Fangchinoline Inhibits Human Esophageal Cancer by Transactivating ATF4 to Trigger Both Noxa-Dependent Intrinsic and DR5-Dependent Extrinsic Apoptosis

Esophageal squamous cell carcinoma (ESCC) is a recalcitrant cancer. The Chinese herbal monomer fangchinoline (FCL) has been reported to have anti-tumor activity in several human cancer cell types. However, the therapeutic efficacy and underlying mechanism on ESCC remain to be elucidated. In the present study, for the first time, we demonstrated that FCL significantly suppressed the growth of ESCC both in vitro and in vivo. Mechanistic studies revealed that FCL-induced G1 phase cell-cycle arrest in ESCC which is dependent on p21 and p27. Moreover, we found that FCL coordinatively triggered Noxa-dependent intrinsic apoptosis and DR5-dependent extrinsic apoptosis by transactivating ATF4, which is a novel mechanism. Our findings elucidated the tumor-suppressive efficacy and mechanisms of FCL and demonstrated FCL is a potential anti-ESCC agent.


INTRODUCTION
Esophageal squamous cell carcinoma (ESCC) is the major histologic subtype of esophageal cancer, and its incidence and fatality keep rising at an alarming rate worldwide (1). Despite the considerable progress in diagnosis and treatment of ESCC, the present therapeutic strategies, including chemotherapy, radiation and surgery, still have high recurrence and metastasis rates (2). Moreover, the developments of therapeutic targets and targeted drugs remain ineffective (3). Therefore, safe and effective therapeutic approaches for ESCC are urgently needed.
Currently, Chinese herbal medicinal agents have made great progress in the treatment of human cancers due to the relatively high efficacy and few side effects (4). The Chinese herbal monomer fangchinoline (FCL), extracted from the traditional Chinese herbal alkaloid tetrandrine root, characterizing as a new compound sharing structural features with tetrandrine ( Figure 1A) (5). FCL has been shown to have a wide range of pharmacological activities such as anti-inflammation, anti-oxidation and anti-thrombosis activities (6)(7)(8)(9). Remarkably, FCL exerts substantial anti-tumor (H) Mouse body weight was recorded every other day during the whole experiment. *denotes P < 0.05, **denotes P < 0.01, ***denotes P < 0.001, ****denotes P < 0.0001, n.s. denotes not significant. efficacy on many types of human tumor cells by arresting cell cycle, inhibiting metastasis, as well as triggering apoptosis (10)(11)(12). For example, it was reported that FCL inhibited cell growth in lung cancer cells and melanoma cells by targeting the FAK pathway (13,14). Furthermore, FCL induced apoptosis of breast cancer cells and glioblastoma cells by activating the PI3K/Akt/GSK-3b pathway (15,16). However, the anti-tumor efficacy of FCL on ESCC and it's underlying mechanism has not been investigated.
In the present study, for the first time, we reported that FCL effectively suppressed the tumor progression of ESCC by triggering cell-cycle arrest and apoptosis. More importantly, we reported a novel mechanism by which FCL transactivated ATF4 to trigger both Noxa-dependent intrinsic and DR5-dependent extrinsic apoptosis. Our study revealed the tumor suppressive efficacy of FCL on ESCC, and validated FCL as a potential anti-ESCC agent.

Reagents
Fangchinoline was purchased from MCE (MedChem Express, Shanghai, China), and the purity of the compounds was ≥99.92%. FCL was dissolved in dimethyl sulfoxide (DMSO) and stored at −80°C for the in vitro study. For the in vivo study, FCL was dissolved first in 5% DMSO and then in 10% 2-hydroxypropyl-bcyclodextrin (Sangon Biotech, Shanghai, China).

Cell Viability and Clonogenic Survival Assay
Cells were seeded in black 96-well plates with 2×10 3 cells per well in triplicate and cultured overnight. Cells were treated with DMSO or FCL at the indicated concentrations for 0, 24, 48 and 72 hours. At the end of the incubation, the cell viability was measured by ATPlite luminescence assay (PerkinElmer, Norwalk, CT, USA) according to the manufacturer's protocol. For clonogenic survival assay, cells were plated into six-well plates (300 cells per well) in triplicate and allowed to adhere overnight. Cells were treated with the indicated concentrations of FCL and cultured for 12 days. Cells were stained with crystal violet and the colony number was counted. Colonies with more than 50 cells each were counted and photographed with a gel imager (GelDoc XR System, Bio-rad, USA).

Cell Cycle Analysis
For cell cycle analysis, cells were treated at the indicated concentrations of FCL for 24 hours. FCL-treated cells or control cells were harvested and fixed in 70% ethanol at -20°C overnight. Then, the fixed cells were stained with propidium iodide (PI, 36 mg/mL; Sigma, St. Louis, MO, USA) at 37°C for 15 min, and performed for fluorescence activated cell sorting (FACS) analysis by Flow Cytometry (BD FACSVerse ™ , New Jersey, USA). Data were analyzed with FlowJo 7.6 software.

Apoptosis Assay
For apoptosis analysis, cells were treated at the indicated concentrations of FCL for 24 hours. FCL-treated cells or control cells were collected and washed with cold PBS, and then stained with an AnnexinV-FITC and PI Apoptosis Kit according to manufacturer's instructions (Yuheng Biotechnology, Suzhou, China). Apoptotic cells were analyzed by Flow Cytometry (BD FACSVerse ™ , New Jersey, USA). Data were analyzed with FlowJo 7.6 software.

Subcutaneous Transplantation Tumor Model
BALB/c nude female mice were purchased from Lingchang Biological Technology Co., Ltd. (Shanghai, China). All mice were kept and bred in a specific pathogen-free environment in the animal facility of Longhua hospital. The mice were maintained in a temperature−controlled room (22 ± 2°C) with a 12−hours light/12−hours dark cycle and a relative humidity of 40−60%, and were given free access to sterilized food and water. Animal experiments were performed in accordance with the National Guidelines for Experimental Animal Welfare, with approval from the Institutional Animal Care and Use Committee of Longhua hospital, Shanghai University of Traditional Chinese Medicine.
Briefly, 4×10 6 Kyse150 cells were subcutaneously injected into the bilateral flank of each mouse, and mice were randomly assigned to control and FCL-treatment groups (five mice per group). Each mouse was treated with either b-cyclodextrin crystalline (vehicle control) or FCL (100 mg/kg) via intraperitoneal injection once a day for 13 consecutive days. The day of tumor appearance was designated day 1 (6 days after xenografting). Tumor size was measured with a caliper and tumor volume was calculated using ellipsoid volume formula (length×width2)/2. The body weights of the mice were measured with an electronic scale every other day. Tumor tissues were harvested, photographed, and weighed at the end of the experiment.

Statistical Analysis
The statistical significance of differences between groups was assessed using GraphPad Prism7 software (GraphPad Software, Inc., San Diego, CA, USA). All data were presented as mean ± Standard Error of Mean. The student's t-test was used for the comparison of parameters between two groups. P-value of P < 0.05 was significant, n.s.=not significant. For all tests, four levels of significance (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001) were used.

Fangchinoline Suppressed the Tumor Growth of ESCC In Vitro and In Vivo
We first evaluated the efficacy of FCL on normal human esophageal epithelial cell and ESCC cells. Our results showed that the IC50 values of FCL for the normal human esophageal epithelial cell line HET-1A and ESCC cell lines EC1, ECA109, Kyse450, Kyse150 were 8.93, 3.042, 1.294, 2.471 and 2.22 mM, respectively ( Figure 1B). Furthermore, we found a time and dose-dependent growth inhibition in the four ESCC cell lines ( Figure 1C). FCL inhibited colony formation of ESCC cells in a dose-dependent manner ( Figures 1D, E). These findings indicated that FCL suppressed the viability of ESCC cells. To further assess the efficacy of FCL, we established a subcutaneoustransplantation tumor model of human esophageal cancer in mice by using Kyse150 cells. As shown, FCL significantly inhibited tumor growth over time compared with the control group (P < 0.01, Figure 1F). Notably, FCL-treated group mice developed smaller tumors than the control group by tumor weight analysis (P < 0.01, Figure 1G). During the whole experiment, there was no substantial change in the body weights of mice between the control group and FCL treatment group, suggesting no general toxicity of FCL treatment ( Figure 1H). Collectively, our findings indicated that FCL inhibited the tumor growth of ESCC both in vitro and in vivo.

Fangchinoline Induced G1-Phase Cell-Cycle Arrest of ESCC Cells
To further explore the inhibitory mechanism of FCL on the viability of ESCC cells, the effect of FCL on cell cycle was determined. We found that cell populations in G0/G1 phase of cell cycle were significantly increased in EC1, ECA109, Kyse150 and Kyse450 cells in a dose-dependent manner (Figures 2A, B). Owing to CyclinE and Cyclin-dependent kinases 2, 4 and 6 (CDK2/4/6) are key regulators in the G1 phase, we next determined the expression levels of indicated regulators in FCLtreated ESCC cells (17). Our data showed that FCL treatment obviously dropped the protein levels of CyclinE and CDK2/4/6 in both EC1 and ECA109 cells ( Figure 2C), suggesting that FCL prevented G1 to S phase progression of ESCC cells.
In addition, we found that the cell cycle inhibitors p21 and p27, which inhibit CDK/Cyclin complexes (18,19), were significantly accumulated upon FCL treatment in EC1 and ECA109 cells ( Figure 2C). To further define the role of p21 and p27 in FCL-induced cell-cycle arrest, the expression of p21 or p27 was downregulated by siRNA silencing in FCL-treated EC1 cells. As shown in Figures 2D, E, p21 or p27 knockdown by siRNA significantly rescued the EC1 cells from FCL-induced G1 phase arrest. Taken together, our findings demonstrated that p21 and p27 played a crucial role in controlling G1 phase cell-cycle arrest elicited by FCL.

Fangchinoline Triggered Apoptosis in ESCC Cells
After revealing that FCL disturbed the ESCC cells in G1 phase, we next examined the cellular responses to FCL treatment. We observed that FCL-treated ESCC cells presented the notable feature of apoptosis-shrunk morphology (Supplementary Figure 1A). PI and Annexin-V-FITC staining analysis confirmed that the number of Annexin V-positive cells (apoptosis marker) increased significantly after FCL treatment ( Figures 3A, B). Furthermore, FCL-treated ESCC cells had increased levels of cleaved PARP, a classical marker of apoptosis ( Figure 3C). Therefore, our findings demonstrated that FCL triggered apoptosis in ESCC cells.

Fangchinoline-Induced Intrinsic Apoptosis Mediated by Noxa
To further characterize the mechanism underlying apoptosis in FCL-treated ESCC cells, we determined the expression of cleaved CASP9, a marker of intrinsic apoptosis. As shown in Figure 4A, FCL induced obvious accumulation of cleaved CASP9, as well as the upregulation of classical apoptotic hallmark cleaved CASP3 in EC1 and ECA109 cells, indicating that intrinsic apoptosis of ESCC cells was triggered by FCL. To explore the mechanism for activation of intrinsic apoptosis upon FCL treatment, we determined the expression of classical proapoptotic protein (Noxa, Bax and Bid). Strikingly, Noxa expression was obviously increased in both EC1 and ECA109 cells while Bax and Bid were downregulated ( Figure 4B). Mechanistic studies showed that Noxa was transactivated by FCL ( Figure 4C).
To further determine the potential role of Noxa in FCLinduced intrinsic apoptosis, the expression of Noxa was downregulated via siRNA silencing. Our data showed that Noxa knockdown with siRNA significantly suppressed FCLinduced intrinsic apoptosis, as evidenced by (i) the attenuated percentage of Annexin V-positive cells ( Figures 4D, E), and (ii) the reduction of the cleaved fragments of PARP ( Figure 4F), demonstrating that FCL induced Noxa-dependent intrinsic apoptosis in ESCC cells. Given that Noxa could be transactivated by ATF4 (20,21), we, therefore tested the potential involvement of ATF4 in FCL-induced Noxa expression in ESCC cells. As shown in Figures 4G, H, downregulation of ATF4 significantly inhibited the induction of Noxa at both mRNA ( Figure 4G) and protein levels ( Figure 4H) in EC1 and ECA109 cells, indicating that ATF4 transactivated Noxa upon FCL treatment.

Fangchinoline Activated Extrinsic Apoptosis via the ATF4-DR5 Axis
Next, we examined the expression of cleaved CASP8, the initiator caspase of extrinsic apoptosis, to investigate whether FCL activated extrinsic apoptosis. Indeed, FCL stimulated the expression of cleaved CASP8 in both EC1 and ECA109 cells ( Figure 5A). To further define the potential mechanism of FCLinduced extrinsic apoptosis, the expression of death receptor family members Fas, DR3, and DR5 were determined. Our results showed that FCL significantly induced the expression of death receptor DR5 both at protein and mRNA levels ( Figures 5B, C), indicating that DR5 was involved in extrinsic apoptosis upon FCL treatment. To support this notion, the expression of DR5 was downregulated via siRNA silencing. We found downregulation of DR5 with siRNA significantly reduced the FCL-induced extrinsic apoptosis, along with a reduction in cleaved PARP expression (Figures 5D-F). These results highlighted the key role of DR5 in extrinsic apoptosis triggered by FCL.
Previous studies reported that transcription factor CHOP, a classical downstream target of ATF4, could transactivated DR5 (22)(23)(24). Therefore, we determined whether the induction of ATF4 and CHOP expression was responsible for the FCLinduced DR5 expression. Our study showed that FCL induced the obvious up-regulation of ATF4 and CHOP in EC1 and ECA109 cells (Figure 5G), along with an increase at the mRNA level of ATF4 ( Figure 5H). To further examine whether DR5-induced extrinsic apoptosis upon FCL treatment was ATF4 dependent, ATF4 expression was downregulated by siRNA silencing. We found that downregulation of ATF4 significantly rescued the induction of DR5 both at the mRNA ( Figure 5I) and protein levels ( Figure 5J), demonstrating the crucial role of ATF4 in the induction of DR5 upon FCL stimulation. As a result, ATF4 siRNA dramatically diminished the expression of cleaved PARP and cleaved CASP8 ( Figure 5J). Collectively, these results indicated that FCL activated the extrinsic apoptosis via ATF4-DR5 axis in ESCC cells.

DISCUSSION
ESCC is one of the most aggressive human malignancies with high incidence and mortality (25). However, few achievements have been achieved in the development of novel anti-ESCC strategies and effective drugs in the past few years (25). Recently, a variety of Chinese herbal extracts and isolated compounds exhibited the substantial anti-tumor efficacy in esophageal cancer cells, and some are candidates for clinical development (26). In the present study, FCL was shown to be a promising anti-ESCC agent with inhibited effects in four ESCC cell lines and in nude mouse xenograft. In mechanisms, FCL-treated ESCC cells arrested in the G1 phase of the cell cycle, which in a p21 and p27-induction manner. Furthermore, FCL transactivated ATF4 to coordinatively trigger Noxa-dependent intrinsic apoptosis and DR5-dependent extrinsic apoptosis ( Figure 6).
The acceleration of cell cycle process contributes to sustained proliferation and rapid growth of cancer cells. Cyclin dependent kinases (CDKs), such as Cyclin D/E and CDK2/4/6, which are involved in promoting cell cycle progression, are often overexpressed, while cyclin-dependent kinases (CDKIs), such as p21 and p27, are generally downregulated in cancer cells (27). Therefore, suppressing cell cycle progression by controlling cell cycle regulators is considered as an effective strategy to halt tumor growth. FCL was demonstrated to induce G1-S arrest by suppressing the expression of Cyclin D/E and CDK2/ 4/6 in several human cancers (12,16,28). Furthermore, it was reported that FCL restrained the cell cycle progression by inducing the accumulation of p21 and p27 in most malignancies, such as breast cancer cells, prostate carcinoma cancer cells and glioblastoma cells (16,28,29). However, the potential role of p21 and p27 in FCL-elicited cell cycle inhibition was unclear. In our study, we found that FCL arrested cell cycle progression at G1 phase by inducing the accumulation of cell cycle inhibitors p21 and p27. Rescue experiments further revealed that additional p21 or p27 knockdown reversed the FCL-induced G1 phase arrest. Transcriptional regulation of ATF4 on DR5 was analyzed by real-time PCR. *denotes P < 0.05, **denotes P < 0.01, ***denotes P < 0.001, **** denotes P < 0.0001, n.s. denotes not significant.
Therefore, for the first time, we demonstrated that FCL-induced cell-cycle arrest in ESCC is dependent on p21 and p27. Endoplasmic reticulum (ER) stress, as the common trigger of apoptosis, has been reported to induce CHOP-mediated DR5 transcription and CASP8-mediated extrinsic apoptosis in human cancer cells (30,31). Previous studies showed that FCL significantly upregulated the ER stress markers including CHOP and ATF4 (32). Therefore, we determined whether FCL activated apoptosis through ATF4-DR5 axis. In this study, we found that, in ESCC cells, FCL induced DR5-mediated extrinsic apoptosis. Moreover, DR5-induced extrinsic apoptosis is ATF4 dependent since downregulation of ATF4 significantly reduced FCL-induced apoptosis. In addition to extrinsic apoptosis, we showed that FCL triggered intrinsic apoptosis in a Noxa-dependent manner. FCL-induced Noxa up-regulation was also ATF4-dependent. However, knockdown of ATF4 did not completely rescue FCL-induced Noxa accumulation ( Figure 4G, H). Considering several transcription factors (ATF3, p53, NF-kB, and c-Myc, etc) that are known to mediate Noxa gene expression, except for ATF4 (33,34). The precise regulatory mechanism of Noxa induction elicited by FCL needs further exploration. Furthermore, our study showed that FCL transactivated ATF4 in ESCC cells. It has been reported that some Chinese herbal medicinal agents transactivate ATF4 by inducing ER stress. For example, Zerumbone and Parthenolide activated eIF2a through ER stress, thus inducing the transcription of ATF4 in human colon cancer cells and lung cells (35,36). Therefore, FCL may also transactivate ATF4 through ER stress. Future studies will be performed to elucidate the mechanism by how FCL transactivates ATF4 in esophageal cancer cells.
In conclusion, our study highlighted a pivotal role of FCL in suppressing the tumor progression of ESCC both in vitro and in vivo, and discovered a novel mechanism of FCL induction of both intrinsic and extrinsic apoptosis in ESCC, suggesting that FCL was a potential anti-ESCC agent.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by Animal Experimental Ethics Committee of Longhua hospital, Shanghai University of Traditional Chinese Medicine.

AUTHOR CONTRIBUTIONS
YJZ, SW, and LJ designed the study. YJZ and SW conducted most of the experiments and wrote the manuscript. YJZ and SW analyzed the data and contributed to the manuscript completion. YC, JZ, JY, JX, and LL offered technical support. HZ and YMZ provided the reagents. RMH, YMZ, and LJ revised the manuscript. All authors contributed to the article and approved the submitted version.