therapy and magnetic hyperthermia as revealed from an in vitro study (Yao et al., 2017). The Fan group loaded IR780 on folic acid functionalized GQD for targeted photothermal therapy. Upon irradiation with an 808 nm laser for 5 min, the temperature at the tumor site of the IR780/GQD-FA treated mice increased abruptly to 58.9◦C and in vivo antitumor study exhibited a clear suppressive effect on tumor growth, and the tumor had almost dissipated by the 15th day (Li et al., 2017). Other studies based on GQDs for different cancer therapies are listed in **Table 1**.

# TOXICITY OF CARBON NANOMATERIALS

Carbon nanomaterials are a novel class of materials that are widely used in biomedical fields including the delivery of therapeutics, biomedical imaging, biosensors, tissue engineering and cancer therapy. However, they still suffer from their toxic effect on biological systems. Until now, various investigations have been carried out on the toxicity of CNT (Liu et al., 2013; Madani et al., 2013; Allegri et al., 2016; Kobayashi et al., 2017). From numerous studies it has been revealed that several factors contribute to the toxicity of CNT. The effect of metal impurities in CNT could have a substantial impact on toxicity (Koyama et al., 2009; Vittorio et al., 2009; Aldieri et al., 2013). The impurities, such as metal ions, were incorporated inside the CNT during synthesis and caused toxicity to the cells. The length of CNT has a great impact on the toxicity of CNT only due to the failure of their cellular internalization (Kostarelos, 2008). Some groups have prepared CNT with different sizes and studied their toxic behavior on cells or DNA (Smart et al., 2006; Raffa et al., 2008). The Donaldson group described that long-term retention of long CNT led to severe inflammation, which caused progressive fibrosis (Murphy et al., 2011). Moreover, the higher diameter with equal average length of CNT exhibits greater toxicity (Kolosnjaj-Tabi et al., 2010). Owing to the difference in size, structure and chemical surface states between SWCNT and MWCNT, they delivered different toxicity effects on cells (Fraczek et al., 2008; DiGiorgio et al., 2011). Moreover, the solubilizing agents played an important role in the toxicity of CNT (Nam et al., 2011; Kim et al., 2012). The individual CNTs tend to bundle in presence of some natural dispersants and led to toxicity. Interestingly, surface functionalization of CNT triggered toxicity in cells. The Jos group found that –COOH functionalized SWCNT induced higher toxicity compared to the non-functionalized SWCNT in the HUVEC cell lines (Praena et al., 2011). On the other hand, Li et al. (2013) demonstrated that strongly cationic functionalized MWCNT has greater potential for lysosomal damaging due to their high cellular uptake and NLRP3 inflammasome activation in comparison to the carboxyl group-functionalized or moderately amine group-functionalized MWCNT, as can be observed by confocal imaging (**Figure 5A**; Li et al., 2013). Like CNT, graphene has also limitations to biomedical application due to its toxicity. Ou et al. (2016) thoroughly described in their recent review article the toxicity of graphene in different organs. Numerous studies have been conducted on the toxicity of graphene in animals and cells (Shareena et al., 2018). It was stated that several parameters, including concentration, lateral dimension, surface property and functional groups, greatly influence its toxicity in biological systems (Seabra et al., 2014; Alshehri et al., 2016). Li et al. (2014)

observed that GO at a concentration of 100 mg/L induced reactive oxygen species (ROS) production in GLC-82 cells upon incubation for 24 h and caused toxicity (**Figure 5B**). To overcome the toxic effect of GO in various biomedical applications, many research groups have designed GO with various biological molecules. The Zhou group modified a graphene sheet by coating it with blood protein to reduce its toxic effect (Chong et al., 2015). Among different materials of the carbon family, GQDs contain some exciting properties and these have thus been extensively used for biological applications as discussed above. The toxicity of GQDs is different from graphene and GO, thus it is an imperative and serious issue that ought to be addressed. After many investigations, it has been implied that various parameters govern the toxicity of GQDs. It seems that the smaller size of GQDs is an advantage over GO or CNT in terms of toxicity. More importantly, Wang et al. (2016) showed a cell viability mapping curve for various cells under the same conditions and concluded that GQDs with a size below 10 nm possess extremely high cell viability. No doubt, the concentration of nanomaterials is a dominating factor in toxicity. For GQDs, the concentration tolerance of the cells to different GQDs is contradictory. The Shen group showed theoretically that the potential cytotoxicity of GQDs depends on their size and concentration (Liang et al., 2016). They observed that in the 100 ns scale simulation, GQDs with relatively small size could permeate into the POPC membrane (**Figure 5C**). The permeation of GQDs could affect the thickness of the POPC lipid membrane. At the starting point, angles between GQDs and lipid membrane were 0◦ in all cases. During simulation, smaller-size GQDs permeated the POPC membrane and created an angle in the range between 45◦ and 70◦ . GQDs with larger sizes were only absorbed on the lipid membrane surface and formed an angle in the range of 0 ◦ to 10◦ (**Figure 5D**). Moreover, it has been observed that the surface functional groups of nanomaterials have a great impact on the toxicity of nanomaterials. The Shang group reported after an investigation that hydroxylated-GQDs have significant toxicity on A549 and H1299 cells (Tian et al., 2016). In contrast, Nurunnabi et al. (2013) claimed that carboxylated GQDs had no acute toxicity on different cancer cells such as KB, MDA-MB231, A549 and the normal cell line such as MDCK. Furthermore, after a long-term in vivo study they did not find notable damage to the organs. Regrettably, we have not yet found any article that gives clear information based on the effect of different functional groups in the toxicity of GQD nanomaterials.

#### REFERENCES


### FUTURE PROSPECTIVE AND CONCLUSION

Over the last two decades, widespread research efforts have been conducted on CBNs as one of the most widely used classes of nanomaterials. Having their inherent mechanical, optical, electrochemical and electrical properties, CBNs have been extensively used in multiple areas. In addition, owing to their versatile surface properties, size and shape over the past decade, CBNs have drawn great attention in biomedical engineering. Interestingly, CBNs are becoming promising materials due to the existence of both inorganic semiconducting properties and organic π–π stacking characteristics. Hence, it could effectively interact with biomolecules and response to the light simultaneously. By taking advantage of such aspects in a single entity, CBN-based nanomaterials could be used for developing biomedical applications in future. Concerning their toxic effect in the biological system, several chemical modification strategies have been developed and successfully used in bio-applications including drug delivery, tissue engineering, detection of biomolecules and cancer therapy. This review article provides some achievements in the use of CBNs for biomedical applications. Moreover, in this paper we also focus on some recently found key features of CBNs and their utilizations for superior bio-applications. However, as CBNs still contain toxicity, more systematic studies are needed to determine the toxicity and pharmacokinetics of CBNs.

# AUTHOR CONTRIBUTIONS

DM and XT wrote the manuscript. KY and XM revised the manuscript.

# FUNDING

This work was partially supported by National Natural Science Foundation of China (31822022, 81471716, 81672430, and 81570198), a Jiangsu Natural Science Fund for Outstanding Youth Science Foundation (BK20180094), a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and Zhejiang Medical Technology Plan Project (WKJ-ZJ-1709).


Banerjee, S. S., Todkar, K. J., Khutale, G. V., Chate, G. P., Biradar, A. V., Gawande, M. B., et al. (2015). Calcium phosphate nanocapsule

fphar-09-01401 March 7, 2019 Time: 17:52 # 12


anti-cancer drug delivery, release, and response. ACS Appl. Mater. Interfaces 9, 27396–27401. doi: 10.1021/acsami.7b08824



graphene oxide drug delivery systems. J. Mater. Chem. B 4, 7441–7451. doi: 10.1039/c6tb00746e




**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Maiti, Tong, Mou and Yang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# pH-Responsive Dual Drug-Loaded Nanocarriers Based on Poly (2-Ethyl-2-Oxazoline) Modified Black Phosphorus Nanosheets for Cancer Chemo/Photothermal Therapy

Nansha Gao<sup>1</sup> , Chenyang Xing<sup>1</sup> , Haifei Wang<sup>1</sup> , Liwen Feng<sup>2</sup> , Xiaowei Zeng<sup>2</sup> , Lin Mei<sup>2</sup> and Zhengchun Peng<sup>1</sup> \*

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Kun Zhou, Boston Children's Hospital, United States Houliang Tang, Southern Methodist University, United States Ruinan Yang, Pharmaceutical Product Development, United States

> \*Correspondence: Zhengchun Peng zcpeng@szu.edu.cn

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 27 January 2019 Accepted: 04 March 2019 Published: 19 March 2019

#### Citation:

Gao N, Xing C, Wang H, Feng L, Zeng X, Mei L and Peng Z (2019) pH-Responsive Dual Drug-Loaded Nanocarriers Based on Poly (2-Ethyl-2-Oxazoline) Modified Black Phosphorus Nanosheets for Cancer Chemo/Photothermal Therapy. Front. Pharmacol. 10:270. doi: 10.3389/fphar.2019.00270 <sup>1</sup> Key Laboratory of Optoelectronic Devices and Systems of Ministry of Education, College of Optoelectronic Engineering, Shenzhen University, Shenzhen, China, <sup>2</sup> School of Pharmaceutical Sciences (Shenzhen), Sun Yat-sen University, Guangzhou, China

Synergistic cancer therapy, such as those combining chemotherapeutic and photothermal methods, has stronger treatment effect than that of individual ones. However, it is challenging to efficiently deliver nanocarriers into tumor cells to elevate intracellular drug concentration. Herein, we developed an effective pH-responsive and dual drug co-delivery platform for combined chemo/photothermal therapy. An anticancer drug doxorubicin (DOX) was first loaded onto the surface of black phosphorus (BP). With poly(2-ethyl-2-oxazoline) (PEOz) ligand conjugated onto the polydopamine (PDA) coated BP nanosheets, targeted long circulation and cellular uptake in vivo was significantly improved. With another anticancer drug bortezomib (BTZ) loaded onto the surface of the nanocapsule, the platform can co-deliver two different drugs. The surface charge of the nanocapsule was reversed from negative to positive at the tumor extracellular pH (∼6.8), ionizing the tertiary amide groups along the PEOz chain, thus facilitating the cell internalization of the nanocarrier. The cytotoxicity therapeutic effect of this nanoplatform was further augmented under near-infrared laser irradiation. As such, our DOX-loaded BP@PDA-PEOz-BTZ platform is very promising to synergistic cancer therapy.

Keywords: synergistic cancer therapy, black phosphorus, co-delivery, pH-responsive, charge reversal

# INTRODUCTION

Black phosphorus (BP), as a novel 2D material, has attracted global attention owing to outstanding optoelectronic properties and wide applications (Han et al., 2017; Zhou et al., 2017). Bulk BP can be easily exfoliated into nanosheets (NSs) with different thicknesses (Churchill and Jarillo-Herrero, 2014; Yasaei et al., 2015). Compared with other 2D materials such as graphene and MoS2, BP has a larger specific surface area to adsorb large amounts of theranostic agents or antitumor drugs, thereby being potentially applicable to drug delivery (Tao et al., 2017). Unlike other 2D materials, the bandgap voltage of BP is related with its number of layers, ranging from 0.3 eV (a bulk value) to

∼2.0 eV (a monolayer value), (Liu et al., 2015) so it has absorptions in both UV and near-infrared (NIR) regions. Thus, BP NSs have unique optoelectronic performance (Shao et al., 2016) to work as either an excellent nano-optoelectronic device or an effective photothermal agent for photothermal therapy (PTT) due to high photothermal conversion efficiency and NIR extinction coefficient (Guo et al., 2015; Sun et al., 2015, 2016; Xing et al., 2017, 2018b). Therefore, antitumor drug doxorubicin (DOX) or paclitaxel are often loaded onto BP NSs to fabricate multifunctional drug delivery systems for synergistic cancer chemotherapy/PTT.

Nevertheless, the application of BP is greatly hindered, because it is prone to degradation into PxOy in the air and aqueous solutions (Liu et al., 2014; Doganov et al., 2015). To this end, researchers have endeavored to stabilize BP NSs through surface modification strategies such as ligand surface coordination (Zhao et al., 2016; Guo et al., 2017), covalent aryl diazonium functionalization (Ryder et al., 2016) and capping layer protection (Wood et al., 2014), which, however, are unsuitable for human drug delivery for either introducing toxic substances or weakening photothermal outcomes. As a surfaceadherent biomimetic material formed through the oxidative self-polymerization of dopamine under alkaline conditions, polydopamine (PDA) is inspired by marine mussel and has been widely used a coating on nanomaterial surfaces owing to high biodegradability, biocompatibility and pH responsiveness at low pH values (Lee et al., 2007; Liu et al., 2013; Cheng et al., 2017a). We have previously elevated the stability of BP NSs in aqueous solution by simply modifying their surface with PDA safely and effectively, without attenuating the photothermal effects (Gao et al., 2018), based on the photothermal conversion efficiency of PDA (Liu et al., 2013; Cheng et al., 2017b; Peng et al., 2018).

Besides, BP NSs can be readily phagocytosed and cleared by the mononuclear phagocytic system (MPS) after being injected in vivo. Conventionally, the in vivo circulation of nanocarriers is prolonged through surface modification with hydrophilic or zwitterionic polymers among which polyethylene glycol (PEG) is most investigated and utilized due to excellent biocompatibility and hydrophilicity, especially for the polymers based drug delivery (Chen et al., 2017a; Feng et al., 2017; Guo et al., 2018; Wang et al., 2018). This technique is also known as PEGylation (Zeng et al., 2013; Cheng et al., 2017c; Nie et al., 2017; Zhao et al., 2017; Xiao et al., 2018; Tang et al., 2019). However, it suffers from the following issues. First, PEGylated therapeutic agents, after being administered repeatedly, cannot fully escape from being phagocytosed by cells in MPS, and the immunogenicity is bound to induce obvious humoral immune response. Therefore, they are recognized and removed by the immune system. Second, PEGylated liposome and nanoparticles can be immunoreactive to induce an accelerated blood clearance (ABC) phenomenon (Wang et al., 2007; Tsai et al., 2018). Third, the in vivo stability of PEG is affected, because its polyether main chain easily undergoes oxidative degradation. Burt et al. (1999) found the cleared PEG and PDLLA fragments of (ethylene glycol)-block-poly(D,L-lactide) (PEG-b-PDLLA) micelles in mouse urine. Furthermore, it is difficult to conjugate the surface of PEGylated nanocarriers with functional ligands because PEG has limited reactive groups, thus requiring an alternative technique to stabilize BP NSs for in vivo biomedical applications (Wu et al., 2018). In recent years, poly(2 ethyl-2-oxazoline) (PEOz) has been verified as a high-molecular weight, long-chain polymer with high water solubility, flexibility and biocompatibility, and approved by the United States Food and Drug Administration. Meanwhile, PEOz is capable of long circulation in vivo, inhibiting protein adsorption and decreasing blood clearance, as a qualified substitute for PEG. Compared with PEG, PEOz has a more stable main chain which facilitates the introduction of various active groups and provides a chemical basis for further linking to target molecules. Notably, with unique tertiary amide groups in the main chain, PEOz has a similar pKa to that of physiological pH, which can be adjusted by varying the molecular weight. At pH lower than its pKa, PEOz is reversed from negatively to positively charged through ionization of tertiary amide groups along the PEOz chain. As a result, PEOz-modified drug nanodelivery system can be enriched and charge-reversed in the weakly acidic environment of tumor tissue (pH ∼6.8), allowing endocytosis and pHresponsive drug release after being induced by the low pH (∼5.0) of endosomes and lysosomes. Finally, the drug release rate is controllable and the tumor-targeting ability is improved, managing to enhance the antitumor activities and to reduce the side effect simultaneously (Gao et al., 2015a,b,c; Zhao et al., 2015; Wang et al., 2017).

As a broad-spectrum antitumor agent, DOX can be intercalated into DNA of tumor cells to suppress nucleic acid synthesis, exerting therapeutic effects on acute leukemia and a variety of solid tumors. Nevertheless, it is highly cytotoxic and easily degradable, without selectivity or specificity (Chen et al., 2017b; Xing et al., 2018a; Zhang et al., 2018). Bortezomib (BTZ), on the other hand, is a common clinical antitumor agent applicable to patients with multiple myeloma, which can inhibit tumor cell growth by binding the threonine residues of active sites of several proteases and induce apoptosis mainly via the mitochondrial pathway. However, the therapeutic effects of BTZ on many types of solid tumors are limited, because it is non-specifically bound to normal cell proteins and can thus be rapidly cleared by the blood, accompanied by dose-related cytotoxicity (Su et al., 2011). Consequently, we herein designed a nanodrug BP-DOX@PDA-PEOz-BTZ carrying DOX and BTZ simultaneously for the chemo/photothermal combination therapy of breast cancer. The PDA layer enhanced the system stability before reaching the tumor site, and maintained the remarkable photothermal effects for subsequent modification. Afterward, PEG was replaced by PEOz to prolong in vivo drug circulation and to increase cellular uptake. In the meantime, a pH-targeted controlled release trigger system was constructed to remedy the deficiency of chemical drugs in solid tumor therapy, and to further boost the antitumor effects relying on high photothermal conversion efficiency. This system is conducive to chemo/photothermal combination therapy by not only raising the drug loading content, cellular uptake and pH-responsive release rate, but also exhibiting high photothermal activity against tumor cells.

#### MATERIALS AND METHODS

fphar-10-00270 March 16, 2019 Time: 17:5 # 3

#### Materials

Bulk BP was purchased from Smart-Elements (Austria) and stored in a 4◦C refrigerator. Dopamine hydrochloride was bought from Sigma-Aldrich (St. Louis, MO, United States). H2N-PEG was obtained from Shanghai Yare Biotech, Inc. (China). H2N-PEOz was bought from Xi'an Ruixi Biological Technology Co., Ltd. (China). The molecular weights (Mw) of PEG and PEOz were both 2 kDa. DOX was purchased from Dalian Meilun Biology Technology Co., Ltd. (China). BTZ was obtained from Beijing Zhongshuo Pharmaceutical Technology Development Co., Ltd. (China). Other analytical-grade reagents and solvents were used as received. HyPure Molecular Biology Grade Water (HycloneTM) was used to prepare all solutions.

#### Preparation of BP NSs

Black phosphorus NSs were fabricated by simply exfoliating the corresponding bulk BP sample in liquid. Briefly, 20 mg of BP was dispersed in 20 mL of 1-methyl-2-pyrrolidinone (NMP) which was argon-bubbled to decrease oxidation by eliminating dissolved oxygen molecules during exfoliation. The mixture was thereafter sonicated in an ice water bath for 8 h (amplifier: 25%, on/off cycle: 5 s/5 s). The low system temperature was kept by the ice water bath. Afterward, unexfoliated bulk BP was removed by centrifuging the brown dispersion at 2,000 rpm for 10 min, and the supernatant containing BP NSs was carefully collected for further use.

#### DOX Loading Onto BP NS Surface

Doxorubicin (2 mg) was mixed with 2 mL of 1 mg mL−<sup>1</sup> BP NSs solution in water, and the solution pH was adjusted to 8.5 with sodium hydroxide. After being stirred vigorously in dark overnight, the obtained DOX-loaded BP (BP-DOX) NSs were collected by centrifugation and washed by water. The LC (%) of DOX were calculated using the following equation.

$$\text{Drug LC (\%)} = \frac{\text{weight of drug in the nanopparticles}}{\text{weight of non-partitions}} \times 100\%$$

# PDA Coating on BP NS Surface

BP-DOX NSs were dispersed in 2 mL HyPure Molecular Biology Grade Water at 1 mg mL−<sup>1</sup> . Then pH was adjusted to 8.5 by adding sodium hydroxide, and the solution was added 10 µL dopamine hydrochloride (100 mg mL−<sup>1</sup> ) and stirred for 2.5 h in dark at room temperature. Finally, BP-DOX@PDA particles were collected by 10 min of centrifugation at 12,000 rpm and washed by deionized water.

# Conjugation of H2N-PEOz or H2N-PEG Onto BP-DOX@PDA Surface

PDA-coated NSs (2 mg) were first resuspended in 2 mL of HyPure Molecular Biology Grade Water with pH adjusted to 8.5 by an appropriate amount of sodium hydroxide. After 2 mg of H2N-PEOz was added into the BP@PDA suspension, the mixture was vigorously stirred for 3 h in dark at room temperature. Then the H2N-PEOz-modified NPs (BP-DOX@PDA-PEOz) were purified by 10 min of centrifugation at 12,000 rpm and washed by deionized water. With a similar procedure, BP-DOX@PDA-PEG was fabricated by using H2N-PEG instead of H2N-PEOz.

# BTZ Loading Onto PDA-Coated NPs

In brief, 50 mg of BP-DOX@PDA-PEOz NPs were suspended at 1 mg mL−<sup>1</sup> in deionized water with pH adjusted to 8.5 by sodium hydroxide, and 6 mg of BTZ powders were dispersed in 200 µl of DMSO. Under stirring, the latter solution was then dropwise added to the former solution. Afterward, the mixture was stirred overnight and centrifuged with the same process as that described above. After 48 h of lyophilization, the product was referred to as BP-DOX@PDA-PEOz-BTZ.

#### Characterizations of BP NSs

Transmission electron microscopy (TEM) images were acquired by Tecnai G2 F30 transmission electron microscope (FEI, Hillsboro, OR, United States). BP NSs were observed after being dropped onto a copper grid-coated carbon membrane and air-dried. Fourier transform infrared (FTIR) spectra were recorded with Nicolet iS 50 spectrometer (Thermo Scientific, United States). Raman spectra were recorded at room temperature by LabRAM HR800 high-resolution confocal Raman microscope (HORIBA, United States). X-ray photoelectron spectroscopy was performed with Axis HSi X-ray photoelectron spectroscope (Kratos Ltd., United Kingdom) employing Al Kα radiation (150 W, 1486.6 eV photons) as the excitation source. Zeta potential and size were measured by Malvern Mastersizer 2000 particle size analyzer (Zetasizer Nano ZS90, Malvern Instruments Ltd., United Kingdom). All measurements were conducted three times independently and averaged.

#### Photothermal Effects of Different BP NSs

The aqueous solutions (0.1 mg mL−<sup>1</sup> ) of different NSs (BP NSs, BP@PDA, BP@PDA-PEOz) were added into microfuge tubes. With the same process, water was utilized as control. The middle of each solution was irradiated at 808 nm with KS-810F-8000 fixed fiber-coupled continuous semiconductor diode laser (Kai Site Electronic Technology Co., Ltd., Xi'an, China) at the power density of 1.0 W cm−<sup>2</sup> . To evaluate the effects of concentration changes, BP@PDA-PEOz solutions with various concentrations of NSs (10–200 µg mL−<sup>1</sup> ) were tested by recording the temperature changes under the above-mentioned irradiation. Also, BP@PDA-PEOz solution (100 µg mL−<sup>1</sup> ) was tested at various power densities (0.5–2.0 W cm−<sup>2</sup> ) to monitor the temperature changes. Ti450 IR thermal imaging camera (Fluke, United States) was used for temperature recording.

#### pH and Photothermal-Induced Drug Release Profiles

To evaluate the DOX release profile of BP-DOX@PDA-PEOz-BTZ NSs, 5 mg of NSs were resuspended in 1 mL of phosphatebuffered saline (PBS, pH = 5.0, 6.8 or 7.4, containing 0.1% w/v Tween 80). Subsequently, the dispersion was transferred

into a dialysis membrane bag [MWCO = 3500, Sangon Biotech (Shanghai) Co., Ltd., China] that was then incubated in 10 mL of PBS at pH 5.0, 6.8 or 7.4 in an orbital water bath and shaken at 37◦C. At dedicated time points, 0.5 mL of the solution outside was collected to detect the amount of released DOX with UV–vis spectrometer at 490 nm, which was supplemented by 0.5 mL of fresh PBS. Under identical conditions, photothermaltriggered drug release was tested at pH 5.0 with 6 min of 808 nm laser irradiation at the power density of 1.0 W cm−<sup>2</sup> . In vitro BTZ release and photothermal-triggered drug release from BP-DOX@PDA-PEOz-BTZ NSs were detected by LC 1200 HPLC system (Nie et al., 2017) (Agilent Technologies, Santa Clara, CA, United States). Compounds were separated by a reverse-phase C18 column (5 µm, 150 × 4.6 mm; Agilent Technologies, Santa Clara, CA, United States) using a mobile phase comprising deionized water and acetonitrile (20/80 for BTZ, v/v). The flow rate was 1.0 mL/min and the injection volume was 20 µl. BTZ amount was measured by UV–vis spectroscopy at 270 nm. Finally, the accumulative release versus time profiles of BTZ and DOX were plotted.

#### Cell Culture Assays

Breast cancer cell line MCF-7 was chosen to study the endocytic behaviors of the above NPs. The cells were incubated in Dulbecco's modified Eagle's medium (DMEM) (Life Technologies, Carlsbad, CA, United States) containing 100 µg mL−<sup>1</sup> streptomycin, 100 U mL−<sup>1</sup> penicillin and 10% (v/v) fetal bovine serum at 37◦C with 5% CO2.

#### Cellular Uptake of NPs

MCF-7 cells were inoculated into 20 mm glass-bottomed Petri dishes and thereafter incubated for 24 h. BP@PDA-PEG and BP@PDA-PEOz loading 10 µg mL−<sup>1</sup> DOX were added into the wells simultaneously at pH 6.8 or 7.4, followed by 4 h of incubation at 37◦C. The cells were washed three times by PBS, and observed with Fluoview FV-1000 confocal laser scanning microscope (Olympus, Tokyo, Japan) at 488 and 590 nm as the excitation and emission wavelengths, respectively.

#### Cytotoxicity Assay

The cytotoxicities of BP@PDA-PEG and BP@PDA-PEOz were determined by the MTT assay. MCF-7 cells were seeded into a 96-well plate at the density of 5 × 10<sup>3</sup> and thereafter incubated for 24 h. Afterward, the medium was replaced by 100 µL of fresh medium containing different concentrations of BP@PDA-PEG or BP@PDA-PEOz (10, 25, 100 µg mL−<sup>1</sup> ), and the cells were further incubated for 48 h. Then the medium was replaced by MTT solution in DMEM (5 mg mL−<sup>1</sup> , 100 µL), followed by another 4 h of incubation. The supernatant was removed from each well into which 100 µL of DMSO was then added to dissolve the formed formazan crystals. The absorption of each well was detected at 570 nm by Model 680 microplate reader (Bio-Rad, United Kingdom).

# In vitro PTT Study

MCF-7 cells were inoculated into a 96-well plate and incubated for 24 h. After the medium in each well was refreshed, the cells were incubated with various concentrations (10, 25, 50 µg mL−<sup>1</sup> ) of BP@PDA-PEG and BP@PDA-PEOz at 37◦C for 4 h, irradiated for 10 min by 808 nm laser (1.0 W cm−<sup>2</sup> ) and incubated again for 12 h. The cell viability was assessed by the MTT assay.

# In vitro Cell Viabilities at Different pH Values

MCF-7 cells were incubated overnight after being inoculated into 96-well plates. Subsequently, they were incubated by BP-DOX@PDA-PEG/BP-DOX@PDA-PEOz/BP-DOX@PDA-PEOz-BTZ with equivalent DOX concentrations at pH 6.8 or 7.4. The MTT solution was added after 48 h, and the cell viability was also detected with the microplate reader at 570 nm. The viability of untreated cells was set at 100%.

#### In vitro Combined Antitumor Therapy

MCF-7 cells were inoculated into a 96-well plate at 1 × 10<sup>4</sup> /well, and incubated overnight. The adherent cells were treated with DOX, BTZ, DOX + BTZ (1:1), BP-DOX@PDA-PEG, BP-DOX@PDA-PEOz, BP-DOX@PDA-PEOz-BTZ, BP-DOX@PDA-PEG + NIR, BP-DOX@PDA-PEOz + NIR, and BP-DOX@PDA-PEOz-BTZ + NIR (808 nm laser irradiation, 1.0 W cm−<sup>2</sup> ) at 0.25, 1, 2.5, and 5 µg mL−<sup>1</sup> equivalent DOX concentrations for 24 and 48 h. After treatment with the NSs for 24 or 48 h in the presence or absence of NIR laser irradiation, the cell viability was tested by the MTT assay. The optical density of each well at 570 nm was measured by the microplate reader. The viability of untreated cells was set at 100%, and the absorbance of the control group was zero.

# In vivo IR Thermal Imaging

All animal experiments have been approved by the Administrative Committee on Animal Research in Sun Yatsen University, and in vivo experiments were conducted according to corresponding guidelines. Female severe combined immunodeficient (SCID) mice aged 5–6 weeks old were provided by the Guangdong Medical Laboratory Animal Center and given free access to water and food. Each mouse was subcutaneously injected with 100 µL PBS suspension of MCF-7 cells (∼2 × 10<sup>6</sup> ) on the dorsal side to induce tumors. Every 2 days, the length and width of tumors were measured with a digital vernier caliper to estimate the volumes using the formula: 0.5 × (length) × (width)<sup>2</sup> . SCID mice bearing MCF-7 tumors were employed as the animal model. IR thermal imaging was carried out when the tumor volumes reached approximately 180 mm<sup>3</sup> . The mice were thereafter intravenously injected with PBS, BP@PDA-PEG, BP@PDA-PEOz or BP@PDA-PEOz-BTZ, and the BP dose is 5 mg/kg in 100 µl PBS. Twenty-four hours later, the tumor sites were irradiated for 5 min by 808 nm laser (1.5 W cm−<sup>2</sup> ). IR thermographic maps and temperature changes were recorded with the IR thermal imaging camera.

#### Statistical Analysis

Unless otherwise stated, all experiments were performed at least in triplicate, and the data were represented as

mean ± standard deviation. Statistical analysis was carried out by using SPSS 22.0 software for one-way analysis of variance and subsequent Bonferroni test. <sup>∗</sup>P < 0.05 and ∗∗p < 0.01 were considered statistically significant and extremely statistically significant, respectively.

# RESULTS AND DISCUSSION

#### Morphology and Characterizations

The entire synthesis procedure of the BP-based drug delivery platform, including DOX loading, PDA coating, PEOz conjugation and BTZ loading, is shown in **Figure 1**. According to the modified liquid exfoliation technique reported by our group previously (Tao et al., 2017), BP NSs were prepared from bulk BP in NMP. Then DOX was absorbed onto the corrugated surface of BP NSs by non-covalent bonding such as Van der Waals force and electrostatic attraction. The DOX loading capacity of BP can reach over 300% in a weakly alkaline condition (Gao et al., 2018). In a weakly alkaline solution, dopamine monomer underwent oxidative self-polymerization into PDA, ultimately adhering to the surface of NPs (Lee et al., 2007; Postma et al., 2009; Hong et al., 2012). PDA coating enhanced the system stability in physiological medium, probably as a universal bond between NSs

synergistic therapy of cancer.

spectrum and (D) N1s spectrum.

and ligand by being reactive with thiol and amine groups. The terminal amine group of NH2-PEOz conjugated to NSs coated with PDA through a simple Michael addition reaction. The oxidative self-polymerization mechanism of dopamine and the conjugation mechanism between H2N-PEOz and PDA coating were shown as **Supplementary Figure S1A**. As a long-chain molecule, PEOz was introduced into the system as a substitute for PEG to maintain long-term circulation, to attenuate the ABC phenomenon and to allow pH-responsive drug release. In alkaline solutions, the boronic acid active site in BTZ reacted with catechol in PDA, which may suppress the activity of BTZ and thus decrease non-specific cellular drug endocytosis. Additionally, BTZ was released under acidic conditions, possibly also facilitating drug release at the tumor site and selectively augmenting antitumor activity (**Supplementary Figure S1B**) (Su et al., 2011). In short, DOX-loaded BP@PDA-PEOz-BTZ may be applicable to synergistic chemotherapy/PTT by prolonging in vivo circulation as well as elevating cellular uptake efficiency, pH responsiveness and dual drug loading capacity.

TEM images (**Figures 2A–D**) exhibit that BP NSs have sheetlike morphology, and bare BP and modified BP NSs have the lateral sizes of approximately 200–250 nm, being consistent with the results of dynamic light scattering analysis. After PDA and PEOz coating, the NS surface became rough and slightly thickened.

**Figure 3A** presents the FTIR spectra of BP NSs, BP@PDA, BP@PDA-PEG, and BP@PDA-PEOz. The adsorption peak at ∼1,625 cm−<sup>1</sup> represents P=O stretching vibration (Shen et al., 2015). After PDA coating, a broad and intense band between 3,150 and 3,600 cm−<sup>1</sup> appears, corresponding to N-H/O-H stretching vibration. The peak at ∼1,500 cm−<sup>1</sup> can be assigned


PDI, polydispersity index; ZP, zeta potential, n = 3.

to the bending vibrations of benzene ring and N-H in PDA. In the spectrum of BP@PDA-PEG, the peak at about 2,900 cm−<sup>1</sup> represents C-H stretching vibration, suggesting that PEG had been successfully modified. The peak at 1,640 cm−<sup>1</sup> is related to the C=O stretching vibration of imide bond in PEOz (Qiu et al., 2013). Moreover, the spectrum of BP@PDA-PEOz shows a peak at ∼2,900 cm−<sup>1</sup> corresponding to C-H stretching vibration (Yu et al., 2014), indicating successful modification of PEOz.

The structures of PDA- and PEOz-modified BP NSs were studied by Raman spectroscopy (**Figure 3B**). In the spectrum of bare BP, there are three obvious peaks at ∼360.7, 437.5, and 466.1 cm−<sup>1</sup> which correspond to the A<sup>1</sup> <sup>g</sup>, B2g, and A 2 <sup>g</sup> modes of BP, respectively. The peaks of modified NSs (BP@PDA, BP@PDA-PEG, BP@PDA-PEOz, BP@PDA-PEOz-BTZ) shift toward lower wavenumber slightly, which can be ascribed to the mild ultrastructural changes after PDA coating and further modification, demonstrating successful modification of PDA and PEOz. The surface modification of NSs loading BTZ was confirmed by X-ray photoelectron spectroscopy (**Figures 3C,D** and **Supplementary Figure S2**). As evidenced by the intensity increase of nitrogen peak (N1s) at 399.49 eV (**Figure 3D**), both PDA coating and PEOz/BTZ loading were successful. The P2p peak (129.6 eV) intensities of bare BP, BP@PDA, BP@PDA-PEG, and BP@PDA-PEOz gradually drop (**Supplementary Figure S2A**) because of the coverage of P element. Collectively, corresponding compounds had indeed been successfully modified (Wang et al., 2015).

The results of dynamic light scattering are listed in **Table 1**. The hydrodynamic sizes increased slightly owing to layer-bylayer modification with PDA, PEG or PEOz, being in accordance with the TEM results. Additionally, the appropriate size and narrow size distribution may be beneficial to NSs accumulation in tumors through the EPR effect (Torchilin, 2011). Besides, the zeta potential of bare BP NSs was −18.9 mV, while that after surface modification with PDA became −16.6 mV, probably because phenolic hydroxyl groups on the PDA layer were deprotonated at neutral pH (Cheng et al., 2017c; Nie et al., 2017). After NH2-PEOz and NH2-PEG modification, the zeta potentials of BP@PDA-PEOz and BP@PDA-PEG in deionized water were measured to be −10.5 mV and −13.2 mV, respectively.

#### Effects of pH on Size and Zeta Potential

To further evaluate the influence of pH, drug-free and DOXloaded BP@PDA-PEG/BP@PDA-PEOz NSs (1 mg mL−<sup>1</sup> ) were exposed in 10 mM PBS at pH 5.0, 6.8 or 7.4 and sonicated for 30 min at 37◦C before their sizes and zeta potentials were measured. As shown in **Figure 4A**, the sizes of all NSs barely change with reducing pH. BP@PDA-PEOz became slightly negatively charged under physiological conditions (pH 7.4), with the zeta potential of −10.2 mV (**Figure 4B**). As pH decreased from 7.4 to 5.0, the surface charge was reversed from negative to positive, and the zeta potential rose to 2.4 mV at pH 6.8 and 5.2 mV at pH 5.0. The charge reversal can be attributed to ionization of the amide groups from PEOz in the outer layer, inducing partial charge neutralization. Taken together, the surface charge of BP@PDA-PEOz and BP-DOX@PDA-PEOz are positive while the pH values decreased from 7.4 to 6.8 and 5.0.

#### In vitro Photothermal Effects

To clarify the photothermal performance of the prepared co-delivery platform, the temperature variations were tested under 808 nm laser irradiation for 10 min. As displayed in **Figure 5A**, the temperatures of bare BP NSs, BP@PDA and BP@PDA-PEOz solutions (0.1 mg mL−<sup>1</sup> ) all soar compared with that of distilled water under identical irradiation conditions. The photothermal efficiency of BP@PDA (1T = 24.1◦C) exceeded that of bare BP (1T = 18.1◦C), and the temperature of BP@PDA-PEOz was elevated by 22.9◦C. PDA coating may be responsible for the augmented photothermal response of BP@PDA, accompanied by considerable photothermal

5.0 (n = 3).

various power intensities. (D) Heating of a suspension of the BP@PDA-PEOz in water for five laser on/off cycles with an 808 nm NIR laser at power density

conversion efficiency. Furthermore, the photothermal properties of PDA-coated BP NSs were both concentration- and laser power-dependent (**Figures 5B,C**). After five cycles of NIR laser irradiation, the temperature no longer changed evidently (**Figure 5D**), so the sample was highly photostable. Moreover, the photostability of BP@PDA-PEOz NSs surpassed that bare BP NSs, without obviously losing photothermal conversion efficiency in 1 week (**Supplementary Figure S3**). Hence, PDA coating boosted the photothermal performance of BP NSs, and rendered them suitable for PTT due to high photostability and photothermal conversion efficiency.

# In vitro pH- and Photo-Responsive Drug Release Profiles

The sustained and controlled DOX release profiles of BP-DOX@PDA-PEOz-BTZ were tested at pH 7.4 for simulating normal physiological microenvironment, pH 6.8 for simulating tumor extracellular microenvironment and pH 5.0 for simulating the acidic microenvironment of tumor endosome/lysosome, in the presence or absence of NIR laser irradiation. At pH 5.0, nearly 30% of DOX was released from BP-DOX@PDA-PEOz-BTZ within 48 h, whereas only 11% of DOX was released at pH 7.4 (**Figure 6A**), which may be ascribed to the pH sensitivity of PEOz coating after tertiary amide groups along the PEOz chain were ionized at a pH value lower than its pKa (Nie et al., 2017; Wang et al., 2017). The positive charges on the nitrogen atoms of PEOz main chains may result in electrostatic repulsion, which loosened the outer shell in the slightly acidic tumor cell microenvironment (Wang et al., 2017) and accelerated the release of inner hydrophobic anticancer drugs into tumor tissues while reducing that into the normal blood circulation. As a result, the anticancer effect was increased, and the side effects of common anticancer drugs were relieved (Gao et al., 2015a; Zhao et al., 2015; Wang et al., 2017).

of 1.0 W/cm−<sup>2</sup>

.

Moreover, the photo-responsive drug release behaviors were studied. After 808 nm laser irradiation (1.0 W cm−<sup>2</sup> , 6 min for each pulse), the temperature of BP-DOX@PDA-PEOz-BTZ increased gradually at pH 5.0, which significantly raised the cumulative DOX release amount, reaching above 40% after irradiation four times. Presumably, the PDA layer decomposed and released the loaded drug after NIR laser irradiation. In the meantime, BP decomposed gradually due to NIR exposure, further inducing drug release (Zeng et al., 2018).

Furthermore, we investigated the drug release behaviors of BTZ from BP-DOX@PDA-PEOz-BTZ NSs at different pH values with or without IR irradiation (**Figure 6B**). Merely approximately 25% of BTZ was released at pH 7.4 after 24 h. At pH 6.8 and 5.0, BTZ release was significantly accelerated. After NIR laser irradiation for 6 min, about 95% of BTZ was released from DOXloaded BP@PDA-PEOz-BTZ within 48 h at pH 5.0. In other words, the pH sensitivity of catechol-BTZ bond contributed to BTZ accumulation at tumor sites, so the treatment outcomes were improved. The drug BTZ was loaded onto the surface of nanoplatform through the reversible covalent bond between catechol and phenylboronic acid. However, the drug DOX was absorbed onto the corrugated surface of BP NSs by non-covalent bond. And the DOX-loaded BP NSs was covered by the PDA. That is, the drug DOX is at the interior of the nanoplatform and BTZ is at exterior. Therefore, the release rate for DOX is in general much lower compared to the release rate of BTZ. Similar results were reported by our previous research (Nie et al., 2017). This release behavior of BTZ highlights the pH sensitivity of the catechol-BTZ bond contributing to the accumulation of BTZ at the tumor sites. The release pattern was very important for better tumor killing effect, as it reduces the drugs leakage during the circulation in blood and increases the drugs enrichment in tumor sites or endosomes. Overall, the pH-sensitive drug release triggered by NIR laser irradiation markedly enhanced the antitumor efficacy and minimized the side effect.

#### Cellular Uptake of NSs

The uptake of BP-DOX@PDA-PEG and BP-DOX@PDA-PEOz NSs by MCF-7 cells in the weakly acidic tumor microenvironment (pH 7.4 and 6.8) was observed by confocal laser scanning microscopy. After treatment with different NSs for 4 h, the intracellular fluorescent intensities of BP-DOX@PDA-PEG NSs at pH 6.8 and 7.4 were similar, but the intensity of BP-DOX@PDA-PEOz NSs at pH 6.8 significantly exceeded that at pH 7.4 (**Figure 7**). The results validated the hypothesis that PEOz promoted the cellular uptake of DOX compared with PEGylated copolymer did in the mildly acidic endosomal/lysosomal and tumor extracellular environment, which can be attributed to the charge reversal of PEOz after tertiary amide groups along the PEOz chain were ionized (Cheng et al., 2017b; Wang et al., 2017; Zeng et al., 2017).

#### Cell Viability

fphar-10-00270 March 16, 2019 Time: 17:5 # 10

The in vitro cytotoxicities of drug-free BP@PDA-PEG and BP@PDA-PEOz as well as DOX-loaded BP@PDA-PEG, BP@PDA-PEOz and BP@PDA-PEOz-BTZ NSs were evaluated by the MTT assay. Drug-free BP@PDA-PEG and BP@PDA-PEOz NSs were also tested to eliminate the potential toxic characteristics of drug delivery capsule. Given that all drug-free BP-based NSs displayed negligible cytotoxicities against MCF-7 cells (**Supplementary Figure S4**), they were highly biocompatible.

The photothermal cytotoxicities of different NSs were also evaluated by the MTT assay. Cell growth was barely affected by individual NIR laser irradiation, whereas BP@PDA-PEG and BP@PDA-PEOz NSs exerted concentration-dependent photothermal effects (**Figure 8A**). Moreover, over 80% of MCF-7 cells were killed in the presence of 50 µg mL−<sup>1</sup> BP@PDA-PEOz under 808 nm laser irradiation. Taken together, BP@PDA-PEOz nanocapsule may be an effective PTT agent with desirable biocompatibility.

Furthermore, the BP-DOX@PDA system exhibited pH-dependent cytotoxicity after PEOz modification (**Figure 8B**). BP-DOX@PDA-PEOz was significantly more toxic at pH 6.8 than at pH 7.4, but BP-DOX@PDA-PEG had almost the same inhibitory effects at pH 7.4 and 6.8, potentially allowing selective killing of cancer cells that were more acidic than normal cells/tissues in vivo. Collectively, PEOz modification was conducive to cellular uptake and pH-sensitive drug release in the mildly acidic tumor microenvironment, thereby promoting tumor inhibition and alleviating side effects during chemotherapy.

In addition, MCF-7 cells were treated by DOX, BTZ, DOX + BTZ (1:1) and drug-loaded NSs with the DOX concentrations of 0.25, 1, 2.5, and 5 µg/mL for 24 or 48 h (**Figures 8C,D**). First, the cytotoxicities of free DOX, BTZ, DTX + BTZ (1:1) and drug-loaded NSs were time- and dose-dependent. Second, compared to individually administered DOX or BTZ, directly co-administering DTX + BTZ was more cytotoxic, as suggested by the raised inhibition rate. Furthermore, the survival rate of cells treated with DOXloaded BP@PDA-PEOz NSs was apparently lower than that of the DOX-loaded BP@PDA-PEG NSs group after incubation for 24 or 48 h. Therefore, PEOz was more conducive to long-term drug circulation in vivo than PEG, extending the half-lives of drugs and enriching pH-sensitive drugs at tumor sites by recognizing the acidic tumor microenvironment. Most importantly, the group treated by DOX-loaded BP@PDA-PEOz-BTZ in combination with 808 nm laser irradiation (1.0 W cm−<sup>2</sup> ) had the lowest survival rate after 48 h of incubation, demonstrating that chemotherapy plus PTT exerted the strongest cytotoxic effects. In short, the antitumor effects of drug were effectively boosted with this pH-sensitive release pattern triggered by NIR laser irradiation, accompanied by minimal side effects.

#### In vivo IR Thermal Images

The photothermal efficacy of this BP-based nanoplatform was further studied by acquiring IR thermal images (**Figure 9A**). 24 h after intravenous injection of BP@PDA-PEG, BP@PDA-PEOz, and BP@PDA-PEOz-BTZ, the tumor sites were irradiated for

IR camera under 808 nm laser (1.5 W cm−<sup>2</sup> ).

5 min by 808 nm laser (1.5 W cm−<sup>2</sup> ), and the temperatures in tumor site were measured every 15 s. The tumor surface temperatures significantly increased after NIR irradiation, which can be attributed to the remarkable photothermal effects of PDA coating and BP. The tumor temperatures of BP@PDA-PEOz group and BP@PDA-PEOz-BTZ group rapidly increased to 51.4 and 50.1◦C, respectively, which were sufficiently high for effective ablation. The in vivo local tumor temperatures of BP@PDA-PEOz and BP@PDA-PEOz-BTZ groups changed more obviously than that of the BP@PDA-PEG group did, because PEOz underwent charge reversal from negative to positive upon tertiary amide group ionization along the PEOz chain in tumor tissue with a low pH, ultimately benefiting the uptake by cancer cells. On the contrary, the temperature of PBS-treated tumor did not rise evidently after irradiation under identical conditions, so cancer cells remained intact. Based on long-term circulation in vivo, the PEOz-modified, BP-based NS drug delivery platform was photothermally active, being capable of pH-triggered targeting of tumor tissues. The quantified tumor temperature variations are presented in **Figure 9B**.

### CONCLUSION

In summary, we have successfully developed an effective pHresponsive and dual drug co-delivery nanoplatform (DOXloaded BP@PDA-PEOz-BTZ) for combined chemotherapy and PTT. Not only did we show PDA coating can enhance both biostability and photothermal activity of the BP NS, we also demonstrated that PEOz conjugation can improve the targeted, long circulation in vivo as well as pH- and photo-responsive

#### REFERENCES


drug release, indicating that PEOz is an excellent substitute for PEG. Given the high drug encapsulation efficiency, the strong cellular uptake and cytotoxicity, together with the photo-responsive, rapid drug release triggered by low pH, our versatile PDA- and PEOz-modified, BP-based dual drug co-delivery nanoplatform has great potentials for synergistic cancer treatment.

#### AUTHOR CONTRIBUTIONS

ZP designed the research project. NG, CX, and LF had full controlled the experiments, data analysis, and preparation of article. HW, XZ, and LM were involved in planning the analysis and drafting the article. The final draft article was approved by all the authors.

# FUNDING

This work was supported by National Natural Science Foundation of China (61671308 and 81701819) and the Science and Technology Innovation Commission of Shenzhen (JCYJ20170817094728456 and JCYJ20170302153341980).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2019.00270/full#supplementary-material



and cisplatin to osteosarcoma. Adv. Sci. 5, 1700821. doi: 10.1002/advs. 201700821


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Gao, Xing, Wang, Feng, Zeng, Mei and Peng. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*,* 

# Targeted Delivery of Chlorin e6 *via* Redox Sensitive Diselenide-Containing Micelles for Improved Photodynamic Therapy in Cluster of Differentiation 44-Overexpressing Breast Cancer

*Chan Feng1 , Donglei Zhu1 , Lv Chen1 , Yonglin Lu1 , Jie Liu1 , Na Yoon Kim2 , Shujing Liang1 Xia Zhang1 , Yun Lin1 , Yabin Ma3 \* and Chunyan Dong1 \**

#### *Edited by:*

*Wei Tao, Harvard Medical School, United States*

#### *Reviewed by:*

*Jiang Ouyang, Central South University, China Yuling Xiao, Wuhan University, China Tianjiao Ji, Boston Children's Hospital and Harvard Medical School, United States*

#### *\*Correspondence:*

*Yabin Ma baishecao@163.com Chunyan Dong cy\_dong@tongji.edu.cn*

#### *Specialty section:*

*This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology*

*Received: 20 February 2019 Accepted: 25 March 2019 Published: 16 April 2019*

#### *Citation:*

*Feng C, Zhu D, Chen L, Lu Y, Liu J, Kim NY, Liang S, Zhang X, Lin Y, Ma Y and Dong C (2019) Targeted Delivery of Chlorin e6 via Redox Sensitive Diselenide-Containing Micelles for Improved Photodynamic Therapy in Cluster of Differentiation 44-Overexpressing Breast Cancer. Front. Pharmacol. 10:369. doi: 10.3389/fphar.2019.00369*

*1 Cancer Center, Shanghai East Hospital, Tongji University, Shanghai, China, 2 Department of Chemical Engineering, Northeastern University, Boston, MA, United States, 3 Pharmacy Department, Shanghai East Hospital, Tongji University, Shanghai, China*

The off-target activation of photosensitizers is one of the most well-known obstacles to effective photodynamic therapy (PDT). The selected activation of photosensitizers in cancer cells is highly desired to overcome this problem. We developed a strategy that enabled diselenide bonds to link hyaluronic acid (HA) and photosensitizer chlorin e6 (Ce6) to assemble the micelles (HA-sese-Ce6 NPs) that can target cancer and achieve a redox responsive release of drugs to enhance the PDT efficiency in breast cancer. The HA was used to form a hydrophilic shell that can target cluster of differentiation 44 (CD44) on the cancer cells. The selenium-containing core is easily dissembled in a redox environment to release Ce6. The triggered release of Ce6 in a redox condition and the positive feedback release by activated Ce6 were observed *in vitro*. In cytotoxicity assays and *in vitro* cellular uptake assays, the increased PDT efficiency and targeted internalization of HA-sese-Ce6 NPs in the cells were verified, compared to a free Ce6 treated group. Similar results were showed in the therapeutic study and *in vivo* fluorescence imaging in an orthotopic mammary fat pad tumor model. In addition, a significant inhibition of metastasis was found after the HA-sese-Ce6 NPs treatment. In general, this study promises an ingenious and easy strategy for improved PDT efficiency.

Keywords: chlorin e6, redox sensitive, diselenide, photodynamic therapy, cluster of differentiation 44, targeted delivery, breast cancer

# INTRODUCTION

Photodynamic therapy (PDT) is a promising noninvasive, localized therapeutic method that has a variety of advantages for cancer treatment, especially for tumors located close to the skin such as breast cancer (Agostinis et al., 2011; Wang et al., 2018). One of the most widely used photosensitizers in PDT is chlorin e6 (Ce6), a second-generation photosensitizer with high efficacy and low dark toxicity (Martynenko et al., 2015; Du et al., 2016). The therapeutic effect of PDT is based on the activation of photosensitizers. When photosensitizers are exposed

**280**

to a certain wavelength of light, they release singlet oxygen ( 1 O2) that can kill cancer cells. However, there are limitations to this therapy. Due to the short half-life (<40 ns) of 1 O2 during PDT, each 1 O2 molecule can only have therapeutic effects in the area of diameter less than 20 nm (Li et al., 2017; Zhang et al., 2018). This makes the precise delivery of photosensitizers highly desirable. Moreover, nonspecific activation of photosensitizers will cause potential cytotoxicity in normal tissue. Thus, selected delivery of photosensitizers to tumor sites is high desired. Beyond the previously mentioned challenges, the vast majority of photosensitizers, including Ce6, have a poor water solubility, which leads to undesired pharmacokinetics (Liu et al., 2017).

In the past few decades, various advanced nanoscale drug delivery systems have been developed to optimize pharmacokinetics by selectively delivering hydrophobic drugs and photosensitizers to achieve better efficacy and less off-target side effects in cancer treatment (Ji et al., 2015, 2017; Ouyang et al., 2018a,b). In addition, smart drug release has been extensively investigated for precise drug release (Ji et al., 2016a,b). By utilizing stimuli-responsive linkages, drug release can be triggered by various specific tumor biological/endogenous stimuli, such as pH (Webb et al., 2011), redox (Go and Jones, 2008), hypoxia (Brown and Wilson, 2004), or enzymes (de la Rica et al., 2012). However, a complicated stimuli-responsive design leads to a larger proportion of drug carrier and limited drug loading capacities. Nowadays, the clinical translation of nanoparticle-based drug delivery has several limitations. One of most controversial problems is the increasing cost of biosafety due to the risk of using synthetic materials in drug carriers. Therefore, designing drug carriers that are simple and made of biocompatible materials is highly desired.

Selenium, one of the essential dietary elements in higher animals, plays an important role in cell growth and functions (Fang et al., 2018). A few studies even reported anticancer activity of selenium (Yu et al., 2015; Baskar et al., 2018). Interestingly, the relatively low electronegativity and large atomic radius give selenium unique chemical properties, such as high reactivity and sensitivity (Xia et al., 2016, 2018b; Sun et al., 2017). Due to their high sensitivity to both oxidation and reduction, diselenide-containing polymers have been gaining attention as attractive drug delivery candidates that can perform controlled drug release in tumor microenvironment with rich redox stimuli (Ma et al., 2010; Gong et al., 2017; Zhou et al., 2017, 2018). Another characteristic of diselenide bonds that is particularly advantageous in PDT drug delivery is their sensitivity to both 1 O2 and 600 nm or higher wavelength light (Xia et al., 2016; Sun et al., 2017). If diselenide bonds in a drug delivery system could be cleaved by 1 O2 and a laser stronger than 600 nm and release drug molecules that produce more 1 O2, the positive feedback on drug release is theoretically possible.

Hyaluronic acid (HA) is a natural anionic hydrophilic polysaccharide in the human body and is especially overexpressed in the tumor matrix (Choi et al., 2011; Agrawal et al., 2018; Huang and Huang, 2018a). In many cancers of epithelial origin, such as breast cancer, cluster of differentiation 44 (CD44) is a main up-regulated HA receptor on the cancer cell surface (Huang and Huang, 2018b; Liao et al., 2018; Maudens et al., 2018). HA can regulate cancer cell proliferation and migration *via* CD44 (Safdar et al., 2017; Tian et al., 2018). Thus, HA has gained attention as a promising cancer targeting ligand for anti-cancer drug delivery. In addition, HA has high water solubility, desirable biocompatibility, biodegradability, and nonimmunogenicity and can be easily functionalized (Choi et al., 2011; Xia et al., 2018a). Many HA-drug conjugates and HA-based micelles have been developed.

Here, we developed a minimalist photosensitizer delivery system, incorporating diselenide bonds into a self-assembled micelle. HA was chosen as the hydrophilic shell and grafted onto a hydrophobic core, Ce6, *via* diselenide bonds. The amphiphilic hyaluronic acid-chlorin e6 (HA-SeSe-Ce6) polymers were synthesized and formed micelles by self-assembly (**Schemes 1, 2**). In our *in vitro* study, the redox-responsive and positive feedback modulated release of Ce6 was observed

when the diselenide bonds were cleaved in redox condition and in the presence of 1 O2. The PDT efficacy was investigated in a breast cancer cell line and orthotopic mammary fat pad tumor model.

# MATERIALS AND METHODS

#### Materials

Chlorin e6 was purchased from J&K Scientific, Ltd. 2-(N-morpholino), Selenocystamine dihydrochloride (C4H12N2Se2·2HCl), N-hydroxysuccinimide (NHS, 98%), and 1-ethyl-3-(3(dimethylamino)propyl) carbodiimide hydrochloride (EDC·HCl, 98.5%) were obtained from Sigma-Aldrich (Shanghai). Hyaluronic acid (mW ≈ 20 k) was purchased from Ruixi biotechnology Co., Ltd. Singlet oxygen sensor green (SOSG) was obtained from Life Technologies. Hydrogen peroxide solution (H2O2, 30 wt. % in H2O), glutathione (GSH, 98%), dimethyl sulfoxide (DMSO, 99.9%), and NaOH (AR, 96%) were obtained from Aladdin Chemistry Co., Ltd. Dulbecco's modified eagle medium (DMEM), penicillin-streptomycin, fetal bovine serum (FBS), and trypsin were purchased from Gibco Invitrogen Corp. CCK-8 and 4, 6-diamidino-2-phenylindole (DAPI) were obtained from Beyotime Institute of Biotechnology. Paraformaldehyde (4%) was obtained from DingGuo Biotech. Co., Ltd. One step TUNEL apoptosis assay kit was purchased in Beyotime biotechnology Co., Ltd.

#### Methods

#### Synthesis of HA-sese-Ce6 Micelles

In a small glass bottle, 20-mg HA, 10-mg C4H12N2Se2·2HCl, 5.75-mg EDC·HCl, 3.45-mg NHS, and 10-ml PBS (pH 7.4) were mixed and stirred at room temperature for 2 h. Ce6 was dissolved in 20-mg/ml DMSO solution. Then, 0.9 ml of Ce6 liquid, 5.75 mg of EDC·HCl, and 3.45 mg of NHS were added dropwise to the mixed liquid and stirred for 4 h. The solution was transferred to a 100 kDa mol. Cutoff centrifugal ultrafiltration tube (Pall Corporation, USA) was centrifuged at 4500 rpm at 25°C for 20 min. It was washed three times with deionized water to remove PBS, unassembled HA, C4H12N2Se2·2HCl, and Ce6.

#### Characterization of HA-sese-Ce6 Micelles

Fluorescence spectra were performed on a Hitachi F2500 luminescence spectrometer. Ultraviolet-visible (UV) spectra were recorded on a UV spectrophotometer (Varian). The size distribution of the micelles was characterized by Nano-ZS 90 Nanosizer (Malvern Instruments, Worcestershire, UK) *via* dynamic light-scattering analysis. The morphology of micelle was studied *via* high-resolution transmission electron microscopy (HRTEM, FEI Tccnai G2 F20 S-Twin). About 1% uranyl acetate was used for negative staining.

#### Chlorin e6 Release Behavior of HA-sese-Ce6 Micelles *in vitro*

HA-sese-Ce6 (containing 5 mg of Ce6) was dissolved in 5 ml of PBS buffer (10 mM, pH 7.4). The solution was transferred to five dialysis tubes (1 ml each) that were immersed in 200 ml of PBS buffer, 200 ml PBS buffer with 10 Mm GSH, 200 ml PBS buffer with 100 Mm GSH, 200 ml PBS buffer with 1 Mm H2O2, and 200 ml PBS buffer with 10 Mm H2O2 at 37°C, stirred at 200 r min−1. To evaluate the positive release of Ce6 from HA-sese-Ce6 micelles, the 1 ml of HA-sese-Ce6 (containing 1 mg Ce6) in PBS buffer was exposed to a 650-nm laser that has the light intensity of 20 mW/cm2 for 10 min. Then, the solution was transferred to a dialysis tube and immersed in 200 ml of PBS buffer. At predetermined time points, 1 ml of the buffer solution outside the dialysis tube was taken out, and the Ce6 release was measured on a UV spectrophotometer at 404 nm. Ce6 release curves of six groups were drawn.

#### Measurement of Singlet Oxygen (1 O2) Generation

One milliliter of the buffer solution outside the dialysis tube was taken out from both 200 ml of PBS buffer with 10 Mm GSH and 200 ml of PBS buffer with 1 Mm H2O2, which were previously described in method 3. To measure the 1 O2 generated from activated Ce6, 1 O2 detecting reagent SOSG was added to the solution. The final concentration of SOSG in the solution was 1 μM. The fluorescence emission spectra were recorded from 490 to 700 nm, and the excitation wavelength was 488 nm.

#### *In vitro* Cellular Uptake Assay

The mouse breast cancer cell line 4T1 cells were purchased from ATCC. 4T1 cells were cultured in DMEM containing 10% FBS. The cell cultures were maintained in 5% carbon dioxide at 37°C. To investigate the targeted uptake of HA-sese-Ce6 micelles by 4T1 cells, the cellular uptake was analyzed by confocal laser scanning microscopy (CLSM) and flow cytometry (FCM). 4T1 cells (1 × 105 cells/well) were cultivated in confocal dishes for 24 h. Then, the cells were treated with serum-free DMEM containing Ce6 and HA-sese-Ce6 micelles (Ce6 concentration of 2 μM). After 4 h, the medium was removed. The cells were washed with PBS and fixed with paraformaldehyde (4%) for 10 min. Afterward, cells were stained with DAPI for 5 min and washed three times. The dishes were measured by confocal laser scanning microscopy (Leica TCS SP5 II, Germany). 4T1 cells (1 × 105 cells/well) were seeded on six wells and cultivated for 24 h. The medium in the dishes was removed, and Ce6 and HA-sese-Ce6 micelles in serum-free DMEM medium were added. Then cells were harvested twice, 4 and 12 h after incubation, and resuspended in 400 μl of PBS and were analyzed through flow cytometry.

#### *In vitro* Phototoxicity Test

In 96-well-plates, 1 × 104 cells/well 4T1 cells were planted and treated with different concentrations (0.25, 0.5, 1, and 2 μM) of Ce6 and HA-sese-Ce6 micelles in serum-free DMEM. In the control group, the same volume of serum-free DMEM was added. After the 24-h incubation period, the medium in the plates was removed, and the fresh medium was added. Half of the cells were exposed to 650 nm laser (20 mW/cm2 ) for 5 min, while the other half was cultured in the dark. After the 24-h incubation period, 10 μl of cck8 reagent was added to each well to measure cell proliferation. Three hours later, the absorbance at 450 nm was measured by the plate reader.

#### *In vivo* Fluorescence Imaging

This study was carried out in accordance with the recommendations of Tongji University Animal Ethics guidelines. The protocol was approved by Tongji University Animal Ethics Committee.

The 5 × 105 4T1 cells were injected subcutaneously into 5-week-old female BALB/c mice. After 2 weeks, tumor tissues were excised and cut into 1 × 1 mm2 tissue blocks to plant in the left mammary fat pad of 5-week-old female BALB/c mice. When the tumor size was large enough, Ce6 and HA-sese-Ce6 micelles were injected into the tail vein of the mice bearing a 4T1 tumor. Fluorescence imaging was performed by a Night OWL LB 983 *in vivo* imaging system 1 and 2 h after the injection.

#### *In vivo* Photodynamic Therapy

Tumor tissues were planted in the mammary fat pad of 5-weekold female BACB/c mice on day 0. On day 13, the volume of the tumors reached 500 mm3 , and the mice bearing a tumor in the mammary fat pad were randomly assigned to three groups (*N* = 6). These mice were treated with Ce6 and HA-sese-Ce6 micelles (Ce6 dose of 2.5 mg/kg) in PBS every 2 days over the course of 10 days. Two hours after injection, tumors were exposed to 650 nm laser at the intensity of 20 mW/cm2 for 30 min. The body weight and tumor size were recorded before each injection, and tumor volumes were calculated by the following formula:

$$\text{Turnor volume} = \frac{\text{Length} \times \text{Width}^2}{2}$$

On day 30, one of mice was randomly picked from each group, and the major organs (heart, liver, spleen, lungs, and kidneys) and tumors were harvested. The collected samples were fixed in 4% paraformaldehyde overnight, dehydrated in graded ethanol solution, and embedded in paraffin. Paraffin sections were prepared to perform the H&E and TUNEL staining. The percent survival of mice (*N* = 5) was recorded until day 40. Tumors weights were recorded upon the death of mice.

#### Statistical Analysis

All experiments were performed in three independent experiments. One-way single factorial analysis of variance (ANOVA) was used for determining the statistical significance of the data, which were expressed as *p* \* **≤** 0.05, \*\* **≤** 0.01, \*\*\* **≤** 0.001.

#### RESULTS AND DISCUSSION

#### Synthesis of HA-sese-Ce6 Micelles

The chemical structure and synthetic process of HA-sese-NH2 and HA-sese-Ce6 were shown in **Scheme 2**. HA-sese-NH2 was prepared by conjugating HA (mW ≈ 20 k) to selenocystamine dihydrochloride (C4H12N2Se2·2HCl) (molar mass proportion 1:1) *via* amino-carboxyl reaction. Subsequently, Ce6 was conjugated to the terminal amino group of HA-sese-NH2 *via* amino-carboxyl reaction. Due to its amphiphilic character, HA-sese-Ce6 can form micelles by self-assembly in water. As shown in **Scheme 1,** HA acts as a hydrophilic coat, and Ce6 acts as a hydrophobic core of the micelle.

The characteristics of HA-sese-Ce6 micelles were analyzed. The hydrodynamic diameter of the micelles was measured *via* DLS. The diameter was 250 nm, and the size had a narrow distribution. The TEM pictures showed the spherical shapes of the micelles. As the samples were dried during the TEM analysis, the size measured in the TEM analysis was smaller than that from the DLS analysis.

#### *In vitro* Redox Sensitivity of HA-sese-Ce6 Micelles

HA-sese-Ce6 micelles were designed to be redox sensitive due to their diselenide component. The bond between diselenide breaks when it is exposed to the redox environment. To

demonstrate the redox sensitivity of HA-sese-Ce6 micelles, they were treated with different concentrations of GSH and H2O2. The size changes of these micelles were recorded at predetermined time points. As shown in **Figure 1A**, the size of micelles tends to be larger when they were treated with higher concentration of GSH. A slight increase in size was observed between the micelles treated with 1 mM GSH and the micelles treated with 10 mM GSH. However, the micelles treated with 100 mM GSH showed a dramatic increase in size.

The size of the micelles treated with different concentrations of H2O2 (0.1, 1, 10, and 100 mM) were analyzed. The size of micelles treated with 0.1, 1, and 10 mM of H2O2 , increased and then decreased over time (**Figure 1B**). The micelles treated with 100 mM GSH had the smallest diameter, and the size consistently decreased over time. Our results could be explained by the effect the diselenide bond cleavages on the size of the micelles. Fewer cleavages of diselenide bonds would cause increase in size, whereas more cleavages of diselenide bonds would cause decrease in size, and redox sensitivity of HA-sese-Ce6 micelles can be demonstrated by the size changes in the reducing and oxidizing conditions.

#### Chlorin e6 Release Behavior of HA-sese-Ce6 Micelles *in vitro*

To further evaluate the redox sensitivity of HA-sese-Ce6 micelles, the Ce6 release behavior was measured in different concentrations of GSH and H2O2 at 37°C. The results of GSH treated groups were shown in **Figure 1C**. Overall, the cumulative Ce6 release increased as the concentration of GSH increased. In the control group, which was treated with PBS, the cumulative release of Ce6 reached a plateau at 30% in 20 h. This could be explained by the physical adsorption effect of micelles on free Ce6. In the group treated with 10 mM GSH, the cumulative release of Ce6 reached a higher plateau at 45% in 11 h, while the group treated with 10 mM GSH reached the higher plateau 65% in 4 h.

As shown in **Figure 1D**, the micelle groups treated with H2O2 generally showed more Ce6 release than those treated with GSH. In the group treated with 100 mM H2O2, the cumulative release of Ce6 reached a highest plateau at 80%, which is higher than the highest plateau (65%) in the GSH treated group.

As shown in **Figure 2**, increased singlet oxygen generation was observed in the micelle groups that were treated with GSH and H2O2. After Ce6 is released from HA-sese-Ce6 micelles due to redox stimuli, Ce6 gets activated and produces more singlet oxygen. This causes a positive feedback on the release of Ce6 because singlet oxygen triggers HA-sese-Ce6 micelles to release more Ce6. The results confirm the high sensitivity of HA-sese-Ce6 micelles to both oxidation and reduction. This suggests that HA-sese-Ce6 micelles would achieve smart drug release in tumor tissues with rich redox stimuli. As shown in **Figure 1E**, the higher drug release was observed when the micelles were treated with a laser. This could be explained by the indirect effect of the singlet oxygen generated from the activated Ce6 and direct effect of the 650 nm laser.

#### Targeted Cellular Uptake and *in vitro* Cytotoxicity

Targeted cellular uptake of HA-sese-Ce6 micelles by cancer cells was investigated by CLSM and FCM. 4T1 cancer cells were treated with free Ce6 and HA-sese-Ce6 micelles for 4 and 8 h, respectively. The CLSM analysis was shown in **Figures 3A,B**. The stronger fluorescence was observed in the cells treated with HA-sese-Ce6 micelles, which indicates markedly higher intracellular uptake of HA-sese-Ce6 micelles when compared to free Ce6. As shown in **Figures 3C,D**, the FCM analysis showed the same results. The number of cells that

internalized Ce6 was higher in the HA-sese-Ce6 micelles treated group than the free Ce6 treated group.

To evaluate the anticancer efficacy of HA-sese-Ce6 micelles, 4T1 cancer cells were seeded in 96-well plants and treated with free Ce6 or HA-sese-Ce6 micelles. After the 24-h incubation, half the cells were exposed to a 650-nm laser (20 mW/cm2 ) for 5 min, and the other half of the cells were kept in the dark as a control. As the data shown in **Figure 4**, the cytotoxicity significantly increased as the Ce6 concentration increases in both groups that had the laser treatment. The cells treated with HA-sese-Ce6 micelles exhibited lower cell viability than free Ce6 treated cells at all concentrations. The targeted redox responsive delivery of HA-sese-Ce6 might be the explanation for lower viability of cancer cells. Both free Ce6 and HA-sese-Ce6 micelles treated cells without the laser exposure exhibited no significant toxicity.

#### *In vivo* HA-sese-Ce6 Micelles Biodistribution in Breast Cancer Bearing Mice

To assess the efficient tumor accumulation of HA-sese-Ce6 *via* targeted delivery, the mice bearing 4T1 tumors in the mammary fat pad were injected with free Ce6 and HA-sese-Ce6 micelles *via* tail veins, respectively. *In vivo* Ce6 fluorescence imaging was performed at 1 and 2 h after intravenous injection. As shown in **Figure 5B**, the accumulation of HA-sese-Ce6 micelles was shown in the liver and cancer cells. This suggested an effective targeted delivery of Ce6 to tumor tissues and the role of the liver in drug clearance. The free Ce6 treated group showed a high liver and kidney accumulation but significant less accumulation of Ce6 in tumors. This indicated a poor drug delivery to tumor tissues and the role of the kidney in drug clearance.

execution (day 40).

#### Anticancer Effect of HA-sese-Ce6 Micelles in Tumor Bearing Animal Model

**Figure 5A** shows the scheme of the photodynamic therapy in a 4T1 orthotopic mammary fat pad tumor growth model in BALB/c female mice. The day tumor blocks were planted in the mice was considered as day 0; after five times of photodynamic therapy (from day 13 to day 21, every 2 days), survival period of mice was recorded until day 40. In **Figure 5C**, the HA-sese-Ce6 micelle treated group exhibited the highest anticancer effect (tumor volume on Day25 was similar to the original tumor volume before the treatment), when compared to the free Ce6 treated group (5 fold original tumor volume) and PBS group (10 fold original tumor volume). The tumor growth inhibition effect is likely due to the HA-based target delivery of Ce6 and diselenide-based responsive Ce6 release. Moreover, the mice treated with

HA-sese-Ce6 micelles showed the longest survival period among all groups (**Figure 5E**). This is consistent with the results of the tumor volume change. Tumor weights of the HA-sese-Ce6 treated group were also lighter than those of the free ce6 treated group and the control group, as shown in **Figure 5F**. In addition, no obvious different in body weight was observed (**Figure 5D**).

To further investigate the effect of HA-sese-Ce6 micelles in promoting apoptosis and inhibiting metastasis, the sections of tumor tissues and other major organs tissues (heart, spleen, kidney, liver, lung) were prepared. TUNEL staining of tumors is shown in **Figure 6A**, and the greatest number of the apoptosis cells (green) was found in the HA-sese-Ce6 micelles group, when compared to the free Ce6 treated and the PBS treated group. In addition, metastasis in major organs was observed by H&E staining of the heart, spleen, kidney, liver, and lung tissues. The decreased metastasis in the liver and lung was found in the HA-sese-Ce6 micelles treated group (**Figure 6B**). These results further confirmed that the HA-sese-Ce6 micelles treatment showed a significantly higher anti-cancer effect due to targeted delivery and smart release of Ce6.

# CONCLUSIONS

In this study, we developed a minimalist photosensitizer delivery system. HA-sese-Ce6 micelles showed targeted, redox sensitive delivery of Ce6 to 4T1 breast cancer cells. The therapeutic effect of this method could be maximized *via* positive feedback because the activated Ce6 generates singlet oxygen molecules, which helps to break more diselenide bonds on the micelles. These characteristics were confirmed in 4T1 mice breast cancer cells and *in vivo* 4T1 tumor bearing mice models. This unique HA-sese-Ce6 micelles exhibited a great anti-cancer effect and metastasis inhibition. We believe that this can be a promising new strategy for improved photosensitizer delivery in breast cancer treatment.

#### ETHICS STATEMENT

This study was carried out in accordance with the recommendations of 'Tongji University Animal Ethics guidelines, name of committee'. The protocol was approved by Tongji University Animal Ethics Committee.

# AUTHOR CONTRIBUTIONS

CF, CD, and YM designed the experiments. CF, DZ, LC, LL, and JL carried out the experiments. NK, SL, XZ, and YL helped to analyze the experimental results. CF wrote the manuscript.

# FUNDING

This work was supported by National Natural Science Foundation of China Regional Project (81860547), National Natural Science Foundation of China Project (81573008), Key Cross-cutting Projects of Central Universities (150721907), Pilot Talent Training Program of Shanghai East Hospital (No. 201701), Research project of Shanghai Association for Science and Technology (16441901004), Key Disciplines Group Construction Project of Pudong Health Bureau of Shanghai (Grant No. PWZxq2017-13).

#### REFERENCES


as reagent for photodynamic therapy. *Nanotechnology* 26:055102. doi: 10.1088/0957-4484/26/5/055102


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2019 Feng, Zhu, Chen, Lu, Liu, Kim, Liang, Zhang, Lin, Ma and Dong. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Synergistic Effect of Retinoic Acid Polymeric Micelles and Prodrug for the Pharmacodynamic Evaluation of Tumor Suppression

#### Edited by:

Jianxun Ding, Changchun Institute of Applied Chemistry (CAS), China

#### Reviewed by:

Youmei Bao, Yale University, United States Xianzhu Yang, South China University of Technology, China Ruinan Yang, Pharmaceutical Product Development, United States

#### \*Correspondence:

Li-Feng Hang hanglf@ustc.edu.cn

†These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Received: 26 February 2019 Accepted: 08 April 2019 Published: 16 May 2019

#### Citation:

Zhu Y-H, Ye N, Tang X-F, Khan MI, Liu H-L, Shi N and Hang L-F (2019) Synergistic Effect of Retinoic Acid Polymeric Micelles and Prodrug for the Pharmacodynamic Evaluation of Tumor Suppression. Front. Pharmacol. 10:447. doi: 10.3389/fphar.2019.00447 Yan-Hua Zhu<sup>1</sup>† , Ning Ye<sup>1</sup>† , Xin-Feng Tang<sup>1</sup> , Malik Ihsanullah Khan<sup>1</sup> , Hong-Liang Liu<sup>2</sup> , Ning Shi<sup>2</sup> and Li-Feng Hang<sup>1</sup> \*

<sup>1</sup> School of Life Sciences, University of Science and Technology of China, Hefei, China, <sup>2</sup> Shandong Liangfu Pharmaceutical Co., Ltd., Jining, China

All-trans retinoic acid (ATRA) is an effective agent that induces differentiation, inhibits cell proliferation, and acts as an anticancer agent. ATRA was successfully conjugated with Pluronic F127 via esterification to enhance its anticancer effects. Pluronic-ATRA showed high cytotoxicity and inhibitory concentrations (IC50) 50% lower than those of ATRA in various breast cancer cell lines (4T1:31.16–8.57 µg/mL; EMT6: 50.48–7.08 µg/mL; MDA-MB-231:37.58–8.99 µg/mL; BT474:25.27–9.09 µg/mL). In combination with chemotherapy, Pluronic-ATRA synergistically enhanced the cytotoxic effects of cisplatin (CDDP). Pluronic-ATRA combined with CDDP effectively suppressed breast tumor growth in vivo. The results of this study demonstrate the potential of Pluronic-ATRA as an anticancer agent that can be used in combination therapy against solid tumors.

Keywords: all-trans retinoic acid, Pluronic F127, cisplatin, breast cancer, combination

# INTRODUCTION

Breast cancer results in the development of destructive tumors and leads to the highest rate of cancer-related deaths in females around the world (Wang et al., 2017; Zhang et al., 2017). Breast cancer therapy is currently based on clinical and pathological findings but shows limited therapeutic efficacy and is restricted to sequential chemotherapy, radiation, and surgery (Misra et al., 2010). Cisplatin (CDDP) is generally used as a first-line therapeutic agent against blood cancer (lymphoma and myeloma) and solid tumors (small cell lung, non-small cell lung, ovarian, stomach, bladder, and particularly testicular cancers) because of its ability to cross-link with DNA after entering the cancer cell (Wang and Lippard, 2005). CDDP binds DNA double-helix strands at atom N7 of guanine bases, thereby impeding double strands from uncoiling and separating. This prevents cell division and leads to programmed cell death and apoptosis (Pasini and Zunino, 1987; Galanski et al., 2003; Aryal et al., 2009). Replacing cytotoxic drugs with combinations (Chen et al., 2017a; Zhang et al., 2018) involving other therapeutic agents such as photosensitizers and neovasculature disruption agents as well as traditional drugs represents an effective alternative to cancer therapy

(Li et al., 2017). Combinations of two or more therapeutic agents or strategies present unique advantages, including the inhibition of different signaling pathways, target-specific actions, improved drug efficacy, and reduced off-target toxicity (Bang et al., 2010).

Retinoic acid (RA) is a natural derivative of vitamin A whose anticancer activity has received much attention (Zhu et al., 1997; Siddikuzzaman et al., 2011). Although RA does not generally inhibit cell growth, it has been proposed that it suppresses tumor growth by inducing cell differentiation, inhibiting cell proliferation, and exerting anti-migration and invasion effects on tumor cells (Siddikuzzaman et al., 2011). RA has different natural and synthetic compound derivatives called retinoids, which affect cell differentiation, growth, and apoptosis. Retinoids have the ability to induce differentiation and apoptosis in cancer cells via cytotoxic and anti-oxidant activities, suggesting their potential as chemotherapeutic agents against cancer (Chen et al., 2014). All-trans retinoic acid (ATRA), the most abundant natural analog, displays activity against various cancers such as lymphoma, leukemia, neuroblastoma, and lung, cervical, and kidney cancers, among others. ATRA controls cancer growth via unique mechanisms, such as the induction of cell differentiation and anti-proliferation, -migration, and invasion effects (Zhu et al., 1997; Park et al., 2011; Siddikuzzaman et al., 2011; Chen et al., 2014). ATRA suppressed the growth of human breast cancer cells (MCF-7) in vitro when used at concentrations >10 nM. Other studies have shown that ATRA inhibited the growth of hepatocellular carcinoma (Zhu et al., 1997). Clinical studies also showed the chemo-preventive effects of RA: ATRA, used as adjutant, effectively suppressed the occurrence of secondary cancers in patients with early-stage skin, head and neck, breast, and hepatocellular cancers (Sano et al., 2003; Clerici et al., 2004; Khuri et al., 2006). Certain studies also reported the synergistic therapeutic effects of ATRA and its derivatives in combination with chemotherapeutics such as doxorubicin, CDDP, and paclitaxel in inducing receptormediated cytotoxicity and inhibiting cell growth factors (Liu et al., 2012; Yao et al., 2013; Sun et al., 2015). Therefore, the use of RA in combination with chemotherapeutic agents contributes to enhancing the therapeutic efficacy and decreasing side effects. However, the clinical applications of ATRA are limited by low water solubility and plasma concentrations, systemic side effects (such as acute retinoid resistance, mucocutaneous dryness, hypertriglyceridemia, and headache), physicochemical instability, and few effective delivery systems (Emami et al., 2018; Wang et al., 2018a; Yu et al., 2019).

Nanotechnology-based medicines have shown significant advantages for cancer treatment in the last two decades, thus improving the shortcomings of current therapeutic agents by prolonging drug exposure through blood circulation, improving pharmacokinetic parameters, and enhancing tumor accumulation and cellular uptake (Peer et al., 2007; Davis et al., 2010a; Doane and Burda, 2012; Yuan et al., 2012). Nanomedicine improves therapeutic agent efficiency in cases of cancer treatment limitations including drug resistance, offtarget effects, and metastasis by ensuring targeted and multimode efficacy (Davis et al., 2010b; Meng et al., 2010; Sarfati et al., 2011; Yang et al., 2014). Nano-based delivery systems, including liposomal (Ozpolat et al., 2003), polymeric (Cho et al., 2001; Sun et al., 2014; Chen et al., 2015; Xiao et al., 2018), and lipid nanoparticles (Lim and Kim, 2002; Castro et al., 2007), have been used to deliver RA to cancer cells. However, the RA loading rate remains mostly low and unstable, easily resolving during storage or when entering the blood circulation after injection. In a recent study, new hybrid nanoparticles composed of polymeroil-based nano-carriers were used to stabilize RA and increase its stability during delivery (Narvekar et al., 2012). Although this new delivery system showed more effective anti-cancer activity compared to that of free RA, its preparation is complicated by low solubility, and these nanoparticles cannot be used for co-delivery of RA with other lipophilic drugs. In other studies, RA was conjugated to different polymers to increase loading efficiency and control the RA release rate (Hou et al., 2012; Zhang et al., 2015; Cao et al., 2018). For example, chitosan oligosaccharide was conjugated with RA to prepare polymer nanoparticles for co-delivery of RA and paclitaxel (Zhang et al., 2015). However, RA-conjugated nanoparticles did not display effective in vitro cytotoxicity because of the low degree of RA conjugation.

In the current study, we developed a polymeric micelle (Feng et al., 2017; Wang et al., 2018b) by conjugating ATRA to Pluronic F127 to form mixed micelles for efficient delivery. Pluronics are polymers composed of poly(ethylene oxide) poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO), in which PPO is hydrophobic and PEO is a hydrophilic (Moebus et al., 2009). Of all the Pluronic agents, Pluronic F127 is widely considered the most appropriate for biomedical applications (Akash and Rehman, 2015). Pluronic F127 has the ability to provide a hydrophobic core (Chen et al., 2017b) structure in aqueous solutions, inside which hydrophobic drugs such as ATRA can be solubilized and stabilized, so the low loading efficacy and instability of ATRA was overcome in this study by forming pluronic-ATRA micelles. Moreover, the combination of ATRA and CDDP effectively inhibited cancer cell proliferation and suppressed tumor growth in vivo. This work focused on the synthesis of F127-ATRA nanoparticles and the cytotoxic assessment of F127-ATRA micelles on different breast cancer cell lines in vitro. Furthermore, we tested whether lowdose F127-ATRA micelles and the traditional chemotherapeutic agent CDDP showed synergistic cytotoxic effects against breast carcinoma both in vitro and in vivo.

#### MATERIALS AND METHODS

#### Materials

Pluronic F127 and ATRA were obtained from Sigma-Aldrich (St. Louis, MO, United States) and CDDP was purchased from Shandong Boyuan Pharmaceutical Co., Ltd. (Jinan, China). The 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) and 4-(dimethylamine)-pyridine (DMAP) were purchased from Aladdin Industrial Co., Ltd. (Ontario, CA, United States). Dichloromethane (DCM) was distilled under reduced pressure before use. Water was ultra-purified using a Milli-Q water system (Millipore, Burlington, MA, United States) consisting of a carbon filter cartridge, two ion-exchange filter cartridges, and an organic removal cartridge. All other solvents and reagents were used as received unless specified.

#### Synthesis of Pluronic-ATRA

fphar-10-00447 May 15, 2019 Time: 16:33 # 3

Pluronic-ATRA was synthesized via esterification. ATRA (120 mg), Pluronic (1000 mg), EDC·HCl (115.2 mg), and DMAP (24.4 mg) were mixed in 10 mL DCM. The mixture was stirred in the dark for 48 h at 20◦C. Excess ATRA was removed by precipitation in ice-cold ether. To completely remove free EDC·HCl and DMAP, the solution was transferred to a dialysis membrane (Spectra/Por <sup>R</sup> , Float-A-Lyzer <sup>R</sup> , molecular weight cut-off (MWCO) = 1 kDa, Sigma-Aldrich) and dialyzed against ultra-purified water three times for 24 h at 4◦C, before further drying using a freeze dryer (Labconco, Kansas City, MO, United States) to completely remove the water. The final product Pluronic-RA was a yellow powder with a percentage grafting of 83%. The product was analyzed via nuclear magnetic resonance (NMR) and Fourier-transform infrared (FTIR) spectroscopy.

#### Cell Culture

The breast cancer cell lines 4T1, MDA-MB-231, EMT6, and BT474 were purchased from the American Type Culture Collection (ATCC, Manassas, VA, United States) and cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, Thermo Fisher Scientific, Waltham, MA, United States) supplemented with 10% fetal bovine serum (FBS, ExCell Bio, Shanghai, China)

(ATRA) (in CDCl3, ppm), (C) Fourier-transform infrared (IR) spectrum of Pluronic-ATRA.

and 1% penicillin/streptomycin (Gibco). All cells were incubated at 37◦C in a humidified atmosphere containing 5% CO<sup>2</sup> and were harvested via centrifugation at 800 × g for 5 min to stimulate propagation.

#### In vitro Cytotoxicity Assays

fphar-10-00447 May 15, 2019 Time: 16:33 # 4

For the toxicity study, MDA-MB-231, BT474, 4T1, and EMT6 cells (5 × 10<sup>3</sup> ) in 100 µL DMEM supplemented with 10% FBS were seeded in 96-well plates and incubated for 24 h at 37◦C in a 5% CO<sup>2</sup> atmosphere. The next day, the culture medium was removed and 100 µL fresh DMEM containing serial dilutions of ARTA, Pluronic F127, and Pluronic-ATRA (in triplicates). After an additional 48 or 72 h incubation, 20 µL 3-[4,5 dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium (MTT) solution (5 mg/mL) was added to each well. After further incubation for an additional 4 h, the MTT medium was replaced with 150 µL extraction buffer [dimethyl sulfoxide (DMSO) at 20◦C] before gently tapping the 96-well plate to dissolve the dye and incubating for 10 min at 37◦C. The absorbance of the each well was measured at 490 nm using a Bio-Rad 680 microplate reader (Bio-Rad, Hercules, CA, United States). Cell viability was normalized to that of the untreated control group, which served as an indicator of 100% cell viability.

# Combination of Pluronic-ATRA and CDDP

MDA-MB-231 and BT474 cells (2 × 10<sup>3</sup> ) as well as 4T1 and EMT6 cells (1 × 10<sup>3</sup> ) were seeded in 96-well plates in 100 µL DMEM and incubated at 37◦C in a 5% CO<sup>2</sup> atmosphere for 24 h. Cells were then divided into three groups: two pretreatment groups and a non-pretreatment group. For the pretreatment groups, the culture medium was replaced by 100 µL fresh medium containing Pluronic-ATRA (10−<sup>5</sup> M). After 48 h, one of the pretreatment groups was treated with CDDP in 100 µL DMEM and the other group was treated with Pluronic-ATRA (10−<sup>5</sup> M) and CDDP in 100 µL DMEM and incubated for an additional 72 h. For the non-pretreatment group, the cells were treated with CDDP alone for 72 h. Next, 100 µL fresh medium (containing 20 µL 5 mg/mL MTT stock solution in Milli-Q Water) was substituted for the old medium, with the exception of the wells used as blanks, to which the same volume of PBS was added and cultured for a further 4 h. After this, 150 µL extraction buffer (DMSO at 37◦C) was added to the wells and incubated for 10 min at 37◦C. The absorbance of the solution was measured at 490 nm using a Bio-Rad 680 microplate reader (Bio-Rad), and cell viability was normalized to that of cells cultured in culture medium with PBS treatment.

#### Xenograft Tumor Model

fphar-10-00447 May 15, 2019 Time: 16:33 # 5

Female mice (6 weeks old) were purchased from the Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All animals received care provided according to the regulations of animal care at the University of Science and Technology of China (USTC, Hefei, China). The study was approved by the USTC Animal Care and Use Committee. All animal experimental protocols conformed to the guidelines outlined in the Guide for the Care and Use of Laboratory Animals by the Laboratory Animal Center in USTC.

The different breast tumor xenograft models were generated using EMT6 cells (1 × 10<sup>6</sup> ) with 20% Matrigel (BD Biosciences, Franklin Lakes, NJ, United States) in 100 µL PBS injected into the mammary fat pat of female BALB/c mice. Tumor-bearing mice were used when the volume of tumors reached 80–100 mm<sup>3</sup> .

#### In vivo Tumor Suppression Evaluation

When the tumor reached a volume of 80–100 mm<sup>3</sup> , mice were randomly divided into four groups (n = 4 per group). Mice in the four groups were treated with PBS (control), Pluronic-ATRA (3 mg/kg), CDDP (2 mg/kg), or Pluronic-ATRA (3 mg/kg)/CDDP (2 mg/kg) via intravenous (i.v.) injection every 3 days, respectively. The volume of the tumor was measured using a venire caliper every second day and the volume was calculated according to the following formula:

$$volume \, = \, \frac{1}{2} \times length \times width^2$$

Changes in the body weights of mice were determined using an electronic scale. At the end of the in vivo therapeutic efficacy study, the mice were sacrificed and tumor tissues were harvested, weighed, and photographed.

#### Statistical Analysis

Statistical significance was assessed via one-way analysis of variance (ANOVA) using SPSS Statistics software (Version 22.0, SPSS Inc., Chicago, IL, United States). P-values <0.05 were considered to indicate statistical significance (95% confidence level).

#### RESULTS

#### Synthesis and Characterization of Pluronic-ATRA

The synthesis of Pluronic-ATRA is depicted in **Figure 1A**. In this study, Pluronic-ATRA was synthesized through a condensation reaction. The reaction was conducted in DCM

at 20◦C with EDC/DMAP as the catalyst. The catalyst is advantageous as it can be used in small doses and can react at 20◦C. The yield was approximately 80%. The structure of the final conjugation was confirmed via <sup>1</sup>H NMR and FTIR spectroscopy (**Figures 1B,C**).

The <sup>1</sup>H NMR spectrum of Pluronic-ATRA is shown in **Figure 1B**. The following resonances were assigned to the <sup>1</sup>H NMR spectrum (CDCl3, 400 MHz) : δ 1.02 (24H, c, c'), 1.06–1.22 (12H, q), 1.46 (4H, d), 1.61 (4H, e), 1.97 (4H, f), 1.71 (6H, g), 6.10–6.18 (4H, h, m), 6.24–6.32 (4H, I, k), 1.98 (6H, j), 7.0 (2H, l), 2.35 (6H, n), 5.81 (2H, o), 3.58–3.84 (516H, p, p', q), 3.30–3.43 (56H, r), 3.43–3.59 (168H, s), 4.26 (4H, t) ppm.

As shown in **Figure 1C**, the chemical structure of Pluronic-ATRA was further confirmed via FTIR spectroscopy. The presence of the strong absorption bands at 1700 and 1626 cm−<sup>1</sup> (**Figure 1C**, II) for the newly formed ester C = O moieties is further proof of the successful synthesis of the desired conjugate, thus confirming the presence of newly attached groups to F127, which were not seen in **Figure 1C**, I. The Pluronic-ATRA micelles were formed by the traditional strategy, and the average size of micelles is approximate 30 ± 2 nm (**Supplementary Figure S1**).

#### In vitro Cytotoxicity of ATRA in Breast Cancer Cells

Next, we evaluated the cytotoxicity of free ATRA and Pluronic-ATRA in vitro. We selected different breast cancer cell lines (4T1, MDA-MB-231, EMT6, and BT474) and evaluated cell viability via MTT assay. As shown in **Figure 2**, ATRA inhibited cell growth and caused cell death in different breast cancer cells. The cytotoxic effects of ATRA against estrogen receptornegative (ER-) 4T1, EMT6, and MDA-MB-231 cells were less potent, and the half maximal inhibitory concentrations (IC50) were relatively high compared to those in estrogen receptorpositive (ER+) BT474 cells. The IC<sup>50</sup> values in 4T1, MDA-MB-231, and EMT6 cells (**Figures 2A–C**) were 31.16, 37.58, and 50.48 µg/mL, respectively, while that in BT474 was 25.27 µg/mL (**Figure 2D**). This suggests that ATRA cytotoxicity depends on the expression level of the RA receptor (RAR) in different breast cancer cells, resulting in different sensitivities to ATRA. This indicates that ATRA may be used as a targeted therapeutic agent for anticancer treatment.

#### In vitro Cytotoxicity of Pluronic-ATRA in Breast Cancer Cells

After preparing Pluronic-ATRA, we carried out standard Pluronic F127 and Pluronic-ATRA cell toxicity tests before performing further biological experiments. The cytotoxicity of Pluronic alone in 4T1, MDA-MB-231, EMT6, and BT474 cancer cells after 48 h incubation was investigated (**Figure 3**), and results revealed that Pluronic F127 exhibited negligible cytotoxicity in cancer cells.

The viability of 4T1, MDA-MB-231, EMT6, and BT474 cells after incubation with Pluronic-ATRA for 48 h is shown in **Figure 4**. Pluronic-ATRA exhibited higher cytotoxicity in all cells than did free ATRA and Pluronic F127. A significant

decrease (four to sixfold) was observed in IC<sup>50</sup> values in MDA-MB-231, 4T1, and EMT6 cells (8.99, 8.57, and 7.08 µg/mL, respectively) compared with those of free ATRA treatment (50.48, 31.19, and 37.58 µg mL, respectively), as shown in **Supplementary Table S1**. Cytotoxicity was not enhanced in ER+ BT474 cells compared to ER- cells, with IC<sup>50</sup> values of 9.09 and 25.27 µg/mL, respectively, after treatment with free ATRA. These results prove that Pluronic-ATRA monotherapy enhanced

inhibitory effects on breast cancer cells compared to those of free ATRA, and that Pluronic F127 itself showed very limited or even no effect on cells. This suggests that Pluronic-ATRA would be a more effective therapeutic agent than free ATRA and may be used with other drugs in cancer treatment.

### The Combination of Pluronic-ATRA and CDDP Enhanced Cytotoxicity in Breast Cancer Cells

We further evaluated the in vitro anticancer effects of Pluronic-ATRA in combination with CDDP in 4T1, MDA-MB-231, EMT6, and BT474 cells. As depicted in **Figures 5A,B**, the cytotoxic effects of CDDP in combination with Pluronic-ATRA were enhanced compared to those of CDDP alone. Moreover, pretreatment with a lower concentration of Pluronic-ATRA (10−<sup>5</sup> M) followed by exposure to CDDP resulted in higher cytotoxicity compared to that after continuous Pluronic-ATRA treatment with CDDP. In addition, ER+ BT474 cells were more sensitive than ER-breast cancer cells. It is speculated that Pluronic-ATRA entered the cells via the RAR and inhibited cell proliferation, further enhancing CDDP efficacy.

#### In vivo Therapeutic Efficacy

Following the promising results obtained with combined Pluronic-ATRA-CDDP therapy in cancer cells in vitro, we investigated the therapeutic effects of Pluronic-ATRA in solid tumors. To assess whether Pluronic-ATRA-CDDP chemotherapy enhanced the therapeutic effects of CDDP in vivo, we established a tumor suppression experiment in female BALB/c mice bearing EMT6 tumors. After randomly dividing the mice into four groups, Pluronic-ATRA and Pluronic-ATRA/CDDP groups were pre-injected with Pluronic-ATRA (3 mg/kg) for 3 days before administering CDDP chemotherapy (**Figure 6A**). As shown in **Figures 6B,C**, tumor volume and weight varied among the different treatment groups. As illustrated in **Figure 6B**, PBS injection showed no inhibitory effects on tumor volume, leading to 14.2-fold volume increases after 21 days. Meanwhile, Pluronic-ATRA and CDDP alone displayed slight inhibitions of tumor growth compared to the control PBS group, although single therapy showed low efficacy. More noticeably, compared with single treatment, the combination of Pluronic-ATRA-CDDP showed the highest tumor growth inhibition and reduced tumor volume and weight, when compared to those in the other treatment groups. These results suggest that the combination of Pluronic-ATRA-CDDP chemotherapy showed higher antitumor efficacy than CDDP chemotherapy and Pluronic-ATRA therapy alone. During treatment, the body weight of mice was measured as an indicator of treatment-induced toxicity. No significant weight loss was observed in any of the treatment groups, suggesting that Pluronic-ATRA-CDDP was safe and had no adverse effects on the mice's health (**Figure 6D**), thus making it a safe, novel, and targeted synergistic therapeutic agent for solid tumor therapy.

# DISCUSSION

We designed and conjugated an RA derivative with Pluronic F127 via esterification and successfully prepared Pluronic-ATRA, which showed excellent biocompatibility and high ATRA loading content. In in vitro cell experiments, Pluronic-ATRA significantly increased the cytotoxic effects of free ATRA and further enhanced inhibition of different breast cancer cell lines when combined with CDDP. Additionally, tumors were efficiently suppressed by combination therapy of Pluronic-ATRA-CDDP in vivo. Therefore, the combination of Pluronic-ATRA and CDDP represents a promising strategy against cancer.

# DATA AVAILABILITY

The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.

#### ETHICS STATEMENT

All animals received care provided according to the regulations of animal care at University of Science and Technology of China (USTC, Hefei, China). The study was approved by the USTC Animal Care and Use Committee. All animal experimental protocols conformed to the guidelines outlined in the Guide for the Care and Use of Laboratory Animals by the Laboratory Animal Center in USTC.

# AUTHOR CONTRIBUTIONS

Y-HZ and NY performed the experiments, analyzed the data, and wrote the manuscript. X-FT, MK, H-LL, and NS participated in the experimental design and analyzed the data. L-FH participated in the experimental design and wrote the manuscript.

# FUNDING

This work was supported by the China Postdoctoral Science Foundation (2018M633141) and the Joint Fund for New Medical Sciences of USTC.

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphar. 2019.00447/full#supplementary-material

# REFERENCES

fphar-10-00447 May 15, 2019 Time: 16:33 # 10


second primary tumors in stage I and II head and neck cancer patients. J. Natl. Cancer Inst. 98, 441–450. doi: 10.1093/jnci/djj091



**Conflict of Interest Statement:** H-LL and NS were employed by the company Shandong Liangfu Pharmaceutical Co., Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Zhu, Ye, Tang, Khan, Liu, Shi and Hang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.