# ADVANCES IN LIVER INFLAMMATION AND FIBROSIS DUE TO INFECTIOUS DISEASES

EDITED BY : M. Victoria Delpino, Guillermo Hernán Giambartolomei, Jorge Quarleri, Sergio C. Oliveira and Gary Splitter PUBLISHED IN : Frontiers in Immunology and Frontiers in Cellular and Infection Microbiology

#### 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 978-2-88966-166-4 DOI 10.3389/978-2-88966-166-4

#### 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

# ADVANCES IN LIVER INFLAMMATION AND FIBROSIS DUE TO INFECTIOUS DISEASES

Topic Editors: M. Victoria Delpino, CONICET Institute of Immunology, Genetics and Metabolism (INIGEM), Argentina Guillermo Hernán Giambartolomei, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina Jorge Quarleri, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina Sergio C. Oliveira, Federal University of Minas Gerais, Brazil Gary Splitter, University of Wisconsin-Madison, United States

Citation: Delpino, M. V., Giambartolomei, G. H., Quarleri, J., Oliveira, S. C., Splitter, G., eds. (2020). Advances in Liver Inflammation and Fibrosis due to Infectious Diseases. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88966-166-4

# Table of Contents

*04 Editorial: Advances in Liver Inflammation and Fibrosis Due to Infectious Diseases*

Sergio C. Oliveira, M. Victoria Delpino, Guillermo Hernán Giambartolomei, Jorge Quarleri and Gary Splitter

*07 Tpl2 Protects Against Fulminant Hepatitis Through Mobilization of Myeloid-Derived Suppressor Cells*

Jing Xu, Siyu Pei, Yan Wang, Junli Liu, Youcun Qian, Mingzhu Huang, Yanyun Zhang and Yichuan Xiao


Qian Wang, Dehai Li, Jing Zhu, Mingyue Zhang, Hua Zhang, Guangchao Cao, Leqing Zhu, Qiping Shi, Jianlei Hao, Qiong Wen, Zonghua Liu, Hengwen Yang and Zhinan Yin

*84 The Genus* Alistipes*: Gut Bacteria With Emerging Implications to Inflammation, Cancer, and Mental Health* Bianca J. Parker, Pamela A. Wearsch, Alida C. M. Veloo and

Alex Rodriguez-Palacios

*99 Role of Kupffer Cells in Driving Hepatic Inflammation and Fibrosis in HIV Infection*

Lumin Zhang and Meena B. Bansal

# Editorial: Advances in Liver Inflammation and Fibrosis Due to Infectious Diseases

Sergio C. Oliveira<sup>1</sup> \* † , M. Victoria Delpino2,3 \* † , Guillermo Hernán Giambartolomei 2,3 \* † , Jorge Quarleri 3,4 \* † and Gary Splitter <sup>5</sup> \* †

<sup>1</sup> Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, Brazil, <sup>2</sup> Facultad de Farmacia y Bioquímica, Instituto de Inmunología, Genética y Metabolismo (INIGEM), Universidad de Buenos Aires, Buenos Aires, Argentina, <sup>3</sup> Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina, <sup>4</sup> Facultad de Medicina, Instituto de Investigaciones Biomédicas en Retrovirus y Sida (INBIRS), Universidad de Buenos Aires (UBA), Buenos Aires, Argentina, <sup>5</sup> Department of Pathobiological Sciences, University of Wisconsin-Madison, Madison, WI, United States

Keywords: fibrosis, HSC, liver, inflammasome, kupffer

**Editorial on the Research Topic**

#### Edited and reviewed by:

Ulrich Emil Schaible, Research Center Borstel (LG), Germany

#### \*Correspondence:

Sergio C. Oliveira scozeus1@gmail.com M. Victoria Delpino mdelpino@ffyb.uba.ar Guillermo Hernán Giambartolomei ggiambart@ffyb.uba.ar Jorge Quarleri quarleri@fmed.uba.ar Gary Splitter gary.splitter@wisc.edu

†These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Microbial Immunology, a section of the journal Frontiers in Immunology

Received: 14 May 2020 Accepted: 30 June 2020 Published: 05 August 2020

#### Citation:

Oliveira SC, Delpino MV, Giambartolomei GH, Quarleri J and Splitter G (2020) Editorial: Advances in Liver Inflammation and Fibrosis Due to Infectious Diseases. Front. Immunol. 11:1760. doi: 10.3389/fimmu.2020.01760 Liver inflammation is a common trigger of hepatic disease, and it is considered the main driver of tissue damage (1). Although the liver is able to regenerate, chronic liver inflammation leads to tissue damage with concomitant fibrosis, frequently leading to cirrhosis, and carcinogenesis (2).

**Advances in Liver Inflammation and Fibrosis Due to Infectious Diseases**

The etiology of chronic liver inflammation may be infectious or not. Several microorganisms including bacteria, parasites, fungi, and virus could be involved. Although, other metabolic and immune disorders may also participate. Several proinflammatory cytokines including IL-1α, IL-1β, TNF-α, and IL-6 has been involved in liver disease (3). They are produced by resident macrophages (Kupffer cells) and also by recruited macrophages and neutrophils (4). The common mediators of fibrosis are IL-10 family cytokines, VEGF, EGF, and TGF-β that constitute the main profibrotic cytokines (5). TGF-β stimulate hepatic stellate cell trans-differentiation from a quiescent, vitamin A storing cell to a proliferative myofibroblast that is the central driver of fibrosis (6).

In this Research Topic, a series of articles provide new insights about the current view of liver pathology, inflammation and fibrosis, with original articles about the role of myeloid suppressor cells in fulminant hepatitis, inflammasome activation in brucellosis, new insights in preventing NALDF progression, and the role of NLRP6 in granuloma formation and liver disease associated with schistosomiasis.

This Research Topic also features three review articles regarding the role of glucocorticoids in therapy of liver failure, the interaction of liver and immune cells in brucellosis, the role of T cells in hepatic schistosomiasis, as well as the role of beneficial bacterial in liver disease progression. Also included is a perspective article with proposed mechanisms of interaction between human immunodeficiency virus (HIV) and Kupffer cells.

The original research articles included in this Research Topic were performed involving both in vitro, and in vivo experiments in the murine model. Such model has acquired particular relevance during the study of non-invasive biomarkers able to assess fibrosis in patients with chronic liver disease.

The liver is the most commonly affected organ in patients with active brucellosis. Accordingly, clinical and biochemical records of liver involvement have been observed in up to 50% of patients with active disease (7). However, the molecular mechanisms involved in liver damage has been

**4**

recently started to be elucidated. Giambartolomei and Delpino report current understanding of the interaction between liver structural cells including hepatocytes and hepatic stellate cells and immune system cells during Brucella infection. They highlight the role of the type IV secretion system (T4SS) and the effector protein BPE005 in the activation of hepatic stellate cells to induce fibrosis. It is likely that BPE005 could participate in granuloma formation that might act as a reservoir of bacteria contributing to the disease chronicity. Prior to the development of liver fibrosis may occur events of inflammation and cell death. The contribution of the inflammasomes activation due to B. abortus infection on liver fibrosis was explored in an original article. Arriola Benitez et al. demonstrate in a series of experiments performed using the hepatic stellate cell line LX-2, that inflammasomes NLRP3 and AIM2 are involved in the induction of a fibrotic phenotype in hepatic stellate cells during B. abortus infection in a mechanism that involved IL-1β. These experiments were further corroborated using knock out (KO) mice, and strongly suggest the main contribution of this inflammasomes in the liver fibrosis during B. abortus infection.

In the liver fibrosis induced by Schistosoma mansoni infection, inflammasomes has also involved. Sanches et al. have unraveled using the murine model, the crucial role of the NLRP6 inflammasome in the schistosomiasisassociated hepatic granulomas. Hence, the absence of NLRP6 correlates with a significant reduction in inflammation and collagen deposition, as well as α-SMA and IL-13 levels as fibrotic markers. Both articles stand out the role of inflammasomes as central contributors to liver fibrosis triggered by infectious hepatic diseases. Moreover, parasite worms such as Schistosoma mansoni and S. japonicum induce a dramatic granulomatous response in liver and intestines. Subsequently, infection may further develop into significant fibrosis and portal hypertension. Zheng et al. extensively review previous reports to elucidate the contribution of T lymphocytes and their secreted cytokines in the immunopathology of schistosomiasis. They clearly indicate the different roles developed by the T-cell subsets for regulating the pathological progression of schistosomiasis in the local microenvironment. Moreover, they highlight recent findings about Tfh and Th9 cells as promoters of liver granulomas and fibrogenesis.

The tumor progression locus 2 (TPL2) is a serine/threonine kinase acting as a key mediator in liver and systemic metabolic disorders with an inflammatory component (8). However, the function of TPL2 in regulating hepatocyte function and liver inflammation during fulminant hepatitis is poorly understood, Xu et al. shed light on the TPL2 role associated with the myeloid derived suppressor cells (MDSC)-mediated protection against fulminant hepatitis. Using the murine model, they demonstrate that TPL2 deficiency suppresses IL-25-induced chemokine CXCL1/2 expression in hepatocytes, thus impairing MDSC recruitment into the liver, and increasing the infiltrate CD4<sup>+</sup>

lymphocyte proliferation that enhances fulminant hepatitis development. These findings strongly suggest that TPL2 is a potential target for the fulminant hepatitis treatment since it plays a critical role in MDSC recruitment.

The interaction between HIV and liver cell population has been previously explored. However, the relevance is still unclear in well-controlled patients on ART. In this context, the perspective article, Zhang and Bansal propose that during HIV infection, changes in the biology of Kupffer cells would create a microenvironment that drives hepatic pathology during microbial translocation. Targeting this pathway could help to improve liver-related consequences in HIV patients.

Liver damage associated to sterile inflammation was evaluated in a model of Non-alcoholic fatty liver disease (NAFLD). This pathology is associated to an excessive storage of fatty acids in the form of triglycerides in hepatocytes and it is considered the main causes of cirrhosis and major risk factors for hepatocellular carcinoma and liver-related death. Wang et al. reveal in a mice model that perforin regulates the abundance of hepatic IFN-γ-producing CD4<sup>+</sup> T cells with concomitant decrease of macrophage accumulation in liver. This indicate that perforin can act as an immune regulator to prevent NAFLD, suggesting its potential use to prevent hepatic steatosis and other related liver metabolic disorders.

Finally, the last two review article focus on the potentials therapy against liver pathology.

Excessive systemic inflammation is considered as the trigger of liver failure. Glucocorticoids (GCs) can rapidly suppress excessive inflammatory reactions and immune response. In a review article, Xue and Meng describe the current knowledge regarding glucocorticoid therapy in liver failure including emerging information. Although, the topic was addressed from basic research and clinical trials, the current understanding remains inconclusive for the application of GS treatment during liver failure.

The beneficial effect of probiotics has been extended to liver function in cirrhosis, nonalcoholic fatty liver disease, and alcoholic liver disease. Parker et al. review the current knowledge of bacteria of the genus Alistipes on its protective role during liver fibrosis among other diseases in animal models.

In summary, this Research Topic highlights the immunopathology of liver damage due to different pathologies. Knowledge acquired from articles contained in this special issue are relevant to the discovery of new targets for controlling liver inflammation and fibrosis.

We wish to express our appreciation to all the authors and reviewers who have participated in this Research Topic.

## AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

## REFERENCES


infection: a study of 530 cases. Medicine (Baltimore). (1996) 75:195– 211. doi: 10.1097/00005792-199607000-00003

8. Gong J, Fang C, Zhang P, Wang PX, Qiu Y, Shen LJ, et al. Tumor progression locus 2 in hepatocytes potentiates both liver and systemic metabolic disorders in mice. Hepatology. (2019) 69:524–44. doi: 10.1002/hep.29820

**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 © 2020 Oliveira, Delpino, Giambartolomei, Quarleri and Splitter. 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.

# Tpl2 Protects Against Fulminant Hepatitis Through Mobilization of Myeloid-Derived Suppressor Cells

Jing Xu1,2, Siyu Pei <sup>2</sup> , Yan Wang<sup>2</sup> , Junli Liu<sup>2</sup> , Youcun Qian<sup>2</sup> , Mingzhu Huang<sup>3</sup> \*, Yanyun Zhang1,2 \* and Yichuan Xiao<sup>2</sup> \*

*<sup>1</sup> The First Affiliated Hospital of Soochow University, Institutes for Translational Medicine, State Key Laboratory of Radiation Medicine and Protection, Key Laboratory of Stem Cells and Medical Biomaterials of Jiangsu Province, Medical College of Soochow University, Soochow University, Suzhou, China, <sup>2</sup> CAS Key Laboratory of Tissue Microenvironment and Tumor, Shanghai Institute of Nutrition and Health, Shanghai Institutes for Biological Sciences, University of Chinese Academy of Sciences, Chinese Academy of Sciences, Shanghai, China, <sup>3</sup> Department of Medical Oncology, Fudan University Shanghai Cancer Center, Shanghai, China*

#### *Edited by:*

*Sergio C. Oliveira, Federal University of Minas Gerais, Brazil*

#### *Reviewed by:*

*Nagarkatti Prakash, University of South Carolina, United States Patricia Lalor, University of Birmingham, United Kingdom*

#### *\*Correspondence:*

*Mingzhu Huang mingzhuhuang0718@163.com Yanyun Zhang yyzhang@sibs.ac.cn Yichuan Xiao ycxiao@sibs.ac.cn*

#### *Specialty section:*

*This article was submitted to Inflammation, a section of the journal Frontiers in Immunology*

*Received: 27 March 2019 Accepted: 05 August 2019 Published: 20 August 2019*

#### *Citation:*

*Xu J, Pei S, Wang Y, Liu J, Qian Y, Huang M, Zhang Y and Xiao Y (2019) Tpl2 Protects Against Fulminant Hepatitis Through Mobilization of Myeloid-Derived Suppressor Cells. Front. Immunol. 10:1980. doi: 10.3389/fimmu.2019.01980* Myeloid derived suppressor cells (MDSC) in the liver microenvironment protects against the inflammation-induced liver injury in fulminant hepatitis (FH). However, the molecular mechanism through which MDSC is recruited into the inflamed liver remain elusive. Here we identified a protein kinase Tpl2 as a critical mediator of MDSC recruitment into liver during the pathogenesis of *Propionibacterium acnes*/LPS-induced FH. Loss of Tpl2 dramatically suppressed MDSC mobilization into liver, leading to exaggerated local inflammation and increased FH-induced mortality. Mechanistically, although the protective effect of Tpl2 for FH-induced mortality was dependent on the presence of MDSC, Tpl2 neither directly targeted myeloid cells nor T cells to regulate FH pathogenesis, but functioned in hepatocytes to mediate the induction of MDSC-attracting chemokine CXCL1 and CXCL2 through modulating IL-25 (also known as IL-17E) signaling. As a consequence, increased MDSC in the inflamed liver specifically restrained the local proliferation of infiltrated pathogenic CD4<sup>+</sup> T cells, and thus protected against the inflammation-induced acute liver failure. Together, our findings established Tpl2 as a critical mediator of MDSC recruitment and highlighted the therapeutic potential of Tpl2 for the treatment of FH.

Keywords: hepatitis, myeloid-derive suppressor cells (MDSCs), TPL2, IL-25, chemokine

#### INTRODUCTION

Fulminant hepatitis (FH) is a dreaded disease characterized by rapid development of hepatocellular dysfunction, leading to the failure of hepatic regeneration (1). Published studies have shown that the pathogenesis of FH is associated with huge liver infiltration of immune cells, which secrete a large number of pro-inflammatory cytokines and thus induce acute inflammatory necrosis of hepatocytes (2, 3). The bacterial infection has been considered as a key factor that contribute to the development of FH pathology (4). Indeed, mice injected with heat-killed Propionibacterium acnes followed by lipopolysaccharide (LPS), one of the most commonly used FH animal models, phenocopy the inflammatory infiltration in hepatic parenchyma and finally lead to the acute liver failure (5–8). Although the pathogenesis of FH has been extensively investigated, there is no proper therapeutic strategies for this disease, leading to high mortality if there is no supportive management and/or liver transplantation (9).

**7**

Myeloid derived suppressor cells (MDSC) are a heterogeneous group of immune cells derived from bone marrow and have been implicated to play important immunosuppressive and protective roles in human hepatitis, hepatocellular carcinoma or various mouse hepatitis models through different mechanism. For example, MDSC inhibited T cell proliferation and IFN-γ production in chronic HCV patients (10), and suppressed NK cell function during the pathogenesis of human hepatocellular carcinoma (11). In hepatitis mouse models, MDSC also exhibited immunosuppressive function through inhibiting the T cells proliferation, activation and secretion of pro-inflammatory cytokines, and thus protected against hepatic inflammation and fibrosis through different mechanisms (12–14). Therefore, increasing the number of MDSC in the liver may help to inhibit the occurrence of local inflammation of the liver and protect against FH. Indeed, administration of IL-25 dramatically prevented and reverses acute liver damage through promoting the recruitment of the MDSC into liver in FH mouse (15).

IL-25, also known as IL-17E, belongs to IL-17 cytokine family, and was initially found to be highly expressed in T helper (Th) 2 cells and promote the proliferation of Th2 cells and eosinophils (16–18). In addition, it has been reported that IL-25 exhibited inhibitory effect of the proliferation of Th1 and Th17 cells and further suppressed the occurrence of autoimmune diseases in mice (19, 20). However, it is not clear how IL-25 initiates the signal pathway to mediate MDSC recruitment into liver during FH pathogenesis. Published study has identified that IL-25 can bind to the heterodimer receptor composed of IL-17RA and IL-17RB, which then recruit Act1 to activate downstream NF-κB and MAPK (21–23), suggesting a similarity with IL-17A-induced signaling pathway. Our previous study has demonstrated that the serine/threonine protein kinase Tpl2 is a key component in regulating the IL-17A signaling pathway, in which the activated Tpl2 directly bound to and phosphorylated TAK1 and further induce the activation of downstream NF-κB and MAPK (24, 25). Based on the similarity of IL-17A- and IL-25-induced signaling and the critical protective role of IL-25 in FH, we speculated that Tpl2 may also regulated the FH pathogenesis through modulation of IL-25 signaling.

In the present study, we found that Tpl2 protected against FH-induced acute liver injury and mouse mortality. Loss of Tpl2 in hepatocytes suppressed IL-25-induced chemokine CXCL1/2 expression, which impaired the recruitment of MDSC into the liver, leading to promoted proliferation of liver-infiltrating CD4<sup>+</sup> T cells and enhanced FH pathology.

#### RESULTS

#### Tpl2 Protected Against *P. acne*/LPS-Induced FH

To investigate the in vivo role of Tpl2 during FH pathogenesis, we induced a FH model by intravenously injecting the mice with heat-killed P. acnes and followed by LPS. In this model, only P. acnes priming is not lethal for the mice, and P. acnes priming plus LPS injection 7 days later will strongly induce acute liver damage, leading to FH-related mortality. However, P. acnes priming-induced liver inflammation is necessary and the reason for the mortality after LPS injection in this FH model (6, 7). As shown in **Figure 1A**, low dose of P. acnes/LPS priming provoked a non-lethal moderate hepatitis in wildtype (hereafter termed WT) mice. In contrast, Tpl2-deficient mice that induced with FH by using the same dose of P. acnes/LPS developed a much severer disease, leading to 86% lethality within 8 h (**Figure 1A**). Consistently, we observed the increased production of serum aspartate aminotransferase (AST) and higher ratio of AST/aminotransferase (ALT) levels, which is a hallmark of hepatitis-induced liver failure, in Tpl2-deficient mice accordingly (**Figure 1B**). In addition, histological analysis showed that there was more inflammatory infiltration observed in the Tpl2-deficient liver tissues on day 7 after P. acnes priming (**Figures 1C,D**). These results collectively suggested an important beneficial role of Tpl2 in protecting P. acnes/LPS-driven acute liver damage.

#### *Tpl2* Deficiency Increased the Liver Infiltration of Pathogenic CD4<sup>+</sup> T Cells

The exaggerated FH in Tpl2-deficient mice promoted us to examine the cellular mechanism by which Tpl2 protect against liver failure during FH pathogenesis. We firstly examined the peripheral immune activation after P. acnes priming, and the results revealed that Tpl2-deficient and WT control mice had similar frequencies and absolute numbers of CD4<sup>+</sup> T cells, CD8<sup>+</sup> T cells, CD11c<sup>+</sup> dendritic cells, B220<sup>+</sup> B cells and CD4+Foxp3<sup>+</sup> regulatory T cells (Treg) in the spleens 7 days after challenged with P. acnes (**Figures 2A–D**). In addition, the frequencies and absolute numbers of IFN-γ- and TNF-α-producing T helper (Th)1 CD4<sup>+</sup> T cells in the spleens were also comparable between the WT and Tpl2-deficient mice (**Figures 2E,F**). These data suggested Tpl2 does not affect peripheral immune activation during FH pathogenesis.

It is known that P. acnes priming promoted the liver infiltration of CD4<sup>+</sup> T cells, which contributed to the inflammation-induced liver injury (8). Although the frequencies and absolute cell numbers of CD4<sup>+</sup> T cells in the spleens or peripheral blood were comparable in WT and Tpl2-deficient mice during P. acnes-primed process (**Figures 2G–J**), loss of Tpl2 dramatically increased the frequencies and absolute numbers of CD4<sup>+</sup> T cells in the livers as compared with that in WT liver, whereas didn't affect the liver infiltration of CD8<sup>+</sup> T cells, dendritic cells, B cells and Treg cells at day 7 after P. acnes priming (**Figures 3A–E**). Since the Th1 cells are the major pathogenic contributor of P. acnes-induced liver injury (26), we next examined the TNF-α and IFN-γ production among the infiltrated CD4<sup>+</sup> T cells. Interestingly, Tpl2 deficiency didn't affect the ability of liver-infiltrating CD4<sup>+</sup> T cells to produce TNF-α and IFN-γ, as reflected by comparable frequencies of IFN-γ- and TNF-α-producing Th1 cells in the inflamed livers of WT and Tpl2-deficient mice (**Figures 3F,G**). However, the absolute cell numbers of the pathogenic Th1 cells were dramatically increased in the inflamed livers of Tpl2-deficient mice (**Figures 3F,G**). Moreover, Tpl2 deficiency gradually increased the liver infiltration of CD4<sup>+</sup> T cells, notably

at day 5 and 7 after P. acnes priming (**Figures 3H,I**). These data collectively suggested Tpl2 may inhibited the liver infiltration of CD4<sup>+</sup> T cells during P. acnes-induced liver injury.

## Tpl2 Specifically Restricted the Proliferation of Liver-Infiltrating CD4<sup>+</sup> T Cells

To confirm the pathogenic role of liver-infiltrating CD4<sup>+</sup> T cells after P. acnes priming in the FH model, we injected different dose of P. acnes and examined the survival rate and liver infiltration of CD4<sup>+</sup> T cells. The results revealed that after challenged with a single shot of same dose of LPS, higher dose of P. acnes priming dramatically increased the mouse mortality rate as compared with that primed with lower dose of P. acnes (**Figure 4A**). As expected, higher dose of P. acnes priming significantly increased the infiltration of CD4<sup>+</sup> T cells in the livers, along with decreased frequencies of liver infiltration of MDSC (**Figures 4B,C**), which is known to restrain the local inflammation in inflamed liver microenvironment through inhibiting T cell proliferation (15, 27). Therefore, we speculated that Tpl2 may regulate the proliferation of liver-infiltrating CD4<sup>+</sup> T cells, and then injected bromodeoxyuridine (BrdU), a synthetic nucleoside that could be incorporated into newly synthesized DNA to monitor the cell proliferation, into WT or Tpl2-deficient mice before P. acnes priming. The flow cytometric analysis revealed that Tpl2 deficiency specifically increased the frequencies of BrdU<sup>+</sup> CD4<sup>+</sup> T cells that isolated from the inflamed livers, whereas the frequencies of BrdU<sup>+</sup> CD4<sup>+</sup> T cells were comparable in the spleens between WT and Tpl2-deficient mice after P. acnes priming (**Figures 4D,E**), suggesting Tpl2 may indirectly regulate the proliferation of CD4<sup>+</sup> T cells in the inflamed liver microenvironment. Indeed, Tpl2 is dispensable for the in vitro CD4<sup>+</sup> T cell proliferation, as characterized by comparable proliferation rate of naïve WT and Tpl2-deficient CD4<sup>+</sup> T cells upon the stimulation of anti-CD3 plus anti-CD28 antibodies (**Figure 4F**).

To further exclude the possibility that Tpl2 may directly function in CD4<sup>+</sup> T cell to regulate FH pathogenesis, we adoptively transferred WT or Tpl2-deficient CD4<sup>+</sup> T cells into T cell-deficient Rag1-KO mice, which were then induced the FH model by injecting P. acnes/LPS. Expectedly, the FHinduced mortality rates are comparable in recipient Rag1-KO

Flow cytometry analysis of the frequencies and absolute numbers of CD4<sup>+</sup> T cells, CD8<sup>+</sup> T cells, B220<sup>+</sup> B cells, CD11c<sup>+</sup> dendritic cells (A,B), CD4+Foxp3<sup>+</sup> Treg cells (C,D), and IFN-γ- and TNF-α-producing pathogenic Th1 cells (E,F) in the spleens of WT and *Tpl2*-KO mice at day 7 after *P. acnes* priming. Data are presented as representative plots of the frequencies of immune cell subpopulations (A,C,E) and a summary graph of the cell frequencies or absolute cell numbers (B,D,F). (G–J) Flow cytometry analysis of the frequencies and absolute numbers of CD4<sup>+</sup> T cells in the spleens (G,H) or peripheral blood (I,J) of WT and *Tpl2*-KO mice at day 3, 5, and 7 after *P. acnes* priming. Data are presented as representative plots of the frequencies of immune cell subpopulations (G,I) and a summary graph of the cell frequencies or absolute cell numbers (H,J). Results are mean ± SD from three independent experiments.

mice that either transferred with WT or Tpl2-deficient CD4<sup>+</sup> T cells (**Figure 4G**). Collectively, these results suggested that Tpl2 specifically restricted the proliferation of liver-infiltrating CD4<sup>+</sup> T cells through an indirect mechanism during FH pathogenesis.

## Tpl2 Mediated the Recruitment of MDSC Into Liver

Considering the indirect function of Tpl2 in regulating CD4<sup>+</sup> T cell proliferation, we examined the infiltration status of MDSC in inflamed liver. In contrast to the increased infiltrating rate of CD4<sup>+</sup> T cells in inflamed liver (**Figures 3H,I**), Tpl2 deficiency gradually decreased the frequencies and absolute numbers of liver-infiltrating CD11b+Gr-1<sup>+</sup> MDSC, notably at day 5 and 7 after P. acnes priming, as compared with that of WT mice (**Figures 5A,B**). The immunofluorescence analysis also confirmed the decreased infiltration of MDSC in the hepatic parenchyma of Tpl2-KO mice at day 7 after P. acnes priming (**Figure 5C**). However, loss of Tpl2 neither altered the frequencies and absolute numbers of MDSC in the spleens nor affected the peripheral distribution of MDSC in the circulation system during P. acnes-primed process (**Figures 5D–G**). In addition, in vitro proliferation assay revealed that Tpl2-deficient MDSC exhibited similar ability as WT MDSC to suppress either WT or Tpl2-deficient T cell proliferation after cocultured with CD4<sup>+</sup> T cell that stimulated with anti-CD3 plus anti-CD28 antibodies (**Figures 5H,I**). These data suggested that Tpl2 deficiency suppressed the liver recruitment of MDSC without affecting their immunosuppressive function.

We next examined whether the impaired liver recruitment of MDSC contributed to the enhanced mortality in Tpl2 deficient FH mice. To this end, we specifically deleted the MDSC by injection of anti-Ly-6G neutralizing antibody

\**P* < 0.05; \*\**P* < 0.01.

(**Supplementary Figures 1A,B**), and challenged mice with P. acnes/LPS to induce FH. As expected, in vivo MDSC depletion dramatically increased the mortality rate of WT FH mice, and largely abolished the difference of the survival rate between WT and Tpl2-deficient mice that injected with P. acnes/LPS (**Figure 5J**). These results collectively established Tpl2 as a critical mediator of MDSC mobilization into liver to protect inflammation-induced liver injury during FH pathogenesis.

#### Tpl2 Functioned in Liver-Resident Cells to Protect Against FH

To figure out which type of cells in vivo was directly targeted by Tpl2 to protect against FH-induced liver failure, we generated the mixed bone marrow (BM) chimeric mice by reconstituting the lethally irradiated WT mice with WT or Tpl2-deficient BM, which were then challenged with P. acnes/LPS to induce FH. Unexpectedly, Tpl2-deficient BM reconstituted chimeric mice were totally resistant to FH-induced death (**Figure 6A**), suggesting that Tpl2 deficiency in myeloid cells (including macrophages and MDSC) does not contribute to the aggregation of FH-induced mortality. However, when reconstituting the WT BM into lethally irradiated WT or Tpl2-deficient mice, we found that WT BM failed to induce FH-mediated death in WT recipient mice, whereas dramatically promoted the mortality rate of Tpl2 deficient recipient mice (**Figure 6B**). These data collectively suggested that Tpl2 didn't target myeloid cells, but functioned in liver-resident cells to protect against FH pathology.

Next, we examined the proinflammatory cytokine and chemokine induction in the livers at the early priming phase of FH model. After 3 days of P. acnes challenge, the expression of Th1 cytokine genes Ifng and Tnf in the Tpl2-deficient inflamed livers were much higher than that in WT livers (**Figure 6C**), which suggested that the increased liver infiltration of pathogenic Th1 cells as shown in **Figure 3G** may contribute to the enhanced expression of these pro-inflammatory genes. Accordingly, the expression of MDSC-attracting chemokine genes Cxcl1 and Cxcl2 were dramatically suppressed in the Tpl2-deficient livers as compared with that in WT livers (**Figure 6C**). In addition, the

WT C57BL/6 mice at day 7 after different dose of *P. acnes* priming. Data are presented as representative plots of the frequencies of immune cell subpopulations (B) and a summary graph of the frequencies or absolute cell numbers (C). (D,E) Flow cytometry analysis of CD4<sup>+</sup> T cell proliferation in the spleens and livers of WT and *Tpl2*-KO mice at day 7 after *P. acnes* priming (*n* = 4 mice/group). Data are presented as a representative histogram (D) and a summary bar graph (E) showing the frequencies of proliferating BrdU<sup>+</sup> CD4<sup>+</sup> T cells. (F) Proliferation of WT and *Tpl2-*KO CD4<sup>+</sup> T cells in the absence (NT) or presence of anti-CD3/CD28 antibodies, then assessed by [3H] thymidine incorporation. (G) Cumulative survival rates of age- and sex-matched *Rag1*-KO mice that adoptively transferred with WT and *Tpl2*-KO CD4<sup>+</sup> T cells and then subjected to *P. acnes*/LPS-mediated FH induction (*n* = 5 mice/group). Results are mean ± SD from three independent experiments. Two-tailed Student's *t*-tests were performed. \**P* < 0.05.

expression of the genes that encoding DC-recruiting chemokine MIP-1α (6) and other two MDSC-attracting chemokine CCL17 and CCL19 (15) were not affected in the livers of P. acnes-primed Tpl2-deficient mice (**Supplementary Figures 1C,D**). Moreover, the enzyme-linked immunosorbent assay confirmed that Tpl2 ablation inhibited P. acnes-induced CXCL1 chemokine protein production in the liver parenchyma as compared with that in WT livers (**Figure 6D**). Together, these results suggested that Tpl2 directly functioned in liver-resident cells, but not in peripheral immune cells, to mediate MDSC recruitment, and thus protect against FH pathology.

## Tpl2 Regulated IL-25 Signaling in Hepatocytes

Published study has suggested that IL-25 is highly produced in both human and mouse livers, and it is critical for the liver recruitment of MDSC in D-Gal/LPS-induced FH mouse model (15). In addition, we have previously demonstrated that Tpl2 mediates the activation of signaling pathway induced by IL-17A, which belongs to the same IL-17 family as IL-25 (24, 25). Therefore, we speculated that Tpl2 may potentially modulates IL-25 signaling in the liver-resident cells to regulate FH pathogenesis. To test this hypothesis, we firstly examined the IL-25 production in the livers, and found that there is no significant difference of IL-25 levels in the liver homogenates between WT and Tpl2-deficient mice that were either under physiological condition or challenged with P. acnes (**Figure 7A**), suggesting Tpl2 is dispensable for the IL-25 secretion in the liver tissue. In addition, we observed that IL-25 production in the livers of P. acnes-primed mice were comparable with that of naïve mice (**Figure 7A**), which is different from the D-Gal/LPS-induced FH model that IL-25 levels are decreased in the livers of FH mice (15). Next, we generated the Tpl2/Il25 double knockout mice and examined the potential in vivo link between Tpl2 and IL-25. Expectedly, IL-25 deletion under WT or Tpl2-KO background both dramatically increased the mortality rate of FH mice, and suppressed the expression of MDSCrecruiting chemokine genes Cxcl1 and Cxcl2 in P. acnes-primed livers (**Figures 7B,C**), implying IL-25 is also critical for the liver recruitment of MDSC and thus protect against P. acnes/LPSinduced FH. Interestingly, Tpl2/Il25 double knockout mice didn't further exaggerated the disease severity of P. acnes/LPS-induced FH when compared with Tpl2-deficient mice, as reflected by comparable mortality rate of these two strains of FH mice (**Figure 7B**), suggesting Tpl2-mediated prevention of FH is indeed through IL-25 signaling.

Next, we examined the cellular source of CXCL1 and CXCL2 in the P. acnes-primed livers. The results revealed that Tpl2 deficiency didn't affect the expression of Cxcl1 and Cxcl2 in P. acnes-primed liver CD11b+F4/80+GR-1<sup>−</sup> kuffer cells and macrophages (**Supplementary Figure 1E**). However, loss of Tpl2 or IL-25 both significantly suppressed these two genes' expression in the hepatocytes that isolated from P. acnes-primed livers (**Figure 7D**). In addition, the RNA sequencing analysis showed that the expression of Cxcl1, Cxcl2, and other IL-25 response genes are dramatically decreased in IL-25-stimulated Tpl2-deficient primary mouse hepatocytes as compared with that of WT cells (**Supplementary Figures 2A–D**, **Figure 7E**). Moreover, the quantitative PCR resulted also confirmed that Tpl2 is indispensable for the constitutively and IL-25-induced expression of Cxcl1 and Cxcl2 in primary mouse hepatocytes (**Figure 7F**). Collectively, these results suggested that Tpl2 mediated IL-25 signaling in hepatocyte to protect against FH.

#### DISCUSSION

FH is a life-threatening disease and liver transplantation is the only definitive treatment for the acute liver injury. However, the obvious side-effects of transplantation, such as donor shortage, immune rejection, detrimental effect of immunosuppressive drugs, etc., suggested an urgent to develop novel therapeutic strategies (1, 9, 28). Recently, accumulating evidences suggested that MDSC is critical to maintain the immunosuppressive niche in inflamed liver during the pathogenesis of various kinds of human hepatitis and related mouse models, and

FH induction (*n* = 4 mice/group). Results are mean ± SD from three independent experiments. Two-tailed Student's *t*-tests were performed. \**P* < 0.05; \*\**P* < 0.01.

FIGURE 6 | Tpl2 functioned in liver-resident cells to protect against FH. (A,B) Cumulative survival rates of the lethally irradiated bone marrow (BM) chimeric WT mice that reconstituted with WT or *Tpl2*-KO BM (A) or the BM chimeric WT or *Tpl2*-KO mice that reconstituted with WT BM (B), and then subjected to *P. acnes*/LPS-mediated FH induction (*n* = 7 or 5 mice/group). (C) QPCR analysis to determine the relative mRNA expression level of proinflammatory genes in livers of WT and *Tpl2*-KO mice at day 3 after *P. acnes* priming (*n* = 6 mice/group). Data were normalized to a reference gene, *Actb*. (D) Enzyme-linked immunosorbent assay of CXCL1 cytokine secretion in the supernatants of liver homogenates from WT and *Tpl2*-KO naïve or *P. acnes*-primed mice at day 3 (*n* = 4). Results are mean ± SD from three independent experiments. Two-tailed Student's *t*-tests were performed. \**P* < 0.05; \*\**P* < 0.01.

FIGURE 7 | Tpl2 regulated IL-25 signaling in hepatocytes. (A) Enzyme-linked immunosorbent assay showing IL-25 cytokine secretion in the supernatants of liver homogenates from WT and *Tpl2*-KO naïve or *P. acnes*-primed mice at day 3 (*n* = 3 or 4). (B) Cumulative survival rates of WT, *Tpl2*-KO, *Il25*-KO, *Tpl2*/*Il25* double knockout mice that subjected to *P. acnes*/LPS-mediated FH induction (*n* = 3 or 4 mice/group). (C,D) QPCR analysis of *Cxcl1* and *Cxcl2* genes' expression of livers (C) or hepatocytes (D) in WT, *Tpl2*-KO, *Il25*-KO, *Tpl2*/*Il25* double knockout mice at day 3 after *P. acnes* priming (*n* = 6 mice/group). (E) Scatter diagram showing the RNA-sequencing analysis of the gene expression pattern in WT and *Tpl2*-KO primary hepatocytes that stimulated by IL-25 for 8 h. The blue dots indicated the IL-25-induced up- or down-regulated genes, and the red dots indicated the down-regulation of *Cxcl1* and *Cxcl2* in *Tpl2*-KO hepatocytes. (F) QPCR determining the relative mRNA expression level of *Cxcl1* and *Cxcl2* in WT and *Tpl2*-KO hepatocytes that left untreated (NT) or stimulated with IL-25 for 8 h. Data were normalized to a reference gene, *Actb*. Results are mean ± SD from three independent experiments. Two-tailed Student's *t*-tests were performed. \**P* < 0.05; \*\**P* < 0.01.

increased infiltration of MDSC effectively attenuated the liver inflammation and protected FH-induced acute liver failure (10– 15, 27). However, the molecular mechanism through which driving MDSC mobilization into inflamed liver remain elusive. Here we identified the protein kinase Tpl2 as an essential mediator to mobilize MDSC into liver during FH pathogenesis, and thus Tpl2 effectively protected the mice against FH-induced acute liver failure and mortality. Therefore, Tpl2 may have therapeutic potential for the treatment of FH.

Tpl2 is a protein kinase that was initially identified as protooncogene due to the tumor promoting function of its Cterminal truncation (29, 30). The expression of Tpl2 is universal and it is found to expressed in both innate and adaptive immune cells and in diverse tissues, including the liver, lung, and intestines (30–33). The immune-regulatory function of Tpl2 is largely attributed to its activation of the MEK/ERK pathway in tolllike receptor (TLR), interleukin-1 receptor (IL-1R), or tumor necrosis factor receptor (TNFR) signaling (34, 35). In addition, Tpl2 also modulate the activation of p38, JNK, protein kinase B, and mammalian target of rapamycin in a context-dependent manner (25, 36). We previously found that Tpl2 functions in astrocytes to mediate IL-17A-induced chemokine (Cxcl1/2) expression through promoting TAK1 phosphorylation and its downstream NF-κB, p38, and JNK activation, whereas ERK activation is not affected (24). Therefore, it is not surprising we found in the present study that Tpl2 functioned in hepatocyte to modulate Cxcl1/2 expression, which then modulated the recruitment of MDSC into liver during FH pathogenesis. A recent study has suggested that Tpl2 exhibited neutrophil intrinsic function to mediate the trafficking of this type of immune cells (37), so it is also possible that Tpl2 may functioned directly in MDSC to promote its liver mobilization in FH mice. However, the increased mortality was only observed in Tpl2 germline knockout FH mice or Tpl2-deficient recipient chimeric FH mice that adoptively transferred with WT BM, but not in the WT recipient chimeric FH mice that adoptively transferred with either WT or Tpl2-deficient BM, suggesting the liver MDSC mobilization during FH pathogenesis is not attributed to the direct intrinsic function of Tpl2 in MDSC, but in hepatocytes.

Although IL-25 is one of the IL-17 family protein, there is no functional similarity of IL-25 as compared with the proinflammatory IL-17A (16, 17). For example, IL-25 augments type 2 immune responses and promote the airway inflammation of patients with asthma (38). A recent study suggested that IL-25 is highly expressed in both human and mouse liver, and plays a critical function in maintaining the homeostasis and limiting local inflammation through recruiting the immunosuppressive MDSC (15). Nevertheless, the molecular mechanism through which IL-25 recruit MDSC into liver is not clear. Our present study provided a Tpl2 link between IL-25 and MDSC mobilization, and established Tpl2 as a key mediator of IL-25 induced signaling that contribute to the MDSC recruitment. In addition, during P. acnes/LPS-induced FH pathogenesis, Tpl2 seemed specifically mediate IL-25-induced expression of CXCL1/2, but not affected the induction of CCL17, a previously reported MDSC-attracting chemokine that could be induced

by IL-25 administration in D-Gal/LPS-induced FH mice (15), suggesting Tpl2 modulated IL-25-induced chemokine expression in a context-dependent manner.

Our previously study has suggested that Tpl2 critically regulate IL-17A-induced signaling in astrocytes to mediate autoimmune inflammation, here we also demonstrated that Tpl2 is a key modulator in IL-25-induced signaling in hepatocyte to restrain hepatitis. This functional controversy may be due to Tpl2 regulates the function of different cells upon different stimulus, and suggest Tpl2 have a dual role in promoting or restraining inflammatory processes in a context-dependent manner.

In conclusion, our findings demonstrated that Tpl2 effectively attenuated the severity of acute liver injury and increased the survival rate of FH mice. Mechanistically, Tpl2 functioned in hepatocytes to mediate IL-25-induced CXCL1/2 chemokines, which promoted the recruitment of MDSC to suppress Th1 mediated local inflammation, resulting in the amelioration of FH (**Figure 8**). Our data not only highlighted a novel function of Tpl2 in mediating IL-25 signaling, but also raised the possibility to develop Tpl2-based therapeutic strategies against this dreaded disease.

#### MATERIALS AND METHODS

#### Mice

Tpl2-deficient mice (C57BL/6 background) were described as previously (24). The Tpl2+/<sup>−</sup> mice were bred to generate age-matched Tpl2−/<sup>−</sup> (Tpl2-KO) and Tpl2+/<sup>+</sup> (WT) mice. The Il25-defcient mice (C57BL/6 background) were provided by Dr. Y. Qian (Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences). In some experiments, Tpl2−/<sup>−</sup> mice were crossed with Il25−/<sup>−</sup> mice to generate Tpl2+/−Il25+/<sup>−</sup> mouse, which were then bred to generate age-matched Tpl2+/+Il25+/+, Tpl2−/−Il25+/+, Tpl2+/+Il25−/−, and Tpl2−/−Il25−/<sup>−</sup> mice. Rag1−/<sup>−</sup> mice (NM-KO-00069) were purchased from Shanghai Model Organisms Center. Mice were maintained in a specific pathogen-free facility, and all animal experiments were in accordance with protocols approved by the institutional Biomedical Research Ethics Committee, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences.

#### Induction of FH Mouse Model

For the induction of FH model, the age- and sex-matched mice were intravenously injected with 0.5 mg of heat-killed P. acnes suspended in 200 µl of phosphate-buffered saline (PBS) and after 7 days mice were injected intravenously with 1 µg of LPS and were monitored the survival rate. In some experiments, the WT and Tpl2-deficient mice were intravenously injected with 200 µg anti-Ly6G antibody for 3 times to deplete MDSC in vivo, or the Rag1-KO mice were adoptively transferred with WT or Tpl2 deficient CD4<sup>+</sup> T cells, or the lethally irradiated mice that were reconstituted with WT or Tpl2-deficient bone marrows, and then these mice were injected with P. acnes/LPS to induce FH and were monitored the survival rate.

#### Antibodies and Reagents

APC conjugated anti-mouse CD4 (17-0041-83), PB conjugated anti-mouse CD4 (48-0042-82), PE-cy7 conjugated anti-mouse CD8 (25-0081-82), APC-cy7 conjugated anti-mouse CD11b (47- 0112-82), PB conjugated anti-mouse CD11c (48-0114-82), PerCP conjugated anti-mouse Ly6G (46-9668-82), FITC conjugated anti-mouse B220 (11-0452-85), PE conjugated anti-mouse CD45 (12-0451-83), APC conjugated anti-mouse IFNγ (17-7311- 82), PE conjugated anti-mouse TNFα (12-7321-82), PE-cy7 conjugated anti-mouse CD25 (25-0251-82), APC conjugated anti-mouse Foxp3 (17-5773-82), PE conjugated anti-mouse CD44 (12-0441-83), FITC conjugated anti-mouse CD62L (11- 0621-86), anti-mouse CD3 (16-0031-86), and anti-mouse CD28 (16-0281-86) antibodies were purchased from eBioscience. BrdU Flow Kits (559619) were purchased from BD Biosciences. Antimouse Ly6G (BE0075) antibody was purchased from Bioxcell. Alexa Fluor 488-conjugated anti-Rat IgG secondary antibody (A-21210) was from Thermo Fisher. Mouse anti-CD4 (L3T4, 130-049-201) and mouse anti-Ly-6G (130-092-332) Micro Beads were purchased from Miltenyi Biotec. Murine IL-25 (1399) were purchased from R&D. Lipopolysaccharides (LPS, L3129) were purchased from Sigma. P. acnes were prepared as previously described (7).

#### Flow Cytometry

The infiltrated immune cells from WT and Tpl2-deficient inflamed livers were prepared through 33% Percoll gradient as previously described (6). The collected liver-infiltrated immune cells or splenic cell suspensions were stained with the indicated antibodies and were subjected to flow cytometry analyses as previously described by using a Beckman Gallios flow cytometer (39). For the intracellular staining of TNF-α, IFNγ, and Foxp3, the cells were fixed and permeabilized by fixation/permeabilization buffer (Thermo Fisher) before staining these antibodies, and then detected by flow cytometer. The absolute numbers of splenic and liver-infiltrating immune cells subpopulations were calculated based on their frequencies and the total number of isolated splenic and liver immune cells, and the data were presented as the average numbers of immune cell subpopulations per one spleen or liver of one mouse.

### Histology and Immunofluorescence Analysis

Liver specimens were fixed in 4% paraformaldehyde and paraffinembedded. Deparaffinized sections (8µm) were stained with hematoxylin and eosin. Semiquantitative analysis of the status of liver inflammation was performed in a blinded manner as previously described (40). Briefly, the H&E stained liver slides were scored by a pathologist in a "blinded fashion" to determine the degree of inflammatory condition as follows: 0 = no infiltration, 1 = minimal/slight infiltration, 2 = moderate infiltration, 3 = severe infiltration. For immunofluorescence staining, the frozen sections (10µm) from liver specimens were incubated with rat anti-mouse Ly6G (BE0075, Bioxcell) and were then labeled with Alexa Fluor 488-conjugated rabbit anti-rat IgG (A21210, Invitrogen), and the nuclei were stained by using DAPI (28718-90-3, Sangon Biotech).

#### Bone Marrow Chimeras

The bone marrow cells were prepared from WT or Tpl2-deficient mice and adoptively transferred into lethally irradiated (137Cs, γ-ray, 950 rad) WT or Tpl2-deficient mice (around 7-week-old; 10<sup>7</sup> cells/mouse) as previously described (41). The lethal-dose irradiation would eliminate the bone marrow and peripheral immune cells without affecting the radioresistant liver-resident cells, and the bone marrow chimeric mice would thus have their peripheral immune system reconstituted. After 8 weeks, the chimeric mice were applied for the indicated experiments.

#### Mouse Hepatocyte Isolation

The mouse primary hepatocytes were prepared as previously described (42). In brief, the livers were sequentially perfused with Earle's balanced salt solutions (EBSS) without Ca<sup>+</sup> and Mg<sup>+</sup> containing EGTA, EBSS with Ca<sup>+</sup> and Mg<sup>+</sup> containing Hepes, EBSS with Ca<sup>+</sup> and Mg<sup>+</sup> containing Hepes and Collagenase IV, and the liver cells were squeezed out to obtain cell suspension in DMEM medium, which were then applied for centrifugation over a mixture of 9 ml Percoll, 1 ml 10 × EBSS and 10 ml DMEM. The precipitated hepatocytes were suspended and cultured with DMEM complete medium in a collagen-coated culture dish. Cell viability was determined by using Trypan blue exclusion assay.

#### *In vivo* BrdU Incorporation

Seven days after P. acnes priming, 2 mg BrdU (559619, BD) in 200 µl PBS was intraperitoneally injected into WT or Tpl2-deficient mice. The mice were sacrificed 2 h after the BrdU administration, and the immune cell suspensions from livers or spleens were prepared for flow cytometric analysis.

#### T Cell Proliferation Assay

The WT CD4<sup>+</sup> T cells were purified by MACS sorting and were labeled with 5µM CFSE. The labeled cells were then seeded in the anti-CD3/CD28 antibodies-pre-coated plates and cocultured with MDSCs that isolated from WT or Tpl2 deficient bone marrow at the indicated ratio for 72 h. The cell proliferation was then determined by flow cytometry. In some experiment, WT or Tpl2-deficient CD4<sup>+</sup> T cells were seeded in anti-CD3/CD28 antibodies-pre-coated plates with 3 replicates and cultured for a total 72 h. The cell proliferation was recorded based on the [3H] thymidine labeling 8 h before examination.

#### RNA-Seq Analysis

Total RNA isolated from WT and Tpl2-KO hepatocytes stimulated with IL-25 were subjected to RNA-sequencing analysis. RNA sequencing was performed by BGI Tech Solutions. Transcriptomic reads from the RNA-Seq experiments were mapped to a reference genome (build mm 10) by using Bowtie. Gene expression levels were quantified by using the RSEM software package. Significant genes were defined by the p-value and false discovery rate of cutoff of 0.05 and fold changes ≥1.5. Differentially expressed genes were analyzed by the IPA and DAVID bioinformatics platform.

#### Quantitative RT-PCR

Liver tissues or cell samples were homogenized in Trizol reagent (Invitrogen). The cDNA was synthesized from 500 ng of extracted total RNA using M-MLV Reverse Transcriptase kit (Takara) according to the manufacturer's instructions. Quantitative PCR was performed with SYBR-Green premix ExTaq (Roche) and detected by a Real-time PCR System by using gene-specific primers. Gene expression was assessed in triplicate and normalized to a reference gene, β-actin. The gene-specific PCR primers are listed in **Supplementary Table 1**.

#### Quantification and Statistical Analysis

Statistical analyses were measured by GraphPad Software. Except where otherwise indicated, all the presented data are representative results of at least three independent repeats. Data are presented as mean ± SD, and the P-values were determined by two-tailed Student's t-tests. The P-values <0.05 were considered statistically significant.

#### REFERENCES


## DATA AVAILABILITY

The RNA-Sequencing data have been deposited into the Gene Expression Omnibus (accession code GSE125764). All other data supporting the findings of this study are available from the corresponding author on reasonable request.

#### ETHICS STATEMENT

This study was carried out in accordance with the recommendations of animal protocols that approved by Biomedical Research Ethics Committee, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences. The protocol was approved by the Biomedical Research Ethics Committee, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences.

#### AUTHOR CONTRIBUTIONS

JX designed and performed the experiments, prepared the figures, and wrote part of the manuscript. SP, YW, and JL contributed to part of the experiments. YQ provided the Il25-deficient mice. MH and YZ supervised the work and contributed to data analysis. YX designed and supervised the work, prepared the figures, and wrote the manuscript.

## FUNDING

We thank Dr. Shao-Cong Sun (Department of Immunology, MD Anderson Cancer Center, The University of Texas) for providing the Tpl2-deficient mice. This research was supported by the grants from the National Natural Science Foundation of China (81770567 and 81571545), the National Key R&D Program of China (2018YFA0107201, 2018YFA0902700), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDPB10), the Key Arrangements of the Chinese Academy of Sciences (KFZD-SW-216), the Thousand Young Talents Plan of China, CAS Key Laboratory of Tissue Microenvironment and Tumor.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu. 2019.01980/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 © 2019 Xu, Pei, Wang, Liu, Qian, Huang, Zhang and Xiao. 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.

# The Management of Glucocorticoid Therapy in Liver Failure

Ran Xue1,2 and Qinghua Meng<sup>1</sup> \*

<sup>1</sup> Department of Critical Care Medicine of Liver Disease, Beijing You-An Hospital, Capital Medical University, Beijing, China, <sup>2</sup> Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education), Department of Gastrointestinal Oncology, Peking University Cancer Hospital & Institute, Beijing, China

Liver failure is characterized by rapid progression and high mortality. Excessive systemic inflammation is considered as the trigger of liver failure. Glucocorticoids (GCs) can rapidly suppress excessive inflammatory reactions and immune response. GCs have been applied in the treatment of liver failure since the 1970s. However, until now, the use of GCs in the treatment of liver failure has been somewhat unclear and controversial. New research regarding the molecular mechanisms of GCs may explain the controversial actions of GCs in liver failure. More results should be confirmed in a larger randomized clinical trial; this can aid the discovery of better definitions in terms of treatment schedules according to different clinical settings. Meanwhile, the timing and dosing of GCs in the treatment of liver failure should also be explored.

#### Edited by:

M. Victoria Delpino, CONICET Institute of Immunology, Genetics and Metabolism (INIGEM), Argentina

#### Reviewed by:

Oscar Bottasso, National University of Rosario, Argentina Carolina Demarchi Munhoz, University of São Paulo, Brazil

#### \*Correspondence:

Qinghua Meng meng\_qh@126.com

#### Specialty section:

This article was submitted to Inflammation, a section of the journal Frontiers in Immunology

Received: 16 April 2019 Accepted: 04 October 2019 Published: 23 October 2019

#### Citation:

Xue R and Meng Q (2019) The Management of Glucocorticoid Therapy in Liver Failure. Front. Immunol. 10:2490. doi: 10.3389/fimmu.2019.02490 BACKGROUND

Keywords: glucocorticoids, liver failure, timing, dosing, mechanism

Liver failure is a life-threatening clinical syndrome with heterogeneous etiology that can cause serious disorders, such as coagulation disorders, icteria, hepatic encephalopathy (HE), and ascites (1, 2). Despite significant advances in artificial liver support system (ALSS) and liver transplantation (LT), these techniques are still difficult to apply more widely due to many restrictions, such as the amount of plasma, the limitation of liver donors, and the patient's economic situation, and so the mortality of liver failure is still high (3–5). It is therefore essential to develop more effective therapies for liver failure.

Glucocorticoids (GCs) have been applied to the clinical treatment of liver failure for many years. The first paper on GCs therapy for liver failure was published in the 1960s. Nowadays, many basic and clinical studies have explored the feasibility of GCs treatment in liver failure (6–12), but they remain inconclusive for the application of GCs treatment in liver failure.

## THE APPLIED STATUS OF GCS THERAPY IN LIVER FAILURE

Among the different liver diseases, the most authoritative clinical indication of GCs therapy is autoimmune hepatitis (AIH) (7). However, in patients with suspected drug-induced AIH who are undergoing GCs therapy, withdrawal of treatment once the liver injury has resolved should be accompanied by close monitoring (13). A recent report from APASL ACLF Research Consortium Working Party defined the histopathological, clinical spectrum, and role of GCs therapy in patients with AIH-ACLF. It was shown that early stratification to LT or GCs therapy (hepatic encephalopathy in ≥F3, MELD>27) would improve outcomes and reduce ICU stay in patients with AIH-ACLF (14).

GCs therapy is also recommended as a first-line treatment strategy in patients with severe alcoholic hepatitis, hepatic encephalopathy, or maddrey discriminant function ≥32 (6). Meanwhile, GCs would not increase occurrence of or mortality from bacterial infections in patients with severe alcoholic hepatitis (15). However, a recent meta-analysis showed that it could not determine whether GCs had a positive or negative effect on people with alcoholic liver disease because available data were still insufficient to produce robust results, trials were small, and the included participants differed in severity of disease (16).

Drug-induced liver failure requires evidence of immunopathogenicity to reverse the condition through GCs blocking immune responses. A recent study showed that short-term use of GCs was strongly recommended for severe DILI patients with hyperbilirubinemia (TBil >243 µmol/L) (17). However, Wan et al. found that prednisone was not beneficial for the treatment of severe drug-induced liver injury (18). The newest EASL clinical practice guidelines for drug-induced liver injury consider how GCs are often given when all else fails to procedure results (19). Early trials of GCs therapies, for all forms of ALF, demonstrated limited benefits (10, 20). GCs are also applied to treat drug-induced cholestatic hepatitis, especially in patients with allergic manifestations such as fever, eosinophilia, and rash. Liver injury caused by antiepileptic drugs are commonly related to features of hypersensitivity and may respond to GCs (21).

There exist significant differences in the etiology of liver diseases between the East and West. HBV is the leading cause of chronic liver disease in the Asia-Pacific region, including China and India (2). HBV-activated immune response and immune pathology caused by liver cell inflammation and necrosis are the initiated factors of liver failure. Although a large number of studies reported that GC therapy is effective in liver failure (22, 23), GC therapy is only recommended for the treatment of early stages of liver failure, and there is little evidence to support its effectiveness.

However, with the arrival of nucleoside analogs (NAs), more and more guidelines have recommended NAs to be used in patients with acute exacerbation of chronic HBV infection. The early combined use of NAs and GCs could be a good option to reverse the potential deterioration in patients with HBV-related liver failure. A recent study reported that early combination therapy with corticosteroid and NAs induces rapid resolution of inflammation in ALF due to transient HBV infection (24). It has been shown that with sufficient doses of NAs, GCs cannot affect the replication of HBV (12). However, Huang et al. (12) investigated retrospectively the efficacy of GCs in patients with hepatitis B virus-related acuteon-chronic liver failure (HBV-ACLF). It was indicated that GC treatment did not improve transplant-free survival in patients with HBV-ACLF.

It is not rare for GCs to be abused in the treatment of liver failure as "reduced bilirubin drugs." Therefore, its use in terms of liver failure therapy should not be exaggerated, although some patients with liver failure can indeed benefit from GCs therapy. As a "double-edged sword," the timing, dosage, and clinical indication of GC therapy are the key points to better definitions in terms of treatment schedules according to different clinical settings in the future.

## THE TIMING OF GC THERAPY IN LIVER FAILURE

In the Asia-Pacific region, the most common type of liver failure is HBV-ACLF. The clinical stage of HBV-ACLF can be divided into four stages: early stage of ascending period, late stage of ascending period, platform period, and recovery period. Immune injury is the major event in the early stage of ascending period. The pathogenesis in the late stage of ascending period is involved in ischemia, immune injury, and hypoxia injury (25). During the platform period, body conditions achieve an immunosuppression state.

Endotoxemia is an important factor during the initiation of liver injury. Recent studies have shown that there was an inflammatory cascade in the early period of HBV-ACLF (26, 27). The sooner systemic inflammatory response syndromes (SIRS) occurred, the higher the mortality rate would be. GCs can inhibit inflammation, stabilize the liver cell membrane, and prevent further necrosis of liver cells (28). Therefore, early application of GC therapy can inhibit immune responses. The inhibition of systemic inflammation delays rapid progression and improves the survival rate of patients with ACLF.

Zhao et al. (11) found that patients responding best to GCs were those with less severe liver failure and a higher risk of rapid disease progression, with lower HE grades and MELD scores but extremely high ALT levels. The optimal time of intervention with GCs was within 14 days of the onset of symptoms.

We consider that the efficacy of GC treatment is primarily associated with the timing of GC administration. Meanwhile, the first-time physician, age, basic condition, and complications should also be considered for GC administration. Patients with some specified indicators can benefit more from GC therapy; these could be indicators such as ALT>1,000 U/L, TBIL in the 10∼20 × ULN, PTA≥30%, MELD score<28, no obvious signs of infection, hepatic encephalopathy Stage<II, no liver and kidney syndrome trends, as well as overactive immunological responses. However, until now, there has been a lack of accurate quantitative indicators for GC therapy. Therefore, it is particularly important for doctors to accumulate more and more clinical experience.

#### THE DOSE SELECTION FOR GC THERAPY IN LIVER FAILURE

Today, the ideal choices regarding GC type and does remains inconclusive. Based on current clinical reports, GC dose is generally controlled in 1∼2 g/kg/d (methylprednisolone). Kotoh et al. (23) explored the feasibility of large doses of GC treatment for the treatment of liver failure. They divided 34 patients with ALF into two groups; 17 patients were given methylprednisolone

**Abbreviations:** ALF, acute liver failure; GCs, glucocorticoids; HBV, hepatitis B virus; HBV-ACLF, hepatitis B virus-related acute-on-chronic liver failure; HE, hepatic encephalopathy; INF-α, interferon-α; LPS, lipopolysaccharide; SALF, subacute liver failure; SIRS, systemic inflammatory response syndromes; TNF-α, tumor necrosis factor-α.

1,000 mg/d via hepatic artery continuously for 3 days. As a result, 13 patients were cured, 2 patients died, and 2 patients underwent LT without serious complications. Fujiwara et al. (29) discussed the value of high-dose GCs in the treatment of HBV-related liver failure. It was found that the survival rate and liver regeneration in the GC-treated group showed a slim advantage, but there was no statistical difference, while patients with HBV infection and a poor basic condition had an unfavorable prognosis.

When the efficacy of GCs therapy cannot be determined in clinic, it is required that possible side effects of GCs are kept within a controllable range based on the principle of safety. GCs can significantly inhibit the presence of phagocytic cells to the antigen, promote the destruction and disintegration of lymphocytes, and develop the removal of lymphocytes from blood vessels so as to reduce the number of lymphocytes in circulation (30). Small doses of GCs mainly inhibit cellular immunity, while high doses of GCs can suppress humoral immune function by inhibiting B cells and antibody production (31).

The number of liver surface glucocorticoid receptors (GRs) may be reduced in liver failure (27). If greater doses of GCs are given, the GCs cannot play a role during the presence of receptor saturation, but may increase the incidence of side effects of GCs. Therefore, patients with liver failure, especially those with cirrhosis, are not recommended to use high-dose GCs. Although GCs can increase the incidence of infection and upper gastrointestinal bleeding, as well as other complications, the side effects of GCs are controllable. Therefore, it is essential to screen and monitor the side effects during GCs therapy in patients with liver failure.

## THE MECHANISM OF THE POTENTIAL BENEFIT OF GC THERAPY IN LIVER FAILURE

#### The Core Pathogenesis of Liver Failure

Currently, it is widely accepted that "endotoxin-macrophagecytokine storm" is the core pathogenesis of liver failure, combined with the immune injury as the initial factor in the development of liver failure, especially in the early stage of liver failure (27, 32).

The chemical essence of endotoxin is lipopolysaccharide [LPS, recognized by the pattern-recognition receptor toll-like receptor 4 (TLR4)] (33). With the interaction of LPS-binding proteins, it binds to a variety of cell membranes with receptor CD14, transmitting signals from the outside of the cell to nucleus and stimulating the synthesis and release of cytokines, which involves tumor necrosis factor-α (TNF-α), interferon-α (INF-α), IL-1, and IL-6 and simultaneously induces macrophages to secrete nitric oxide and large amounts of oxygen-free radicals (34– 37). The liver cells are injured by delayed type hypersensitivity, oxidative stress, and apoptosis. If the immune response cannot be suspended in time, it would lead to a vicious cycle, resulting in significant liver cell necrosis, apoptosis, and liver failure (38–40). Peripheral blood mononuclear cells (PBMCs) and monocytes from patients with cirrhosis respond stronger to LPS stimulation (41). Heat shock proteins (HSPs) are well-known as protective proteins that make cells resistant to stress-induced cell damage. However, simultaneous activation of TLR4 by HSPs causes enhanced tissue injury (42).

Immune injury is considered as the first blow in the "triple hit theory" of liver failure, and timely suspension of its excessive immune response may reduce or even reverse its condition (43, 44). As the most commonly used anti-inflammatory and immunosuppressive agents, GCs can inhibit macrophage phagocytosis and antigen treatment and suppress the production of inflammatory cytokines. Therefore, GCs have the theoretical basis for the treatment of liver failure.

#### The Anti-inflammatory Mechanisms of GCs

Aside from rapid non-genomic effects, GCs exert genomic effects by binding to the glucocorticoid receptor (GR), a member of the nuclear receptor family of transcription factors (45). Upon ligand binding, the GR translocates to the nucleus, where it acts either as a monomeric protein that affects transcription with other transcription factors or as a homodimeric transcription factor, which binds glucocorticoid response elements (GREs) in promoter regions of GC-inducible genes (46). Some reports have clearly showed that GR dimerdependent transactivation is essential in the anti-inflammatory activities of GR (47–49). GRdim/dim mutant mice were used to show reduced GR dimerization, and hence GC cannot control inflammation (50, 51).

#### THE POTENTIAL MECHANISM FOR CONTROVERSIAL ACTIONS OF GC THERAPY IN LIVER FAILURE

#### The Pro-inflammatory Mechanisms of GC

Emerging studies have shown that GCs have a two-way regulation for inflammatory and immune responses (52). The basal state of the immune system and the type of exposure to GCs are significant factors influencing the effects of GCs (53). For instance, while chronic exposure to GCs seems to be immunosuppressive, acute exposure increases the peripheral immune response (54).

It was found that GCs can induce the expression of several innate immune-related genes, including several members of the Toll-like receptor (TLR) family, such as TLR2 and TLR4 (55–57). The activation of TLRs via the repression of NF-κB and AP-1 or via the induction of GC-induced leucine zipper (GILZ) or MKP-1 is a hallmark feature of inflammation (55).

The GR signaling interplays with the TLRs signaling pathway via several mechanisms (58). Hermoso et al. (59) found that dexamethasone increased TNF-α induction of TLR2 through the activation of GR, supporting the existence of positive feedback between the activation of the TLR signaling pathway and GC secretion. Meanwhile, GCs may exert pro-inflammatory actions through interactions with inflammatory cytokines such as TNF-α and acute phase protein serpinA3 (60). Besides, some studies indicated that GCs can work synergistically with proinflammatory mediators to enhance the defense mechanisms to ensure removal and clearance of pathogens in the hepatic acutephase response (58, 61). GC-mediated activation of NLRP3, TLR2, and P2Y2R and the potentiation of LIF and TNF-α regulated pro-inflammatory genes (58, 62). All these results provide a potential explanation for the controversial actions of GC therapy in liver failure. More studies are required to characterize the liver-specific effects of the anti- and proinflammatory roles of GR signaling.

#### GC Resistance (GCR)

There are two types of GCR, inherited or familial GCR and acquired GCR (63). It is accepted that a pro-inflammatory environment can negatively affect GR sensitivity (64, 65). The mechanisms contributing to reduced GC responsiveness are heterogeneous as they involve various cytokines and cell types. The mechanisms of GCR are still unclear. As GCR occurs in many inflammatory diseases, it is widely considered that GCR is a heterogeneous phenomenon with multiple underlying mechanisms (66). Some of these involve problems with the GR protein itself, but many others are independent from GR and involved in mutations in GR-induced genes and problems with chaperones or cofactors (63).

Meanwhile, the down-regulation of the GR protein is associated with GRC, involving many different mechanisms such as reduced transcription and homologous down-regulation (67), GR protein degradation (68), and decreased stability of GR mRNA (the involvement of AUUUA motifs in the 3′ UTR of GR mRNA) (69). Moreover, post-translational modifications of GR also contribute to a reduced GC response, such as ubiquitination of the GR (Lys-426 within a PEST element) and phosphorylation of the GR (68, 70). Besides, some research also showed miRNAs have a prominent role in the regulation of GR mRNA turnover and the occurrence of GCR (71, 72).

In addition to the non-genomic and genomic actions of GCs, GR signaling also relies on the existence of post-translational modifications (PTMs) and multiple receptor isoforms. GR is transcribed from a single gene, NR3C1; however, alternative splicing of this gene generates GRα and GRβ isoforms (73). The GRβ isoform also participates in the GCR. It was found that upregulation of the dominant negative GRβ isoform was correlated with GR insensitivity via inhibiting GR-induced transactivation and GR nuclear translocation (74).

#### The Possible Factors of GC Refractoriness in Liver Failure: Sepsis

Sepsis is a common complication of ACLF, which is an acute systemic inflammatory disease (75). However, GCs are hardly useful in sepsis (63). Thus, sepsis is considered a GCR disease. GCR is an essential problem in sepsis and leads to: lack of

#### REFERENCES

1. Ran X, Zhonghui D, Haixia L, Li C, Hongwei Y, Meixin R, et al. A novel dynamic model for predicting outcome in patients with hepatitis B virus related acute-on-chronic liver transport and removal of bile acids in the liver, resulting in cholestasis; increased production and reduced removal of L-lactate, resulting in lactic acidosis; GCs having no antiinflammatory effects.

Our previous study proposed that the diagnostic criteria of sepsis are not suitable for patients in HBV-ACLF with sepsis, because patients with underlying chronic liver disease and cirrhosis may have deranged clinical parameters (76). Therefore, it is essential to establish compatible diagnostic criteria for sepsis in patients with ACLF. When sepsis occurred, the serum TBiL level and WBC count elevated significantly while PLT count decreased significantly. We argue that when sepsis occurs during the process of liver failure GCs are not recommended for patients.

#### CONCLUSIONS AND FUTURE PERSPECTIVES

The idea of using GCs during acute liver failure has circulated for so many years, but, so far, no meaningful work has provided conclusive evidence of its therapeutic efficacy, except in the field of autoimmune etiology. Beyond the east/west demarcation, current data availing GC's use in liver failure revealed benefits that appeared marginal and were no longer present upon adjustment (10), came from evidence recorded in non-randomized studies (22), or were other ones carried out in small groups of patients (23, 24). More results should be confirmed using a larger randomized clinical trial to in order to arrive at better definitions in terms of treatment schedules according to different clinical settings.

Meanwhile, due to the complicated pathophysiology of liver failure, the exploration of immunological manifestations with different etiology and different clinical staging of patients with liver failure is needed urgently. This is a prerequisite for the feasibility and safety of GC applications. With an in-depth study, we can find the accurate timing, dosage, and clinical indicators of GC therapy for the clinical management of liver failure, so that clinicians can make timely treatment options so as to obtain the greatest benefits for patients.

#### AUTHOR CONTRIBUTIONS

RX wrote this manuscript. QM designed this manuscript.

#### FUNDING

This study was funded by the National Natural Science Foundation of China (No. 81470877).

failure. Oncotarget. (2017) 8:108970–80. doi: 10.18632/oncotarget. 22447

2. Bernal W, Jalan R, Quaglia A, Simpson K, Wendon J, Burroughs A. Acute-on-chronic liver failure. Lancet. (2015) 386:1576– 87. doi: 10.1016/S0140-6736(15)00309-8


neuroinflammation and neurodegeneration. J Neuroinflammation. (2014) 11:166. doi: 10.1186/s12974-014-0166-7


**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 Xue and Meng. 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.

# Immunopathogenesis of Hepatic Brucellosis

#### Guillermo Hernán Giambartolomei and María Victoria Delpino\*

Instituto de Inmunología, Genética y Metabolismo (INIGEM), Universidad de Buenos Aires, Buenos Aires (UBA), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina

The hepatic immune system can induce rapid and controlled responses to pathogenic microorganisms and tumor cells. Accordingly, most of the microorganisms that reach the liver through the blood are eliminated. However, some of them, including Brucella spp., take advantage of the immunotolerant capacity of the liver to persist in the host. Brucella has a predilection for surviving in the reticuloendothelial system, with the liver being the largest organ of this system in the human body. Therefore, its involvement in brucellosis is practically invariable. In patients with active brucellosis, the liver is commonly affected, and the most frequent clinical manifestation is hepatosplenomegaly. The molecular mechanisms implicated in liver damage have been recently elucidated. It has been demonstrated how Brucella interacts with hepatocytes inducing its death by apoptosis. The inflammatory microenvironment and the direct effect of Brucella on hepatic stellate cells (HSC) induce their activation and turn these cells from its quiescent form to their fibrogenic phenotype. This HSC activation induced by Brucella infection relies on the presence of a functional type IV secretion system and the effector protein BPE005 through a mechanism involved in the activation of the autophagic pathway. Finally, the molecular mechanisms of liver brucellosis observed so far are shedding light on how the interaction of Brucella with liver cells may play an important role in the discovery of new targets to control the infection. In this review, we report the current understanding of the interaction between liver structural cells and immune system cells during Brucella infection.

#### Edited by:

Nesrin Ozoren, Bogaziçi University, Turkey

#### Reviewed by:

Kai Yang, Indiana University School of Medicine—Lafayette, United States Thorsten Maretzky, The University of Iowa, United States

> \*Correspondence: María Victoria Delpino mdelpino@ffyb.uba.ar

#### Specialty section:

This article was submitted to Microbes and Innate Immunity, a section of the journal Frontiers in Cellular and Infection Microbiology

Received: 27 August 2019 Accepted: 28 November 2019 Published: 20 December 2019

#### Citation:

Giambartolomei GH and Delpino MV (2019) Immunopathogenesis of Hepatic Brucellosis. Front. Cell. Infect. Microbiol. 9:423. doi: 10.3389/fcimb.2019.00423 Keywords: liver, fibrosis, inflamación, hepatocite, stellate cells (Ito cells)

## BRUCELLOSIS

Brucellosis is a common zoonotic disease distributed around the world. The disease is rare in industrialized countries because of the routine screening of domestic livestock and animal vaccination program (Pappas et al., 2006; Mancini et al., 2014). Recently, case studies and review of published cases have been carried out to determine the prevalence of brucellosis in endemic regions. However, it should be borne in mind that because many areas with endemic brucellosis have poor infrastructure, it is likely that the incidence of the disease is underestimated. There are a reported 500,000 incident cases of human brucellosis per year. However, true incidence is estimated to be 5,000,000 to 12,500,000 cases annually (Godfroid et al., 2013; Hull and Schumaker, 2018). The clinical disease is common in the Middle East, Asia, Africa, South and Central America, the Mediterranean Basin, and the Caribbean (Pappas et al., 2006; Mancini et al., 2014; Cross et al., 2019).

At present, 12 species of Brucella genus have been described, each of which has its natural host, including domestic and farm animals, as well as wild animals, such as camels, bison, foxes, cetaceans, among others. Brucella is an expanding genus, and the most recent species were isolated from amphibians (Eisenberg et al., 2012; Fischer et al., 2012). The bacteria are transmitted from animals to human by ingestion of infected food products (meat or raw milk), direct contact with infected animals or their tissues, or inhalation of aerosols. Humans are accidental hosts, but brucellosis continues to be a major public health concern worldwide and is the most widespread zoonotic infection.

Brucella genus does not exhibit classic virulence factors, such as exotoxins, exoproteases, cytolysins, or other exoenzymes (Moreno and Moriyon, 2002). Observed tissue harm is a result of inflammatory immune responses through the activation of host immune responses after recognition of brucellar antigens by immune receptors, such as Toll-like receptors (TLR) and inflammasomes (Campos et al., 2004, 2017; Giambartolomei et al., 2004; Zwerdling et al., 2008, 2009; García Samartino et al., 2010; de Almeida et al., 2011, 2013; Delpino et al., 2012; Gomes et al., 2013). Notwithstanding, with its intracellular lifestyle, Brucella limits exposure to innate and adaptive immune responses and leads the clinical manifestations of the disease and pathology. Brucella takes advantage of intracellular destruction by restringing fusion of Brucella-containing vacuoles with lysosomal compartments in a mechanism mediated by the type IV secretion system (T4SS) (Comerci et al., 2001). Also, Brucella inhibits the apoptosis of infected macrophages and prevents the development of adequate adaptive immune response by the inhibition of antigen presentation (de Figueiredo et al., 2015; Barrionuevo and Giambartolomei, 2019).

### CLINICAL FEATURES OF BRUCELLA LIVER INFECTION

The liver is the most commonly affected organ in patients with active brucellosis. Accordingly, clinical and biochemical records of liver involvement have been observed in up to 50% of patients with active disease (Colmenero et al., 1996). Histopathological analyses of liver biopsies from a large number of patients with brucellosis have revealed liver parenchyma lesions due to inflammation, including focal areas of cellular inflammation with minimal necrosis of liver cells or the presence of granulomas with different localizations in parenchymal tissue and portal space (Akritidis et al., 2007). The pathology report on liver granulomas in Brucella infection usually shows necrotizing granulomas with a peripheral halo of epithelioid cells, lymphocytes, and plasma cells, as well as polymorphonuclear infiltrate in the necrotic area (Colmenero Jde et al., 2002; Villar et al., 2002). There are differencesin the histological evaluation of liver manifestations in human brucellosis due to several causes. One of them is that most reports are retrospective and lack bacteriological confirmation. Also, previous reports with bacteriological confirmation did not always specify the species of Brucella involved. Moreover, not all of them used the same criteria to define granuloma (for example, presence of epithelioid cells or presence of giant cells) (Adams, 1976).

Liver biopsies also presented evidence of hepatitis (Akritidis et al., 2007). The liver conducts the metabolism of carbohydrates, proteins, and fats. Some of the enzymes and the end products of these metabolic pathways may be utilized as biochemical markers of liver dysfunction because they are very sensitive to any abnormality that takes place. Despite the abovementioned, the liver function is frequently normal in Brucella-infected individuals. The total serum bilirubin may be slightly increased, and the total serum protein, albumin, and globulin are usually normal. The most frequent abnormalities are shown by the increase in transaminases and alkaline phosphatase, although they are non-specific (Young, 1995; Madkour, 2001). In these patients, the presence of inflammatory infiltrates is also common, and most of them present parenchymal necrosis (Madkour, 2001; Akritidis et al., 2007).

## THE LIVER AS AN IMMUNE ORGAN

The liver is usually regarded only as a non-immunological organ involved mainly in metabolic, nutrient storage, and detoxification activities. However, it is a member of the immune system responsible for the production of acute-phase proteins, chemokines, cytokines, complement proteins, and it also carries diverse populations of resident immune cells (O'Farrelly and Crispe, 1999; Crispe, 2009; Nemeth et al., 2009). Notwithstanding, the spleen is the main critical mediator in the clearance of blood pathogens, which overlaps with the function of the liver (Robinson et al., 2016).

Lymphocyte populations are present in the parenchyma and the portal tracts of the liver. These populations consist of conventional and unconventional lymphocytes of innate immunity, NKT and NK cells, as well as cells of the adaptive immune system, T and B cells (Freitas-Lopes et al., 2017). Resident antigen-presenting cells are also abundant in the liver. These cells are able to capture antigens that enter through the liver or are released by dead or infected hepatocytes. The group of resident antigen-presenting cells comprises Kupffer cells (members of the reticuloendothelial system) (Gale et al., 1978), liver sinusoidal endothelial cells (LSEC) (a particular type of vascular endothelial cells) (Steffan et al., 1986), and dendritic cells (DC) (Prickett et al., 1988; Lau and Thomson, 2003). The presence of these cells is essential for the maintenance of liver tolerance in physiological conditions (Robinson et al., 2016).

Kupffer cells are the liver resident macrophages that adhere to sinusoidal endothelial cells inside the sinusoids (Smith, 2013). This localization is adequate to perform its function as a scavenger removing protein complexes, small particles, senescent red blood cells, and cell debris from the portal blood via pattern recognition receptors (PRRs) (Petrasek et al., 2012).

LSEC are specialized endothelial cells that by the absence of diaphragm and lack of basement membrane are the most permeable endothelial cells in the body. In physiological conditions, these cells maintain hepatic stellate cell (HSC) quiescence inhibiting fibrosis development. These cells have a high phagocytic capacity and the molecules that promote antigen processing and presentation with efficacy similar to that of DC (Steffan et al., 1986; Lohse et al., 1996; Knolle et al., 1999).

Resident hepatic DC are located around the central veins and portal tracts. The presence of interleukin (IL)-10 and transforming growth factor (TGF)-β secreted by Kupffer and LSEC cells in the absence of infection provides a microenvironment capable of generating tolerant resident DC. The activation of these cells increases their capacity to migrate via the space of Disse to the lymphatic vessels in the portal tracts and then to the extrahepatic lymph nodes (Matsuno et al., 1996; Kudo et al., 1997).

Additionally, HSC and hepatocytes contribute to the immune homeostasis of the liver. Quiescent HSC exert immunoregulatory roles secreting chemokines, chemokine receptors, macrophage inflammatory proteins (MIPs), and TLR and also function as antigen-presenting cells (Friedman, 2008; Hernandez-Gea and Friedman, 2011). Hepatocytes can also participate as antigenpresenting cells, but they do not express the costimulatory molecules CD80 and CD86 (Bertolino et al., 1998). Therefore, hepatocytes induce T-cell functional activation but fail to promote survival (Bertolino et al., 1998). In addition, hepatocytes can also express PD-L1 in response to interferon (IFN) of type I and II with the consequent induction of T-cell apoptosis contributing to liver tolerance (Mühlbauer et al., 2006). This immunotolerant capacity of the liver is due to its structure, with resident immune cells in constant stimulation and the hepatic blood supply that creates a unique cytokine and growth factor milieu. This microenvironment determines the balance between tolerance and inflammation in the healthy liver. The complex of cytokine milieu in adult liver, in the absence of pathological inflammation, includes basal expression of proinflammatory cytokines IL-15, IL-7, IL-2, IL-12, IFN-γ, and the anti-inflammatory cytokines IL-10, IL-13, and TFG-β (Golden-Mason et al., 2004; Kelly et al., 2006). The tolerogenic environment is maintained by regulatory myeloid populations, such as myeloid-derived suppressor cells (MDSC), that mediate their suppressive activity through the production of IL-10 and TGF-β (Gabrilovich and Nagaraj, 2009). The transmigration of monocytes leads to MDSC differentiation and activation, contributing to the immunoregulatory capacity of the liver (Sander et al., 2010; Zimmermann et al., 2015). In particular, the impact of IL-10 in Brucella persistence and establishment of chronic infection through the modulation of macrophages has been demonstrated previously by using IL-10-deficient mice (Xavier et al., 2013).

Nevertheless, this immune tolerance, the hepatic immune system is able to induce rapid and controlled responses to pathogenic microorganisms and tumor cells (Robinson et al., 2016). Accordingly, most of the microorganisms that reach the liver through the blood are eliminated. Although the liver has several mechanisms to resist and to eliminate infectious agents, some of them, such as Brucella spp, take advantage of the immunotolerant capacity of the liver to escape from the immune response and persist in the host. Besides, because the liver is the organ with the largest number of resident macrophages (Heymann and Tacke, 2016), we can speculate that the liver constitutes a place for Brucella persistence.

#### BRUCELLA AND HSC IN FIBROSIS

HSC reside in the liver between the hepatocytes and the small blood vessels. These cells are characterized as containing intracellular lipid droplets and protrusions that extend around the blood vessels. During liver damage, these cells are activated and secrete collagen with the formation of scar tissue, leading to chronic fibrosis or cirrhosis (Xu et al., 2012).

Brucella is an infectious stimulus that causes HSC activation that involves the conversion of quiescent cells into myofibroblasts, as revealed by the increase in α-smooth muscle actin (SMA) expression, the increase in collagen secretion, the inhibition of matrix metalloproteinase (MMP)-9 secretion, the induction of the tissue inhibitor of metalloproteinases (TIMP)-1, and the secretion of the master regulator of fibrosis TGF-β (Arriola Benitez et al., 2013). TGF-β was identified as a main driver of HSC activation, and several approaches targeting the TGF-β signaling pathway were successfully used to engage fibrosis in animal models of chronic liver diseases (Dooley and ten Dijke, 2012; Puche et al., 2013).

HSC secrete the chemokines IL-8 and the monocyte chemotactic protein (MCP)-1 in response to Brucella abortus infection that could attract to the site of infection, neutrophils and monocytes, respectively (Arriola Benitez et al., 2013). It has been demonstrated that B. abortus-infected monocytes could not reverse the fibrotic phenotype in B. abortus-infected HSC. However, B. abortus-infected monocytes inhibit collagen deposition and induce MMP-9 secretion in uninfected HSC (Arriola Benitez et al., 2013). This may explain, at least in part, why many patients with brucellosis have an inflammatory infiltrate, but only some of them develop cirrhosis. Moreover, in patients with brucellosis and cirrhosis, other causes such as viral hepatitis or alcoholic cirrhosis, have not been ruled out (Madkour, 2001).

Until now, it is unclear that Brucella infection can directly or indirectly modulate HSC activity and, consequently, the deposition of extracellular matrix. As in other liver diseases, HSC activation during B. abortus infection could contribute to granuloma formation by laying down a ring of collagen to encapsulate the granuloma (Chuah et al., 2014) (**Figure 1**).

### BRUCELLA TYPE IV SECRETION SYSTEM IN FIBROSIS

T4SSs are a multiprotein complex involved in the translocation of nucleoproteins and/or protein substrates across the bacterial cell envelope to the host cell (Zechner et al., 2012). In Brucella, T4SS is encoded by the virB operon, which consists of 12 genes (virB1-12) located on chromosome II. The transcription of the virB operon is controlled by the promoter upstream of virB1 (O'Callaghan et al., 1999; Sieira et al., 2000). T4SS protein substrates have been shown to modulate several cellular processes in the host cell, such as apoptosis, vesicle trafficking, ubiquitination, and

(IL)-8 (chemoattractant of monocytes and neutrophils, respectively) in response to B. abortus infection. (2) Resident macrophages or monocytes attracted to the site of infection could be infected by B. abortus and then modulate the expression of collagen and MMP-9 depending on the status of HSC (infected or not). In non-infected HSC, B. abortus-infected monocytes induce an inflammatory phenotype characterized by the increase in MMP-9 expression and the inhibition of collagen deposition. In contrast, B. abortus-infected monocytes induce the inhibition of matrix metalloproteinase (MMP)-9 secretion and the increase in collagen deposition in infected HSC. (3) B. abortus infection of HSC induces HSC activation by inducing a fibrotic phenotype characterized by the expression of α-smooth muscle actin (α-SMA), collagen deposition, and inhibition of MMP-9 through a mechanism that is dependent on the increase in transforming growth factor (TGF)-β and tissue inhibitor of metalloproteinase (TIMP)-1. (4) Fibrosis induction by B. abortus infection was dependent on the Type IV secretion system (T4SS) and its secreted effector BPE005 in a mechanism that depends on the cyclic AMP (cAMP)/protein kinase A (PKA) signaling pathway. (5) Fibrosis induction is also dependent on autophagy pathway induction characterized by the increase in the expression of LC3II and Beclin-1 and the inhibition of p62 expression. (6) Then, B. abortus infection induces cell death by apoptosis of activated HSC in a mechanism that is dependent on a functional T4SS and caspase 3 cleavage.

so on (Ninio and Roy, 2007; Franco et al., 2009). In Brucella, T4SS has been shown to be involved in the modulation of the immune system during infection (Roux et al., 2007; Rolán and Tsolis, 2008; Gomes et al., 2013); Brucella protein effector (BPE)005 is particularly involved in liver fibrosis (Arriola Benitez et al., 2016). The predicted structure BPE005 suggests that it might have an effect on cyclic AMP (cAMP) signaling pathways (Marchesini et al., 2011). In the HSC, BPE005 may mediate its effect by blocking the binding between cAMP and protein kinase A (PKA) with the concomitant fibrosis induction, as demonstrated in studies performed in vitro (Arriola Benitez et al., 2016). Interestingly, the role of BPE005 in the modulation of the fibrotic phenotype during B. abortus infection was confirmed in an in vivo model in mice infected with a B. abortus BPE005 mutant. The histological analysis by Masson's trichrome staining revealed that the level of fibrotic patches is lower in mice infected with BPE005 mutants than in those infected with the wild-type counterpart. Accordingly, levels of collagen and TGFβ were lower in mice infected with the B. abortus BPE005 mutant (Arriola Benitez et al., 2016).

The role of the cAMP/PKA signaling pathway in the liver has been reported previously. The involvement of this pathway was demonstrated in various metabolic functions with the main effect in the facilitation of carbohydrate and lipid metabolism. cAMP plays a major role in the modulation of the HSC function by inhibiting profibrogenic pathways in HSC (Lopez-Sanchez et al., 2014).

Autophagy is involved in the fibrotic response during chronic hepatic lesion caused by hepatitis virus infection, alcohol abuse, and nonalcoholic steatohepatitis (Song et al., 2014).

Autophagy is a catabolic intracellular pathway that targets defective or excessive organelles to the lysosomes for degradation into amino acids, free fatty acids, or other small molecules used for material recycling or energy harvesting (Mao and Fan, 2015). Autophagy, usually stimulated by energy restriction, stress, or inflammation, is regarded as a survival mechanism that plays a critical role in maintaining cellular homeostasis, which is involved in many human disorders, such as fibrotic disease (Yin et al., 2008). In fibrosis, autophagy is mostly a cell survival mechanism that attenuates hepatic inflammatory injury and ultimately induces liver fibrosis (Mao and Fan, 2015). In addition, and supporting the role of BPE005 in the induction of HSC activation to a fibrotic phenotype, it has been demonstrated that autophagy is involved in the fibrotic response due to B. abortus, which depends on a functional T4SS and its effector BPE005 (Arriola Benitez et al., 2018). Autophagy was revealed by the upregulation of the LC3II/LC3I ratio and Beclin-1 expression and by the inhibition of p62 expression in infected cells (Arriola Benitez et al., 2018). However, further studies are necessary to determine if BPE005 could be a therapeutic target because this factor induces a fibrogenic phenotype that could help not only the host in the response to the inflammatory damage but also in the formation of granulomas likely favoring the persistence of bacterium.

Of note, the presence of liver cirrhosis during hepatic brucellosis is a debatable issue (Madkour, 2001). In addition, in murine models, cirrhosis was not observed upon infection with B. abortus (Arriola Benitez et al., 2013). In line with previous observations, B. abortus infection induces the clearance of activated HSC by apoptosis through a mechanism that is dependent on a functional T4SS and involved caspase 3 cleavage, triggering the recovery of liver fibrosis (Arriola Benitez et al., 2018) (**Figure 1**).

#### BRUCELLA AND HEPATOCYTES

Hepatocytes are the main cells of the liver parenchyma tissue and occupy around 70–85% of the liver volume (Kmiec, 2001 ´ ). Hepatocytes are known to be involved in the synthesis of proteins, glycoproteins, cholesterol, bile salts, and phospholipids. On the other hand, they contribute to the detoxification and excretion of substances. However, these cells also participate in the immune response against pathogens. Hepatocytes respond to viral, bacterial, and parasitic infections by secreting proinflammatory cytokines and chemokines as mentioned elsewhere (Santos et al., 2005; McCord et al., 2006; Costa et al., 2008; Heydtmann and Adams, 2009), and B. abortus is not the exception. B. abortus infects hepatocytes and induces the secretion of IL-8, the main chemoattractant for neutrophil (Delpino et al., 2010), which correlates with the neutrophil infiltration observed in the liver of Brucella-infected patients (Hunt and Bothwell, 1967; Colmenero Jde et al., 2002). B. abortus infection also induces intercellular adhesion molecule (ICAM-1) and MMP-9 secretion by hepatocytes (Delpino et al., 2010). These molecules ensure the influx of neutrophils to the tissues. ICAM-1 could facilitate the interaction of hepatocytes with neutrophils, and MMP-9 promotes neutrophil transmigration through degradation of extracellular matrix. These neutrophils could be also infected by B. abortus, and in response to this infection, they can contribute to the inflammatory reaction through the secretion of IL-8 and MMP-9 by inducing the expression of ICAM-1 by hepatocytes (Delpino et al., 2010).

Presence of liver parenchymal necrosis has been demonstrated in patients with hepatic manifestations of brucellosis (Akritidis et al., 2007), and hepatocytes occupy most of the parenchymal tissue. The analysis of the effects of B. abortus infection on hepatocyte viability revealed the induction of cell death by apoptosis and cytotoxic release of lactate dehydrogenase (LDH) in a T4SS-dependent manner (Delpino et al., 2010). Neutrophils also contribute to the induction of apoptosis and LDH release by hepatocytes as revealed by the stimulation with culture supernatants from B. abortus-infected neutrophils (Delpino et al., 2010).

B. abortus-infected hepatocytes can also modulate HSC function. Supernatants from B. abortus-infected hepatocytes induce MMP-9 secretion and inhibit collagen deposition in LX-2 cells (Arriola Benitez et al., 2013). However, when stimulation was performed on B. abortus-infected HSC, supernatants from B. abortus-infected hepatocytes were unable to reverse the inhibitory effect of B. abortus infection in the inhibition of MMP-9 secretion and the induction of collagen deposition on HSC (Arriola Benitez et al., 2013). The status of HSC (infected or not) defines if B. abortus-infected hepatocytes induce a fibrotic or an inflammatory phenotype in these cells (**Figure 2**).

#### IN VIVO STUDIES FOR LIVER BRUCELLOSIS

The absence of a suitable animal model that can reproduce, after experimental infection, the variety of disease manifestations of human brucellosis has determined the slow progress in most of the pathobiology of focal forms of the disease. However, while laboratory rodents do not mimic all of the spectrum of clinical signs and symptoms of human disease, there are focal manifestations of liver brucellosis (Olsen and Palmer, 2014). This allows characterizing at least in part, the pathophysiological and immune manifestations of hepatic brucellosis.

Studies performed in mice infected with Brucella melitensis stated that most of the infected cells in the liver expressed F4/80 myeloid marker. In the peak of infection, granuloma formation occurs. Granulomas are mainly composed by CD11b+ F4/80+ MHC-II+ cells expressing iNOS/NOS2 enzyme (Copin et al., 2012). A population of these cells also expressed

CD11c marker, indicating the presence of inflammatory DC. However, further analyses are necessary to determine if Brucella will remain in these cells or will migrate to others that protect them from the immune response (Copin et al., 2012). The delay in conducting these studies is due to the fact that the murine model does not represent all of the clinical manifestations of the disease, and that the liver biopsy is not a medical practice for the diagnosis of brucellosis.

In vivo studies also revealed that IL-10 production by CD25+CD4<sup>+</sup> T cells that modulate macrophage function enhances bacterial survival and persistence in liver through the modulation of the balance between pro-inflammatory and anti-inflammatory cytokines (Xavier et al., 2013). In addition, liver histopatology from infected IL-10−/ <sup>−</sup> mice at 3 weeks revealed multifocal granulomas and liver necrosis, as occurs in wild-type mice. However, at 6 weeks postinfection, the number of granulomas was reduced in IL-10−/ <sup>−</sup> mice with respect to wild-type controls. This reduction in liver pathology was accompanied by the increase of CD4+CD25<sup>+</sup> foxp3<sup>+</sup> T cells with the expression of TGF-β (Corsetti et al., 2013). We can speculate that this reduction in liver pathology despite the increase in TGFβ can be attributed to the reduction of bacterial number and the consequent reduction in liver damage. However, further studies would be necessary to determine if the TGF-β levels produced are capable of causing liver fibrosis in these mice.

In addition, intraperitoneal infection of mice with B. abortus allows corroborating in vitro findings, among them the role of BPE005 in the induction of liver fibrosis revealed by the increase in collagen deposition in liver and the expression of TGF-β in HSC, as well as the presence of lymphocytic inflammatory infiltrate (Arriola Benitez et al., 2013, 2016).

Despite the limitations of the rodent model, it allowed to corroborate the most relevant in vitro findings related to human clinical outcomes.

#### CONCLUDING REMARKS

Liver involvement in active brucellosis has ranged from 5 to 52% or more (Madkour, 2001). While the clinical and imaging aspect of hepatic brucellosis has been widely described, the molecular and cellular mechanisms involved in the damage and the immune response in the liver have only been partially clarified over the past 10 years. The findings summarized in this review attempt to answer the key questions about the soluble mediators, local and infiltrating cells involved in inflammatory damage during liver brucellosis.

The infection through the T4SS effector, BPE005, could modulate HSC activation to control the fibrotic process. It is likely that BPE005 could participate in granuloma formation that might act as a reservoir of bacteria contributing to the disease chronicity. Additionally, and in line with the presence of liver

#### REFERENCES


parenchymal necrosis observed in patients (Akritidis et al., 2007), B. abortus induces cell death of hepatocytes in a mechanism that also depends on a functional T4SS (Arriola Benitez et al., 2018). Altogether, this knowledge could frame the first approach toward the discovery of a new therapeutic target leading to new treatments that could be coadministrated with antibiotics aimed at reducing liver damage.

#### AUTHOR CONTRIBUTIONS

GG helped draft and correct the manuscript. MD drafted the manuscript.

#### FUNDING

This work was supported by grants PICT 2014-1111 and PICT 2015-0316 from Agencia Nacional of Promoción Científica y Tecnológica (ANPCYT, Argentina), by grant PIP 2015-2017- 0200 from Consejo Nacional de Investigaciones científicas y técnicas (CONICET).

molecular diagnostic approach. Diagn. Microbiol. Infect. Dis. 42, 159–167. doi: 10.1016/S0732-8893(01)00344-3


**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 Giambartolomei and Delpino. 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.

# *Brucella abortus* Infection Elicited Hepatic Stellate Cell-Mediated Fibrosis Through Inflammasome-Dependent IL-1β Production

Paula Constanza Arriola Benitez <sup>1</sup> , Ayelén Ivana Pesce Viglietti <sup>1</sup> , Marco Tulio R. Gomes <sup>2</sup> , Sergio Costa Oliveira<sup>2</sup> , Jorge Fabián Quarleri <sup>3</sup> , Guillermo Hernán Giambartolomei <sup>1</sup> and María Victoria Delpino<sup>1</sup> \*

#### *Edited by:*

Juarez Antonio Simões Quaresma, Instituto Evandro Chagas, Brazil

#### *Reviewed by:*

Leticia Cedillo-Barron, Center for Research and Advanced Studies (CINVESTAV), Mexico Rebecca Leigh Schmidt, Upper Iowa University, United States

> *\*Correspondence:* María Victoria Delpino mdelpino@ffyb.uba.ar

#### *Specialty section:*

This article was submitted to Microbial Immunology, a section of the journal Frontiers in Immunology

*Received:* 13 October 2019 *Accepted:* 11 December 2019 *Published:* 21 January 2020

#### *Citation:*

Arriola Benitez PC, Pesce Viglietti AI, Gomes MTR, Oliveira SC, Quarleri JF, Giambartolomei GH and Delpino MV (2020) Brucella abortus Infection Elicited Hepatic Stellate Cell-Mediated Fibrosis Through Inflammasome-Dependent IL-1β Production. Front. Immunol. 10:3036. doi: 10.3389/fimmu.2019.03036 1 Instituto de Inmunología, Genética y Metabolismo (INIGEM), Universidad de Buenos Aires, CONICET, Buenos Aires, Argentina, <sup>2</sup> Department of Biochemistry and Immunology, Institute of Biological Sciences, Federal University of Minas Gerais, Belo Horizonte, Brazil, <sup>3</sup> Instituto de Investigaciones Biomédicas en Retrovirus y Sida (INBIRS), Universidad de Buenos Aires, CONICET, Buenos Aires, Argentina

In human brucellosis, the liver is frequently affected. Brucella abortus triggers a profibrotic response on hepatic stellate cells (HSCs) characterized by inhibition of MMP-9 with concomitant collagen deposition and TGF-β1 secretion through type 4 secretion system (T4SS). Taking into account that it has been reported that the inflammasome is necessary to induce a fibrotic phenotype in HSC, we hypothesized that Brucella infection might create a microenvironment that would promote inflammasome activation with concomitant profibrogenic phenotype in HSCs. B. abortus infection induces IL-1β secretion in HSCs in a T4SS-dependent manner. The expression of caspase-1 (Casp-1), absent in melanoma 2 (AIM2), Nod-like receptor (NLR) containing a pyrin domain 3 (NLRP3), and apoptosis-associated speck-like protein containing a CARD (ASC) was increased in B. abortus-infected HSC. When infection experiments were performed in the presence of glyburide, a compound that inhibits NLRP3 inflammasome, or A151, a specific AIM2 inhibitor, the secretion of IL-1β was significantly inhibited with respect to uninfected controls. The role of inflammasome activation in the induction of a fibrogenic phenotype in HSCs was determined by performing B. abortus infection experiments in the presence of the inhibitors Ac-YVAD-cmk and glyburide. Both inhibitors were able to reverse the effect of B. abortus infection on the fibrotic phenotype in HSCs. Finally, the role of inflammasome in fibrosis was corroborated in vivo by the reduction of fibrotic patches in liver from B. abortus-infected ASC, NLRP, AIM2, and cCasp-1/11 knock-out (KO) mice with respect to infected wild-type mice.

Keywords: *Brucella*, inflammasome, fibrosis, hepatic stellate cells, IL-1β

## INTRODUCTION

Human brucellosis is a zoonosis that induces a chronic and debilitating disease caused by Brucella species that manifests itself with a broad clinical spectrum (1, 2). Liver involvement in human brucellosis is usually documented, given the well-characterized tropism of Brucella for the reticuloendothelial system (1, 2). The incidence of liver involvement in active brucellosis has ranged from 5 to 53% or more (2).

Inflammasome activation has been documented in several liver diseases. Accordingly, it has been postulated that the upregulation of IL-1β and IL-18 secretion leads to myofibroblast differentiation with concomitant increase of collagen and TGF-β expression (3). In addition, it was established that inflammasome components are present in hepatic stellate cells (HSCs) and could regulate their function (3). The consequences of activation of inflammasome pathway were also confirmed in vivo, demonstrating its key role in liver fibrosis (4).

The activation and release of IL-1β and IL-18 requires two distinct signals. TLR engagement by pathogen or endogenous signal induces the expression of the precursor forms of these cytokines (pro–IL-1β and pro–IL-18), after which NLRdependent activation of caspase-1 regulates their proteolytic processing and release (5). Activation of inflammasome by Brucella abortus infection has been previously demonstrated in bone marrow-derived macrophages and dendritic cells (6, 7). In these cells, B. abortus induces the secretion of IL-1β, in a process in which NLRP3 is necessary for activation of ASC inflammasome and the concomitant activation of caspase-1 and maturation and secretion of IL-1β (6, 7). In addition, ASC inflammasomes are also essential for IL-1β secretion induced by B. abortus infection in astrocytes and microglia (8). The first signal can be triggered by various pathogen-associated molecular patterns (PAMPs) via TLR activation. In the case of B. abortus infection inflammasome activation, the second signal involved the presence of a functional type 4 secretion system (T4SS) and DNA-sensing inflammasome receptor AIM2, in bone marrowderived macrophages, and Mal/TIRAP and TLR-2 are the main signaling involved in astrocytes and microglia (8).

Previously, we have demonstrated that upon infection of HSCs, B. abortus triggers a profibrotic response characterized by inhibition of MMP-9 secretion inducing concomitant collagen deposition and transforming growth factor (TGF) β1 secretion in a way that involves a functional T4SS and its effectors protein BPE005 (9). Taking into account that inflammasome has been documented to be necessary to induce activation to a fibrotic phenotype of HSCs, we hypothesized that Brucella infection might create a microenvironment that would promote inflammasome activation and concomitant profibrogenic phenotype in HSCs. The results of the study are presented here.

## MATERIALS AND METHODS

#### Bacterial Culture

Brucella abortus S2308 DsRed-expressing B. abortus S2308 or the isogenic B. abortus virB10 polar mutants were grown overnight in 10 ml of tryptic soy broth (Merck, Buenos Aires, Argentina) with constant agitation at 37◦C. Bacteria were harvested and the inocula were prepared as described previously (10).

To obtain heat-killed B. abortus (HKBA), bacteria were washed five times for 10 min each in sterile PBS, heat killed at 70◦C for 20 min, aliquoted, and stored at −70◦C until they were used. The total absence of B. abortus viability after heat killing was verified by the absence of bacterial growth on tryptose soy agar.

All live Brucella manipulations were performed in biosafety level 3 facilities located at the Instituto de Investigaciones Biomédicas en Retrovirus y SIDA (INBIRS).

## Cell Culture

LX-2 cell line, a spontaneously immortalized human HSC line, was kindly provided by Dr. Scott L. Friedman (Mount Sinai School of Medicine, New York, NY). LX-2 cells were maintained in DMEM (Life Technologies–Invitrogen, Carlsbad, CA, USA) and supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100µg/ml streptomycin, and 2% (v/v) fetal bovine serum (FBS; Gibco–Invitrogen, Carlsbad, CA, USA). All cultures were grown at 37◦C and 5% CO2.

## Cellular Infection

LX-2 cells were seeded in 24-well-plates and infected with B. abortus S2308, DsRed-expressing B. abortus S2308, or its isogenic mutants at multiplicities of infection (MOI) of 100 and 1000. After the bacterial suspension was dispensed, the plates were centrifuged for 10 min at 2,000 rpm and then incubated for 2 h at 37◦C under a 5% CO<sup>2</sup> atmosphere. Cells were extensively washed with DMEM to remove extracellular bacteria and incubated in medium supplemented with 100µg/ml gentamicin and 50µg/ml streptomycin to kill extracellular bacteria. LX-2 cells were harvested at different times to determine cytokine production, MMP secretion, and collagen deposition.

#### Neutralization Experiments

Neutralization experiments were performed with 5µM of Bay 11-7082, an inhibitory compound of the nuclear factor-κB (NFκB), 50µM of glybenclamide (glyburide), an inhibitor of the NLRP3 inflammasome, 50µM of general caspase inhibitor Z-VAD-FMK, or 50µM of caspase 1 inhibitor Y-VAD-FMK (all inhibitors were from Sigma-Aldrich). The cells were treated for 1 h with each inhibitor before infection.

AIM2 inflammasome complex formation was prevented using 3µM of A151, a DNA sequence that inhibits in a competitive manner the immunostimulatory DNA. A151 (5-TTAGGGTTAG GGTTAGGGTTAGGG-3) and the control C151 (5-TTCAA ATTCAAATTCAAATTCAAA-3) constructs were synthesized with a phosphorothioate backbone. To determine the implication of IL-1β, neutralization experiments were performed by adding 50 ng/ml of ANAKINRA, the inhibitor of IL-1 receptor, and the natural antagonist IL-1Ra (R&D Systems). Recombinant human IL-1β (rIL-1β, R&D Systems) at a concentration of 50 ng/ml was used as a positive control.

#### mRNA Preparation and Quantitative PCR

RNA from LX-2 cells was isolated using the Quick-RNA MiniPrepKit (Zymo Research) and 1µg of RNA was subjected to reverse transcription using Improm-II Reverse Transcriptase (Promega). PCR analysis was performed with StepOne real-time PCR detection system (Life Technology) using SYBR Green as fluorescent DNA binding dye. The primer sets used for amplification were as follows: β-actin sense: 5′ -AACAGTCCGC CTAGAAGCAC-3′ , β-actin antisense: 5′ -CGTTGACATCCG TAAAGACC-3′ ; NLRP3 sense: 5′ -CCACAAGATCGTGAGAA AACCC-3′ ; NLRP3 antisense: 5′ -CGGTCCTATGTGCTC GTCA-3′ ; IL-1β sense: 5′ -AGCTACGAATCTCCGACCAC-3′ ; IL-1β antisense: 5′ -CGTTATCCCATGTGTCGAAGAA-3′ ; ASC sense: 5′ -TGGATGCTCTGTACGGGAAG-3′ ; ASC antisense: 5′ - CCAGGCTGGTGTGAAACTGAA-3′ ; Capase-1 sense: 5′ -TTT CCGCAAGGTTCGATTTTCA-3′ Caspase-1 antisense: 5′ - GGCATCTGCGCTCTACCATC-3′ ; AIM2 sense: 5′ -TGGCAA AACGTCTTCAGGAGG-3′ ; AIM2 antisense: 5′ -AGCTT GACTTAGTGGCTTTGG-3′ .

The amplification cycle for Caspase-1, ASC, and β-actin was 95◦C for 15 s, 58◦C for 30 s, and 72◦C for 60 s; the amplification cycle for IL-1β, NLRP3, and AIM2 was 95◦C for 15 s, 61◦C for 30 s, and 72◦C for 60 s. All primer sets yielded a single product of the correct size. Relative expression levels were normalized against β-actin.

#### Immunofluorescence

LX-2 cells were infected with B. abortus, and after 24 h, cells were fixed in 4% paraformaldehyde for 10 min at room temperature, permeabilized with 0.3% Triton X-100 (Roche Diagnostics GmbH, Mannheim, Germany) for 10 min, and blocked with PBS containing 1% BSA for 1 h. Infected cells were stained with mouse anti-ASC (Santa Cruz Biotechnology) diluted in 0.1% PBS–Tween 20 overnight at 4◦C. Cells then were incubated with rabbit anti-mouse Alexa Fluor 488 (Molecular Probes, Life Technologies) diluted in 0.1% PBS–Tween for 4 h at room temperature. 4,6-Diamidine-2 phenylindole (DAPI) was used for nuclear staining, and cells were stained for 30 min at room temperature. After washing in PBS, cells were mounted and then analyzed by fluorescence microscopy. Confocal images were analyzed using FV-1000 confocal microscope with an oil immersion Plan Apochromatic 60× NA1.42 objective (Olympus).

#### Zymography

Gelatinase activity was assayed by the method of Hibbs et al. with modifications, as described (11–13). Briefly, a total of 20µl of cell culture supernatants from infected LX-2 cells cultured in the presence or not of the inhibitors Bay 11-7082, glyburide, Y-VAD-FMK, A151, control C151, and ANAKINRA at the concentrations mentioned above was mixed with 5µl of 5× loading buffer [0.25 M Tris (pH 6.8), 50% glycerol, 5% SDS, and bromophenol blue crystals] and loaded onto 10% SDS-PAGE gels containing 1 mg/ml gelatin (Sigma-Aldrich, Buenos Aires, Argentina). After electrophoresis, gels were washed with a solution containing 50 mM Tris–HCl (pH 7.5) and 2.5% Triton X-100 (buffer A) for 30 min and with buffer A added with 5 mM CaCl<sup>2</sup> and 1µM ZnCl<sup>2</sup> for 30 min and were later incubated with buffer A with an additional 10 mM CaCl<sup>2</sup> and 200 mM NaCl for 48 h at 37◦C. This denaturation/renaturation step promotes MMP activity without the proteolytic cleavage of pro-MMP. Gelatin activity was visualized by the staining of the gels with 0.5% Coomassie blue. Unstained bands indicated the presence of gelatinase activity, and their positions in the gel indicate the molecular weights of the enzymes involved.

#### Measurement of Cytokine Concentrations

Secretion of TGF-β1 and IL-1β in the supernatants was quantified by ELISA (BD Biosciences).

## Assessment of Collagen Deposition—Sirius Red Staining

Collagen deposition was quantified using Sirius red (Sigma-Aldrich, Argentina SA), a strong anionic dye that binds strongly to collagen molecules (14). Sirius red staining was performed as was described (15). Briefly, Sirius red was dissolved in saturated aqueous picric acid at a concentration of 0.1%. Bouin's fluid (for cell fixation) was prepared by mixing 15 ml saturated aqueous picric acid with 5 ml of 35% formaldehyde and 1 ml of glacial acetic acid. Cell layers were fixed with 1 ml of Bouin's fluid for 1 h. Afterwards, culture plates were washed three times with deionized water. Culture dishes were air dried before adding 1 ml of Sirius red dye reagent. Cells were stained for 18 h with mild shaking. The stained cell layers were extensively washed with 0.01 N hydrochloric acid to remove all unbound dye. The stained material was dissolved in 0.2 ml of 0.1 N sodium hydroxide by shaking for 30 min. The dye solution was transferred to microtiter plates, and OD was measured using a microplate reader (Thermo Scientific) at 550 nm against 0.1 N sodium hydroxide as a blank.

## Lipoproteins and LPS

Brucella abortus lipidated outer membrane protein 19 (LOmp19) and unlipidated Omp19 (U-Omp19) were obtained as described (16). Both contained <0.25 endotoxin U/µg of protein as assessed by Limulus Amebocyte Lysates (Associates of Cape Cod). B. abortus S2308 LPS and Escherichia coli O111k58H2 LPS were provided by I. Moriyon. The synthetic lipohexapeptide (tripalmitoyl S-glyceryl-Cys-Ser- Lys4-OH [Pam3Cys]) was purchased from Boehringer Mannheim (Mannheim, Germany).

#### DNA From *B. abortus*

Brucella abortus DNA was purified using the kit Wizard <sup>R</sup> Genomic DNA (Promega) following the instructions of the manufacturer. Brucella DNA was measured spectrophotometrically. Transient transfections of LX-2 cells with 2µg/ml of B. abortus DNA were carried out using Lipofectamine 2000 (Invitrogen), following the manufacturer's instructions. After the purification step, an aliquot of 100 µg of DNA was treated with DNase I (1 U/mg) DNA (Zymo Research) according to the instructions of the manufacturer.

#### Hepatic Fibrosis in a Mouse Model

Mouse strains used in this study included apoptosis-associated speck-like protein containing a CARD (ASC), Nod-like receptor (NLR) containing a pyrin domain 3 (NLRP3), absent in melanoma 2 (AIM2), and caspase-1 (Casp-1)/11 knock-out (KO) mice, as described previously (8), and C57BL/6 wildtype (WT) mice (provided by Federal University of Minas Gerais, Belo Horizonte, Brazil). Six- to eight-week-old mice were infected through the intraperitoneal route with 5 × 10<sup>5</sup> CFU of B. abortus S2308. Mice were born from breeding pairs that were housed under controlled temperature (22 ± 2 ◦C) and artificial light under a 12-h cycle period. Mice were kept under specific pathogen-free conditions in positive-pressure cabinets and provided with sterile food and water ad libitum.

All animal procedures were performed according to the rules and standards for the use of laboratory animals of the National Institutes of Health. Animal experiments were approved by the Institutional Committee for the Care and Use of Laboratory Animals (CICUAL, permit number: 287/2015). Histological examination of liver was carried out at week 4 post-infection after routine fixation and paraffin embedding. Five-micrometer-thick sections were cut and stained with Masson's trichrome stain. Masson's trichrome staining was conducted according to the manufacturer's instructions (Sigma-Aldrich). Collagen-positive areas were visualized by light microscopy and quantified using Image Pro-Plus 6.0 software (Media Cybernetics, Inc.).

## Statistical Analysis

Statistical analysis was performed with one-way ANOVA, followed by post-hoc Tukey test (a single-step multiple comparison statistical test that finds means that are significantly different from each other) using GraphPad Prism 4.0 software. Data were presented as mean ± SEM.

#### RESULTS

#### *B. abortus* Infection Induces IL-1β Secretion by LX-2 Cells via T4SS

It has been established that inflammasome components are present in HSCs and could regulate their function (3). To determine if B. abortus infection induces inflammasome activation, LX-2 cells were infected with B. abortus and the secretion of IL-1β was evaluated in culture supernatants by ELISA 24 h post-infection. B. abortus infection induces the secretion of IL-1β by LX-2 cells (**Figure 1A**). The T4SS encoded by virB genes was first involved in the capacity of Brucella to establish an intracellular replication niche in several cell types (17). In addition, this system has been involved in the induction of inflammatory response during B. abortus infection (18) and also in the inflammasome signaling activation (7). Therefore, experiments were conducted to determine if the T4SS could be involved in the secretion of IL-1β induced by B. abortus in LX-2 cells. To this end, LX-2 cells were infected with B. abortus and its

isogenic B. abortus virB10 mutant, and IL-1β secretion induced by B. abortus was dependent on the expression of a functional T4SS, since the levels or IL-1β did not differ significantly between LX-2 cells infected with B. abortus virB10 mutant and uninfected controls (**Figure 1A**). Taken together, our results indicated that B. abortus infection induces IL-1β secretion in a mechanism that is dependent on the presence of a functional T4SS.

## IL-1β Secretion Induced by *B. abortus* Infection Is Dependent on Caspase-1 and NLRP3

Caspase-1 plays a fundamental role in innate immunity as the protease that activates the pro-inflammatory cytokines pro-IL-1β and pro-IL-18. Caspase-1 itself is activated in different inflammasome complexes; however, activation of the NLRP3 inflammasome has been frequently implicated in the development of fibrosis (19). To determine the role of caspase-1 and NLRP3 in IL-1β secretion by B. abortus-infected LX-2 cells, we performed the infection of LX-2 cells in the presence of specific pharmacological inhibitors. Inhibition of caspase-1 using the general caspase inhibitor Z-VAD-FMK or the specific caspase-1 inhibitor Ac-YVAD-cmk completely abrogated the secretion of IL-1β secretion induced by B. abortus infection of LX-2 cells (**Figure 1B**). When infection experiments were performed in the presence of Bay compound that inhibits NFκB, the secretion of IL-1β was significantly inhibited. However, glyburide, a compound that inhibits NLRP3, partially inhibits IL-1β secretion with respect to untreated cells (**Figure 1B**). Taken together, these results indicated that caspase-1 and NLRP3 are involved in the secretion of IL-1β by B. abortus-infected LX-2 cells.

### IL-1β Secretion Induced by *B. abortus* Infection Is Also Dependent on AIM2

AIM2 inflammasome has been previously involved in the induction of IL-1β secretion in a T4SS-dependent manner during B. abortus infection of bone marrow derived macrophages and dendritic cells (6, 7). Since the inhibition of NLRP3 did not

cells (N.I.).

completely abrogate the secretion of IL-1β in response to B. abortus infection, experiments were conducted to determine whether AIM2 inflammasome is also involved in caspase-1 activation. To this end, LX-2 cells were infected in the presence of A151, the oligodeoxinucleotide sequence that inhibits AIM2 or the sequence control C151. Our results indicated that AIM2 inflammasome contributes to IL-1β production by B. abortusinfected LX-2 cells, since the secretion of IL-1β was significantly inhibited when cells were treated with A151 with respect to untreated cells or cells treated with C151 (**Figure 1C**). When infection experiments were performed in the presence of A151 and glyburide administered in conjunction, the production of IL-1β was completely abrogated (**Figure 1D**). Taken together, these results indicated that NLRP3 and AIM2 inflammasomes are involved in the secretion of IL-1β by B. abortus-infected LX-2 cells.

#### *B. abortus* Infection Induces ASC, NLRP3, AIM2, and Caspase-1 mRNA Expression and ASC Speck Formation in LX-2 Cells

The basal AIM2 expression was sufficient to initiate inflammasome activation (20), but NLRP3 upregulation is necessary to initiate the activation of inflammasome (21, 22). Then, experiments were conducted to determine whether expression of inflammasome components could be upregulated during B. abortus infection. To this end, we determine the mRNA transcription of ASC, NLRP3, AIM2, and caspase-1 by RT-qPCR. Our results indicated that B. abortus infection induces an increase in ASC, NLRP3, AIM2, and caspase-1 mRNA transcription in LX-2 cells (**Figures 2A–D**). Most inflammasomes require oligomerization of ASC and thus the presence of ASC specks is a direct evidence of inflammasome activation. After infection with B. abortus, the formation of ASC specks was detected using specific antibodies by a fluorescence microscope. ASC specks were formed in B. abortusinfected LX-2 cells, but these structures were not detectable in non-infected cells. This indicates that B. abortus infection induces inflammasome assembly and consequently ASC speck formation (**Figure 2E**).

## *Brucella* DNA Induces IL-1β Secretion in LX-2 Cells

IL-1β secretion was dependent on bacteria viability since stimulation of LX-2 cells with heat-killed B. abortus (HKBA) was unable to induce IL-1β (**Figure 3A**).

It has been recently demonstrated that Brucella DNA is involved in IL-1β secretion via activation of caspase-1 through the AIM2 inflammasome in macrophages and dendritic cells (6, 7). Therefore, we decided to test whether Brucella genomic DNA could be the putative ligand for AIM2 in the context of the inflammasome activation in LX-2 cells. To this end, Brucella DNA was transfected into LX-2 cells using lipofectamine or added to the culture medium to determine IL-1β secretion. Brucella DNA induced IL-1β secretion by LX-2 cells when it was added to the culture medium and also in transfected cells (**Figure 3A**). To determine if AIM2 inflammasome is involved in the secretion of IL-1β induced by Brucella DNA, experiments were performed in the presence of A151, the oligodeoxinucleotide that inhibits AIM2 or its oligodeoxinucleotide control C151. AIM2 is involved in the secretion of IL-1β induced by Brucella DNA, since A151 abrogated its secretion (**Figure 3B**). Additionally, DNase I treatment significantly reduced or abrogated Brucella DNA-induced IL-1β secretion (**Figure 3B**), demonstrating that

bacterial DNA participates or is a major agonist that activates the inflammasome.

## PAMPs Associated to Inflammasome Activation

For the production of IL-1β, PAMPs via TLRs and NLRs function in concert. PAMPs induce the expression of the precursor form of this cytokine (pro-IL-1β), and NLR-dependent CASP-1 activation induces its proteolytic processing and release (5). Hence, to assess the role of Brucella PAMPs, LX-2 cells were stimulated with HKBA and mRNA expression of IL-1β was determined by RT-qPCR. Our results indicated that HKBA was able to induce IL-1β expression at the mRNA level (**Figure 4A**) without IL-1β release, which demonstrated that viability is crucial for IL-1β protein production.

Previous observations indicated that LPS and lipoproteins from B. abortus are crucial for inflammatory responses induced by B. abortus in different models in vivo and in vitro (16, 23– 26). Cells were then incubated with LPS from B. abortus and

lipidated Omp19 (L-Omp19) as a Brucella lipoprotein model, and the expression of IL-1β mRNA was determined by RT-qPCR. Our results indicated that L-Omp19 and LPS induce an increase in IL-1β mRNA expression in LX-2 cells. IL-1β mRNA expression induced by Omp19 was dependent on the lipid moiety of the molecule because unlipidated Omp19 (U-Omp19) did not induce IL-1β mRNA expression. The requirement for lipidation was further supported by the fact that Pam3Cys, a lipohexapeptide with an irrelevant peptide sequence, also induced the production of mRNA of IL-1β (**Figure 4A**). As expected, the presence of IL-1β at the protein level was not detected in supernatants of cultures of LX-2 cells treated with LPS or L-Omp19 (**Figure 4B**). Taken together, these results indicated that B. abortus lipoproteins and LPS induce mRNA of IL-1β in LX-2 cells.

## The Inflammasome Pathway Is Involved in the Profibrogenic Response of LX-2 Cells Upon *B. abortus* Infection

It has been demonstrated that inflammasome activation has a variety of functional consequences for HSCs, including enhanced collagen 1 and TGF-β expression (4). Previously, we have demonstrated that upon infection of LX-2 cells, B. abortus inhibits MMP-9 secretion and induces concomitant collagen and TGF-β1 secretion (15). Therefore, experiments were conducted to determine the role of the inflammasome in the induction of a fibrogenic phenotype in LX-2 cells during B. abortus infection. To this end, the levels of secretion of MMP-9, TGF-β, and collagen deposition were determined in LX-2 cells infected with B. abortus in the presence of different inflammasome inhibitors. When infection experiments were performed in the presence of YVAD, glyburide, or A151, the effect of B. abortus infection on MMP-9 expression, collagen deposition, and TGF-β secretion on LX-2 cells was partially reversed with respect to untreated infected cells (**Figure 5**). These results indicated that NLRP3 and AIM2 inflammasome are involved in the induction of profibrogenic phenotype in LX-2 cells.

#### IL-1β Is Involved in the Induction of a Profibrogenic Phenotype

Caspase-1 is required not only for IL-1β secretion but also for IL-18 secretion. To determine the role of IL-1β in the induction of a fibrotic phenotype, infection experiments were performed in the presence of the inhibitor of IL-1 receptor, the natural antagonist IL-1Ra (ANAKINRA). As shown in **Figure 6**, ANAKINRA abrogated the ability of B. abortus to inhibit MMP-9 and to induce collagen deposition. These results indicate that IL-1β could be a key cytokine during inflammasome activation involved in the fibrogenic phenotype triggered by B. abortus infection.

### NLRP3 and AIM2 Influence Liver Fibrosis in Livers From *B. abortus*-Infected Mice

Finally, to verify the in vivo significance of our hypothesis, Casp-1, ASC, NLRP3, and AIM2 KO mice and WT mice, as control, were infected with B. abortus, and 4 weeks later, animals were sacrificed to determine the role of inflammasome in the liver fibrosis. Accordingly with our previous results, Masson's

presence or not of ANAKINRA (ANA). Determination of MMP-9 at 24 h post-infection in culture supernatants by zymography (A). Recombinant human IL-1-β, (rIL-1β) (50 ng/ml) was used as a positive control. Quantification of collagen deposition was revealed by Sirius red staining by OD readings at 550 nm at 10 days post-infection (B). Data are given as the means ± SD of duplicates. \*\*P < 0.01; \*\*\*P < 0.001 vs. untreated cells. #P < 0.001 vs. non-infected cells (N.I.).

trichrome staining revealed the presence of fibrotic patch in livers from B. abortus-infected mice with respect to uninfected control (9, 15). In contrast, Casp-1, ASC, NLRP3, and AIM2 KO animals presented a significant reduction in the fibrotic patch (**Figure 7**). No fibrotic patch was observed in mice inoculated with saline (data not shown). These results indicated that inflammasomes NLRP2 and AIM2 play a key role in the modulation of fibrosis during B. abortus infection.

## DISCUSSION

The liver is frequently affected in patients with active brucellosis, as revealed by the presence of histopathology lesions, such as granulomas, inflammatory infiltrates, and necrosis of liver parenchyma (27, 28).

Fibrosis is an intrinsic response to chronic persistent liver injury that results in a wound-healing process to mitigate the damage, but can also lead to scar formation. Inflammasome activation may play an important role in this process (19).

In the present study, we demonstrated that B. abortus infection activates the inflammasome with concomitant secretion of IL-1β in HSC leading to upregulation of a profibrogenic phenotype.

Inflammasomes have emerged as critical signaling molecules of innate immune system involved in liver fibrosis. Inflammasomes are intracellular multiprotein complexes that act as regulators of inflammation and cell destiny. They respond to several danger signals by activating caspase-1 by the release of proinflammatory cytokines IL-1β and IL-18 (29). In particular, the NLRP3 inflammasome has been frequently implicated in the pathogenesis of chronic inflammatory liver diseases that causes liver fibrosis (30).

HSCs are the main cells involved in extracellular matrix deposition during liver fibrosis (31). In this process, the NLRP3 inflammasome has been involved in the functional changes in HSCs, including upregulation of the expression of collagen and TGF-β (4, 32), in findings that were confirmed by performing the knocking in NLRP3 (33).

Previous studies performed in dendritic cells, macrophages, and glial cells indicate that Brucella is sensed by ASC-dependent inflammasomes, mainly NLRP3 and AIM2, that induce caspase-1 activation with pro-inflammatory response (6, 8, 34, 35). Accordingly, in Brucella-infected HSCs, the secretion of IL-1β depends on NLRP3 and AIM2 inflammasomes. In this context, the participation of PAMPs in the activation of the inflammasome and the secretion of IL-1β must be discussed. The activation and release of IL-1β requires two distinct signals. The first signal can be triggered by various pathogen-associated molecular patterns via TLR activation, which induces the synthesis of pro–IL-1β. The second signal is provided by the activation of the inflammasome and caspase-1 leading to IL-1β processing. During B. abortus infection, the induction of IL-1β at mRNA level was independent of bacterial viability and induced at least by two structural bacterial components, including lipoproteins and LPS. However, the second signal requires bacterial viability and the presence of a functional T4SS and B. abortus DNA. The T4SS is encoded by virB genes that play a main role in Brucella intracellular replication (17), and it has also been involved in the immune response to Brucella infection (7, 18). Bacteria utilize the T4SS to deliver effectors to eukaryotic cells. In LX-2 cells, B. abortus virB10 mutant was unable to induce IL-1β secretion, and it could suggest that Brucella T4SS is involved in the transport of effector molecules that act via NLRP3 and/or AIM2 to activate inflammasomes. This is in agreement with our other finding in which we demonstrate that HKBA does not induce the secretion of IL-1β by LX-2 cells.

In vitro studies have shown that IL-1β promotes the proliferation and myofibroblast transdifferentiation of HSCs with substantial increased levels of their fibrogenic markers (36– 38). Activated caspase-1 could also cleave pro-IL-18 into its active form IL-18, and this cytokine has also been involved in fibrosis induction (38, 39). In a murine model of non-alcoholic steatohepatitis, fibrosis was not reversed by IL-1Ra treatment, indicating that other regulators of NLRP3 inflammasome are involved in fibrogenesis promotion (33). However, our experiments performed in B. abortus-infected HSCs in the presence of ANAKINRA, a version of the human interleukin-1 receptor antagonist, indicated that IL-1β secreted by HSCs has a main role in the induction of TGF-β with concomitant collagen deposition and inhibition of MMP-9 secretion. This role of IL-1β in the myofibroblast transdifferentiation with concomitant fibrosis is not exclusive for HSCs; accordingly, it has been previously described in endothelial and epithelial cell transdifferentiation and fibrosis (40–42).

These results and previous findings suggest that the interaction of Brucella with innate immunity in vivo may result in an increase of inflammatory response that results in liver fibrosis. Inflammasomes would dictate this fibrotic phenotype. Consequently, B. abortus infection induces fibrosis in mice that was reduced in mice lacking AIM2 or NLRP3.

Together, these results indicated that upon infection of HSCs, B. abortus triggers AIM2 and NLRP3 inflammasome activation with concomitant IL-1β secretion in a mechanism that is dependent on a functional T4SS and DNA. This IL-1β is implicated in the induction of fibrotic phenotype in HSC.

#### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the article/supplementary material.

#### REFERENCES


#### ETHICS STATEMENT

This animal study was reviewed and approved by CICUAL-Facultad de Medicina, Universidad de Buenos Aires.

#### AUTHOR CONTRIBUTIONS

MD conceived and designed the experiments, supervised experiments, interpreted the data, and wrote the manuscript. PA, MG, and AP performed the experiments. PA analyzed the data and wrote sections of the manuscript. SO, JQ, and GG supported the work with key suggestions and helped with data interpretation. All authors reviewed the manuscript.

### FUNDING

This work was supported by grants PICT 2014-1111 and PICT 2015-0316 from Agencia Nacional of Promoción Científica y Tecnológica (ANPCYT, Argentina) and by grant PIP 2015-2017- 0200 from Consejo Nacional de Investigaciones científicas y técnicas (CONICET). Funding agencies had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

#### ACKNOWLEDGMENTS

We thank Horacio Salomón and the staff of the Instituto de Investigaciones Biomédicas en Retrovirus y Sida (INBIRS) for their assistance with biosafety level 3 laboratory uses. PA and AP are recipients of a fellowship from CONICET. JQ, GG, and MD are members of the Research Career of CONICET.

ASC inflammasome-dependent IL-1beta production. J Immunol. (2016) 196:3794–805. doi: 10.4049/jimmunol.1500908


effects of transforming growth factor beta2 and ascorbic acid 2 phosphate. Histochem Cell Biol. (1999) 112:271–6. doi: 10.1007/s004180 050447


**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 © 2020 Arriola Benitez, Pesce Viglietti, Gomes, Oliveira, Quarleri, Giambartolomei and Delpino. 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.

# T Lymphocyte-Mediated Liver Immunopathology of Schistosomiasis

Bing Zheng1,2, Jianqiang Zhang<sup>1</sup> , Hui Chen<sup>1</sup> , Hao Nie1,2, Heather Miller <sup>3</sup> , Quan Gong1,2 \* and Chaohong Liu<sup>4</sup> \*

*<sup>1</sup> Department of Immunology, School of Medicine, Yangtze University, Jingzhou, China, <sup>2</sup> Clinical Molecular Immunology Center, School of Medicine, Yangtze University, Jingzhou, China, <sup>3</sup> Department of Intracellular Pathogens, National Institute of Allergy and Infectious Diseases, Bethesda, MD, United States, <sup>4</sup> Department of Pathogen Biology, School of Basic Medicine, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, China*

#### Edited by:

*Sergio C. Oliveira, Federal University of Minas Gerais, Brazil*

#### Reviewed by:

*Sharvan Sehrawat, Indian Institute of Science Education and Research Mohali, India Rashika El Ridi, Cairo University, Egypt Miguel J. Stadecker, Tufts University School of Medicine, United States*

\*Correspondence:

*Quan Gong gongquan1998@163.com Chaohong Liu chaohongliu80@126.com*

#### Specialty section:

*This article was submitted to Microbial Immunology, a section of the journal Frontiers in Immunology*

Received: *26 September 2019* Accepted: *10 January 2020* Published: *18 February 2020*

#### Citation:

*Zheng B, Zhang J, Chen H, Nie H, Miller H, Gong Q and Liu C (2020) T Lymphocyte-Mediated Liver Immunopathology of Schistosomiasis. Front. Immunol. 11:61. doi: 10.3389/fimmu.2020.00061* The parasitic worms, *Schistosoma mansoni* and *Schistosoma japonicum*, reside in the mesenteric veins, where they release eggs that induce a dramatic granulomatous response in the liver and intestines. Subsequently, infection may further develop into significant fibrosis and portal hypertension. Over the past several years, uncovering the mechanism of immunopathology in schistosomiasis has become a major research objective. It is known that T lymphocytes, especially CD4<sup>+</sup> T cells, are essential for immune responses against *Schistosoma* species. However, obtaining a clear understanding of how T lymphocytes regulate the pathological process is proving to be a daunting challenge. To date, CD4<sup>+</sup> T cell subsets have been classified into several distinct T helper (Th) phenotypes including Th1, Th2, Th17, T follicular helper cells (Tfh), Th9, and regulatory T cells (Tregs). In the case of schistosomiasis, the granulomatous inflammation and the chronic liver pathology are critically regulated by the Th1/Th2 responses. Animal studies suggest that there is a moderate Th1 response to parasite antigens during the acute stage, but then, egg-derived antigens induce a sustained and dominant Th2 response that mediates granuloma formation and liver fibrosis. In addition, the newly discovered Th17 cells also play a critical role in the hepatic immunopathology of schistosomiasis. Within the liver, Tregs are recruited to hepatic granulomas and exert an immunosuppressive role to limit the granulomatous inflammation and fibrosis. Moreover, recent studies have shown that Tfh and Th9 cells might also promote liver granulomas and fibrogenesis in the murine schistosomiasis. Thus, during infection, T-cell subsets undergo complicated cross-talk with antigen presenting cells that then defines their various roles in the local microenvironment for regulating the pathological progression of schistosomiasis. This current review summarizes a vast body of literature to elucidate the contribution of T lymphocytes and their associated cytokines in the immunopathology of schistosomiasis.

Keywords: T lymphocyte, schistosomiasis, immunopathology, liver fibrosis, soluble egg antigen

## INTRODUCTION

Human schistosomiasis is a zoonotic parasitic disease caused by schistosomes, which are digenetic trematodes. In nature, many species of Schistosoma exist, but the main human pathogens are Schistosoma japonicum, Schistosoma mansoni, and Schistosoma haematobium. There is an estimate of more than 230 million people in tropical countries around the world infected with Schistosoma species (1). Infection occurs when cercariae, the freeliving larval form of schistosomes, are released from freshwater snails and penetrate a human host's skin, where they may remain in the host epidermis for ∼72 h (2). Then, cercariae transform into schistosomula, the parasitic larvae form that enter the vasculature and travel via the pulmonary artery to the lungs, which they are then referred to as lung schistosomula (3). After exiting the lungs, schistosomula re-enter the venous circulation and finally migrate to the perivesicular venules (S. haematobium) or mesenteric venules (S. japonicum, S. mansoni), where they mature into worms and form copulating pairs. Paired adult worms migrate to the intestinal venous vasculature to release their eggs. Of the fertilized eggs released by the female, a portion is discharged via emunctory routes and hatch in freshwater, releasing free-living ciliated miracidia that invade snails and continue on the life cycle. However, a large portion of S. haematobium eggs may become trapped in the bladder and urogenital system, while the eggs of S. japonicum and S. mansoni become lodged in the intestinal wall and liver (4). The eggs then induce a local granulomas inflammatory response. The granulomas mainly consist of lymphocytes, macrophages, and eosinophils, which contain egg proteolytic enzymes to prevent tissue damage; however, the egg-induced granuloma also leads to chronic schistosomiasis. In this review, we will focus on S. japonicum and S. mansoni, both of which have been extensively studied and have also been used in mouse models to study liver immunopathology.

The basic pathological process of schistosomiasis includes acute and chronic phases of the disease. Acute schistosomiasis (sometimes referred to as Katayama fever) occurs most often in schistosome-endemic regions to naive individuals and to those who experience heavy reinfection. The major clinical manifestation of acute schistosomiasis includes a sudden onset of fever, fatigue, urticaria, eosinophilia, and abdominal tenderness. These symptoms occur weeks to months after schistosome infection because of worm maturation, egg production, and egg antigen-induced inflammatory granulomatous response (4, 5). Chronic schistosomiasis involves immune responses to the eggs within the liver and intestine, which causes formation of granuloma (6–9). Over time, the granulomatous response is gradually downregulated, leading to the progression of relatively tolerable chronic intestinal schistosomiasis. However, some patients will develop lifethreatening hepatosplenic schistosomiasis accompanied by extensive hepatic and periportal fibrosis, portal hypertension, ascites, and gastrointestinal hemorrhage.

T lymphocytes are generally classified into CD4<sup>+</sup> T helper (Th) cells and CD8<sup>+</sup> cytotoxic T lymphocytes (CTLs). Th cells are important for host humoral and cellular immune responses against parasitic infections (6). The Th cells are further classified into several distinct Th phenotypes [Th1, Th2, Th17, follicular helper T cell (Tfh), Th9, and regulatory T cells (Tregs)] according to cytokine production and specialized functions. This review focuses on the role of various Th subsets and their associated cytokines in the immunopathogenesis of schistosomiasis.

### THE ROLE OF Th1/Th2 RESPONSES IN THE IMMUNOPATHOLOGY OF SCHISTOSOMIASIS

An accumulation of evidence suggests that Th cells are involved in the immunopathogenesis of schistosomiasis. In athymic nude mice or mice without Th cells resulted in decreased granuloma size after infection by schistosomes (10, 11). The role of Th1/Th2 responses in schistosomiasis has been intensely investigated and reviewed (6, 9, 12). In the acute illness, parasite antigens elicit a moderate Th1 response, which is characterized by increased levels of proinflammatory cytokines, including tumor necrosis factor alpha (TNF-α), interferon gamma (IFNγ), interleukin-1 (IL-1), and IL-6 (13). About 5–6 weeks post-infection, the schistosomula develop into mature worms. Female worms release fertilized eggs that stimulate the immune response via their soluble egg antigens (SEAs). SEA induces resident macrophages to secrete inflammatory cytokines and chemokines that stimulate the influx of lymphocytes, neutrophils, and monocytes, which initiates circumoval granulomatous inflammation (14). A clinical study comparing acutely and chronically infected patients showed that the level of TNF-α was elevated in the plasma of acutely infected patients and their peripheral blood mononuclear cells spontaneously secreted high levels of IL-1 and IL-6 and detectable levels of IFNγ, while chronically infected patients produced little TNF-α or IFN-γ. Thus, higher levels of IFN-γ and proinflammatory cytokines in patients with acute illness show a dominant Th1 response (15). During the acute phase, the SEA-specifically activated CD4<sup>+</sup> Th cells release the Th1-type cytokines, IL-2 and IFN-γ, which mediate the establishment of early granulomas. Immunocytochemical examination in situ have confirmed the presence of Th1-type cytokine-producing cells within local microenvironments of the lesion in early granuloma formation (16). Although the granulomatous response is detrimental to the liver because of subsequent progression of hepatic fibrosis, the egg-induced granuloma is beneficial to the host. If the eggs are not sequestered effectively, continuously secreted egg antigens act as a stimulus and lead to uncontrolled inflammatory responses and permanent tissue damage. For instance, IL-4-deficient mice that cannot mount a normal Th2 response develop an unchecked Th1 response and die earlier than immunity intact mice when infected by S. mansoni (6). Similarly, in a Tamoxifen-induced IL-4 receptor α (IL-4Rα)-deficient mouse model, interrupting IL-4Rα-mediated signaling during the acute stage decreased protective Th2 responses, leading to severe disease and premature death (17). Therefore, moderate Th1 responses are involved in the acute schistosomiasis and early granuloma formation, whereas excessive polarization toward the Th1 response is detrimental to the host. The Th1 response during the early stage of schistosomiasis is downregulated by IL-4 and IL-10. IL-10-/IL-4-deficient mice develop extremely polarized Th1-type cytokine IFN-γ responses that lead to 100% mortality during the acute illness (18).

The immune response and immunopathology of schistosomiasis are a consequence of CD4<sup>+</sup> T-cell sensitization to egg antigens. Some of the major components of egg are implicated in the Th response in schistosomiasis, including glycoprotein IPSE/α1, ω1 (19, 20), lacto-N-fucopentaose III (21), and S. mansoni-p40 (Sm-p40) (22), of which Sm-p40 is the most abundant egg component that can induce a strong Th1-polarized response (23). IPSE/α1 induces a mixed Th1/Th2 response and promote the development of hepatic granuloma (19), while lacto-N-fucopentaose III and ω1 directly act on dendritic cells (DCs) to enhance the Th2 response (21, 24). In addition, Th1 type responses could also be induced by the schistosome vaccine candidates MAP4 (25), egg-derived r38 (22, 26), rSmLy6B, rSmTSP6, and rSmTSP7 (27), and rSjCRT (28).

Compared to the Th1 response, the Th2 response exerts anti-inflammatory effects and regulates Th1-mediated immunopathology. SEA is considered to be the key factor in driving the dramatic transition from a Th1- to a Th2 dominated response. During the Th response transition, the interaction of CD40–CD40L, CD80/86–CD28, and B7-related protein 1-inducible costimulator (ICOS) are required for egg antigen-induced Th2 responses (29–31). In correlation with the Th2 transition, the cytokine profile is changed in that IFN-γ production is decreased, whereas Th2-type cytokines IL-4, IL-5, and IL-13 are increased (32). This has been shown in lymphocytes isolated from liver granulomas at 8 weeks, which secrete IL-4 and IL-5, but not IFN-γ. The mechanism underlying this switch to Th2 response has been deeply investigated. Lymphocytes from S. mansoni ova-infected signal transducer and activator of transcription (STAT6)-deficient mice produce minimal levels of Th2-type cytokines and enhanced IFN-γ production, which greatly reduces the size of pulmonary and hepatic granulomas (33). IL-4 signal is critical for the development of the Th2 response and animals treated with anti-IL-4 showed decreased Th2-type IL-4, IL-5, and IL-10 and increased Th1-type IL-2 and IFN-γ (34). IL-4 is recognized as the dominant cytokine for granuloma development and injection with neutralizing antibodies against IL-4 significantly suppresses splenic cell proliferation and hepatic granulomatous inflammation (35). Nevertheless, other studies showed that infected IL-4 knockout (KO) mice still have the ability to generate egg granulomas. Surprisingly, infected IL-4Rα-deficient mice exhibit only minimal hepatic granulomas and fibrosis, even though Th2-type cytokine production is similar to infected IL-4 KO mice, which demonstrates that the IL-4R signaling pathway, rather than IL-4, may be essential for egg granulomas (36). However, another Th2-type cytokine, IL-13, is known to be involved in granulomatous inflammation and fibrosis through the IL-4Rα signaling. Blocking IL-13 has been shown to be highly effective in treating an established and ongoing S. mansoni infection-induced fibrosis (37, 38). These results were further confirmed by studies using schistosome-infected IL-13 and IL-4-/IL-13-deficient mice (39). Furthermore, IL-13-deficient mice exhibit increased survival time, demonstrating the important role of IL-13 in the pathogenesis of schistosomiasis. In vitro studies revealed that IL-13 stimulates collagen production in fibroblasts and has a direct role in collagen homeostasis of normal human skin and keloid fibroblasts (40, 41). Overall, it is known that IL-4 and IL-13 play redundant roles in the schistosomiasis granulomatous response, but IL-13 is not dependent on profibrotic cytokines or affected by Th1/Th2 cytokines.

The source of Th2-type cytokines are not only produced by Th2 cells but also secreted by other innate lymphocytes, such as type 2 innate lymphoid cells (ILC2s). The three epithelial cell-derived cytokines, IL-33, IL-25, and thymic stromal lymphopoietin (TSLP), act as crucial initiators of Th2 responses to induce ILC2s to produce IL-13, therefore promoting Th2 type immunity (42–44). It has been reported that IL-33 treatment promotes a Th2 response together with increased liver immunopathology, and these effects are prevented by anti-IL-33 monoclonal antibodies. Furthermore, IL-33 is a requisite for IL-13-, but not IL-4, driven Th2 responses during the pathology (45). Subsequent studies have further confirmed that IL-33 is involved in initiating Th2 pathology after schistosome infection via regulating IL-13 expression in hepatic stellate cells (46) and inducing polarization of M2 macrophages (47). However, another study demonstrated an overlapping role of IL-33, IL-25, and TSLP in a schistosome-induced lung granuloma and liver fibrosis model. They showed that simultaneous disruption of IL-33, IL-25, and TSLP signaling inhibited the progression of Th2 cytokine-driven liver fibrosis but that individual disruption of each had no effect (48).

In addition to SEA-induced specific Th2 responses, parasitederived cysteine peptidases are also responsible for Th2 responses. For instance, during acute schistosomiasis, injection of outbred mice with S. mansoni cathepsin B1 (SmCB1) or S. mansoni cathepsin L3 (SmCL3) develops a polarized Th2 type immune environment that is harmful for the development of S. mansoni larvae and leads to significant reduction in worm burden and liver egg counts (49). A similar study also showed the effectiveness of a cysteine-peptidase-based vaccine that protects hamsters from schistosomiasis haematobium by inducing Th2 immune responses (50). Therefore, the cysteinepeptidase-based vaccine has shown great potential to be further used in non-human primates, and even in humans, through inducing protective Th2 immune responses.

Although a strong Th2 response seems detrimental to the host, a Th2 response is indispensable for the survival of the host. Th2-deficient IL-4−/<sup>−</sup> mice are highly susceptible to infection and develop severe acute cachexia followed by death (51). Depletion of the Th2 response against the eggs results in tissue damage and increased host mortality and liver pathology due to proinflammatory Th1-type responses (52, 53). Clinical cases also demonstrate that more severe forms of hepatosplenic schistosomiasis is linked to low levels of Th2-type IL-5 and increased Th1-type IFN-γ and TNF-α (54). Therefore, Th2 immunity acts as a double-edged sword: on the one side, it protects the host to decrease the overall pathology and prevent excessive granulomatous inflammation, but on the other side, it causes liver immunopathological damage. Thus, the Th2 response-mediated egg granuloma is a necessary evil for host survival. Therefore, maintaining the proper balance of Th1/Th2 responses is important to control the excessive pathology of schistosomiasis. Previous studies demonstrated that excessively polarized Th1- or Th2-type cytokine responses induce different but equally detrimental pathologies after infection (18). IL-4/IL-10 double-deficient mice develop highly polarized Th1-type cytokine responses, exhibited by rapid weight loss during egg production and 100% acute mortality by week 9 post-infection. In contrast, IL-12/IL-10 double-deficient mice with highly polarized Th2-type cytokine responses develop increased hepatic fibrosis and mortality during the chronic stages of infection (18). Both Th1 and Th2 phases are downregulated by endogenous IL-10, which is produced by macrophages and T cells (55). Thus, IL-10 acts as an important regulator to prevent excessive Th1 and Th2 responses during the development of schistosomiasis.

### Th17/IL-17 EXACERBATE THE EGG-INDUCED LIVER IMMUNOPATHOLOGY

The first notion of Th17 cells playing a role in schistosomiasis egg-induced granuloma formation came from experiments using knockout mice unable to produce IL-23, which drives the production of IL-17 by Th17 cells. In these experiments, IL-12p40−/<sup>−</sup> mice, incapable of producing IL-12 or IL-23, were highly resistant to liver pathology, whereas IL-12p35−/<sup>−</sup> mice, able to produce IL-23 but not IL-12, developed severe granuloma lesions (56). Granuloma formation is associated with high levels of IL-17 and treatment with anti-IL-17 neutralizing antibodies significantly inhibited hepatic granulomatous inflammation (56). Therefore, the IL-17-producing CD4<sup>+</sup> T-cell population driven by IL-23 was recognized as a separate lineage and designated as Th17 cells that contribute to severe immunopathology in schistosomiasis (56, 57).

The IL-23–IL-1–IL-17 axis plays an essential role in the development of severe forms of schistosomiasis (58). A study conducted on mice lacking the IL-23-specific subunit p19 revealed an impaired liver immunopathology together with a marked decrease in IL-17 in the granulomas, but not in the draining mesenteric lymph nodes (59). Subsequent studies conducted on high pathology-prone CBA mice and low pathology-prone C57BL/6 mice identified IL-23 and IL-1, derived from DCs stimulated by live schistosome eggs, as the critical host factors that drive IL-17 production (60, 61). S. mansoni-infected CBA mice possess more IL-17-producing cells in the spleen and granulomas when compared with C57BL/6 mice (62). It was later determined that egg antigens, but not adult worm antigens, preferentially induce the generation of Th17 cells. Lowering IL-17 levels by neutralizing anti-IL-17 antibodies can increase the parasitespecific antibody levels and supply partial protection against S. japonicum infection in mice (63). In addition, anti-IL-17 antibody markedly inhibited hepatic granulomatous inflammation and hepatocyte necrosis partly through reducing the proinflammatory cytokines/chemokines and infiltrating neutrophils (64). Similar results were obtained with S. japonicum-infected IL-17RA-deficient mice that displayed decreased granulomatous inflammation, hepatic fibrosis, improved liver function, and high survival (65). Rutitzky et al. investigated the role of IL-17 and IFN-γ in schistosomiasis immunopathology using mice lacking either one or both cytokines. They found that IL-17-deficient mice show significantly reduced immunopathology associated with the increased levels of IFN-γ, whereas IFN-γ-deficient mice displayed exacerbated immunopathology as well as increased levels of IL-17. Hence, IL-17 plays a powerful pathogenic role in severe immunopathology in murine schistosomiasis that normally is restrained by IFN-γ (66). It has also been reported that T-bet−/<sup>−</sup> mice have significantly increased egg-induced hepatic immunopathology, with the absence of IFN-γ and increased IL-23p19, IL-17, and TNF-α in granulomas. Thus, T-bet-dependent signaling negatively regulates Th17-mediated schistosomiasis immunopathology (67).

Recent reports found that ICOS is essential for pathogenicinduced Th17 cell development. This was discovered by Wang et al., who found that ICOSL KO mice had lower levels of Th17-associated cytokines (IL-17/IL-21), IL-13, and TGF-β1, which is correlated with improved survival rate, alleviated liver granulomatous inflammation, and hepatic fibrosis development (68). In addition, CD209a (C-type lectin receptor), expressed on DCs, also proved to be essential for the development of Th17 cell responses in murine schistosomiasis (61, 69). CD209a-deficient CBA mice had decreased Th17 responses and developed reduced egg-induced liver immunopathology (69, 70). Clinical studies also demonstrated the positive correlation between the percentage of Th17 cells and bladder pathology in S. haematobium-infected children (62).

The above findings demonstrate that Th17/IL-17 exacerbate the egg-induced liver immunopathology in schistosomiasis. However, it should be noted that a recent study found that acute schistosome infection induced a transient Th17 response to cathepsin B1 cysteine proteases secreted by the worms and that, this early, Th17 response may determine the pathogenic progression of the infection (71).

Although the role of Th17/IL-17 in schistosomiasis liver immunopathology has been defined, the source of IL-17 is not completely clear. Generally, IL-17-producing cells include Th17 cells, CTL cells, γδT cells, and natural killer T cells. Some researchers ignore the source of IL-17, and others found that Th17 cells are the major IL-17-producing cell population that contributes to pulmonary granuloma induced by S. japonicum (72). However, some studies showed that innate γδ T cells are the major IL-17-producing cells that contribute to the formation of granuloma in murine schistosomiasis (73, 74). The above discrepant findings may result from the different granuloma models. In addition, aforementioned ILC2s mediate Th2-type immunity through secreting IL-13. The ILC3s, as counterparts to Th17 cells, engage in Th17 immunity-mediated mucosal homeostasis and defense (75). However, the exact role of ILC3s has not been clarified yet in the pathology of schistosomiasis.

Thus far, numerous studies have promoted our understanding of the basic immunopathogenesis of schistosomiasis; however, it should be noted that schistosome-induced liver immunopathologies are associated with Schistosoma species and host. Examples of this include that the cellular composition of the S. japonicum egg-induced granulomas are mainly neutrophils, whereas S. mansoni-induced granulomas consist of a higher ratio of mononuclear cells and eosinophils, with lower numbers of neutrophils (76, 77). The differences between S. japonicum- and S. mansoni-induced hepatic granuloma could be attributed to the secreted specific leukocyte-associated chemokines at the site of inflammation (77). A host's genetic background also affects infection intensity and pathology of schistosomiasis (78, 79). Nevertheless, most researchers perform their studies using BALB/c and C57BL/6 strain mice. Actually, schistosome-infected BALB/c and C57BL/6 mice only develop mild hepatic granulomatous inflammation; however, CBA and C3H mice develop a severe pathology with larger size and poorly confined granulomas (80). As previously discussed, mouse strain-dependent schistosomiasis pathology may arise from the difference of Ag-specific Th responses (62), such as S. mansoniinfected CBA mice displaying exacerbated granulomatous lesions when compared to C57BL/6 mice because of high ratios of IL-17-producing cells in the granulomas (62).

### Tfh PROMOTES LIVER PATHOLOGY OF SCHISTOSOMIASIS

Tfh cells are a specialized Th subset equipped to provide B cell help (81). Gene microarrays revealed that the transcriptional profile of Tfh cells is different from Th1, Th2, and Th17 cells (82). Tfh cells are mainly located in the periphery of Bcell follicles in secondary lymphoid organs and are identified by expression of various molecules, such as surface receptors CXCR5, programmed death 1 (PD-1), ICOS, the transcription factor B-cell lymphoma 6 (Bcl-6), and the cytokine IL-21 (83–85). Tfh cells were found to differentiate from Th2 cells in germinal centers responding to SEA (86). S. japonicum recombinant protein, SjGST-32, also has the ability to induce the formation of Tfh cells in BALB/c mice, which promotes humoral immune responses and long-lived memory B cells (87). Functional studies on Tfh cells showed that downregulation of their cellular development leads to immune deficiencies and that Tfh cells are closely associated with autoimmune and chronic inflammatory disease (85, 88). S. mansoni eggs induce differentiation of Tfh, which is highly dependent on Notch expression on T cells. Notch-deficient mice show impaired germinal center formation and decreased secretion of highaffinity antibodies (89). Tfh expansion and antibody production in response to schistosome infection are negatively regulated by B7-H1 (programmed death ligand 1) that is expressed on B cells (90). Immunization of B7-H1−/<sup>−</sup> mice with SEA leads to increased numbers of Tfh cells compared to wildtype mice (90). Recent studies have demonstrated that Tfh cells promote liver granulomas and fibrogenesis in mice infected with S. japonicum (91, 92). Wang et al. found that, in murine schistosomiasis, Tfh cells accumulate in the splenic germinal center and that the Tfh phenotypic molecule, Bcl-6, and the Tfh-type cytokine, IL-21, correlate with progression of liver fibrosis (92). In addition, clinical studies have shown that Tfh cells are involved in immune responses for both acute and chronic human schistosomiasis (93, 94). The frequency of circulating PD-1+CXCR5+CD4<sup>+</sup> Tfh cells in the peripheral blood is increased in both acute and chronic schistosomiasis patients relative to healthy controls. However, the difference between acute and chronic schistosomiasis is that there is no correlation between percentages of PD-1+CXCR5+CD4<sup>+</sup> Tfh cells, memory B cells, or the level of immunoglobulin G (IgG) specific to S. japonicum antigen in acute schistosomiasis patients; however, frequency of PD-1+CXCR5+CD4<sup>+</sup> Tfh cells is positively correlated to the levels of IL-21 in sera and the levels of SEA-specific antibody in chronic schistosomiasis patients (93, 94). These findings demonstrate that Tfh cells might exhibit distinct mechanisms to regulate the immune response between acute and chronic schistosomiasis. In addition, IL-4-producing Tfh cells are presumed to be important for naturally acquired resistance to schistosome reinfection (95).

## Th9 CELLS AND EGG-INDUCED HEPATIC GRANULOMATOUS AND FIBROSIS

Th9 cells are another unique subset of CD4<sup>+</sup> T cells recently characterized by their production of cytokine IL-9 after activation. The specific transcription factors for Th9 cells include PU.1 and IRF-4 (96, 97). Before discovery of the Th9 cell, IL-9 was thought to be a Th2-specific cytokine. Until 2008, it was reported that IL-9 could be secreted exclusively by distinct IL-9+IL-10<sup>+</sup> Th cells lacking suppressive function (98). Th9 cells have been implicated in many diseases, such as allergic inflammation, autoimmune disorders, as well as helminth infections (96, 99, 100). Zhan et al. investigated the dynamics of splenic Th9 cells and IL-9 expression in liver and serum and found that the proportion of splenic Th9 cells and levels of IL-9 were significantly higher in S. japonicum-infected mice compared to uninfected controls. Moreover, dynamic changes of Th9, IL-9, and PU.1 levels were consistent with hepatic egg granulomatous inflammation (101). In agreement with these studies, Li et al. additionally showed that anti-IL-9 monoclonal antibody treatment significantly inhibits S. japonicum-induced hepatic granulomatous and fibrosing inflammation (102). The above findings indicate that Th9 cells may be involved in immunopathogenesis in murine schistosomiasis. However, a clinical investigation showed that the level of IL-9 in serum had no significant difference between acute and chronic schistosomiasis patients (103). Therefore, further studies are needed to clarify the defining roles of Th9 cells and IL-9 in the pathology of human schistosomiasis.

### Tregs REGULATE THE GRANULOMATOUS INFLAMMATION AND FIBROSIS

In 1995, Sakaguchi et al. introduced regulatory T cells to the field of immunology (104). Tregs are a separate lineage of T cells that are responsible for maintaining immunological homeostasis, suppressing potentially deleterious activities of Th cells, as well as mediating the magnitude of immunity against invading pathogens (105). It has been conceded that two main types of Tregs exist: one is termed "inducible" Tregs (iTregs), which responds to infectious challenge, and the other is termed "natural" Treg (nTregs), which is an endogenous Tregs (106). The forkhead box protein 3 (Foxp3) is a unique transcription factor that can be used to separate nTregs from iTregs that have similar regulatory properties (107). The nTregs develop from a normal process of maturation in the thymus and serve as an essential subset of T cells in the periphery. The specific markers for nTregs include CD25, the T-cell inhibitory co-receptor CTLA-4, and the glucocorticoid-inducible TNF receptor (108, 109). The iTregs originate from conventional CD4<sup>+</sup> T cells that are exposed to specific stimulatory factors, such as a cocktail of cytokines or drugs (110). Currently, the iTregs identified include IL-10 producing Tr1 cells, TGF-β-producing T helper type 3 (Th3) cells, and regulatory CD4+CD25+Foxp3<sup>−</sup> cells (111, 112). Both iTregs and nTregs exert suppressive/regulatory effects to restrict immune-mediated pathology.

So far, numerous studies have demonstrated that both nTregs and iTregs represent key players in the regulation of schistosomiasis pathology. During schistosome infection, Tregs suppress DC activation, mediate Th2 responses, and inhibit granuloma development and fibrosis. After infection with S. mansoni, the percentage of granuloma nTregs (CD4<sup>+</sup> CD25<sup>+</sup> Foxp3+) has a significant increase at 8 and 16 weeks of the infection (113). Similarly, schistosome eggs show the ability to induce a significant Foxp3<sup>+</sup> Treg cell response, which suggests that SEA may be the most potent inducer for the generation of nTregs during infection (114). Except for SEA, schistosome-derived molecules, such as lysophosphatidylserine (lyso-PS), identified as the TLR2-activating molecule, can also actively induce the development of IL-10-producing Tregs (115). Although the percentage of Tregs are elevated either following infection or egg immunization, the natural ratio between nTregs and effector T cells is remarkably stable during the progress of the egg-induced inflammation, suggesting that the expansion of effector T cells maybe closely regulated by the nTregs response.

With the progression into the chronic stage of egginduced inflammation, the phenotype of nTregs is changed so that the frequency of CD103-expressing nTregs is strongly increased. CD103 binds integrin β7 to form the complete integrin molecule αEβ7, which is an activation marker for nTregs. Thus, increasing the CD103-expressing nTregs is required for immunosuppression during chronic schistosomiasis (116). Although numerous studies have characterized nTregs during schistosome infection, the exact mechanism of how nTregs regulate immunopathology is not yet clear. To address the functional role of Tregs in schistosomiasis, many studies used the CD4 and CD25 sorting of Tregs; however, this makes it difficult to distinguish the function of either Tregs because CD4+CD25<sup>+</sup> compartment includes both nTregs and iTregs. Nevertheless, Baumgart et al. established nTregs depleted mice to evaluate the role of nTregs in egg-induced inflammation. They found that both IFN-γ and IL-4 responses were increased following immunization, demonstrating that nTregs suppress both Th1 and Th2 responses (116), which is likely not associated with IL-10 (114).

In addition to nTregs, schistosomiasis infection also induces the production of iTregs (117, 118). It has been reported that IL-10, secreted from iTregs and Th2 cells, inhibits IL-12 production by CD40 agonist-stimulated DC and iTregs, thus suppressing development of egg-specific Th1 responses during schistosomiasis (118). Furthermore, IL-10 secreting iTregs isolated from the granuloma of chronically infected mice can inhibit CD4<sup>+</sup> T-cell proliferation (118), which is different from nTregs that do not appear to control the proliferation of T cells in vivo (116). Therefore, it is possible that both nTregs and iTregs have the ability to suppress both Th1 and Th2 responses. To investigate the importance of Tregs for controlling schistosomiasis liver pathology, adoptive transfer purified populations of CD25 depleted CD4<sup>+</sup> T cells into RAG-deficient mice (lack of mature T or B lymphocytes) lead to increased weight loss, liver damage, and mortality following infection, suggesting a strong capacity for Tregs to suppress liver pathology (117).

## WHAT ABOUT THE ROLE OF CD8<sup>+</sup> T CELLS?

The role and mechanism of CD8<sup>+</sup> T cells in the process of schistosomiasis has been a somewhat neglected area of study. Early studies conducted in the pulmonary granuloma model demonstrated that CD8<sup>+</sup> T-cell deficiency increased granuloma formation by 70%, which was attributed to CD8<sup>+</sup> T-cell inhibition of Th2 maturation (119). However, a subsequent study found that granuloma formation and hepatic granulomatous reaction were unchanged in CD8<sup>+</sup> T-cell-deficient mice (120). Another study, using major histocompatibility complex (MHC) class II or I mutant mice to examine the role of CD4<sup>+</sup> and CD8<sup>+</sup> T cells in the pathology of schistosomiasis, found that schistosome-infected MHC I mutant mice developed normal granulomatous lesions, while in contrast, MHC II mutant mice failed to form egg granuloma (121). Therefore, CD8<sup>+</sup> T cells are not likely essential for regulating the immunopathology of schistosomiasis.

## INNATE γδ T CELLS IN SCHISTOSOMIASIS

The previously discussed T lymphocyte, the αβ T cell, has TCR composed of two glycoprotein chains, α and β, while γδ T cells, another subset of lymphocytes, are CD4−CD8−CD3<sup>+</sup> T cells with TCR encoded by the γ and δ genes (122). Generally, γδ T cells consist of ∼5% of the circulating peripheral blood T cells, but in some infectious diseases, the enumeration of γδ T cells exceeds 30% of the peripheral blood T cells (123). During schistosomiasis, the response to schistosome antigen is primarily mediated by activated CD4<sup>+</sup> αβ T cells (33). However, it has also been reported that γδ T cells are recruited to egg-induced granuloma in the murine schistosomiasis (124) and that the levels of γδ T cells are increased in schitosome-infected mice and patients (125, 126). However, an earlier study investigating the relative roles of αβ and γδ T cells in the immunopathology of schistosomiasis found that mutant mice lacking γδ T cells display vigorous formation of egg granulomas similar to normal mice, which demonstrates that granuloma formation is not dependent on γδT cells (127). In contrast with this study, another study identified two different γδ T cell subsets, including the Vγ1 γδ T cells that secrete IFN-γ only and the Vγ2 γδ T cells that secrete both IL-17A and IFN-γ (74). During S. japonicum infection, Vγ2 γδ T cells prevent hepatic fibrosis by recruiting neutrophils and secreting IL-17A (74). Thus, the role of innate γδ T cells in the pathology of schistosomiasis requires further investigation.

#### THE IMPACT OF OTHER IMMUNE CELLS ON CD4<sup>+</sup> T CELLS AND REGULATION ON IMMUNOPATHOLOGY OF SCHISTOSOMIASIS

Over the past four decades, immunoregulation has been deeply studied in the context of schistosomiasis. T-cell subsets are influenced by macrophages, B cells, DCs, and eosinophils, of which macrophages represent nearly 30% of the total granuloma cells. Macrophages, similar to Th1 and Th2 cells, can be classified into two major types, classically activated macrophages (CAM/M1) and alternatively activated macrophages (AAM/M2). CAM polarization, stimulated by inflammatory cytokines, such as IFN-γ, IL-12, and IL-18, serve a vital role in the response to intracellular pathogens like Mycobacterium tuberculosis (128), whereas AAM polarization is dependent on Th2-type cytokines, such as IL-4 and IL-13, that induce the expression of arginase-1 (Arg-1), Ym-1, and Fizzl, which is mainly involved in allergic, cellular, and humoral responses to parasites (129). Egg-induced inflammation and liver fibrosis promotes the Th2 response, which in turn increases AAM polarization in granulomas (130). In several models of Th1-polarized mice, such as IL-4/IL-10 deficient, IL-4/IL-13 deficient, and egg/IL-12-immunized or macrophagespecific IL-4 deficient (131, 132), schistosome infection failed to induce the expression of Arg-1, demonstrating that Th2 cytokines are necessary for the development of Arg-1-expressing AAM. Moreover, Th1-polarized mice showed an increased inducible nitric oxide synthase response together with smaller granulomas and elevated mortality (131, 132). Thus, CAM polarization is associated with Th1 responses and Arg-1 expressing AAM is associated with Th2 responses and liver fibrosis progression.

The function of B cells has also been intensely investigated in schistosomiasis. Several studies have demonstrated that the B-cell number in the lymph nodes and spleen significantly increase during the schistosome infection, suggesting that B cells are important for host responses against schistosome infection (133, 134). Mice immunized with radiation-attenuated cercariae of S. mansoni showed reduced protection against challenge infection in B-cell knockout mice compared to wild-type mice (135). IgE antibody against the worms, but not the eggs, has been closely associated with resistance to reinfection, and eosinophil-mediated antibody-dependent cellular cytotoxicity is the main immune mechanism to kill early schistosome larvae (136). Evidence shows that SEA-specific IgG1 is the main isotype released from the early granulomas and that, during the chronic infection stage, granulomas secrete a mixture of IgG1, IgG2, IgG3, and IgA (137). The antibodies produced within intragranulomas mainly function to downregulate the granuloma formation, and it has been found that immune complexes from chronic schistosomiasis patients can inhibit granuloma formation in vitro (138). In addition, research has been done on the regulatory role of B cells in schistosomiasis immunopathology and T-cell effector functions. A previous study has proven that B cells are important in promoting a strong Th2-type response to helminths (139). After SEA stimulation, mesenteric lymph cells from B-cell-deficient mice produce more Th1-type cytokines and less Th2-type cytokines compared to wild-type mice (140, 141). A recent study has shown that the absence of IL-4α signaling in B cells leads to increased mortality and pathology in S. mansoni-infected mice, which is attributed to decreased Th2 responses (142). These findings demonstrate that B cells are critical in directing and mediating Th cells toward Th2 responses. In addition, in a study about S. mansonimediated protection against experimental ovalbumin-induced allergic airway inflammation, the investigators found that splenic marginal zone CD1d<sup>+</sup> B cells from schistosome-infected mice had the ability to secret IL-10 and induce generation of FoxP3<sup>+</sup> Treg cells. Although B cells are essential for Th2 responses and Treg cell generation, B cells are not responsible for granuloma formation (140). However, a contrasting study showed that B-cell deficiency had no effect on Th2 responses but augmented tissue pathology (143). Therefore, further study is needed to clarify the role of B cells in immunopathology of schistosomiasis.

DCs are important antigen-presenting cells that have superior activity to stimulate naive T-cell activation and mediate the polarization of CD4<sup>+</sup> T cells in response to invading pathogens. DCs determine Th differentiation through secreted polarizing cytokines. For example, M. tuberculosis infection induces DCs to produce IL-12 to elicit a strong IFN-γproducing Th1 response, which leads to macrophage activation and killing of intracellular bacteria (144). Depletion of DCs in CD11c-diphtheria toxin receptor mice demonstrated that DCs are required to initiate Th2 responses in the S. mansoni infection (145). However, DC subsets, such as conventional CD11c<sup>+</sup> DCs and some specialized DCs do not produce IL-4, which is important for Th2 polarization. Ma et al. uncovered a novel subset of CD11c+CD49b+FcεRI<sup>+</sup> DCs that can produce IL-4 and subsequently promote Th2 differentiation in an IL-4-dependent manner (146). However, contradictory findings revealed that conventional DCs are critical subsets for Th2 effector cell development during acute S. mansoni infection (147). In fact, more researchers support the notion that DCs do not need to produce IL-4 to promote Th2 development (148–150) because IL-4-deficient DCs still show the ability to induce excellent Th2 responses (148). In addition, a recent study revealed an unrecognized role of type I IFN in the Th2 response, whereby IFN-I signaling is implicated in not only activating DCs but also enhancing DCs effective migration and antigen presentation during the S. mansoni-induced Th2 response (151). Thus, the exact mechanism by which DCs induce Th2 responses needs to be further clarified.

#### SUMMARY AND CONCLUDING COMMENTS

Schistosomiasis is a disease with profound impact on human health. At least 230 million people worldwide are affected by the parasitic disease (1). In the near future, we may face a situation where there is no available drug to treat schistosomiasis, as praziquantel is still the only effective drug being used to treat the disease and praziquantel resistance has been reported in endemic areas and in the laboratory (152, 153). Therefore, it is urgent to develop some alternative drugs to control schistosomiasis, including the use of vaccines (154– 156). Schistosome infection can lead to acute febrile illness and chronic life-threatening hepatosplenic disease. As summarized in **Figure 1**, during the progression of the disease, Th cells are activated and differentiated into distinct effector subsets, including Th1, Th2, Th17, Tfh, Th9, and Treg cells. Acute schistosomiasis is recognized as a Th1-dominated disease. Recent studies showed that Th17 and Tfh are also involved in the immune response of acute cases (71, 94). Chronic disease of hepatic granuloma formation and fibrosis are upregulated by Th2 and Th17 cells, mainly secreting IL-4 and IL-17, respectively (33, 63), and downregulated by Th1 and Treg cells (80, 113). The plasticity of Th cells are affected by the local microenvironment (e.g., cytokines and antigens of schistosomes) and regulated by various immune cells (e.g., macrophages, B cells, and DCs) through a complicated interacting network. Although we have acquired sufficient knowledge about the immunopathology of schistosomiasis, the effective strategies to restrain the development of granulomas and subsequent fibrosis are still lacking. A vaccine designed to skew the immune response toward the Th1 phenotype would be useful to prevent the development of fibrosis (80). However, murine studies have shown that extreme immune deviation toward either Th1 or Th2 results in increased pathology and premature death (18, 157). Most importantly, maintaining the balance of various effector T cells would be critical to prevent excessive pathology. While many researchers are focused on the Tregs-associated suppression and the powerful force of IL-10, successful immunotherapies will only be developed for schistosomiasis if we have a broader view

FIGURE 1 | T helper (Th) subsets differentiation and their roles in immunopathology of the murine schistosomiasis. During the progression of schistosomiasis, the schistosome antigen from schistosomula, adult worms, or eggs are captured and processed by antigen-presenting cells, such as dendritic cells (DCs) and macrophages; this leads to the production of various cytokines, and processed antigens are presented to naive CD4<sup>+</sup> T cell in the form of major histocompatibility complex (MHC) class II-peptide complexes, then activated CD4<sup>+</sup> T cells differentiate into Th1, Th2, Th17, Tfh, Th9, and Tregs, which are strictly regulated by the local microenvironment (e.g., cytokines and antigens of schistosomes) and cross-talk with various immune cells (e.g., macrophages, B cells, and DCs) through a complicated interacting network. It is generally thought that Th2, Th17 responses and Tfh, Th9 cells positively drive the development of pathology of chronic schistosomiasis, whereas the Th1 response counteracts Th2-mediated chronic pathology, and Tregs exert a suppressive effect on the immunopathology of schistosomiasis by inhibiting both Th1 and Th2 responses.

and deeper understanding of the mechanism of T-cell-mediated liver immunopathology.

## AUTHOR CONTRIBUTIONS

CL organized the article. BZ wrote the draft. JZ, HC, HN and HM participated in the conception, discussion and revision of the draft. QG edited the language and figure.

#### REFERENCES


#### FUNDING

This work was supported by National Natural Science Foundation of China (81861138002, 81722002, and 81271872), Joint Foundation of the Health Commission of Hubei Province (WJ2018H173), and Central Funds Guiding the Local Science and Technology Development of Hubei Province (2019ZYYD066).

schistosomiasis via interleukin-4 receptor alpha. PLoS Negl Trop Dis. (2017) 11:e0005861. doi: 10.1371/journal.pntd.0005861


and in combination with glyceraldehyde 3-phosphate dehydrogenase. PLoS Negl Trop Dis. (2017) 11:e0005443. doi: 10.1371/journal.pntd.0005443


regulating Th17-mediated inflammation. Eur J Immunol. (2009) 39:2470–81. doi: 10.1002/eji.200939325


is downregulated by a B cell-dependent mechanism requiring Fc receptor signaling. J Exp Med. (1998) 187:619–29. doi: 10.1084/jem.187.4.619


**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 © 2020 Zheng, Zhang, Chen, Nie, Miller, Gong and Liu. 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.

# NLRP6 Plays an Important Role in Early Hepatic Immunopathology Caused by Schistosoma mansoni Infection

Rodrigo C. O. Sanches<sup>1</sup>† , Cláudia Souza<sup>1</sup>† , Fabio Vitarelli Marinho<sup>1</sup> , Fábio Silva Mambelli<sup>2</sup> , Suellen B. Morais<sup>1</sup> , Erika S. Guimarães<sup>1</sup> and Sergio Costa Oliveira1,3 \*

#### Edited by:

Michael Harrison Hsieh, Children's National Hospital, United States

#### Reviewed by:

Malcolm Jones, The University of Queensland, Australia Zhipeng Xu, Nanjing Medical University, China

\*Correspondence:

Sergio Costa Oliveira scozeus1@gmail.com

†These authors have contributed equally to this work

#### Specialty section:

This article was submitted to Microbial Immunology, a section of the journal Frontiers in Immunology

Received: 05 December 2019 Accepted: 07 April 2020 Published: 05 May 2020

#### Citation:

Sanches RCO, Souza C, Marinho FV, Mambelli FS, Morais SB, Guimarães ES and Oliveira SC (2020) NLRP6 Plays an Important Role in Early Hepatic Immunopathology Caused by Schistosoma mansoni Infection. Front. Immunol. 11:795. doi: 10.3389/fimmu.2020.00795 <sup>1</sup> Departamento de Bioquímica e Imunologia, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil, <sup>2</sup> Departamento de Genética, Ecologia e Evolução, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil, <sup>3</sup> Instituto Nacional de Ciência e Tecnologia em Doenças Tropicais (INCT-DT), CNPq MCT, Salvador, Brazil

Schistosomiasis is a debilitating parasitic disease that affects more than 200 million people worldwide and causes approximately 280,000 deaths per year. Inside the definitive host, eggs released by Schistosoma mansoni lodge in the intestine and especially in the liver where they induce a granulomatous inflammatory process, which can lead to fibrosis. The molecular mechanisms initiating or promoting hepatic granuloma formation remain poorly understood. Inflammasome activation has been described as an important pathway to induce pathology mediated by NLRP3 receptor. Recently, other components of the inflammasome pathway, such as NLRP6, have been related to liver diseases and fibrotic processes. Nevertheless, the contribution of these components in schistosomiasis-associated pathology is still unknown. In the present study, using dendritic cells, we demonstrated that NLRP6 sensor is important for IL-1β production and caspase-1 activation in response to soluble egg antigens (SEA). Furthermore, the lack of NLRP6 has been shown to significantly reduce periovular inflammation, collagen deposition in hepatic granulomas and mRNA levels of α-SMA and IL-13. Livers of Nlrp6−/<sup>−</sup> mice showed reduced levels of CXCL1/KC, CCL2, CCL3, IL-5, and IL-10 as well as Myeloperoxidase (MPO) and Eosinophilic Peroxidase (EPO) enzymatic activity. Consistently, the frequency of macrophage and neutrophil populations were lower in the liver of NLRP6 knockout mice, after 6 weeks of infection. Finally, it was further demonstrated that the onset of hepatic granuloma and collagen deposition were also compromised in Caspase-1−/−, IL-1R−/<sup>−</sup> and Gsdmd−/<sup>−</sup> mice. Our findings suggest that the NLRP6 inflammasome is an important component for schistosomiasis-associated pathology.

Keywords: Schistosoma mansoni, immunopathology, inflammasome, NLRP6, fibrosis

## INTRODUCTION

fimmu-11-00795 May 2, 2020 Time: 21:39 # 2

Schistosomiasis is a debilitating parasitic disease which affects 78 countries worldwide. This disease leads to approximately 200,000 deaths annually and severely compromises the life quality from those affected. Among causative species, Schistosoma haematobium, Schistosoma japonicum, and Schistosoma mansoni stand out as those of major importance to human health (1– 3). Infection occurs through the direct contact of the host with the parasite's larval form. After parasite penetration and sexual development, egg laying begins. The release of eggs in the feces and its hatching in the environment closes the parasite's life cycle (4). However, a significant amount of these eggs is trapped in some of the host's organs, such as liver and intestine, where they induce a granulomatous inflammatory reaction (5, 6). Hepatic granulomatous inflammation arises from the egg-secreted antigens, which perform hepatotoxic and immunological activities capable of recruiting immune cells to the organ and forming periovular granuloma. The composition of the granuloma includes macrophages, eosinophils, neutrophils, T and B lymphocytes and especially fibroblasts, responsible for the fibrotic characteristic of the structure (7, 8).

Although the process of hepatic granuloma formation is extensively studied, all cellular events and key participants have not been fully established yet. The role of intracellular immune receptors in granuloma formation, for instance, was first described in a seminal study conducted by Ritter and colleagues (9). However, a better understanding of cytosolic sensors during S. mansoni infection is required. These intracellular receptors are those responsible for activating the inflammasome pathway. This pathway induces the formation of an intracellular protein complex typically consisting of Nucleotide-binding oligomerization domain (NOD), leucine-rich repeat (LRR) containing protein (NLR) family members, an adapter molecule known as ASC, and the cysteine protease caspase-1 as an effector molecule. Activation of this pathway leads to cleavage of immature forms of IL-1β and IL-18 into their mature forms. It might also induce cell death by pyroptosis. The inflammasome activation takes place in both immune and non-immune cells and is essentially triggered by pathogen-associated molecular patterns (PAMPs) and Danger-associated molecular patterns (DAMP) (10, 11).

It is known that the inflammasome pathway plays an important role during chronic liver diseases (12). Besides fighting pathogens such as bacteria (13) and viruses (14), inflammasome also participates in aggravating sterile liver inflammations such as Alcoholic Liver Disease (ALD) (15) and Non-alcoholic Steatohepatitis (NASH) (16). NLRP3 is the most widely studied receptor in this context since it is activated by several types of insults (17). On the other hand, the participation of other NLR family receptors in hepatic pathological processes, such as NLRP6, is still elusive.

The inflammasome pathway plays an essential role in schistosomiasis-associated liver pathology. It has been demonstrated that NLRP3 is critical for granuloma formation and hepatic stellate cells (HSCs) activation (18) in S. japonicum infections. Regarding S. mansoni infection this same sensor has been shown to be involved in the adaptive immune response and also granuloma formation (9). Recent studies have reported that NLRP3 and NLRP6 expression are simultaneously modulated in some processes, including those occurring in the liver (19, 20). Additionally, the role of NLRP6 in fibrotic diseases has already been described (21, 22). Thus, we decided to investigate whether the NLRP6 sensor plays a role in the course of S. mansoni infection and liver pathology. In this study, we demonstrate that lack of NLRP6 modulates the formation of hepatic granuloma, influencing local chemokine and cytokine production as well as macrophage and neutrophil recruitment into the liver. Also, this receptor is important for promoting collagen deposition.

## MATERIALS AND METHODS

### Ethics Statement

This study was carried out in accordance with Brazilian laws #6638 and #9605 in Animal Experiments. The protocol was approved by the Committee on Ethics of Animal Experiments of the Universidade Federal de Minas Gerais (UFMG) (Permit Number: #367/2017).

### Mice and Parasite

Wild-type C57BL/6 mice were purchased from the Universidade Federal de Minas Gerais (UFMG). Nlrp3−/−, Nlrp6−/−, Casp-1 <sup>−</sup>/−, IL-1R−/−, and Gsdmd−/<sup>−</sup> were described previously (23– 26). The animals were maintained at UFMG and used at 6– 10 week of age. Schistosoma mansoni (LE strain) cercariae at Fundação Oswaldo Cruz – Centro de Pesquisas René Rachou (CPqRR-Brazil) were routinely obtained from infected Biomphalaria glabrata snails exposed to light, inducing the shedding of parasites.

## Eggs, SEA, and SWAP

Eggs were obtained from 50-day-infected Swiss mice livers. Briefly, the liver was blender processed in cold saline (2% NaCl) for 2 min. Next, the material was decanted into a glass goblet for 35 min at low temperature. Part of the decanting supernatant was discarded and the remaining solution was washed with cold saline. Decantation-washing was repeated until reaching a translucent solution. Eggs were recovered by filtration. For the preparation of Soluble Egg Antigens (SEA), eggs were disrupted for 40 min at low temperature in PBS and then the homogenate was centrifuged at 100,000 × g for 1 h at 4◦C. The resulting supernatant was frozen at −80◦C. Soluble adult worm antigen (SWAP) was obtained by mechanical maceration of worms in cold PBS. After centrifugation (13,000 × g – 7 min), the supernatant was collected and stored at −80◦C. The protein concentration of SEA and SWAP was determined using BCATM protein assay kit (Thermo Fisher Scientific, Waltham, MA, United States).

#### BMDC Generation and Activation

To obtain bone marrow-derived dendritic cells (BMDCs), bone marrow cells were cultured in RPMI with 10% FBS, 100 U/mL penicillin, 100 µg/mL streptomycin and 20 ng/mL murine

recombinant GM-CSF (Peprotech, Riberão Preto, Brazil). Petri dishes containing 1 × 10<sup>7</sup> cells were incubated at 37◦C in 5% CO2. At day 3 of incubation, 5 mL of fresh complete medium with GM-CSF was added, and 5 mL of medium was replaced with fresh supplemented medium containing GM-CSF on days 5 and 7. At day 10, non-adherent cells were harvested and seeded in 24-well plates (5 × 10<sup>5</sup> cells/well). Stimulation of BMDCs was performed by priming cells with 1 µg/ml of Pam3Cys (Sigma-Aldrich, St. Louis, MO, United States) for 5 h and then stimulating with 50 µg/mL of SEA for 17 and 24 h. As positive control for inflammasome activation, cells were primed with 1 µg/ml of Pam3Cys (5 h) or 1 µg/ml of LPS (4 h) and stimulated with ATP (5 mM) (50 min) or Nigericin (20 µM) (50 min). Culture supernatants were collected and cells were lysed with M-PER Mammalian Protein Extraction Reagent (Thermo Fisher Scientific) supplemented with 1:100 protease inhibitor mixture (Sigma-Aldrich).

#### Western Blotting

Cell lysates and supernatants from DCs culture were subjected to SDS-PAGE analysis and western blotting. The proteins were resolved on a 15% SDS-PAGE gel, and transferred to nitrocellulose membranes (Amersham Biosciences, Uppsala, Sweden). Membranes were blocked for 1 h in TBS (0.1% Tween-20; 5% non-fat dry milk) and incubated with primary antibodies at 4◦C, overnight. Primary antibody used was mouse monoclonal against the p20 subunit of caspase-1 (Adipogen, San Diego, CA, United States). Monoclonal antibody against β-actin (Cell Signaling Technology, Danvers, MA, United States) was used as a loading control blot (1:1,000). The membranes were washed three times for 10 min in TBS with 0.1% Tween 20. Next, membranes were incubated for 1 h at room temperature with the suitable HRP-conjugated secondary antibody (1:1,000). Immunoreactive bands were visualized using Luminol chemiluminescent HRP substrate (Millipore).

### Splenocyte Culture

Spleen cells were obtained from macerated spleens of individual C57BL/6 and Nlrp6−/<sup>−</sup> mice after 6 weeks of infection with S. mansoni cercariae (n = 5/group). Cells were washed with PBS and the erythrocytes were lysed with a hemolytic solution (155 mM NH4Cl, 10 mM KHCO3, pH 7.2). Cells were adjusted to 1 × 10<sup>6</sup> /well in complete RPMI medium (10% fetal bovine serum, 100 U/mL penicillin and 100 µg/mL streptomycin). Spleen cells were cultured in 96-well plates with medium and stimulated with SWAP (200 µg/mL), SEA (20 µg/mL), Eggs (50/well) or concanavalin A (ConA) (5 µg/mL). Culture supernatants were collected after 24 h for IL-5 and after 72 h for IFN-γ, IL-10, and IL-13 measurements by ELISA.

## Liver Processing

Right lobe of liver from 6-week-infected C57BL/6 and Nlrp6−/<sup>−</sup> mice was collected and 1 mL of cytokine extraction solution (0.4 M NaCl, 0.05% Tween 20, 0.5% BSA, 0.1 mM PMSF, 0.1 mM benzethoniumchloride, 10 mM EDTA and 20 KI aprotinin) was added to each 100 mg of tissue. Ultra-Turrax homogenizerdispenser was used to homogenize solutions containing the organs. Next, the samples were centrifuged at 10,000 × g for 10 min at 4◦C. Non-parenchymal cells from left lobe were used for flow cytometry analysis. Tissue was removed without perfusion, cut into small pieces, incubated in RPMI medium containing 30 µg/ml of Liberase TM (Roche) and 20 U/mL of DNAse I (GE) for 40 min, and passed through a 70 µm poresize cell strainer. After centrifugation, the cells were resuspended in PBS containing 2% fetal bovine serum (FBS) and 5 mM EDTA. Low-speed centrifugation (50 × g – 5 min) was used to remove parenchymal cells. Erythrocytes were lysed with a hemolytic solution (155 mM NH4Cl, 10 mM KHCO3, pH 7.2). The remaining non-parenchymal cells were resuspended in RPMI culture medium.

#### Cytokine Measurements

Cytokine/Chemokine production was evaluated using the Duoset ELISA kit (R&D Diagnostic, Minneapolis, MN, United States) according to the manufacturer's instructions.

## EPO and MPO Activity Assays

Eosinophilic Peroxidase and Myeloperoxidase assays were performed as described by Cançado et al. (27). Right lobe of the liver was homogenized, red blood cells subjected to hypotonic lysis and the remaining liver cells subjected to detergent lysis and freeze-thaw cycles. The enzymatic assay was performed using the suitable substrates and the result was measured on a microplate reader at the appropriate wavelength (492 nm for EPO and 450 nm for MPO). The result was expressed in absorbance units.

#### Flow Cytometry Analysis

Spleen and non-parenchymal liver cells were stained for CD11b, CD11c, Ly6G, F4/80, CD3, and CD4. Briefly, cells were incubated for 20 min with anti-mouse CD16/32 (BD Biosciences) in FACS buffer (PBS, 1% FBS, 1 mM NaN3) and were stained for surface markers for 20 min using: APC-Cy7-conjugated anti-mouse CD11b (1:200, M1/70; BD Biosciences), FITCconjugated anti-mouse CD11c (1:100, HL3; BD Biosciences), PE-conjugated anti-mouse Ly6G (1:200, 1A8; BD Biosciences), biotinylated anti-mouse F4/80 (1:200, BM8; BD Biosciences), PE-Cy7-conjugated anti-mouse CD3 (1:100, BD Biosciences) and APC-conjugated anti-mouse CD4 (1:200, BD Biosciences). The appropriate isotype controls were used. Next, cells were washed and incubated for 20 min at 4◦C in the dark with PerCP-Cy5.5 conjugated streptavidin (1:200 BD Biosciences). Lastly, cells were washed and resuspended in PBS. Attune Flow Cytometer (Applied Biosystems, Waltham, MA, United States) was used for collecting approximately 100,000 events and data were analyzed using FlowJo software (Tree Star, Ashland, OR, United States).

In order to evaluate macrophage polarization, nonparenchymal liver cells were stained as described above using APC-Cy7-conjugated anti-mouse CD11b (1:200, M1/70; BD Biosciences), biotinylated anti-mouse F4/80 (1:200, BM8; BD Biosciences), BB700-conjugated anti-mouse CD197 (1:200, 4B12

BD Biosciences), FITC-conjugated anti-mouse CD80 (1:200, 16-10A1, BD Biosciences), PE-conjugated anti-mouse CD163 (1:200, TNKUPJ, eBioscience) and APC-conjugated anti-mouse CD206 (1:200, MR5D3, BD Biosciences).

#### Quantitative Real-Time PCR

Liver middle lobe of 6-week-infected C57BL/6 and Nlrp6−/<sup>−</sup> mice was used for RNA extraction. The tissue was homogenized in TRIzol (Invitrogen) and total RNA was isolated in accordance with the manufacturer's instructions. Reverse transcription of total RNA was performed and quantitative real-time RT-PCR was conducted in a final volume of 20 µL containing SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, United States), oligo-dT cDNA as the PCR template and 2.5 µM of primers. The PCR reaction was performed with QuantStudio3 real-time PCR instrument (Applied Biosystems). Primers were used to amplify a specific fragment (100–120 bp) corresponding to specific gene targets as follows: 18S Forward (5<sup>0</sup> -CGTTCC ACCAACTAAGAACG-3<sup>0</sup> ) and 18S Reverse (5<sup>0</sup> -CTCAACACGG GAAACCTCAC-3<sup>0</sup> ; α-SMA Forward (5<sup>0</sup> -GTCCCAGACATCAG GGAGTAA-3<sup>0</sup> ) and α-SMA Reverse (5<sup>0</sup> -TCGGATACTTCA GCGTCAG-3<sup>0</sup> ); IL-13 Forward (5<sup>0</sup> -CCTGGCTCTTGCTTGCC-3 0 ) and IL-13 Reverse (5<sup>0</sup> -GGTCTTGTGTGATGTTGCTCA-3 0 ); IL-1β Forward (5<sup>0</sup> -TGACCTGGGCTGTCCAGATG-3<sup>0</sup> ) and IL-1β Reverse (5<sup>0</sup> -CTGTCCATTGAGGTGGAGAG-3<sup>0</sup> ); Casp-1 Forward (5<sup>0</sup> -GGAAGCAATTTATCAACTCAGTG-3<sup>0</sup> ) and Casp-1 Reverse (5<sup>0</sup> -GCCTTGTCCATAGCAGTAATG-3<sup>0</sup> ).

#### Mice Infection and Parasite Burden

Six-to-eight-week-old wild-type and knockout mice were anesthetized with 5% ketamine, 2% xylazine and 0.9% NaCl and then infected with 100 cercariae (LE strain) through exposure of percutaneous abdominal skin, for 1 h. After 6 weeks of infection, mice were euthanized and perfused from the portal veins, the recovered worms were counted and the mean difference between groups of mice was evaluated.

#### Pathological Parameters

Number of eggs was obtained from liver median lobe. The tissue was weighed and digested in an aqueous solution of KOH (5%) for 16 h at 37◦C. After, eggs were washed in saline and centrifuged twice at 270 × g for 10 min and counted using a light microscope. The number of calculated eggs was corrected by considering the mass of the tissue, resulting in number of eggs per gram of liver. The left lobe was fixed with 10% buffered formaldehyde in PBS. Histological sections were performed using microtome at 6 µm and stained on a slide with Hematoxylin-Eosin (HE) and Masson blue. For measurement of granuloma size and collagen deposition, a JVC TK-1270/RBG camera, attached to the microscope (10 × objective lens), was used to obtain the images. Analysis were carried out using ImageJ software (U.S. National Institutes of Health, Bethesda, MD, United States)<sup>1</sup> . Granuloma size was measured, in µm<sup>2</sup> , for all granulomas found in liver sections.

#### <sup>1</sup>http://rsbweb.nih.gov/ij/index.html

#### Statistical Analysis

The statistical tests were performed using Student's t-test, oneway and two-way ANOVA followed by Bonferroni adjustments for comparison between groups. P-values obtained were considered significant if they were <0.05. Statistical analysis was performed using GraphPad Prism 6 (La Jolla, CA, United States).

### RESULTS

#### IL-1β Production and Caspase-1 Activation Are Partially Dependent on NLRP6 in SEA/Eggs-Stimulated Dendritic Cells

During S. mansoni infection, hepatic dendritic cells (DCs) are the main cells responsible for promoting the shift from Th1 to Th2 immune profile, which is triggered by egg antigens (28). Ritter and colleagues (2010) reported that SEA induces inflammasome activation in DCs and described the involvement of NLRP3 (9). Thus, given the relevance of these cells, we decided to use bone marrow-derived dendritic cells (BMDCs) in order to investigate whether the NLRP6 sensor is involved in IL-1β secretion in response to different parasite antigens. Using Pam3Cys (P3Cys) as the first signal, we observed that eggs and their soluble antigens (SEA) induce high levels of IL-1β and this production was partially influenced by NLRP6 (**Figure 1A**). In contrast, IL-1β production induced by soluble adult worm antigens (SWAP) was much lower compared to SEA (**Figure 1A**). Additionally, TNF-α levels were not altered comparing both WT and NLRP6 knockout (KO) mice (**Figure 1B**). Since SEA was sufficient to induce IL-1β production, we used this stimulus to evaluate the role of NLRP6 in caspase-1 (casp-1) activation. **Figures 1C,D** demonstrate that SEA induces activation of casp-1 in WT DCs and this process was clearly inhibited in Nlrp6−/<sup>−</sup> DCs.

#### NLRP6 Influences Granuloma Formation and Collagen Deposition in the Liver

Since the liver is one of the main entrapment tissues for parasite's eggs and once the participation of NLRP6 in the egg and SEAinduced immune response has been observed, we wondered if this sensor could play any role in liver pathology. First, we observed that in livers of 4- and 6-week infected animals the levels of IL-1β and caspase-1 mRNA did not significantly change between Nlrp6−/<sup>−</sup> and WT mice (**Supplementary Figure S1**). Initially, we observed that the number of eggs per gram of tissue was not altered between WT and knockout mice (**Figure 2A**). Consistently, the worm burden recovery was the same comparing both groups (**Supplementary Figure S2**). On the other hand, NLRP6 has been shown to influence the periovular inflammatory response, contributing significantly to granuloma formation (**Figure 2B**). In addition, collagen deposition within the granulomatous structure was reduced in Nlrp6−/<sup>−</sup> mice when compared to WT (**Figures 2B,C,H**). We also evaluated levels of cytokines and fibrotic markers within the tissue, such as the cytokines IL-5, IL-10, IL-13, and the protein Alpha-smooth muscle actin (α-SMA). IL-5 and IL-10 cytokines

and stimulated with SEA (50 µg/mL – 17 h, 24 h), Eggs (100 eggs/well – 24 h), SWAP (200 µg/mL – 24 h) or ATP (5 mM – 50 min). For nigericin control (20 µM), cells were primed with LPS (1 µg/ml – 4 h) and stimulate for 50 min. (A) IL-1β and (B) TNF-α were measured by ELISA. Casp-1 activation was analyzed by western blot in (C) cell lysate and (D) supernatant using antibody against p20 subunit. An asterisk denotes statistically significant differences between NLRP6 versus WT animals (p < 0.05).

in WT and Nlrp6−/<sup>−</sup> mice. (A) Number of eggs per gram of liver, (B) Granuloma size (µm<sup>2</sup> ) and (C) Collagen deposition. (D) IL-5 and (E) IL-10 cytokine levels were detected. (F) α-SMA and (G) IL-13 transcripts were also measured. (H) Representative images of granulomas detected in hematoxylin-eosin, masson blue and the output from ImageJ software, respectively. An asterisk denotes statistically significant differences between NLRP6 versus WT animals (p < 0.05). The bar represents 100 µm. NI stands for Non-infected mice.

levels, as well as, α-SMA and IL-13 mRNA measurements were reduced in Nlrp6−/<sup>−</sup> group when compared to WT (**Figures 2D– G**). These data demonstrate that NLRP6 contributes to the pathology caused by S. mansoni.

## NLRP6 Influences IL-10 and IFN-γ Production by Spleen Cells Activated With Egg Antigens

After determining the impact of NLRP6 on inflammasome activation and granuloma formation in response to egg antigens, we decided to investigate the frequency of dendritic cells and CD4<sup>+</sup> T lymphocytes in spleen cells derived from S. mansoni infected mice. Cells were obtained following the gate strategy described in **Supplementary Figure S3**. We observed that 6-week-infected Nlrp6−/<sup>−</sup> and WT mice presented no significant difference regarding CD11b+CD11c<sup>+</sup> (dendritic cells) and CD3+CD4<sup>+</sup> (CD4<sup>+</sup> lymphocytes) cell populations (**Figures 3A,B**). Interestingly, when spleen cells were stimulated with eggs or SEA, the cytokine production was altered. Both antigens induced increased levels of IL-10 and IFN-γ in splenocyte culture supernatants from Nlrp6−/<sup>−</sup> compared to WT mice (**Figures 3E,F**). Additionally, NLRP6 appears to have no effect on IL-5 and IL-13 production (**Figures 3C,D**). Curiously, when Nlrp6−/<sup>−</sup> and WT spleen cells from S. mansoni infected mice were stimulated with SWAP, no significant difference on IFN-γ, IL-10, IL-5, and IL-13 levels was observed (**Supplementary Figure S4**). These data emphasize the relevance of NLRP6 during egg antigen response regulating IL-10 and IFN-γ production. Furthermore, enhanced IL-10 production in Nlrp6−/<sup>−</sup> may be related to reduced granuloma formation and fibrosis.

#### NLRP6 Mediates Innate Immune Cells Recruitment in Schistosoma-Infected Liver

Since lack of NLRP6 has been shown to modulate hepatic granuloma formation, we decided to investigate how this sensor influences liver pathology. Initially, we evaluated the level of chemokines (CCL2, CCL3, CCL11, and CXCL1) in livers of Schistosoma-infected mice. These chemokines have already been described as related to granuloma formation. In Schistosoma-infected Nlrp6−/<sup>−</sup> mice only the production of CCL11 was not reduced in comparison to WT mice (**Figure 4D**). CCL2, CXCL1, and CCL3 were diminished in Nlrp6−/<sup>−</sup> mice compared to WT animals (**Figures 4A–C**). Additionally, we observed that the enzymes MPO and EPO were also reduced in Nlrp6−/<sup>−</sup> mice when compared to WT (**Figures 4E,F**). Our next step was to evaluate which non-parenchymal cell populations could be altered in Schistosoma-infected Nlrp6−/<sup>−</sup> mice. Cells were obtained following the gate strategy described in **Supplementary Figure S5**. Neutrophils and macrophages were the major cell populations reduced in Nlrp6−/<sup>−</sup> mice when compared to WT (**Figures 4G,H**). The frequency of dendritic cells and CD4<sup>+</sup> T lymphocytes remained unaltered in either mouse groups (**Figures 4I,J**). We also found that the lack of NLRP6 does not affect macrophage polarization, even though there is a strong tendency of reduction in anti-inflammatory macrophages (CD206+CD163+) in Nlrp6−/<sup>−</sup> compared to WT mice (**Supplementary Figure S6**). Therefore, the NLRP6 sensor possibly induces the formation of hepatic granuloma by favoring chemokine production and the recruitment of immune cells to the liver.

### Inflammasome Pathway Is Broadly Relevant to Granuloma Formation

The NLR family receptors perform their functions depending on tissue and cell type, as already demonstrated for NLRP3 and NLRP6 (29, 30). For this reason, we decided to evaluate whether other inflammasome pathway-related molecules, such as Casp-1, GSDMD, and IL-1R, played a role in the formation of hepatic granuloma. Similarly to what we observed here for Nlrp6−/<sup>−</sup> mice, a significant reduction in periovular inflammatory response in Casp-1−/−, Gsdmd−/−, and IL-1R−/<sup>−</sup> mice was observed. In addition, a reduction in collagen deposition was observed in the granulomas of Casp-1−/−, Gsdmd−/−, and IL-1R−/<sup>−</sup> mice when compared to WT (**Figures 5B–D**). This was accompanied by no alteration in the number of eggs in the tissue (**Figure 5A**), and worm burden recovery (**Supplementary Figure S2**) when compared to WT mice.

## DISCUSSION

The role of the inflammasome pathway in the pathogenesis of chronic liver diseases has been investigated in the last few years (31, 32). Hepatic injuries from different sources are capable of leading to inflammasome activation, as described for druginduced damage (33), ischemia-reperfusion (34), alcoholic and non-alcoholic fatty liver disease (15, 16), and viral hepatitis (14). Inflammasome triggers or amplify liver diseases by releasing pro-inflammatory cytokines such as IL-1β, IL-1α, IL-18, and also through other inflammatory mediators such as High Mobility Group Box 1 (HMGB1) (35). Release of such cytokines and DAMPs occur, in part, due to Gasdermin-D cleavage and subsequent pyroptosis of the cell (36). The fibrotic process resulting from chronic liver diseases has also involved the inflammasome pathway. IL-1β and danger signals induce Hepatic Stellate Cells (HSC) to transdifferentiate and perform extracellular matrix remodeling function. In addition, HSCs can internalize pre-formed inflammasome complexes released by other pyroptosis dying cells (37–39).

The participation of the NLRP3 receptor in chronic liver diseases such as non-alcoholic fatty liver disease (NAFLD) is well described (16, 17). On the other hand, the role of the NLRP6 sensor in these liver pathologies is still elusive. Recently, Xiao and colleagues (2018) reported that in the NAFLD model induced by methionine-choline deficient (MCD) diet, NLRP3 and NLRP6 expression is highly detected in the liver. After Lycium barbarum polysaccharides (LBP) treatment, NAFLD condition improves and the expression of both NLR receptors decreases (20). Besides, a previous study with NAFLD obese patients demonstrates that when hepatic portal fibrosis is present, the expression of NLRP6 mRNA in adipose tissues is higher compared to cases when

hepatic portal fibrosis is not observed (21). These findings suggest that NLRP6 might play important role in chronic liver disease and fibrosis. All those findings intrigued us to evaluate the role of this sensor in S. mansoni infection.

Schistosoma mansoni and S. japonicum infections are sources of injury and are able to induce chronic liver disease. The long survival period of Schistosoma worms within the human host implies recurrent inflammation and wound-healing cycles in the liver, triggered by egg antigens which can result in fibrosis, portal hypertension and hepatosplenomegaly (4, 8). Among non-parenchymal liver cells, DCs stand out as key cells during this pathological process. Broadly responsive to egg antigens, DCs are essential for promoting systemic shift in the immune response profile (from Th1 to Th2), which is crucial for host survival upon infection (28, 40). In addition, the first report of inflammasome pathway activation by S. mansoni antigens involved dendritic cells responding to SEA. For these reasons, we initially decided to evaluate the role of NLRP6 in DCs. Our findings suggest that NLRP6 is important for the formation of the inflammasome complex, since it influences IL-1β secretion

and caspase-1 activation in response to SEA. Previous studies demonstrate that NLRP6 structurally has the ability to form the inflammasome complex and does so in response to grampositive bacteria cell wall components, activating both caspase-1 and caspase-11 in the same complex (41, 42).

The activation of intracellular receptors was not expected to occur in response to multicellular parasites such as S. mansoni. Surprisingly, Ritter and colleagues (2010) demonstrated that SEA triggers NLRP3 inflammasome pathway in DCs (9). Following this seminal study, NLRP3 inflammasome role has been investigated, especially in S. japonicum infection. In this context, NLRP3 has been shown to be pivotal for inducing hepatic granuloma formation and collagen deposition in the granulomatous structure (18, 43). In this study, we confirmed that NLRP3 is pivotal for hepatic granuloma formation, but we did not find any alteration in collagen deposition in Nlrp3−/<sup>−</sup> mice when compared to WT (**Supplementary Figure S7**). This observation may be related to the early time point infection of our model (6 weeks), which also could explain the distribution of collagen throughout the granuloma structure and not only peripherally, as typically observed in later granulomas. Surprisingly, we have demonstrated for the first time that NLRP6 also plays an important role in S. mansoni-induced pathology. We found that this sensor influences the formation of hepatic granuloma, altering local chemokine (CCL2, CCL3, and CXCL1) and cytokine (IL-5, IL-10, and IL-13) production, macrophage and neutrophil recruitment into the liver, and also is important for promoting collagen deposition. In Schistosoma egg-induced pathology, chemokine production is essential to modulate granuloma formation (44, 45). CCL3-deficient mice, for instance, showed size reduced granuloma, lower fibrosis and lower EPO activity in the liver (46). Although classically responsible for neutrophil recruitment, CXCL1 also impacts on the recruitment of HSCs, which are responsible for collagen

deposition in the granuloma structure (47). Following injury, hepatic resident cells produce CCL2, important for monocyte and macrophage recruitment (48).

NLRP6 is known as an atypical sensor with wide functional capability, performing activities integrated and/or independent on the inflammasome complex (30). Our findings demonstrate that Casp-1 activation and IL-1β production, in response to SEA, can be regulated by NLRP6. Therefore, our findings suggest that the role of this receptor in hepatic granuloma formation is due to the inflammasome activation, since Nlrp3−/<sup>−</sup> mice have demonstrated a similar phenotype. It is clear that the inflammasome pathway is important for Schistosoma-induced liver pathology, once we have observed granuloma reduction and lower collagen deposition in Casp-1−/−, Gsdmd−/−, and IL-1R−/<sup>−</sup> mice when compared to WT. Previous liver disease studies support our observations for Casp-1−/<sup>−</sup> and IL-1R−/<sup>−</sup> mice. In a high fat diet-induced NASH model, deficient Casp-1 animals showed improvement in hepatic steatosis, inflammation and fibrogenesis (49). Similarly, IL-1R signaling has been shown to be critical for the progression of steatohepatitis and hepatic fibrosis in hypercholesterolemic mice (50). The role of Gasdermin- D and pyroptosis has also been described in chronic liver diseases (51, 52). During human NAFLD/NASH, GSDMD and its N-terminal peptide (GSDMD-N) are upregulated, besides MCD-fed Gsdmd−/<sup>−</sup> mice showed decreased severity of steatosis and inflammation comparing to WT (52). During Schistosoma infection, Liu and colleagues (2019) demonstrated that S. japonicum induces expression of the GSDMD-N in the

liver, and that this expression is modulated by NLRP3 sensor (53). In addition, it has been reported that SEA from S. japonicum eggs induces pyroptosis in HSCs (54). However, no in vivo mouse study has been reported correlating GSDMD deficiency and granuloma formation and fibrosis induced by S. mansoni infection as demonstrated here.

In summary, the data presented here demonstrate that lack of NLRP6 modulates activation of the inflammasome pathway in response to S. mansoni egg antigens. In addition, NLRP6 and the inflammasome components are important in liver pathology induced by S. mansoni infection. Taken together, these data reinforce the relevance of understanding the inflammasome signaling pathway, given its potential to influence the severe pathological conditions induced by this disease.

#### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the article/**Supplementary Material**.

#### ETHICS STATEMENT

The animal study was reviewed and approved by the Committee on Ethics of Animal Experiments of the Federal University of Minas Gerais (Permit Number: #367/2017) and carried out in accordance with Brazilian laws #6638 and #9605 in Animal Experiments.

#### REFERENCES


### AUTHOR CONTRIBUTIONS

RS and SO designed the project and experiments, and wrote the manuscript. RS, CS, FVM, FSM, EG, and SM carried out most of the experiments. RS carried out statistical analysis and prepared the figures. SO submitted this manuscript. All authors reviewed the manuscript.

#### FUNDING

This study was funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico grants #302660/2015- 1 and 465229/2014-0, and Fundação de Amparo à Pesquisa do Estado de São Paulo grant # 2017/24832-6.

#### ACKNOWLEDGMENTS

We thank Dr. Cristina T. Fonseca from Fundação Oswaldo Cruz – Centro de Pesquisas René Rachou (CPqRR-Brazil) for providing S. mansoni cercariae to carry out the experiments.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu. 2020.00795/full#supplementary-material


alkali burn injury. Exp Eye Res. (2018) 176:59–68. doi: 10.1016/j.exer.2018.07. 001


MAPK pathways in rat hepatic stellate cells. Mol Med Rep. (2013) 8:1861–5. doi: 10.3892/mmr.2013.1719


**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 © 2020 Sanches, Souza, Marinho, Mambelli, Morais, Guimarães and Oliveira. 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.

# Perforin Acts as an Immune Regulator to Prevent the Progression of NAFLD

Qian Wang1,2 \* † , Dehai Li 1,2†, Jing Zhu1,2†, Mingyue Zhang1,2, Hua Zhang1,2 , Guangchao Cao1,2, Leqing Zhu1,2, Qiping Shi <sup>3</sup> , Jianlei Hao1,2, Qiong Wen1,2 , Zonghua Liu1,2, Hengwen Yang1,2 \* and Zhinan Yin1,2 \*

*<sup>1</sup> Zhuhai Precision Medical Center, Zhuhai People's Hospital (Zhuhai Hospital Affiliated with Jinan University), Jinan University, Zhuhai, China, <sup>2</sup> The Biomedical Translational Research Institute, Faculty of Medical Science, Jinan University, Guangzhou, China, <sup>3</sup> The First Affiliated Hospital of Jinan University, Guangzhou, China*

#### Edited by:

*Gary Splitter, University of Wisconsin-Madison, United States*

#### Reviewed by:

*Nadia Maria Terrazzini, University of Brighton, United Kingdom Erik Petersen, East Tennessee State University, United States*

#### \*Correspondence:

*Qian Wang wangqian@jnu.edu.cn Hengwen Yang hengwenyang@jnu.edu.cn Zhinan Yin zhinan.yin@yale.edu*

*†These authors share first authorship*

#### Specialty section:

*This article was submitted to Inflammation, a section of the journal Frontiers in Immunology*

Received: *15 October 2019* Accepted: *14 April 2020* Published: *22 May 2020*

#### Citation:

*Wang Q, Li D, Zhu J, Zhang M, Zhang H, Cao G, Zhu L, Shi Q, Hao J, Wen Q, Liu Z, Yang H and Yin Z (2020) Perforin Acts as an Immune Regulator to Prevent the Progression of NAFLD. Front. Immunol. 11:846. doi: 10.3389/fimmu.2020.00846* Non-alcoholic fatty liver disease (NAFLD) is one of the main causes of cirrhosis and major risk factors for hepatocellular carcinoma and liver-related death. Despite substantial clinical and basic research, the pathogenesis of obesity-related NAFLD remains poorly understood. In this study, we show that perforin can act as an immune regulator to prevent the progression of NAFLD. Aged perforin-deficient (Prf−/−) mice have increased lipid accumulation in the liver compared to WT mice. With high-fat diet (HFD) challenge, Prf−/<sup>−</sup> mice have increased liver weight, more severe liver damage, and increased liver inflammation when compared with WT controls. Mechanistic studies revealed that perforin specifically regulates intrinsic IFN-γ production in CD4 T cells, not CD8 T cells. We found that CD4 T cell depletion reduces liver injury and ameliorates the inflammation and metabolic morbidities in Prf−/<sup>−</sup> mice. Furthermore, improved liver characteristics in HFD Prf−/<sup>−</sup> and IFN-γR <sup>−</sup>/<sup>−</sup> double knockout mice confirmed that IFN-γ is a key factor for mediating perforin regulation of NAFLD progression. Overall, our findings reveal the important regulatory role perforin plays in the progression of obesity-related NAFLD and highlight novel strategies for treating NAFLD.

#### Keywords: perforin, NAFLD, CD4 T cells, IFN-γ, inflammation

#### INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) pathogenesis is tightly linked to obesity and therefore is an emerging healthcare problem worldwide (1, 2). NAFLD, along with related inflammation, progressive subtype non-alcoholic steatohepatitis (NASH), fibrosis, and ultimately hepatocellular carcinoma, is becoming one of the leading causes of liver-related morbidity and mortality worldwide (3–5). The pathogenesis of NAFLD remains incompletely understood. It is appreciated that multiple concurrent intrahepatic and extrahepatic events contribute to development and progression of NAFLD, including cell senescence, insulin resistance, and immune system dysfunction (6, 7). Cellular senescence refers to the irreversible arrest of cell growth that occurs when cells are exposed to various stresses (8–10). Recent experimental evidence suggests that hepatocyte senescence is linked to the fibrosis that develops as NAFLD progresses; hepatocyte expression of p21, the universal cell cycle inhibitor, is positively correlated with fibrosis stage in liver sections from 70 NAFLD patients (11). Dysregulated lipid metabolism plays a key role in initiation and progression of hepatic steatosis and is frequently associated with inflammation of the liver (12, 13). Elevated inflammation promotes the development of insulin resistance, which in turn further promotes ectopic fat accumulation in the liver, thus forming a vicious cycle (14, 15). Inflammation and fibrogenesis are regulated by complex immunologic pathways that may present possible new therapeutic targets in the liver for NAFLD (7).

Perforin, which is primarily released by CD8<sup>+</sup> T cells and natural killer (NK) cells, helps eliminate infected or dangerous cells and induce apoptosis (16, 17). Following degranulation, pores formed by perforin enable granzyme entry into cells and subsequent caspase activation. Perforin-mediated cytotoxicity is also involved in the homeostatic regulation of CD4 and CD8 T cells in vivo (18, 19). Recent reports revealed that perforin-mediated exocytosis (but not death-receptor-mediated apoptosis) is essential for immune surveillance of senescent cells, and disruption of this pathway as a result of disease or inflammation can lead to the accumulation of senescent cells in the liver (20). Interestingly, a recent study showed that mice on a high-fat diet (HFD) lacking perforin developed more severe obesity, glucose tolerance, and insulin resistance and had higher triglyceride levels in the liver when compared with wild-type (WT) controls (21). However, the precise role of perforin in the context of HFD-induced NAFLD has not been systematically researched yet.

We show that perforin acts as an important immune regulator to prevent NAFLD progression. Aged Prf−/<sup>−</sup> mice had more severe liver injury and lipid accumulation than did WT control mice. In the condition of HFD-induced NAFLD, we also found that Prf−/<sup>−</sup> mice developed more severe hepatic steatosis with more macrophage and IFN-γ, producing CD4<sup>+</sup> T cell infiltration of the liver. Depletion of CD4<sup>+</sup> T cells in Prf−/<sup>−</sup> mice almost completely rescued the observed phenotypes, suggesting an important regulatory role for CD4<sup>+</sup> T cells. Moreover, when IFN-γ receptor signaling is ablated by using perforin and IFN-γ receptor double knockout mice, both liver injury and lipid accumulation were dramatically diminished, indicating that IFN-γ signaling plays a pivotal role in mediating NAFLD pathogenesis.

Overall, our studies reveal that perforin acts as an important immune regulator for NAFLD progression. This finding expands our understanding of inflammation in regulating NAFLD and may have therapeutic implications for NAFLD in the future.

### MATERIALS AND METHODS

#### Mice

Prf−/<sup>−</sup> and IFN-γR <sup>−</sup>/<sup>−</sup> mice were purchased from the Jackson Laboratory. C57BL/6J mice were purchased from Guangdong Medical Laboratory Animal Center (Guangzhou, China). All mice were males and received either a normal control diet (SFD) or HFD (60 kcal % fat; Research Diets) beginning at an age of 6–8 weeks old. All mice were maintained under specified pathogenfree conditions at Jinan University (Guangzhou, China). Animal procedures were approved by and performed in accordance with the Jinan University's Institutional Laboratory Animal Care and Use Committee guidelines.

## Isolation of Liver Mononuclear Cells

The protocol used for isolating murine liver mononuclear cells (MNCs) was as described previously (22). Liver tissue was obtained from mice, and the tissue was dissociated to procure MNCs. To obtain liver MNCs, murine livers were pressed through a 200-gauge stainless steel mesh and suspended in either RPMI-1640 medium or PBS. The cells were then centrifuged at 50 g for 1 min. The cell suspension was collected and centrifuged again at 974 g for 10 min. The cell pellet containing MNCs was then resuspended in 40% Percoll (GE Healthcare, Uppsala, Sweden), after which the cell suspension was overlaid on 70% Percoll and centrifuged at 1,260 g for 30 min. The resulting cell pellets were collected from the interphase following two additional washings in PBS or RPMI-1640 medium.

## Serum Biochemistry

Mice were fasted overnight. Then, whole blood was collected, and serum alanine aminotransferase (ALT) and cholesterol levels were determined using an automatic biochemistry analyzer (7600-020, Hitachi, Japan).

## Cytokine Detection With ELISA

Mice were fasted overnight, and 0.1 g of liver tissue was harvested from the mice in 1 ml of PBS. Liver tissue was then homogenized by hand and centrifuged at 3,000 rpm for 10 min, after which the supernatant was carefully collected. All steps were performed at 4 ◦C. IL-6, IFN-γ, and TNF-α levels in liver supernatants were determined using a commercially available mouse enzyme-linked immunosorbent assay (ELISA) kit (eBioscience, San Diego, CA, USA) according to the manufacturer's instructions.

## Flow Cytometry Analysis

Non-parenchymal cells were transferred to a new well and treated with 1:1000 GolgiPlug, 1 ng/ml ionomycin, and 50 ng/ml PMA for 4–6 h. Intracellular and cell surface staining was performed as described in the fixation/permeabilization kit (554714; BD) protocol. Cells were stained with the surface markers PEcy7 anti-mouse CD3, PE-anti-mouse NK1.1, FITC-anti-mouse CD4, and PerCPCY5.5-anti-mouse CD8 for 15 min at 4◦C. Cells were stained for cytokines with BV421 anti–mouse IFN-γ and APC-IL-17A for 30 min at 4◦C, washed with PBS, and analyzed using FACS verse flow cytometry (BD). Data were analyzed using FlowJo (TreeStar).

## CD4<sup>+</sup> T Cell Depletion

To deplete CD4<sup>+</sup> T cells, 200-µg doses of anti-CD4 monoclonal antibody (clone: GK1.5; Sungene Biotech) per mouse were intraperitoneally injected weekly during HFD challenge. Sterilefiltered PBS was used as a control.

## Histological Examination

Liver tissue was harvested and fixed in 4% (w/v) paraformaldehyde, and 4 mm-thick sections that had been affinized and rehydrated were stained with hematoxylin and eosin (H&E). Hepatic lipid content was determined using frozen sections embedded in Tissue-Tek O.C.T. compound and stained with Oil Red O (Sigma-Aldrich, St. Louis, MO, USA). Images were acquired on a Leica DM3000 microscope.

#### Immunofluorescence

Liver tissue was harvested, fixed in 4% (w/v) paraformaldehyde, and cut in 4 mm-thick sections. Liver sections were then perfused with 30 ml of 4% paraformaldehyde for fixation. Sections were then incubated with the following dilutions of mouse-specific primary antibodies: 1:200 anti-F4/80 (ab16911, Abcam) and 1:200 iNOS antibody (GTX74171, Gentex). For visualization, 1:200 fluorescent Alexa Fluor 594 and FITC 488 secondary antibodies (Invitrogen Vector) were used for both individual staining and co-staining at room temperature for 2 h. After washing, tissue sections were fixed with Vectashield containing DAPI for visualization. A laser cofocal microscopy (TCS SP8, Leica) was used to capture images and conduct further analysis. For the microscopy images displaying M1 (iNOS+ F4/80+) or total macrophages (F4/80+), 4 slides per mouse liver tissue were prepared and 4 fields were captured from each slide. The quantification of M1 or total macrophages was conducted in these 16 fields and designated as one biological independent sample, and the percentage of M1 in total macrophages was calculated and shown.

### Tissue Triglyceride Quantification

The protocol for quantifying hepatic triglyceride (TG) levels was carried out as described previously (23). Briefly, 20–30 mg of liver tissue was homogenized in 500 µl of PBS and mixed with chloroform/methanol 2:1 (vol/vol). The organic phase was transferred, air-dried overnight, and resuspended in 1% Triton X-100 in absolute ethanol. The concentration of TGs was then quantified using a serum triglyceride determination kit (Sigma, Triglyceride Reagent T2449 and Free Glycerol Reagent F6428).

## RNA Extraction and Quantitative Real-Time PCR

Total liver RNA was isolated using TRIzol Reagent (DP424, Tiangen, China). cDNA synthesis was performed using a Prime Script RT Reagent Kit (Takara, Shiga, Japan). Levels of mRNA expression were quantified by real-time PCR (RT-PCR). RT-PCR was performed using TB Green (Takara). Primer sequences are shown in the **Table 1**.

#### Statistical Analysis

Data are presented as the mean ± SEM. Statistical significance between two groups was evaluated using a two-tailed unpaired Student's t-test. Values of P < 0.05 were considered to be statistically significant. The data shown in each panel of these figures were collected from a single experiment; each experiment was repeated for at least three times and showed consistent results. Moreover, the statistical analysis was conducted on each single experiment.

## RESULTS

### Perforin Deficiency Accelerates Liver Injury and Enhances Lipid Accumulation in 14 Month-Old Mice

NAFLD is common in the elderly, in whom it carries a more substantial burden of hepatic (non-alcoholic steatohepatitis,

#### TABLE 1 | Primers for real-time RT-PCR.


cirrhosis, and hepatocellular carcinoma) and extra-hepatic manifestations and complications (cardiovascular disease, extrahepatic neoplasms) than in younger age groups (24). Aged Prf <sup>−</sup>/<sup>−</sup> mice have been reported to have accumulation of senescent cells and development of chronic systemic and local inflammation in the liver (25, 26). We hypothesized that aged Prf−/<sup>−</sup> mice might also have more severe hepatic morbidities since inflammation correlates with liver dysfunction. To test this hypothesis, we first determined liver weights and liver injury (ALT) levels in aged WT and Prf−/<sup>−</sup> mice at 14 months of age. As expected, the aged Prf−/<sup>−</sup> mice showed significantly increased liver weight (**Figure 1A**), elevated liver damage, and increased lipid accumulation as shown by levels of ALT that trended as increased and significantly increased liver TG levels (**Figure 1B**). Furthermore, liver histological analysis revealed more severe hepatic steatosis and significantly increased accumulation of lipid in aged Prf−/<sup>−</sup> mice compared with WT mice (**Figure 1C**). These results indicated that the perforin deficiency aggravates liver injury and steatosis in aged mice.

results. Means ± SEM, \**p* < 0.05.

## Perforin Deficiency Aggravates HFD-Induced Liver Injury and Steatohepatitis

The role of perforin in NAFLD was then investigated using an HFD-induced NAFLD model. Prf−/<sup>−</sup> and WT mice at 6–8 weeks of age were fed on SFD or HFD for 10 weeks to induce NAFLD. As expected, HFD challenge was associated with elevated body weight and ALT activation in WT mice (**Figures 2A,B**). The increase in body weight in response to HFD challenge was comparable between WT and Prf−/<sup>−</sup> mice; however, Prf−/<sup>−</sup> mice had significantly enlarged livers (**Figure 2A**). Additionally, the Prf−/<sup>−</sup> mice had more severe liver damage as indicated

by higher ALT levels. The livers of HFD Prf−/<sup>−</sup> mice exhibited significantly increased lipid accumulation (**Figure 2C**). Histological analysis of livers indicated that HFD Prf−/<sup>−</sup> mice had more severe hepatic steatosis and lipid accumulation when compared with WT controls (**Figure 2D**). Moreover, RT-PCR analysis of liver samples from HFD mice showed that the expression levels of genes involved in lipid production such as fatty acid binding protein 4 (Fabp4) and peroxisome proliferatoractivated receptor gamma (PPARγ) were significantly increased, whereas expression levels of lipid catabolism-related genes such as carnitine palmitoyl transferase 1 (CPT-1α) and aldehyde oxidase (AOX-1) were significantly decreased in HFD Prf−/<sup>−</sup> mice (**Figure 2E**). These results indicated that perforin deficiency with HFD challenge aggravated liver injury and steatohepatitis.

#### Perforin Deficiency Promotes an Inflammatory Response in the Liver After HFD Challenge

Pro-inflammatory T cells promote M1 macrophage activation and intensively contribute to HFD-induced NAFLD (27). To explore the mechanism that drives more severe NAFLD in Prf−/<sup>−</sup> mice, we analyzed the composition of the immune cell infiltrate in the liver by flow cytometry. Perforin deficiency did not alter the infiltration of CD4, CD8, NK, NK1.1+ T cells,

experiment, and each experiment was repeated for at least three times and rendered consistent results. Means ± SEM, \**p* < 0.05; \*\**p* < 0.01, \*\*\**p* < 0.001.

Wang et al. Perforin Protect Against NAFLD

or total macrophages in the liver (**Figures 3A–C**). However, the cell number of CD11c+ macrophages was significantly increased (**Figure 3D**). We next evaluated inflammatory cytokine production by these immune cell subsets. Interestingly, we observed that IFN-γ production from CD4 T, but not CD8 T cells, NK cells, or NK1.1+ T cells, was significant increased, while IL-17 was barely detectable and largely unaffected in all subsets (**Figures 3E–H**). We characterized the cell number in each category and found no significant difference of CD4 T cells, CD8 T cells, and NK and NK1.1+T cells. The cell numbers of CD11c+ macrophages (M1) and IFN-γ+CD4 T cells were significantly increased in perforin KO liver, which equivalent as the percentage analysis (**Figure 3I**).

We also determined the levels of pro-inflammatory cytokines secreted by the liver in HFD-challenged Prf−/<sup>−</sup> and WT mice. As expected, Prf−/<sup>−</sup> livers produced more IL-6, TNF-α, and IFN-γ compared with livers from WT controls (**Figure 4A**). Immunofluorescence analysis showed that perforin deficiency robustly promoted the enrichment of M1 macrophage in the liver, which was consistent with the previous percentage and cell number analysis (**Figure 4B**). These findings suggest that upon HFD challenge, the IFN-γ level and M1 macrophage-mediated inflammation are enhanced in Prf−/<sup>−</sup> mice.

### Perforin Regulates Fatty Liver Disease Through CD4 T Cells in the Liver

To determine whether increased IFN-γ production from CD4 T cells in Prf−/<sup>−</sup> mice was associated with the exacerbated liver phenotypes that develop after HFD challenge, we depleted CD4 T cells in Prf−/<sup>−</sup> mice and then fed the mice with HFD. As expected, CD4 T cell depletion predisposed Prf−/<sup>−</sup> mice to decreased liver weights, lipid accumulation, and diminished liver damage (**Figures 5A–C**). Notably, levels of the pro-inflammatory cytokine TNF-α, as well as macrophage accumulation, were also significantly decreased following CD4 T cell depletion in Prf−/<sup>−</sup> mice; so was the IFN-γ level, though CD4 T cells are not the only cells producing IFN-γ (**Figures 5D,E**). Furthermore, the mRNA expression levels of genes involved in lipogenesis such as Fabp4, CEBPα, PPARγ, SREBP1c, Chrebpβ, and Scd1 were decreased following CD4 T cell depletion in Prf−/<sup>−</sup> mice, whereas lipolysisrelated genes such as AOX1, CPT1α, and LPL were unchanged following CD4 T cell depletion in Prf−/<sup>−</sup> mice (**Figure 5F**). These findings indicate that CD4 T cells play a critical role in perforinmediated regulation of NAFLD progression.

## Hepatic Steatosis in Prf-/- Mice Is Dependent on IFN-γ-Mediated Inflammation

Since the level of IFN-γ was significantly increased in the livers of Prf−/<sup>−</sup> mice, we next explored whether CD4 T cells contribute to exacerbated NAFLD in these mice via IFN-γ activity. Therefore, we crossed Prf−/<sup>−</sup> mice with IFN-γ receptor-deficient mice to get double knockout mice (IFN-γR <sup>−</sup>/<sup>−</sup> and Prf−/−). Following HFD challenge, IFN-γR <sup>−</sup>/<sup>−</sup> and Prf−/<sup>−</sup> mice gained similar amounts of body weight but had significantly decreased liver weights when compared to Prf−/<sup>−</sup> mice (**Figures 6A,B**). Notably, IFN-γR <sup>−</sup>/<sup>−</sup> and Prf−/<sup>−</sup> mice showed significantly rescued NAFLD symptoms, including diminished hepatic steatosis, cellular ballooning, and lipid accumulation (**Figure 6C**). We also found that IFN-γR <sup>−</sup>/−& Prf−/<sup>−</sup> mice had reduced serum ALT, cholesterol, and liver TG levels, as well as diminished pro-inflammatory cytokine production, while the level of IFNγ was no significantly changed after IFN-γ receptor deficiency (**Figure 6D**). Moreover, the cell number of pro-inflammatory (F4/80+iNOS+) macrophage in the livers of IFN-γR <sup>−</sup>/−Prf−/<sup>−</sup> mice was also dramatically decreased when compared to Prf−/<sup>−</sup> mice (**Figure 6E**). Taken together, these findings strongly support an important role for elevated IFN-γ in promoting NAFLD progression in the context of perforin deficiency, given that ablation of IFN-γ signaling had a protective effect on the liver in an NAFLD mouse model.

### CD4 T Cells Demonstrate Intrinsically Elevated IFN-γ Production in Prf-/- Mice

To define the functional properties of CD4 T cells from Prf−/<sup>−</sup> mice, total spleen lymphocytes from WT and Prf−/<sup>−</sup> mice were stimulated in vitro with anti CD3/anti-CD28 in the presence of Golgi-Stop. CD4 but not CD8 T cells from Prf−/<sup>−</sup> mice showed increased levels of IFN-γ production upon CD3/CD28 stimulation (**Figures 7A,B**). To further study whether elevated IFN-γ production by CD4 T cells in Prf−/<sup>−</sup> mice was an intrinsic property of these mice, naïve CD4 T cells were sorted from WT and Prf−/<sup>−</sup> spleens and directly differentiated into Th1 cells. Interestingly, naïve CD4 T cells from Prf−/<sup>−</sup> mice showed an elevated ability to differentiate into Th1 cells (**Figure 7C**). These findings support the conclusion that CD4 T cells undergo an intrinsic functional change in Prf−/<sup>−</sup> mice.

## DISCUSSION

HFD-induced NAFLD is a well-established mouse model for studying the pathophysiological mechanisms of human fatty liver disease. NAFLD is a prevalent liver disease worldwide that can have severe complications such as liver fibrosis and even development of hepatocellular carcinoma, for which there are no effective therapeutic approaches (28). Numerous factors such as leptin, TNF-α, and IL-6 are involved in the initiation and progression of hepatic steatosis and related metabolic dysfunction (29–31). However, the precise role of perforin, a cytotoxic factor released by T cells, has not been precisely studied in the context of HFD-induced NAFLD. Here we described an important protective role for perforin in regulating NAFLD progression. We found that perforin regulates intrinsic IFN-γ production in CD4 T cells, which influences pro-inflammatory macrophage accumulation to affect the progression of NAFLD.

One major finding of this study is the discovery of the protective role perforin plays in regulating NAFLD progression. Perforin is a ∼67-kDa pore-forming protein that is stored in the secretory vesicles (granules) of CTLs and NK cells (32). Perforin is known to have potent and extensive functions in mediating targeted killing together with various other factors

FIGURE 3 | Perforin deficiency drives inflammatory reactions in liver after HFD challenge. WT and Prf−/<sup>−</sup> mice were fed HFD for 10 weeks. Flow cytometry analysis of the representative histogram (left) and frequency (right) of liver CD4 T cells, CD8 T cells (A), NK cells, NK1.1+ T cells (B), macrophages (C), and CD11c+ macrophages (D) and intracellular staining in the liver for IFN-γ and IL-17 in CD4 T cells (E), CD8 T cells (F), NK cells (G), and NK1.1+ T cells (H). (I) The total number of CD4 T cells, CD8 T cells, NK cells, and NK1.1+ T cells, CD11c+ macrophages and IFN-γ+CD4 T cells. *n* = 4–5 mice per group. The data shown in each panel of the figures were collected from a single experiment, and each experiment was repeated for at least three times and rendered consistent results. Means ± SEM; \**p* < 0.05.

secreted by immune cells (19). Previous studies have shown that perforin-deficient mice are sensitive to obesity-induced insulin resistance as a result of restricted T cell expansion and activation in adipose tissue. Perforin has also been reported to play critical roles in promoting inflammation-mediated diseases, including type 1 diabetes (33), cerebral malaria (34), and viral myocarditis (35). A recent study revealed that perforin expressed in CD8 T cells regulates innate and adaptive immunity in the liver and exerts a protective effect in MCD (methionine/choline-deficient diet) diet-induced NASH models (36). Interestingly, MCD dietinduced non-obese NAFLD displays characteristics distinct from those of obesity-induced NAFLD. The precise role of perforin in liver metabolic disorders such as obesity-induced fatty liver disease has not been systematically researched yet. Using 14 month-old Prf−/<sup>−</sup> mice fed either normal chow or HFD, we demonstrated that perforin played a critical protective role in obesity-induced NAFLD.

In our mouse experiments, we chose male mice fed on HFD (60% fat) for 10 weeks to induce NAFLD and found that Prf−/<sup>−</sup> mice had more liver weight and liver TG accumulation in hepatocytes. These data are seemingly in contrast to a recent paper published by Cuff et al. which showed that after 24 weeks of obesogenic diet [22.6% fat, 23.0% protein, and 40.2% carbohydrate (w/w) supplement with sweetened condensed milk (Nestle) ad libitum], there was no difference in hepatomegaly and liver weight between the wild-type and perforin knockout female mice; otherwise, the fibrosis was significantly lower, and perforin KO mice suffer from less severe NAFLD mediated by NK cells (37). These conflicting findings may be due to the different diets, gender, and feeding time. In Cuff et al.'s paper, they chose 24 weeks as the timepoint so that they could compare the development of fibrosis, which is not usually pronounced at 10 weeks in NAFLD. Several reports indicated that free access to condensed milk induced an increase in serum AST activity

FIGURE 5 | Representative images of liver sections stained with H&E (top) and Oil red O (bottom). (D) Serum TNF-α and IFN-γ levels were determined (*n* = 4–6). (E) Representative immunofluorescence images from liver with DAPI (blue), F4/80 (red), iNOS (green), and co-localization (merged image) (left) and the percentage of M1/M (refers to iNOS+F4/80+cells/F4/80+ cells) was calculated (right). (F) Expression levels of lipogenic genes (left) and fatty acid oxidation genes (right) in the liver relative to Hprt expression levels (*n* = 6–11). The data shown in each panel of the figures were collected from a single experiment, and each experiment was repeated for at least three times and rendered consistent results. Means ± SEM, \**p* < 0.05; \*\**p* < 0.01, \*\*\**p* < 0.001.

measured (*n* = 12–13). (C) Representative images of liver sections stained with H&E (left) and Oil red O (right). (D) TNF-α, IL-6, and IFN-γ levels were detected in liver culture supernatants (*n* = 4). (E) Representative immunohistochemistry of DAPI (blue), F4/80 (red), iNOS (green), and co-localization (merged image) in liver, and the percentage of M1/M (refers to iNOS+F4/80+cells/ F4/80+ cells) was calculated (right). The data shown in each panel of the figures were collected from a single experiment, and each experiment was repeated for at least three times and rendered consistent results. Means ± SEM, \**p* < 0.05; \*\**p* < 0.01, \*\*\**p* < 0.001.

FIGURE 7 | CD4 T cells in Prf−/<sup>−</sup> mice have an intrinsically elevated ability of IFN-γ production. CD4 T cells (A) and CD8 T cells (B) from WT and Prf−/<sup>−</sup> mouse spleens were cultured in media supplemented with GolgiStop in the presence or absence of plate-bound anti-CD3 plus anti-CD28 for 4–6 h and then stained for intracellular IFN-γ (left, percentage; right, cell number, n = 5–6). (C) Naive CD4+ T cells were sorted from the spleen of WT or KO mice and differentiated to Th1 cells for 3 days. The IFN-γ production in these cells was detected by flow cytometry. Representative FACS plots and total cell number of IFN-γ+CD4+T cells were shown. The data shown in each panel of the figures were collected from a single experiment, and each experiment was repeated for at least three times and rendered consistent results. Means ± SEM, \**p* < 0.05; \*\**p* < 0.01.

and type I collagen deposition in the liver (38). NAFLD refers to a spectrum of liver diseases, including non-alcoholic fatty liver, which is characterized by steatosis with no or minor inflammation, and NASH, which is associated with inflammation and ballooning with or without fibrosis, and it may progress to liver cirrhosis and hepatocellular carcinoma (39, 40). The livers from mice fed a high-fat diet lacked fibrosis and showed mild steatosis and focal hepatocellular necrosis and apoptosis (41, 42). These contradictory findings suggest that perforin might have different actions at different stages during the pathogenesis of NAFLD and NASH.

Compared with WT controls, SFD-fed Prf−/<sup>−</sup> mice showed increased liver TG levels at an early age (**Figure 2C**), which suggests that perforin may regulate early liver lipid accumulation independent of diet. After 10 weeks of HFD challenge, Prf−/<sup>−</sup> mice had more IFN-γ-producing CD4 T cells in the liver. Further studies revealed that Prf−/<sup>−</sup> mice had intrinsically increased IFNγ-producing ability in CD4 T cells. However, it is still unclear how perforin, a cytotoxic factor that helps mediate target cell death, could stimulate CD4 T cells to produce IFN-γ. Further studies are needed to better understand this phenomenon.

The promotion of hepatic steatosis resulting from perforin deficiency was associated with a strong increase in hepatic macrophage accumulation and inflammation as evaluated by the expression of TNF-α, IL-6, and iNOS. Traditionally, macrophages are divided into pro-inflammatory (M1) and wound-healing (M2) classes. M1 macrophages, which are induced by IFN-γ and LPS and express pro-inflammatory cytokines such as TNF-α, IL-6, and IL-1β, are implicated in the pathogenesis of chronic liver inflammation. M2 macrophages, which are induced by IL-4, IL-10, and IL-13 and produce IL-10, TGF-β, PDGF, and EGF, have antiinflammatory effects and promote wound healing (43–45). It is well established that macrophages play an important role during NAFLD pathogenesis. Previous studies found that depletion of macrophages with clodronate could significantly reverse NAFLD in mice (3, 46). Liver immune homeostasis is largely regulated within the hepatic sinusoid, where resident macrophages (Kupffer cells) are located as part of the liver reticuloendothelial system (also known as the mononuclear phagocyte system). This system forms a highly active, dynamic, and complex network, constituting the primary line of defense against invading microorganisms along with the involvement of other immune cells such as neutrophils. In different stages of liver disease, resident Kupffer cells and freshly recruited monocyte-derived macrophages play a key role in the regulation of inflammation, fibrogenesis, and fibrolysis (47). In our study, Prf−/<sup>−</sup> mice had more macrophages, especially M1-type macrophage accumulation in liver after HFD challenge when compared with WT controls. However, we did not determine the mechanism of M1-type macrophage accumulation in the liver in this study. Is the increased accumulation due to the proliferation of resident Kupffer cells, or recruitment from peripheral circulatory systems, or the polarization of monocytes? We speculated that the increase in M1-type macrophage accumulation in the liver might be the result of monocyte polarization, since liver injury and lipid accumulation were almost non-existent in IFN-γR <sup>−</sup>/<sup>−</sup> &Prf−/<sup>−</sup> mice with decreased M1-type macrophage accumulation in the liver when compared with Prf−/<sup>−</sup> mice. Further studies are necessary to better understand the mechanism behind this observation.

Depletion of CD4 T cells in Prf−/<sup>−</sup> mice rendered these mice less sensitive to NAFLD, with similar levels of liver TG and macrophage accumulation detected when compared with WT controls. This finding highlights the indispensable role of CD4 T cells, especially Th1 cells, in NAFLD progression. In clinical studies, it was reported that the peripheral CD4 compartment in obese children displayed a Th1-prone phenotype, and pediatric patients with NASH also showed increased expression of IFNγ in the liver. Dysregulated lipid metabolism in NAFLD was reported to cause a selective loss of intrahepatic CD4+ lymphocytes, leading to accelerated hepatocarcinogenesis (48). In this study, we demonstrated that the protective effect of perforin in HFD-induced NAFLD was almost completely dependent on Th1 cells, which is consistent with the existing literature.

In conclusion, we demonstrated that perforin acts as an important immune regulator in NAFLD progression through regulating INF-γ-producing CD4 T cells to decrease macrophage accumulation in the liver. Based on these findings, therapeutic strategies targeting perforin might be a promising approach for the development of novel strategies to prevent or treat hepatic steatosis and related metabolic disorders in the liver.

## DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

## ETHICS STATEMENT

The animal study was reviewed and approved by Laboratory Animal Ethics Committee Jinan University.

## AUTHOR CONTRIBUTIONS

DL and QWa designed the project, performed experiments, and collected and analyzed the data. QWa wrote the manuscript. JZ and MZ worked on the mouse model. GC and JH helped modify and revise the article. QWe, HZ, and ZL maintained and genotyped the mice. LZ and QS performed RT-PCR. ZY, HY, and QWa supervised and coordinated the work, designed the overall research study, and helped write the manuscript. All authors have read, discussed, and approved the final manuscript.

## FUNDING

This work was supported by the National Natural Science Foundation of China (the Key Program, 31830021), and the State Bureau of Foreign Experts Affairs (111 project, B16021) to ZY and HY. This work was also supported by the Science and Technology Department of the Guangdong Province of China (2018B030312008) to HY, the National Natural Science Foundation of China (grant 31800721 to QWa), the China Postdoctoral Fund (grant 2018M633278 to QWa), the National Natural Science Foundation of China (81630025, 31970830 to JH), the Traditional Chinese Medicine Bureau of Guangdong Province (2018071 to JH), the Guangzhou Municipal Science and Technology Bureau (201904010090 to JH) and the Health Commission of Guangdong Province (A2019520 to JH).

#### REFERENCES


#### ACKNOWLEDGMENTS

We thank Tucker Sarah for helping us correct typos and grammar errors and polish the text.


**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 © 2020 Wang, Li, Zhu, Zhang, Zhang, Cao, Zhu, Shi, Hao, Wen, Liu, Yang and Yin. 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.

# The Genus *Alistipes*: Gut Bacteria With Emerging Implications to Inflammation, Cancer, and Mental Health

Bianca J. Parker <sup>1</sup> , Pamela A. Wearsch<sup>1</sup> , Alida C. M. Veloo<sup>2</sup> and Alex Rodriguez-Palacios 3,4 \*

*<sup>1</sup> Department of Pathology, Case Western Reserve University School of Medicine, Cleveland, OH, United States, <sup>2</sup> Department of Medical Microbiology, University Medical Centre Groningen, University of Groningen, Groningen, Netherlands, <sup>3</sup> Division of Gastroenterology and Liver Disease, Case Western Reserve University School of Medicine, Cleveland, OH, United States, <sup>4</sup> Germ-Free and Gut Microbiome, Digestive Health Research Institute, Case Western Reserve University, Cleveland, OH, United States*

#### *Edited by:*

*M. Victoria Delpino, CONICET Institute of Immunology, Genetics and Metabolism (INIGEM), Argentina*

#### *Reviewed by:*

*Valerio Iebba, Istituto Pasteur Italia Cenci Bolognetti Foundation, Italy Alexander R. Moschen, Innsbruck Medical University, Austria*

> *\*Correspondence: Alex Rodriguez-Palacios axr503@case.edu*

#### *Specialty section:*

*This article was submitted to Microbial Immunology, a section of the journal Frontiers in Immunology*

*Received: 21 December 2019 Accepted: 20 April 2020 Published: 09 June 2020*

#### *Citation:*

*Parker BJ, Wearsch PA, Veloo ACM and Rodriguez-Palacios A (2020) The Genus Alistipes: Gut Bacteria With Emerging Implications to Inflammation, Cancer, and Mental Health. Front. Immunol. 11:906. doi: 10.3389/fimmu.2020.00906* *Alistipes* is a relatively new genus of bacteria isolated primarily from medical clinical samples, although at a low rate compared to other genus members of the *Bacteroidetes* phylum, which are highly relevant in dysbiosis and disease. According to the taxonomy database at The National Center for Biotechnology Information, the genus consists of 13 species: *Alistipes finegoldii*, *Alistipes putredinis*, *Alistipes onderdonkii*, *Alistipes shahii*, *Alistipes indistinctus*, *Alistipes senegalensis*, *Alistipes timonensis*, *Alistipes obesi*, *Alistipes ihumii*, *Alistipes inops, Alistipes megaguti, Alistipes provencensis,* and *Alistipes massiliensis. Alistipes communis* and *A. dispar*, and the subspecies *A. Onderdonkii* subspecies vulgaris (vs. *onderdonkii* subsp.) are the newest strains featured outside that list. Although typically isolated from the human gut microbiome various species of this genus have been isolated from patients suffering from appendicitis, and abdominal and rectal abscess. It is possible that as *Alistipes* spp. emerge, their identification in clinical samples may be underrepresented as novel MS-TOF methods may not be fully capable to discriminate distinct species as separate since it will require the upgrading of MS-TOF identification databases. In terms of pathogenicity, there is contrasting evidence indicating that *Alistipes* may have protective effects against some diseases, including liver fibrosis, colitis, cancer immunotherapy, and cardiovascular disease. In contrast, other studies indicate *Alistipes* is pathogenic in colorectal cancer and is associated with mental signs of depression. Gut dysbiosis seems to play a role in determining the compositional abundance of *Alistipes* in the feces (*e.g.*, in non-alcoholic steatohepatitis, hepatic encephalopathy, and liver fibrosis). Since *Alistipes* is a relatively recent sub-branch genus of the *Bacteroidetes* phylum, and since *Bacteroidetes* are commonly associated with chronic intestinal inflammation, this narrative review illustrates emerging immunological and mechanistic implications by which *Alistipes* spp. correlate with human health.

Keywords: *Alistipes finegoldii*, inflammatory bowel diseases, *A. putredinis*, *A. onderdonkii*, *A. shahii*, *A. indistinctus*, *A. senegalensis*, *A. timonensis*

## INTRODUCTION

The human gut microbiome is acquired at birth and as individual go through life, their gut becomes a home to trillions of microorganisms, namely bacteria. With the use of 16S rRNA gene sequencing a diverse profile of bacterial phyla has been identified, with the two most common being Bacteroidetes and Firmicutes. Collectively, the composition of the gut microbiota amounts to a massive number of cells and functions that equate (even surpass) to that of an additional organ (1). The human microbiota plays a critical role in regulation of the immune response, protection against pathogens, aiding in digestion, as well as in neurologic signaling and vascularization (2). Notably, the microbiota has been found to be involved in health and disease. The imbalance, or dysbiosis, of the intestinal microbiota has been attributed to numerous diseases such as cancer (3), cardiovascular disease (4), inflammatory bowel disease (5), and nervous system disorders (6). Investigators have examined various genera and species that make up the microbiome trying to decipher their role as a collective trying to identify individual microbial species with the ability to modulate diseases.

Taxonomically, Alistipes (a. li. sti'pes) is a genus described in 2003 after being discovered in tissue samples of children with appendicitis (7). Alistipes are anaerobic bacteria found mostly in the healthy human gastrointestinal (GI) tract microbiota (8). According to the taxonomy database at The National Center for Biotechnology Information (NCBI, txid239759), as of April 2020, Alistipes consists of 13 species: Alistipes finegoldii, Alistipes putredinis, Alistipes onderdonkii, Alistipes shahii, Alistipes indistinctus, Alistipes senegalensis, Alistipes timonensis, Alistipes obesi, Alistipes ihumii, Alistipes inops, Alistipes megaguti, Alistipes provencensis, and Alistipes massiliensis (the latter strain reported as "unpublished by Lacroix et al., as of March 4, 2004, NCBI:txid265312" remains unpublished under that name). However, a case study in 2017 from the same group of scientists reported a novel species called Tidjanibacter massiliensis Marseille-P3084, isolated from the colon of a person with irritable bowel syndrome, with 92.1% sequence homology to A. putredinis (9), but the strain has not been officially recognized or deposited in public strain biorepositories. To date it is unclear if the Tidjanibacter is the same Alistipes isolate, or what occurred with the publication of the Alistipes massiliensis for which there is no genome or 16S RNA gene sequences available in NCBI. Three newer species were added to the the "List of Prokaryotic names with Standing in Nomenclature" database in January 2020: A. communis (strain 5CBH24; DSM 108979; JCM 32850), A. dispar (5CPEGH6; DSM 108978; JCM 32848), and the sub-speciation scheme for A. Onderdonkii as subsp. vulgaris subsp. nov. and the subsp. onderdonkii subsp. nov. (https://lpsn. dsmz.de/genus/alistipes, which listed only 11 species on May 12, 2020) (10).

Anecdotally, the first species of Alistipes to be discovered was A. finegoldii (fi.ne.gol'di.i) which was then named after Sydney M. Finegold (born in 1921; died on September 27, 2018, age 97), a contemporary American researcher in anaerobic bacteriology and infectious diseases clinician who held several Emeritus academic positions at the UCLA and the Wardsworth VA Hospital in Los Angeles since the year 2000 (7, 11). Several recent articles commemorate Dr. Finegold's career in anaerobic microbiology, who greatly contributed to our understanding of anaerobes, which led to the species of Alistipes being named after him (12). With a remarkable productivity, Dr. Finegold left a legacy with 819 publications on various aspects of medicine and anaerobic microbiology (13). Alongside A. finegoldii, in 2003, Bacteroides putredinis was reclassified as A. putredinis based on 16S rRNA gene sequencing and biochemical features that showed this species did not belong to the genus Bacteroides, but to the genus Alistipes (7). Dr. Finegold's charisma is illustrated with the species A. onderdoinklii, which was named to honour Andrew B. Onderdonk, a contemporary American microbiologist, Professor of Pathology at Harvard Medical School, for his contribution to intestinal and anaerobic microbiology.

In recent years, several studies have investigated the alterations in bacterial abundance for Alistipes in human patients and mouse models during disease. Studies have shown that Alistipes dysbiosis can be either beneficial, or harmful. Alistipes has been implicated in liver fibrosis (14), colorectal cancer (15), cardiovascular disease (16), and mood disorders (17), among other potential diseases. Additionally, the unique way of fermenting amino acids, putrefaction, has implicated Alistipes to play a critical role in inflammation and disease (18). The objective of this review is to expound upon the relationship between Alistipes and several non-communicable diseases and to highlight the implications of Alistipes dysbiosis.

## PHENOTYPING, CULTURE, AND SPECIATION

Alistipes is classified as gram-negative, rod-shaped, anaerobic, and non-spore forming. The species within this genus are nonmotile except for A. obesi. The genus consists of 12-published species (A. massiliensis is not yet published, see above). Alistipes finegoldii (type strain AHN 2437<sup>T</sup> ; CCUG 46020) has circular colonies (0.3–1.0 mm; raised, gray to opaque on sheep blood agar). Despite supplementation, its growth is difficult in liquid media. Alistipes finegoldii is bile-resistant and esculin-negative. This bacterium is catalase-negative and nitrogen-reductase negative; and it cannot liquefy gelatin in liquefaction test. This bacterium can hydrolyze tryptophan to indole. In peptone yeast glucose (PYG) broth the major acid produced is succinic acid with a minor amount of acetic and propionic acid produced. The main fatty acid is 13-metyltetradecanoic acid (iso-C15:0). These strains of A. finegoldii are resistant to vancomycin, kanamycin, and colistin, as it is expected for Bacteroides, according to the Wadsworth manual (19) (named after the hospital where Dr. Finegold worked as a physician and scientist). For the primary identification of anaerobic bacteria, the further use of a bile susceptibility test combined with the antimicrobial resistance helps in differentiating Alistipes species as illustrated for some species below (for comparison, Bacteroides are resistant to bile) (20), although the precise definition of a species is a subject of constant debate (21). The phenotypic characteristics



*Antibiotic Resistance. A. finegoldii, vancomycin, kanamycin, and colistin. A. putredinis, his strain is sensitive to clindamycin, cefoxitin, chloramphenicol, erythromycin, and metronidazole and moderately resistant to tetracycline and doxycycline. A. timonensis, susceptible to penicillin G, amoxicillin* + *clavulanic acid, imipenem and clindamycin; A. timonensis is susceptible to metronidazole. A. senegalensis, susceptible to penicillin G, amoxicillin* + *clavulanic acid, imipenem and clindamycin; A. senegalensis is resistant to metronidazole. A. obesi, susceptible to imipenem, ciprofloxacin, metronidazole, nitrofurantoin, and rifampicin, but resistant to penicillin G, amoxicillin, amoxicillin-clavulanic acid, erythromycin, vancomycin, gentamicin, doxycycline, ceftriaxone, and trimethoprim/sulfamethoxazole. A. ihummi, susceptible to amoxicillin, imipenem, and clindamycin, but resistant to vancomycin. /, not reported; F, fermentative; NF, non-fermentative.*

of 10 Alistipes spp. in culture media are summarized in **Table 1**.

From a clinical perspective, it is reasonable to expect that bileresistant strains are more likely to be present or abundant than bile-susceptible strains in the terminal ileum where most bile reabsorption takes place in several mammalian species including humans (22) and where chronic inflammatory bowel conditions are more likely to occur due to alterations in bile-mediated Tcell immunoreactivity (23). However, with the recent discovery of this genus, and the emerging complexity of the genus speciation, it will be necessary to determine if the resistance to bile acids indeed determines the Alistipes abundance regionally within the GI, or clinically in diseases characterized by alterations in bile production, which originates in the liver, or their recirculation pathways through the small intestine. Also, clinically relevant, some strains produce beta-lactamases (7), which enhance their antimicrobial resistance to beta-lactam antibiotics including penicillin and cephalosporins.

The species A. putredinis (type strain ATCC 29800<sup>T</sup> ) has been isolated from various specimens such as feces, appendiceal tissue of a patient with acute appendicitis, abdominal and rectal abscesses, foot rot in sheep, and even farm soil (7). This species has circular to slightly irregular convex colonies (pinpoint to 0.5 mm; smooth, translucent to gray). This is a non-pigment producing species and is not 20% bileresistant. It is catalase-positive and nitrate reductase- and urease-negative. This bacterium can hydrolyze tryptophan to indole. This species is non-fermentative. In 6-day old PYG broth only trace amount of acid was produced, however, in 24-h chopped meat carbohydrate broth succinic acid was produced as a major acid product with minor amounts of acetic, isobutyric, isovaleric, and propionic acid. This strain is sensitive to clindamycin, cefoxitin, chloramphenicol, erythromycin, and metronidazole and moderately resistant to tetracycline and doxycycline (7).

Alistipes onderdonkii (type strain WAL 8169<sup>T</sup> ; CCUG 48946<sup>T</sup> ; ATCC BAA-1178<sup>T</sup> ) and A. shahii (type strain WAL 8301<sup>T</sup> ; CCUG 48947<sup>T</sup> ; ATCC BAA-1179<sup>T</sup> ) were isolated from abdominal abscess and appendix tissue respectively, as well as urine (24). They have circular colonies (0.5–0.8 mm and 0.5– 1 mm, respectively; gray to opaque on blood agar) and are pigment producing, non-fluorescing when grown on rabbit blood agar. Both species are resistant to 20% bile and hydrolyze tryptophan to indole. They are catalase-, nitrogen reductase-, and urease-negative. Mannose and raffinose are fermented when using the API rapid ID 32A system. Succinic acid is the major metabolic end-product with minor production of acetic and propionic acid. The primary long-chain fatty acid is iso-C15:0 (24). Details for A. onderdonkii subspeciation as vulgaris subsp. nov. (3BBH6T = JCM 32839T = DSM 108977T, isolated from human feces) for can be found at https://lpsn.dsmz.de/ genus/alistipes.

Alistipes indistinctus (type strain YIT 12060<sup>T</sup> ; DSM 22520<sup>T</sup> ; JCM 16068<sup>T</sup> ) was the next species to be discovered. It was isolated from human feces. This species is slightly different, since its shape is more coccoid than rod-shaped, as well as its inability to hydrolyze tryptophan to indole. Its colonies are circular (0.1–0.5 mm slight opaque and gray on modified GAM agar). This bacterium is susceptible to 20% bile, catalase-positive, and urease- and nitrogen reductase-negative. The major products in PYG broth are succinic and acetic acid. The major cellular fatty acid is iso-C15:0 (25).

Two other species, A. senegalensis (type strain JC50<sup>T</sup> ; CSUR P156; DSM 25460) and A. timonensis (type strain JC136<sup>T</sup> ; CSUR P148; DSM 25383), were initially isolated from the fecal flora of healthy patients in Senegal. They have circular colonies (0.2–0.3 mm) and are pigment producing. They are catalase and can hydrolyze tryptophan to indole. A. senegalensis does ferment mannose, and A. timonensis does not. These strains of bacteria are susceptible to penicillin G, amoxicillin plus clavulanic acid, imipenem, and clindamycin. Furthermore, A. senegalensis strains are resistant to metronidazole and A. timonensis strains are susceptible to metronidazole (21, 26).

Alistipes obesi (type strain ph8<sup>T</sup> ; CSUR; P186;DSM 25724) was isolated from the fecal microbiota of a French patient suffering from obesity (27). It is a pigment producing species with colonies that are 0.5 mm in diameter; translucent and light gray on blood enriched Columbia agar. It is catalase-positive, nitrogen reductase-, and urease-negative as well as non-fermentative. This bacterium cannot hydrolyze tryptophan to indole. These strains are susceptible to imipenem, ciprofloxacin, metronidazole, nitrofurantoin, and rifampicin, but resistant to penicillin G, amoxicillin, amoxicillin-clavulanic acid, erythromycin, vancomycin, gentamicin, doxycycline, ceftriaxone, and trimethoprim/sulfamethoxazole (27).

Alistipes ihumii (type strain AP11<sup>T</sup> ; CSUR P204; DSM 26107) was isolated from the fecal microbiota of a female French patient suffering from severe anorexia nervosa. This species forms colonies that are 0.2 mm and translucent on blood-enriched Columbia agar. This bacterium is non-fermentative. It is negative for urease, nitrate reduction, and catalase activity. This bacterium cannot hydrolyze tryptophan to indole. A. ihumii strains are susceptible to amoxicillin, imipenem, and clindamycin, but resistant to vancomycin (28).

The species Alistipes inops (type strain 627<sup>T</sup> ; 5DSM 28863<sup>T</sup> ; VKM B-2859<sup>T</sup> ) was isolated from human feces. The colonies are circular (0.18–0.30 mm; light gray, with a pale brown center). It has scanty growth on liquid media. It is positive for indole production and negative for catalase, nitrate reductase, and urease. This bacterium is nonfermentative. The major metabolic end products in PYG broth are succinic and acetic acid. The major fatty acid is iso C15:0 (8).

Recently, between November 2019–January 2020, four, new species of Alistipes were discovered, A. megaguti, A. provencensis, A. communis (type strain 5CBH24; DSM 108979; JCM 32850, commonly in human feces) and A. dispar (5CPEGH6; DSM 108978; JCM 32848, from human feces). A. megaguti (type strain Marseille-P5997<sup>T</sup> ) was isolated from a fresh fecal sample of a young healthy female. This species forms colonies that are 0.2–0.9 mm in diameter on blood-enriched Columbia agar. A. megaguti are catalase, urease, and oxidase negative. A. provencensis **(**type strain Marseille-P2431<sup>T</sup> ) was isolated from a male patient with hypertension and diabetes. This species forms colonies that are 0.4–0.64 mm in diameter on blood-enriched Columbia agar. This species is oxidase and urease negative, and catalase positive (29).

The phenotypic identification of Alistipes species is feasible (**Table 1**), yet challenging. A more suitable way to identify Alistipes spp. is via NGS (next generation sequencing-based microbiome studies) (30). However, most studies for the classification of new species, which cannot be fully sequenced have been based on Sanger sequencing of the full gene or parts of the 16S rRNA gene. For reference, the specificity of various primers employed for the identification of Alistipes species are listed and referenced in **Table 2**, while the details in nucleotide sequence homology and genetic distances within the genus are depicted in **Figure 1**. Compared to culture techniques, the NGS method is more powerful for bacterial identification. However, it does not provide information on the metabolic functions of the organism. Another form of bacterial identification is through metagenomic studies. In this approach, bacterial DNA is processed and sequenced along with host DNA. A new clinical approach that is faster and more effective than Sanger sequencing for bacterial classification is based on mass-spectrometry. Specifically, new rapid high-through put strategies based on matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) have become widely adapted and common in clinical microbiological settings, which require the constant update of databases to incorporate the spectral of new bacterial strains, primarily of clinical relevance.

#### MASS SPECTROSCOPY FOR IDENTIFICATION OF BACTERIAL SPECIES

Since the innovative use of MALDI-TOF MS to expedite the identification of microbes in clinical microbiology, there has been a major increase in our understanding of the presence and prevalence of Alistipesspp. in clinical medicine. MALDI-TOF MS is a method based on protein spectra, enabling identification of an unknown bacterium within minutes. Due to the introduction of this method in diagnostic microbiology laboratories, fastidious organisms which could previously only be identified using molecular methods, are now identified with ease. With the advancement of MALDI-TOF MS, there has been collaborative initiatives to improve microbial identification, including the European Network of Rapid Identification of Anaerobes (ENRIA), which is multi-national consortium of laboratories in Europe with the goal of unifying the MALDI-TOF MS database for the rapid identification of clinical anaerobes. According to the ENRIA validation study, conducted in 2018 (38), at present 5 Alistipes species are represented in the MALDI-TOF MS database: A. finegoldii, A. onderdonkii, A. shahii, A. indistinctus, and A. putredinis.

TABLE 2 | Referent partial Sanger 16S rRNA gene sequences and primers for *Alistipes* spp.


\**Primers (upper case) used by Weisburg et al. (37) and designed for most eubacteria. f, forward; r, reverse; D, distal; P, proximal. Linker sequences containing restriction sites for cloning are designated in lower case. The "f" series of linkers contain EcoRI and SalI sites, and the "r" series contain HindIJl, BamHI, and XmaI recognition sequences. Reverse primers produce sequences complimentary to the rRNA. Primers rPl, rP2, and rP3 are identical except for the 17th base from the 3*′ *end. The P2 primer was determined as the sequence providing the greatest diversity of bacteria. FASTA files available for download at https://www.ncbi.nlm.nih.gov/search/all/?term=NR\_\_125589* \*\**complete sequence extracted from complete genome available in PATRIC https://docs.patricbrc.org.*

\*\*\**Phylogenetic analysis for speciation can also be conducted for Alistipes using the heat shock protein 60 gene (hsp60) as reported by Sakamoto in 2020. Accession numbers for the whole genome sequences of strains 3BBH6T, 5CBH24T, 5CPEGH6T, 5CPYCFAH4, 5NYCFAH2 and 6CPBBH3 are AP019734-AP019739, respectively.*

To contextualize the current relevance of the genus in clinical diagnostics, we conducted a screening of the number of Alistipes isolates recovered from human clinical samples at the University Medical Center Groningen, Groningen, the Netherlands, within the 7 years prior to submission of this manuscript. Keeping in mind that only species represented in MALDI-TOF MS databases have been identified using this method, it is noteworthy to mention that only 11 (n = 11) clinical isolates belonged to the Alistipes genus. Of relevance, these isolates were identified only as either A. onderdonkii (n = 7) or A. finegoldii (n = 4). The Alistipes isolates were recovered from pus, blood, ascites fluid or from an unspecified specimen. Considering that there have been at least 13 different species identified as distinct according to the NCBI lineage database (as of November 2019), it is interesting that only few species (n = 5) have been identified in clinical settings screened by ENRIA. The lack of identification of the other species (e.g., A. senegalensis, A. timonensis, A. obesi, A. ihumii, and A. inops) in this population indicated that either: (i) Alistipes may have specific patterns of geographically-restricted distribution, (ii) that their prevalence in the environment and patients is rather low, (iii) that they are comparatively more difficult to be cultured from clinical specimens using media that apparently facilitates the recovery of the other Alistipes spp. (namely, A onderdonkii and A. finegoldii), (iv) that the MS profile used

FIGURE 1 | Nucleotide and phylogram of 16S rRNA gene DNA sequences from thirteen strains described in the genus *Alistipes*. (A) *Alsitipes megaguti* complete 16S rRNA gene sequences derived from PATRIC complete genome illustrates slight differences between three gene copies contained within three operons in a single genome. Only the most divergent parts of the genes are shown. (B) Overview of complete alignment for the 13 strains. Notice the *A. obesi* insertion at around 1,100 bp position. (C) Detail of sequence insertion in the *A. obesi* starting at position 1,137 bp position. (D) Overview of the distance tree with 10,000 bootstrap branch values depicted if value >90%. Branching set at 60% threshold. Notice that differences may occur in part due to uneven partial gene sequences. Analysis conducted with CLC genomics viewer. Notice position of *T. massiliensis* vs. *A. inops* and *A. putredinis*.

to differentiate species is still suboptimal to discriminate such species as separate, or (v) that the original species descriptions may not be reproducible descriptions in the context of culture and mass spectrometry profiles.

## ECOLOGY IN DISEASES

According to the NCBI, the full lineage of Alistipes is Bacteria; Bacteroidetes; Bacteroidia; Bacteroidales; Rikenellaceae (NCBI: txid239759). Therefore, Alistipes is a genus member of the Rikenellaceae family (NCBI:t xid171550), which is a small family within the Bacteroidales order that is composed of eight genera: Acetobacteroides, Alistipes, Anaerocella, Millionella, Mucinivorans, Rikenella, Ruminofilibacter, and Tidjanibacter. These genera are commonly found in the intestinal tract of various animals and humans, where they are believed to have symbiotic relationships with the host. One of the most commonly detected species from the genus Alistipes, that can be isolated from human feces, is A. onderdonkii (39).

From an ecological perspective, Alistipes is found primarily in the gut of healthy humans (8). However, Alistipes has also been isolated from the blood stream, as well as appendicular, abdominal, perirectal and brain abscesses highlighting their potential opportunistic pathogenic role in human diseases. Alistipes has been found in other body fluids, namely, urine and peritoneal fluid (8). Of contrasting interest within this genus, A. ihumii has been isolated from the feces of a patient suffering from anorexia nervosa, while A. obesi was isolated from a patient suffering from morbid obesity. Combining the later observations with the fact that Alistipes finegoldii has been considered a growth promoter in broiler chickens (40), and A. putredinis has been observed to increase with cruciferous vegetable intake in humans (41), it is reasonable to assume that different Alistipes species may have different roles in nutrition and health, depending on the host, and the bodily system influenced. Although the NCBI taxonomy database has a disclaimer stating that it is not an authoritative source for nomenclature or classification, the NCBI recommend the database users to consult the relevant scientific literature for the most reliable information, which is summarized below and in **Table 3**.

#### *ALISTIPES* IN LIVER DISEASE AND SHORT-CHAIN FATTY ACIDS

Hepatocellular carcinoma (HCC) is the second deadliest form of cancer worldwide (55).

HCC is often developed from advanced liver fibrosis that is caused by cirrhosis, non-alcoholic fatty liver disease (NAFLD), or non-alcoholic steatohepatitis (NASH). These liver diseases have been associated with the "microbiota-liver axis," indicating that dysbiosis as one of the potential causes (56). In studies done on the microbiota composition and liver fibrosis. It is seen that throughout the advancement of the fibrosis that Alistipes is decreased.

For example, in patients with compensated and decompensated liver cirrhosis (LC), a paired-end metagenomic sequence of the gut microbiome from fresh stool samples of healthy volunteers and patients with various types of LC. This study showed a decrease of A. shahii and A. putredinis when compared to healthy controls. This finding of an increase in the abundance of Alistipes in healthy control patients in comparison to patients with LC has also been seen in studies focusing on feces and biopsies of LC patients (57). Moreover, a decrease of A. indistinctus was observed as the disease progressed from compensated to decompensated (43). Once an individual has decompensated liver cirrhosis, the patient begins to generate a multitude of severe complications, such as hepatic encephalopathy. Another study showed that when comparing stool microbiome between patients suffering from decompensated liver cirrhosis with acute hepatic encephalopathy, Alistipes has a protective role and the decrease in its abundance correlates to an increase of hepatic encephalopathy recurrence (44). Thus, a decrease in Alistipes spp. correlates with the progression of liver cirrhosis into the decompensated state.

Additionally, the reduction in Alistipes abundance in patients with liver fibrosis can be seen in other fibrotic diseases such NASH and NAFLD. Rau et al. (14) showed that patients with NAFLD with substantial fibrosis had a reduction in fecal concentration of acetate and propionate, with no significant difference in butyrate concentration. When comparing healthy controls with patients suffering from NASH, there is a significant reduction in A. finegoldii with the normalized count mean reduced from 542 to 19 (fold-change of −1.829). Moreover, in NAFLD patients with significant fibrosis, a major reduction in A. onderdonkii was observed from 285 to 31 (fold change of −2.566). It is noteworthy that Alistipes spp. decreases in these advanced fibrotic patients as well as fecal acetate and propionate levels. This correlates with a study done by Polansky et al. (58) where they showed that in the cecal microbiota of chickens that Alistipes is a propionate producer expressing methylmalonyl-CoA epimerase, in which the gene for this enzyme is located on an operon with the acetyl-CoA carboxylase gene. It has also been shown that Alistipes is an acetate producer (59). Due to previous studies suggesting that short chain fatty acids (SCFAs) have anti-inflammatory mechanisms, it can be suggested that this decrease in Alistipes contributes to the decrease in SCFA and therefore contributes to the advanced fibrosis seen in these NAFLD patients.

Moreover, a study performed on mice with HCC showed the potential anti-inflammatory effects of a health-beneficial bacteria probiotic. It was believed that T-regs would be induced in patients with HCC thus suppressing the Th17 expression. A probiotic entitled Prohep was used, and after 38 days a reduction in tumor growth and Th17 cells was observed. Moreover, there was an increase in Treg/Tr1 cells, an antiinflammatory cell subset. Alistipes was shown to increase in abundance in the mice cohort receiving the probiotic (45, 60). At the species level it is shown that A. shahii was one of the species that was significantly increased in the probiotic group. They hypothesize that A. shahii was playing a role in tumor suppression similar to what was is seen in cancer immunotherapy (61). Additionally, the investigators noted a significant increase in acetate and propionate metabolic potential in the probiotic group correlating with an increase in SCFAs-producing bacteria, such as Alistipes contributing to the suppression of Th17 cells in the gut, ultimately reducing the recruitment of Th17 cells to the liver (45). Also, in past studies it has been shown that SCFAs derived from fermented dietary fibers increase the levels of propionate found within the portal vein ultimately preventing cancer cell proliferation in liver tissue (62). Ultimately, Alistipes

TABLE 3 | Summary of studies reporting the experimental or observational associations between *Alistipes* spp. and various non-communicable diseases (2003–2019).


\**Protective; overall interpretation of results with respect to outcome for each disease.*

can be seen as a potential SCFA producer and their decrease contributes to these hepatic fibrotic conditions due to a decrease in anti-inflammatory cytokines and inability to suppress Th17 cells. Further studies will be important in the near future to better define the role of this genus in liver pathologies and health.

### *ALISTIPES* IN CARDIOVASCULAR DISEASE, HYPERTENSION, AND THE EPITHELIUM

Cardiovascular disease (CVD) is the leading cause of mortality and morbidity in both developing and developed countries (5). With CVDs projected to rise in the future as the global population ages, the evaluation of the relationship with the gut microbiota has been investigated more extensively. Alistipes has been linked with CVD risk factors such as hypertension, as well as several CVDs such as atrial fibrillation (AF), congestive heart failure (CHF), and atherosclerosis cardiovascular disease (ACVD) (50).

In hypertension, it is believed that Alistipes contributes to inflammation and epithelium alterations. Kim et al. (48) demonstrated the relationship between the gut barrier dysfunction in patients with hypertension and the gut microbiota in humans. They used shotgun metagenomic analysis studying fecal samples from 22 patients with high blood pressure and 18 control patients. The data revealed an increase in A. finegoldii and A. indistinctus that was positively correlated with systolic blood pressure. Additionally, it was shown that A. finegoldii, known to trigger intestinal inflammation, had an increased number and functional genes in the high blood pressure cohort. Overall, the increase in A. finegoldii was observed to be positively correlated with systolic blood pressure (SBP), suggesting that this species is a potential driver for gut barrier dysfunction and inflammation in patients with high blood pressure. Also, it is believed that the hypotensive phenotype of inflammation is caused by the lipopolysaccharide (LPS) in Alistipes, which is known to be pro-inflammatory, leading to an increase in Th17 cells expressing CD161 and CCR6/integrin Beta7, as well as the decrease in butyrate-producing bacteria, which are known to be anti-inflammatory (47). Despite such argument, it is important to highlight that several other types of anaerobic bacteria, including species within the Bacteroides genus, may also contribute to the complexity of disease attributes conferred to date to the Alistipes genus. Further studies will be critical to elucidate more precise mechanisms.

Various studies have indicated that Alistipes plays a protective role in CVDs. Additionally, Alistipes has been associated directly with CVDs such as atrial fibrillation (AF). Atrial fibrillation is the most common arrhythmia and is prevalent in patients with hypertension, heart failure, and obesity (63). Atrial fibrillation is characterized on an EKG by the absence of P waves with irregular R-R intervals due to irregular beating of the atria leading to irregular conduction of impulses to the ventricles (64). In a study by Zuo et al. (16), conducted to quantify the relationship between the gut microbiome and AF, a whole-metagenomic shotgun sequence performed on 100 stool samples from Chinese participants, showed a drastic decrease in Alistipes spp. in the intestinal tract of patients with AF. However, bacteria that drastically increased during AF, such as Streptococcus, were proposed in the study to be the cause of the decline in Alistipes, suggesting a potential antagonistic effect between Alistipes and Streptococcus. This trend was commonly seen in other heart conditions, such as ACVD (50) and CHF (49). Evidence for the involvement of Alistipes in CVD is contradictory; that is, it is unclear if associations are protective or beneficial or pathogenic. Because most CVDs share common pathophysiological characteristics, such as endothelial dysfunction (16), it is possible that the role of Alistipes may depend on disease mechanisms shared across several CDVs. Therefore, more studies on the gut-heart axis could lead to the future understanding of microbiome-related diseases and potential therapies.

## *ALISTIPES* IN GUT INFLAMMATION AND OTHER *BACTEROIDETES*

Due to the diverse microbial community in the gastrointestinal tract, there is a strong correlation between dysbiosis and inflammatory bowel disease (IBD). The most common IBDs within the human population are Crohn's disease (CD) and ulcerative colitis (UC). UC is a chronic inflammatory disease that primarily targets the colon. It has been suggested that A. finegoldii is a protective species against colitis since A. finegoldii is decreased in mice with colitis. Due to this fact, a study was performed in which microbiota-depleted mice were treated with oral DSS to induce colitis. When gavaged with A. finegoldii, the severity of colitis was similar to that of the WT mice (42). Mice developed colitis when A. finegoldii was added with Bacteroides eggerthii, a colitis- inducing bacteria, but the severity of the colitis was significantly decreased compared to that of mice with B. eggerthii alone or added with other bacteria such as Parabacteroides distasonis, or Prevotella falsenii. This further indicates that A. finegoldii is a colitis-attenuating bacteria (42). Of contrasting clinical interest, A. finegoldii has been isolated in association with other Bacteroides from gut-wall cavernous fistulous tract (CavFT) microlesions in severe Crohn's disease. The causal-effect and their prevalence in such lesions are currently under investigation in our laboratory (65–67).

A study by Butera et al. (68) showed 8 weeks old NOD2 knockout mice had an enrichment of Alistipes, anti-inflammatory cytokines (TGF-beta and IL-10), and CD4+LAP+FoxP3<sup>−</sup> regulatory T cells. A possible connection for these observations comes from studies with curcumin, a spice that has been shown to modulate bowel inflammation by increasing CD4+LAP+FoxP3<sup>−</sup> cells via IL-10 (69). To test to see the severity of the NOD2 knockout in mice with colitis, they induced colitis via intra-rectal administration of 2,4,6-Trinitrobenzene sulfonic acid (TNBS) and used the expression of anti-inflammatory cytokine mRNA to determine the severity. It was found that the NOD2 knockout mice had less severe colitis than the wild-type. Also, it was noted that the severity of the colitis was related to the different proportion of CD4+LAP+FoxP3<sup>−</sup> cells observed prior to the TNBS treatment. Moreover, previous studies have shown a common trend of Alistipes abundance in NOD2 knock-out murine microbiota profiles (70). Of interest, Alistipes has been observed increasing among patients taking probiotics in an anti-inflammatory effect background (45). To date, it remains unclear what mechanisms of interaction exist between this genus and the other microorganisms in the gut, including food and probiotic strains, and the intramural fistulizing lessions we reported in surgical patients with CD. Metagenomic studies of fecal samples from the mouse model of spontaneous CD ileitis (SAMP1/YitFc, characterized by a fully penetrant 3D- stereomicroscopic pattern of segmental cobblestone ileitis resembling human CD) revealed an enrichment of Alistipes compared to the parental AKR/J mouse colony (cohabiting for >10 years in the same room), suggesting that Alistipes spp. could be associated with the promotion of segmental ileitis (65, 71, 72).

#### *ALISTIPES* AND CANCER VIA BENEFICIAL IMMUNOMODULATION

Cancer, like cardiovascular disease, causes high rates of mortality and morbidity worldwide. Colorectal cancer (CRC) is one of the most common types of cancer typically targeting older individuals, as well as African Americans. CRC is also a form of cancer that has been linked to dysbiosis of the gut microbiota. Alistipes has been found to contribute to the pathogenesis, acting as a potential pathogen. Moschen et al. (15) showed that A. finegoldii promotes right sided colorectal cancer via the IL-6/STAT 3 pathway. Lipocalin 2 (LCN 2), an anti-microbial protein that binds to siderophores ultimately reduces iron availability (73). In patients with IBD LCN 2 is found in high concentrations in mucosal and fecal samples. Essentially, this can reduce the prevalence of Alistipes, as iron is a regulatory factor for the growth of A. finegoldii. However, Moschen et al. (15) went on to show that A. finegoldii caused intestinal inflammation after 1 week of being orally administered in WT, LCN 2 KO, and IL-10 KO C57BL/6J mice. Therefore, the paper concluded that Alistipes thrives in an inflamed environment that lacks LCN 2 and promotes inflammation and tumor formation. Moreover, they showed that Alistipes finegoldii was found in higher abundance in the ceca than other locations within the large intestine.

Despite the pathogenic effects observed for Alistipes in CRC, this genus has been shown to have a beneficial role in cancer immunotherapy by modulating the tumor microenvironment. One of the main hallmarks of cancers is to evade the immune system. Therefore, one form of anti-cancer treatment has been to manipulate the tumor microenvironment. An example of immunotherapy is to manipulate the microenvironment by inducing tumor necrosis factor (TNF) production by tumorassociated myeloid cells which ultimately leads to tumor eradication. One way to do this is to use a combination of intra-tumoral CpG-oligodeoxynucleotides (ODN) to activate TLR9, and inhibitory IL-10R antibodies. This immunotherapy typically halts tumor growth and induces TNF-dependent hemorrhagic necrosis by tumor-associated myeloid cells leading to tumor suppression.

In a study performed by Iida et al. (61) C57Bl/6 mice were injected subcutaneously with MC38 colon carcinoma cells and pre-treated with antibiotics (vancomycin, imipenem, and neomycin). The study initially determined whether antibiotics affected cancer immunotherapy. The authors found that antibiotics led to a decrease in efficiency of the tumor eradication due to a reduction in TNF production. The investigators then determined if those results depended on the bacterial load in the intestinal tract. Therefore, germ free (GF) mice with the MC38 tumors received anti-IL-10R/CpG-ODN treatment. Treated GF mice produced a significantly lower amount of TNF than specific pathogen free (SPF) mice. This suggests that the tumor-associated innate myeloid cells are primed by microbiota for inflammatory cytokine production in response to anti-IL-10R/CpG-ODN and that the reduced bacterial load from either the antibiotic treatment or germ-free status reduces this response and the TNF-dependent early-tumor necrosis. To better understand the role of antibiotics and the role of the gut microbiota, MC38 tumor bearing mice were gavaged with LPS and the TNF expression was reestablished. When they examined the microbiota involved via fecal microbiota composition, there was a positive correlation between Alistipes genus and a role for TLR4-priming/TNF production. Iida et al. (61) believed the TNF restoration was due to the role of pro-inflammatory gram-negative bacteria, A. shahii, binding to TLR4, priming the expression of TNF production. To further prove their hypothesis, authors then showed a delay in the recovery of A. shahii following antibiotic treatment, which also paralleled an ∼4-week phase of TNF restoration after the antibiotic administration. Furthermore, they showed that when mice that were pre-treated with antibiotic and gavaged with A. shahii, the tumor-associated myeloid cells function to produce TNF was restored. Clinically relevant, the study indicated that when there is a reduction in Alistipes, there is a parallel reduction in optimal responses to cancer immunotherapy (61).

Others have also identified a role for Alistipes in cancer immunotherapy. For instance, non-small cell lung carcinoma (NSCLC) has a poor prognosis and no current therapy. Recently checkpoint inhibitors for PD-1 have been proposed as a potential immunotherapy for NSCLC. However, a major number of patients still persist with poor prognosis. Nivolumab is a fully human Immunoglobulin G4 monoclonal antibody against PD-1, therefore it blocks T cells from binding to the ligand PD-1L, typically expressed by the tumor, preventing T cell exhaustion. A study to find a potential correlation between gut microbiota and patients with NSCLC responding favorably to Nivolumab was conducted that presented evidence that A. putredinis was increased in patients who responded well to Nivolumab correlating with the study mentioned above (74).

#### MENTAL HEALTH

Although Alistipes can be found commonly in the intestinal tract, it has been shown to have a significant effect on diseases with localization outside of the gut such depression, anxiety, chronic fatigue syndrome, autism, cirrhosis, and aging. Dysbiosis within the intestine can affect the gut-brain axis and be used to explain the relationship between the gut microbiota, depression, and other mood disorders such as anxiety.

In a study conducted with BALB/c mice placed in a stressful environment, induced by grid-floor housing, there was a significant increase in Alistipes abundance (17). Furthermore, Alistipes concentration was also found to be increased, almost 4-fold, in Norwegian patients suffering from chronic fatigue syndrome (51). These findings correlate with the evidence of an

increase in Alistipes for patients suffering from depression, since patients with depression typically struggle with fatigue and stress (75). It is believed that this increase in Alistipes disrupts the gutbrain axis because Alistipes is an indole-positive organism, and, thus decreases serotonin availability. Tryptophan is a precursor for serotonin, and a decrease in serotonin is associated with depression (52). Moreover, Alistipes has been shown to express glutamate decarboxylase, an enzyme that metabolizes glutamate into γ-aminobutyric acid (GABA) in chickens (58). This increase in Alistipes abundance could also possibly be related to the increase in GABA. However, studies should be done to show if the GABA is being secreted into the gut lumen (58).

Additionally, there are associations between the brain and gut amongst patients suffering from autism spectrum disorder. It is often found that individuals suffering from autism have frequent GI symptoms, speculating that it could be due to a dysbiosis within the microbiota. Strati et al. (53) found that there was a decrease in Alistipes in patients with autism spectrum disorder. However, another study done on a different form of autism, PDD-NOS, showed a significant abundance of Alistipes in children (54). It has been speculated that this could be from the production of propionic acid, which has shown to have neurobiological effects in rats (76). There is need for more studies on Alistipes and its effects on the gut-brain axis, since there is contradictory evidence regarding its protective/pathogenic role in both systems.

## SULFONOLIPIDS AND BIOCHEMICAL MARKERS

Alistipes is a genus of bacteria with numerous immunological and biochemical pathways that are associated with the diseases mentioned above. One important implication is the promotion of CRC by Alistipes via IL-6/STAT 3 pathway. Thus, future studies could consider the use of Alistipes species as potential biomarkers for CRC, using our understanding based on microbiome DNA based data and the integration of biochemical concepts on disease pathogenesis. A potential method to accomplish this would be to look for sulfonolipids, a unique class of sphingolipids with a sulfonic acid group in the sphingoid base (77). Walker et al. (78) showed that when C57BL/6N mice are fed high fat diets with either safflower oil or lard fat for 3 weeks there is an increase in sulfonolipids, as well as, body weight when compared to mice fed the normal chow diet. A metagenome analysis was performed and screened for bacterial genes involved in sulfonolipid biosynthesis in the cecum of these mice. All species of Alistipes were found to produce sulfonolipids except A. inops (information on the most recent species A. megaguti, A. provencensis, and sulfonolipid production remain unknown). To further prove that sulfonolipids are a product of bacteria and a marker of Alistipes, scientists performed a mono-colonization study of germ-free (GF) mice with Alistipes spp. and detected a significant emergence of sulfonolipids in the cecum of the mono-colonized mice that was previously absent in the GF mice. Therefore, with common risk factors of CRC being high fat diets, obesity, and age (79), in addition to the increased abundance of Alistipes in these colonic conditions, there is the intriguing proposition for the potential of using sulfonolipids as markers for patients at risk for developing CRC, particularly within ethnicities, such as African-Americans with family histories of CRC. Additionally, studies should be conducted to verify if Alistipes abundance is increased in polyps, pre-cancerous vs. cancerous, being polyps another risk factor for CRC.

Moreover, Alistipes has the highest number of putrefaction pathways amongst commensal bacteria. Putrefaction is the fermentation of undigested proteins in the GI tract by the gut microbiota typically leading to bacterial production of harmful metabolites (80). These products have been reported as deleterious and associated with CRC (81). Such products include ammonia, H2S, cresol, indole, and phenol (82). In a study done to identify the main putrefaction pathways used by the gut bacteria, Kaur et al. (18), found that Alistipes contributed to histidine degradation/THF production, indole production, and phenol production. Histidine degradation/THF production have been found to release excess ammonia, which when absorbed, damages the colonic cells (83). Ammonia has also been found to increase intestinal cell proliferation and assist in the growth of cancer cells in CRC (84). Findings of excess ammonia and other Alistipes produced putrefaction products could be of use to clinicians when patients are at risk for developing CRC.

#### CONCLUSION, LIMITATIONS, AND FUTURE DIRECTIONS

Alistipes is a relatively new genus of bacteria isolated from clinical samples, although at a low rate compared to others within the Bacteroidetes phylum. At the protein level, a genomewide protein phylogenetic analysis shown in **Figure 2** illustrates that this genus may have unique functional properties that may enable them to have unique physiological roles, compared to other members within the Bacteroidetes, and/or also affect our ability to isolate these species in vitro. Therein, it is possible that the identification of this genus in clinical samples may be underrepresented as novel MS-TOF methods may not be fully capable to discriminate distinct species as separate. Immunologically, Alistipes has been seen to contribute to disease in both clinical and preclinical studies. Intriguingly, other studies have shown their presence is correlated with the promotion of healthy phenotypes such as Alistipes protective roles in diseases such as colitis, autism spectrum disorder, and various liver and cardiovascular fibrotic disorders. Despite Alistipes role in healthy phenotypes, Alistipes contrastingly has been shown to have a pathogenic role in diseases such as anxiety, myalgic encephalomyelitis/chronic fatigue syndrome, depression, PDD-NOS, and CRC.

Collectively, this review represents a summary of studies where Alistipes has been experimentally tested after inoculation into animal models, or where Alistipes has been found among other abnormally present species in human or animal microbiome studies. Thus, the perspective here presented, indicates that the genus, depending on the study discussed, could

FIGURE 2 | Protein phylogram of 155 complete genomes of the *Bacteroidetes* phylum to illustrate the potential functional distinction of the genus *Alistipes* within the group. The pipeline for genomic phylograms is described in detail as follow by PATRIC, the Pathosystems Resource Integration Center, https://docs.patricbrc.org. In short, the order-level pre-built trees in PATRIC are constructed by an automated pipeline that begins with amino acid sequence files for each genome. For each order-level tree the genomes from that order are used along with a small set of potential outgroup genomes. Branch values are not bootstrap values, which can be overly optimistic for long genomes. Instead, trees are built from random samples of 50% of the homology groups used for the main tree (gene-wise jackknifing). One hundred of these 50% gene-wise jackknife trees are made using FastTree, and the support values shown indicate the number of times a particular branch was observed in the support trees. As of May 19, 2020, there were 140 Alistipes genomes available (11287 unique contigs), of which 10 are complete.

have a leading role in the modulation of diseases, or alternatively could just have either a bystander role, or a co-inducer role (with other gut microbes) of the observed clinical phenotypes. Animal studies will be further needed to decipher the mechanisms that may explain disease modulation alone and as symbiont by this genus across a multitude of complex multimodal diseases, which will benefit from the targeted study of subtype phenotypes, as we have analytically illustrated (9).

The use of germ-free animals and models will be beneficial to understand the role of this genus in disease and health and the interaction with the host immune defense tolerance, such as there should be studies investigating the roles on the SCFAs produced by Alistipes and their effect on the various liver diseases and Alistipes direct role on T-cell differentiation.

#### AUTHOR CONTRIBUTIONS

BP, PW, AV, and AR-P conceived the performed review and wrote the manuscript. All authors discussed and edited the manuscript and approved its final version.

#### REFERENCES


#### FUNDING

This project was primarily supported by NIH grant R21 DK118373 to AR-P, entitled Identification of pathogenic bacteria in Crohn's disease. We acknowledge the Biorepository Core of the NIH Silvio O. Conte Cleveland Digestive Disease Research Core Center (P30DK097948) for supporting the abovementioned R21 project. Partial support also originated for AR-P from a career development award from the Crohn's and Colitis Foundation and National Institutes of Health award 2P01DK091222-06 (Germ-Free and Gut Microbiome Core to Dr. Fabio Cominelli, Case Western Reserve University and AR-P).

#### ACKNOWLEDGMENTS

BP was especially grateful to the Department of Pathology for the scientific and academic opportunities, their mentoring, and their support during the preparation of this manuscript. This project was part of a Master's of Science program in Pathology at Case Western Reserve University and she is currently admitted to the medical school at Wayne State University School of Medicine.


intestinal microbiota isolated from faeces. Int J Syst Evol Microbiol. (2010) 60:1296–302. doi: 10.1099/ijs.0.014571-0


**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 © 2020 Parker, Wearsch, Veloo and Rodriguez-Palacios. 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.

# Role of Kupffer Cells in Driving Hepatic Inflammation and Fibrosis in HIV Infection

Lumin Zhang and Meena B. Bansal\*

*Divison of Liver Diseases, Icahn School of Medicine at Mount Sinai, New York, NY, United States*

While the interactions between HIV and various liver cell populations have been explored, the relevance of these interactions when patients are well-controlled on ART is less clear. Therefore, we focus this perspective on HIV-related alterations that may drive hepatic inflammation and fibrosis in aviremic patients, with a focus on Kupffer cells and Hepatic Stellate Cells. Persistent CD4+ T cell depletion in the gut resulting in increased gut permeability has been postulated to play a role in systemic immune activation in HIV patients. The liver, with its unique location, remains the gatekeeper between the gut and the systemic circulation. The resident liver macrophage, Kupffer cell, is responsible for clearing and responding to these products. We propose that changes in Kupffer cell biology, in the context of HIV infection, creates a mileu that drives hepatic inflammation and fibrosis in response to microbial translocation. Targeting these pathways may be helpful in improving liver-related outcomes in HIV patients.

#### Edited by:

*M. Victoria Delpino, CONICET Institute of Immunology, Genetics and Metabolism (INIGEM), Argentina*

#### Reviewed by:

*Larisa Y. Poluektova, University of Nebraska Medical Center, United States Natalia A. Osna, University of Nebraska Medical Center, United States*

> \*Correspondence: *Meena B. Bansal meena.bansal@mssm.edu*

#### Specialty section:

*This article was submitted to Viral Immunology, a section of the journal Frontiers in Immunology*

Received: *10 February 2020* Accepted: *05 May 2020* Published: *16 June 2020*

#### Citation:

*Zhang L and Bansal MB (2020) Role of Kupffer Cells in Driving Hepatic Inflammation and Fibrosis in HIV Infection. Front. Immunol. 11:1086. doi: 10.3389/fimmu.2020.01086* Keywords: liver fibrosis, HIV - human immunodeficiency virus, hepatic stellate cell (HSCs), Kupffer cells, microbial translocation

## INTRODUCTION

End-stage liver disease is a major cause of non-AIDS related mortality in HIV+ patients even with effective anti-retroviral therapy, accounting for almost 15% of deaths (1–7). As a result of shared routes of transmission, HCV and HBV are the most common liver diseases in HIV-infected patients, although other chronic liver diseases are emerging (8, 9). Most data, therefore, regarding fibrosis progression rates is derived from those with coinfection. These patients have a higher relative risk (RR) of cirrhosis, increased development of decompensated cirrhosis and accelerated fibrosis progression rates compared with those who are only infected with HCV or HBV (10, 11). Furthermore, rapid fibrosis correlates with reduced CD4+ T cell counts and detectable plasma HIV levels. HIV patients are also more susceptible to other liver diseases, which synergize to accelerative liver fibrosis. Alcohol consumption is associated with increased relative risk of fibrosis progression in HIV mono-infected patients (12) while NASH is emerging as a major cause of liver disease, with half of mono-infected patients with unexplained liver enzyme elevations having NASH (13). While many may have an unrecognized chronic liver injury, a higher frequency of liver fibrosis was demonstrated in HIV-1–monoinfected patients (range 11–40.9%) compared with uninfected patients even without coinfection of hepatitis viruses and alcohol abuse, suggesting a correlation between HIV-1 infection and advanced liver fibrosis (14–19). Therefore, persistent HIV-1 infection and viral associated liver immune dysfunction may independently contribute to the progression of liver diseases (20). Lastly, in those on ART, drug-induced liver injury and increased rates of NASH due to both medications and metabolic derangements common in HIV are being observed. While hepatic stellate cells are the downstream effector of liver fibrosis, this perspective focuses on the role Kupffer cells play in promoting a mileu conducive to fibrosis progression in patients with HIV infection, particularly in aviremic patients.

## THE LIVER AS THE GATEKEEPER

Shortly after HIV infection, a severe CD4+ T cell depletion in the gut-associated lymphoid tissues leads to a disruption of the intestinal barrier, consequently promoting translocation of microbial products into the portal circulation. The liver, which derives the majority of its blood flow from the portal circulation, is uniquely positioned to protect the systemic circulation from gut-derived products. In particular, the resident hepatic macrophage, the Kupffer cell, located within the hepatic sinusoid is charged with clearing translocated bacterial products in an immunotolerant manner. However, when products provoke a pro-inflammatory response by Kuppffer cells, a cascade of intrahepatic inflammatory respones is initiated with numerous secreted cytokines, such as IL-1β, TNF-α, and IL-6, serving as major drivers in the progression of liver injury and fibrosis.

### KUPFFER CELLS AT THE NEXUS OF LIVER INFLAMMATORY RESPONSES

Kupffer cells (KCs) are the largest population of resident tissue macrophages in the liver. They reside within the hepatic sinusoid in close proximity to hepatic stellate cells, liver sinusoidal endothelia cells, and intrahepatic lymphocytes. Both the low flow state of the portal circulation and the uniquely fenestrated endothelium create a conducive environment for interaction of KCs with neighboring cells and circulating cells of the immune system. Physiologically, KCs are the first line of defense to eliminate macromolecules, immune complexes, senescent cells, virally-infected cells, and translocated microbial products from the gut to avoid liver injury and systemic immune responses (21). Given the dynamic nature of cell surface receptor expression on macrophage populations and some controversy regarding their origins, CD163 or CD68, CD14 and CD16 are often used to identify human KCs. However, murine KCs display phenotypic patterns characterized by F4/80+, MHCII, and CD11bInt expression. A detailed discussion of markers for various macrophage subpopulations within the liver is beyond the scope of this perspective and discussed elsewhere (22). The focus of this perspective is on the role of CD68+human KCs in promoting liver inflammation and fibrosis in patients with HIV.

The importance of KCs in liver injury and inflammation have been established with depletion studies wherein GdCl<sup>3</sup> was associated with AST reduction and inflammation in an alcohol model of liver injury (23). Crosstalk between KCs and HSCs is also evidenced by KC depletion as mRNA levels of TGF-β, α-SMA and collagen I are significantly decreased (24). Although GdCl3 is not specific to KCs, and thus interpretation is complex, GdCl3 treatment dramatically decreased cytokines predominantly produced by KCs, TNF-α, IL-6, and IL-1β, in response to LPS stimulation in murine livers (25, 26). Similarly, liposome/clodronate can suppress pro-inflammatory responses through the depletion of KCs (27).

In homeostasis, KCs are central to intrahepatic immune tolerance through an antigen-mediated induction of functional arrest of CD4 cells and regulatory T cells. However, in an inflamed microenvironment this delicate equilibrium is disrupted resulting in immune dysfunction and tolerance break (28). Indeed, knockout of TREM-1 (Triggering receptor expressed on myeloid cells), which is highly expressed on KCs in liver fibrosis, reduced liver fibrosis through the inhibition of TNF-α and IL-6 responses in a number of chronic injury models (29). Similarly, knock down of Jun N-terminal kinase ½ (JNK-1/2) from KCs reversed liver fibrosis in a cholinedeficient L-aminoacid-defined (CDAA) model, with a decline in inflammatory responses, including TNF-α, IL6, IL-1β, and TGF-β (30).

While KCs display M1-like features in acute liver injury, with protracted chronic inflammation, due to exhaustion of M1-like macrophages and immune cells, M2-like macrophages emerge and secrete protective cytokines upon chronic cytotoxic stimulation such as IL-4, IL-10, and TGF-β (31, 32). IL-10, an anti-inflammatory cytokine, down-regulates macrophage effector functions and differentiation of neighboring cells to maintain immune microenvironment homeostasis. For example, administration of IL-10 decreased TNF-α produced from LPStreated KCs (33). While very complex, the manipulation of KC mediated immune responses or approaches to limit their stimulation may be exploited therapeutically.

## MICROBIAL TRANSLOCATION AND KUPFFER CELLS

The impact of translocated microbial products on KCs is wellestablished. pretreatment with 2.5% dextran sulphate sodium (DSS) causes increased intestinal permeability and promotes translocation of microbial products into the portal blood in mice. The resulting amplified TLR4 mediated inflammatory responses in KCs resulted in significant livery injury (34). Using a liver slice model, LPS stimulation increased IL-1β and TNFα production compared to the control (35). Consistently, in mouse models, LPS administration rapidly induces the release of inflammatory cytokines in the liver with a higher IL-6 production obtained from LPS stimulated KCs than splenic and alveolar macrophages (36).

## THE ROLE OF TLR4 SIGNALING IN INFLAMMATORY RESPONSES OF KUPFFER CELLS

TLR4, as one member of Toll-like receptors, belongs to the pattern recognition receptor (PRR) family. After stimulation by TLR4 ligands, for example lipopolysaccharides, TLR4 is activated through conformational changes and interaction with TIRdomain-containing adapter proteins via hydrophilic interactions. Intracellular TLR4 signaling is mediated by two classical Zhang and Bansal HIV-1, Kupffer Cells, Liver Fibrosis

pathways: the TIRAP–MyD88-NF-κB pathway and the TRIF– TRAM-interferon regulatory factor-3 (IRF3)-NF-κB pathway. TLR4 signaling participates in the initiation of pro-inflammatory response, especially TRIF mediated TNF-α and synthesis of chemokines and have been reveiwed in detail elswhere (37).

In addition to TNF-α, TLR4 signaling also contributes to the transmission of two priming signals for the IL-1β pathway through the NLRP3 inflammasome. IL-1β is a crucial proinflammatory cytokine in response to microbial infection. IL-1β from LPS-treated KCs can produce a deleterious effect on hepatocytes and promote the secretion of VLDL apo B and lipid (38). IL-1β was also found to inhibit IFN-α induced STAT1 activation in hepatocytes, attenuating the innate immune response to viral infection in hepatocytes (39). In general, NLRP3 mediated-cleavage of caspase 1 is the critical step to promote the maturation of IL-1β. The formation of the NLRP3 inflammasome is initiated by ATP or microbial stimulation (40). Blockage of NLRP3 activation in KCs decreased IL-1β response to Ischemia/Reperfusion induced liver injury and improved survival (41, 42). Administration of MCC950, a small molecule selective inhibitor of NLRP3, suppressed LPS primed IL-1β response in NPC cells, subsequently, decreasing liver injury (43). Given the important role of TLR4 signaling in KCs, the modulation of this pathway in the context of HIV infection and persistent microbial translocation is critical.

### MODULATION OF INFLAMMATORY RESPONSES BY HIV-1 INFECTION IN KCS

In addition to CD4, both CCR5 and CXCR4, HIV-1 coreceptors, are detected on human KCs isolated from non-HIV-1 individuals, suggesting that KCs are permissive for HIV-1 infection. HIV-1 infection of KCs in viremic patients has been shown by in situ hybridization for HIV-1 RNA and PCR for proviral DNA on FACS-purifed KCs from livers of patients with Acquired Immunodeficiency Sydnrome (AIDs) (44–46). Moreover, retrieval of HIV-1 from KCs derived from patients either not on ART (47) or on ART for short durations has been shown and supported by studies in SIVDH12R-infected macaques (48, 49). Recently it has been shown that KCs derived from patients on long term ART, while containing evidence of HIV-1 transcripts, do not secrete replication competent virus (50). While macrophages are known to be able to transmit infectious virus to susceptible CD4+ cells via cell-cell contact (51, 52), the ability of KCs in patients on long-term ART to do so has not yet been explored though warrants investigation.

We have shown that Kupffer cells are highly permissive for HIV-1 infection in vitro with robust and sustained viral replication (53). HIV-1BaL, a laboratory adapted CCR5-tropic HIV, infection rendered KCs more sensitive to LPS treatment through an increase in CD14 and TLR4 expression on the cell surface, resulting in increased secretion of TNF-α and IL-6, which was blocked by a small molecule TLR4 inhibitor. Interestingly, despite AZT and ritonavir abrogated viral replication, KCs maintained their sensitivity to the pro-inflammatory response to LPS. These findings suggest that even in patients on ART, KC biology may be impacted and promote a mileu supporting hepatic inflammation and fibrosis in response to microbial translocation. While no change in IFNα or IFNβ expression in HIV-1 infected KCs was observed, IL-1β mRNA and both intracellular and secreted IL-1β was increased by HIV-1BaL infection. Similar to the IL-6 and TNF-α response, this HIV-related sensitization was found to be TLR4-dependent and further determined to be via the NLRP3-caspase 1 pathway. Immunostaining on liver tissue derived from aviremic HIV+ patients demonstrated an increased expression of IL-1β compared to normal liver with a high degree of colocalization in CD68+ macrophages (54). These studies show that TLR4 mediated NLRP3 activation is critical for the inflammatory responses to microbial products in KCs. Importantly, liver injury and resulting damage-associated molecular patterns (DAMPS) also activate TLR4 signals in KCs and thus may play a role in other forms of liver injury in HIV patients such as drug-induced liver injury. Interestingly, it has also been shown that CCR5 and TLR may co-cluster on monocyte-derived macrophages (MDMs) as secretion of CCL2 and CXCL8 in response to either R5 gp120, recombinant envelope protein from CCR5-tropic HIV-1, or LPS can be blocked by either a CCR5 inhibitor or TLR4 blocking. These results suggest another mechanism for synergistic effects of HIV and LPS on macrophage biology and should be specifically examined in human KCs (55).

#### INFLAMMATORY RESPONSES TO HIV-1 INFECTION IN OTHER LIVER IMMUNE CELLS

While beyond the scope of this perspective, HIV-1 infection impacts a number of other cells critical to the inflammatory response in the liver. In line with circulating CD4+ T cells, HIV infection leads to a depletion of CD4+ T cell in the liver with relative reversal of CD4/CD8 ratio typically seen. Viral infection also makes IL2+ CD4+ T cells dysfunctional and attenuates hepatic immune response to microbial infection (56–58). CD4+ T cells from HIV mono-infected patients exhibit a low regulatory effect on Natural killer (NK). Co-cultured with NK cells, CD4+ T cells from HIV-1+ individuals greatly reduced anti-fibrotic effect of NK cells on HSCs (59). Therefore, reduction in CD4+ T cells influences progression of liver fibrosis in HIV+ patients (10, 60) while increased relative CD8+T cells correlates with a higher fibrosis scores in HIV-1 infected patients (61).

NK cells, which account for up to 30–50% human liver lymphocytes, play an important role in clearing virally infected cells trough NK cell antibody dependent cell cytotoxicity (ADCC). The activation and NK cellular numbers are spontaneously increased early in response to HIV-1 infection but with chronic infection exhaustion results in NK dysfunction with persistent viremia (62).

While the role of DCs in HIV-1 infection and progression and ability to transmit infectious virus to CD4 cells by cell-cell contact has been shown, HIV interaction with DCs in the liver is less studied. TLR7 is constitutively expressed by human pDCs. The delivery of HIV-1 viral nucleic acids in early endosome of pDCs can be blocked by TLR7 inhibitor (63), suggesting that TLR7 is involved in the antigen presentation by pDCs. In addition, mDC express TLR4 and the frequency of mDCs, especially CXCL16 producing mDCs has been shown to be associated with the level of microbial products in the liver of HIV+ patients (64).

#### INTERACTIONS BETWEEN HIV AND HEPATOCYTES AND IMPLICATIONS FOR HEPATIC INFLAMMATION AND FIBROSIS

In vitro studies have shown that the envelope protein, HIV gp 120, which binds either CXCR4 (X4) or CCR5 (R5) on its target cell can promote hepatocyte apoptosis (65) and along with the HCV glycoprotein E2 promote the secretion of the proinflammatory cytokine IL-8 (66, 67). It has been known that Kupffer cells play a primary role in the clearance of apoptotic hepatocytes/cellular debris within the liver and thus play a key role in sterile inflammation and repair (68). More recently, Ganesan et. al demonstrated that ethanol exposure promotes HIV accumulation within hepatocytes, ultimately leading to increased oxidative stress and apoptosis. These apoptotic hepatocytes then stimulate inflammasome activation in KCs and pro-fibrogenic genes in hepatic stellate cells (69). Moreover, hepatic stellate cells can also engulf apoptotic hepatocytes resulting in NAPDH oxidation, stellate cell activation, and fibrogenesis (70). Therefore, effects on HIV on hepatocytes can promote both KC and HSC activation, synergistically driving hepatic inflammation and fibrosis.

### DIRECT INTERACTIONS BETWEEN HIV AND HUMAN STELLATE CELLS

HSCs express both HIV CCR5 and CXCR4 co-receptors. We have shown that HIV and its envelope protein gp120 promote HSC activation, collagen I production, and CCL2 secretion through interactions with CXCR4 (71) and others have shown that the envelope protein on HIV that preferentially uses CCR5 for cellular entry (R5 gp120) promotes HSC chemotaxis and CCL2 secretion (72). While HIV can infect HSCs in vitro, infection in vivo has not been established. Similar to what has recently been shown for KCs, in vitro infected HSCs do not secrete replication competent virus though, like DCs, may be able to transmit virus by cell-cell contact (71, 73, 74). Similar to what has been shown on MDMs, CCR5 and TLR4 seem to cocluster on HSCs and result in increased CCL2 and CXCL8 in response to gp120 and LPS, with effects of ligands blocked by inhibiting either receptor alone (55). As CCL2 is an important chemokine for attracting circulating monocytes into the liver, this may be important for propagating hepatic inflammation. R5 gp120 also promotes IL-6 secretion from HSCs through Jun-NF-kB activation (75). These studies suggest that HIV promotes inflammation and fibrosis by interacting with CXCR4 and CCR5 via gp120 and synergizes with TLR4 signals. While the

latter is important in viremic patients, relevance for those on ART are not clear. For those on ART, the impact of HIV on microbial translocation and KC biology may be more important.

#### INTERPLAY BETWEEN KUPFFER CELLS AND HEPATIC STELLATE CELLS

While the ultimate effector cell in liver fibrosis is the hepatic stellate cell, the signals generated by KCs are critically important in promoting the activation of HSCs and then perpetuating the activated state. TLR4 activation on HSCs results in downregulation of the TGFβ1 pseudoreceptor, BAMBI, which sensitizes HSCs to the pro-fibrogenic effects of TGFβ1 (76), much of which is derived by KCs. Therefore, in the context of HIV-1 and associated microbial translocation, the effects on both KCs and HSCs are compounded and drive hepatic inflammation and fibrosis (**Figure 1**). Overall association between microbial translocation and liver fibrosis progression has been shown in a variety of liver diseases and thus HIV simply compounds this effect.

### CONCLUSION

As patients with HIV live longer, liver disease will continue to emerge as a leading cause of morbidity and mortality.

#### REFERENCES


Understanding how HIV may set the stage for hepatic injury, inflammation and fibrosis may lead to novel therapeutic strategies. While treatment of underlying diseases, ranging from viral hepatitis to NASH or alcohol, remains the most imporant strategy, the alterations unique to this population need to be kept in mind. With this perspective, HIV related alterations in KC biology and microbial translocation may be at the nexus of creating a milue conducive to hepatic fibrosis. Targeting either the KC response to TLR4 ligands, PAMPs or DAMPS, decreasing the burden of microbial products from reaching the portal circulation, or blocking downstream pro-inflammatory or pro-fibrogenic effects on stellate cells are important to consider. Much will be learned from current treatments undergoing investigation for non-HIV related liver fibrosis that may be additionally leveraged for this special population.

### AUTHOR CONTRIBUTIONS

LZ and MB wrote and reviewed this manuscript.

#### FUNDING

This work was supported by research grants NIH R01DK108364 to MB.

C virus coinfected patients. The Multivirc Group Hepatol. (1999) 30:1054–8. doi: 10.1002/hep.510300409


of proteasome-dependent pathway. J Immunol. (2000) 165:3959–65. doi: 10.4049/jimmunol.165.7.3959


in humanized mice infected with HIV-1. J Exp Med. (2007) 204:705–14. doi: 10.1084/jem.20062411


**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 © 2020 Zhang and Bansal. 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.