Molecular Mechanisms Underlying Microrna-122 Mediated Suppression of Liver

Molecular mechanisms underlying microRNA-122 mediated suppression of liver

inflammation, fibrosis, and carcinogenesis

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

Kun-Yu Teng

Graduate Program in Molecular, Cellular and Developmental Biology

The Ohio State University

2017

Dissertation Committee:

Dr. Kalpana Ghoshal, Advisor

Dr. Samson T. Jacob

Dr. Jianhua Yu

Dr. David R. Brigstock

Kun-Yu Teng

2017

Abstract

MicroRNA-122 (miR-122) is a liver-specific microRNA that maintains liver homeostasis by regulating lipid metabolism, cell differentiation and viral infections in vertebrates. In clinics, hepatocellular carcinoma (HCC) patients with low miR-122 levels are usually associated with poor prognosis, implying miR-122 functions as a tumor suppressor. This notion is supported by the phenotypes of miR-122 genetic knockout (KO) mouse that develops spontaneous hepatitis, steatosis, fibrosis and HCC with age. Although several studies have shown the importance of miR-122 in maintaining liver homeostasis, the mechanisms by which loss of miR-122 contributes to these liver pathological processes remains largely unknown. In the current study, we explored the role of miR-122 in regulating liver inflammation and fibrosis by combining molecular, biochemical, and bioinformatic analysis.

We demonstrated that the increased expression of the chemokine CCL2 in the liver is one of the causes of liver inflammation upon deprivation of miR-122. Blocking CCL2 using specific neutralizing antibody (CCL2 nab) ameliorates liver inflammation and tumorigenesis through decreasing the population of CD11b+/Gr1+ cells and their corresponding downstream pathways such as the IL-6-Stat3-cMYC axis and TNF-α-NF-κB axis. Along with the shrinking tumors in the CCL2 nab treated liver, CCL2 nab also activated natural killer (NK) cells and increased their cytotoxicity toward tumor cells. Besides its robust role in inhibiting liver inflammation and HCC tumors, miR-122 also has a strong anti-fibrosis

ii function. Analyzing database that contained both clinical and expression profiles of liver cirrhosis patients revealed downregulation of miR-122 in the cirrhotic liver tissues compared to normal livers. Ectopic expression of miR-122 in LX-2, an immortalized human hepatic stellate cell (HSC) cell line, reduced cell proliferation and fibrotic gene expressions.

Moreover, co-culture of miR-122 expressed HCC cells with miR-122 non-expressing LX-2 cells resulted in miR-122 expression and reduced expression of fibrotic genes, suggesting extracellular miR-122 could be delivered from hepatocytes to HSCs. We next searched hepatic microarray and RNA-seq data in the wild type and miR-122 KO mice to identify miR-122 downstream molecules that promote liver fibrosis. We focused on genes that have been shown to be pro-fibrotic and overexpressed in the human fibrotic livers as well as in the miR-122 KO livers. One such fibrotic gene, B cell lymphoma 2 (BCL2), was found to be upregulated in the HBV- and HCV-infected cirrhotic livers compared to the normal livers

(data retrieved from GEO and EMBL-EBI). Besides, BCL2 expression is significantly increased in the mouse livers depleted of miR-122 as demonstrated by microarray, RNA-seq and immunoblot analysis. Venetoclax, an FDA approved BCL2 inhibitor for chronic leukemia lymphoma (CLL), inhibited LX-2 cell proliferation and expression of pro-fibrotic markers, and reduced liver fibrosis in the miR-122 KO mice. Collectively, this study provides a new mechanistic insight on miR-122 mediated suppression of liver inflammation, fibrosis, and HCC. Moreover, our data demonstrated that targeting miR-122 downstream molecules (e.g. CCL2, and BCL2) could be efficient in blocking spontaneous liver inflammation, fibrosis and HCCs developed in miR-122 depleted livers.

iii

Dedication

This document is dedicated to my wife, my son, and my parents.

Acknowledgments

I would like to give my sincere gratitude to my advisor Dr. Kalpana Ghoshal for her great support and mentorship during my Ph.D. career. Dr. Ghoshal inspired me to do good science by her great passion and extensive experience in conducting research. The training I received from her has prepared me well for my future career.

I am sincerely grateful to my co-advisor Dr. Samson T. Jacob who gave me the opportunity to join the lab and learn from him. Dr. Jacob always provided a deep insight not only into my research but also to my career development. His great success in the science field provides me a perfect role model to follow.

I would also like to thank my committee members Dr. Jianhua Yu and Dr. David R.

Brigstock. I have learned how to address scientific questions from Dr. Yu. His expertise in immunology also provided me a great opportunity to study the immune responses in the liver. I also gained my expertise in molecular experiments from Dr. Brigstock. He and his colleagues taught me the procedures of isolating primary hepatic stellate cells from mouse livers, which is a great technique and model for studying liver fibrosis.

I owe special thanks to Dr. Sarmila Majumder for her training and supports in my early career in science. Dr. Majumder opened the door of science for me and guided me to master

v my molecular techniques in conducting research, which has been invaluable throughout my science career.

I would like to extend my enormous gratitude to all the lab members, collaborators, and friends who have helped me in my research. I appreciate the discussions during the lab meetings and all the suggestions for my experiments. I also thank for the technical assistance from all of them, which contributed to my publications.

I would like to sincerely thank my family. I thank my parents from the deepest of my heart for their selfless dedication to their children’s education, which gave me a chance to accomplish my Ph.D. I also like to thank my brother Ming-Feng for his greatest support from the other side of the earth. I especially thank him for taking care of my parents while I was pursuing my Ph.D. in the states. Last but not the least, I would like to thank Sheng-Wei and Darren, the best wife and son in the land in my opinion. There is not a perfect word or phrase that could precisely express my thanks to both of them. Their supports have been tremendous throughout my life. It would be impossible to achieve my Ph.D. without their understanding and love. I thank them from the bottom of my heart.

Vita

2008 ...... B.S. Biological Science and Technology, China

Medical University, Taiwan

2012 to present ...... Graduate Research Associate, Molecular,

Cellular, and Developmental Biology Program,

The Ohio State University

Publications

1. Chowdhary V, Teng KY, Zhang B, Yang D, Thakral S, Wani NA, Lin CH, Kutay H,

James L, Sharma AD, Bruschweiler R, Lee WM and Ghoshal K. miR-122 protects mice

and human hepatocytes from acetaminophen toxicity by suppressing CYP2E1 and

CYP1A2. American Journal of Pathology. 2017; in press

2. Luna JM, Barajas JM, Teng KY, Sun HL, Moore MJ, Rice CM, Darnell RB and

Ghoshal K. Argonaute CLIP defines a deregulated miR-122 bound transcriptome that

predicts patient survival in human liver cancer. Mol Cell. 2017 Aug 3;67(3):400-410.e7.

3. Teng KY, Han J, Zhang X, Hsu SH, He S, Wani N, Barajas J, Snyder LA, Frankel WL,

Caligiuri MA, Jacob ST, Yu J, Ghoshal K. Blocking the Ccl2-Ccr2 axis using Ccl2

vii

neutralizing antibody is an effective therapy for hepatocellular cancer in a mouse model.

Mol Cancer Ther. 2017 Feb;16(2):312-322.3.

4. Sun HL, Cui R, Zhou J, Teng KY, Hsiao YH, Nakanishi K, Fassan M, Luo Z, Shi G,

Tili E, Kutay H, Lovat F, Vicentini C, Huang HL, Wang SW, Kim T, Zanesi N, Jeon YJ,

Lee TJ, Guh JH, Hung MC, Ghoshal K, Teng CM, Peng Y, Croce CM. ERK Activation

Globally Downregulates miRNAs through Phosphorylating Exportin-5. Cancer Cell.

2016 Nov 14;30(5):723-736.

5. Hsu SH, Delgado ER, Otero PA, Teng KY, Kutay H, Meehan KM, Moroney JB, Monga

JK, Hand NJ, Friedman JR, Ghoshal K, Duncan AW. MicroRNA-122 Regulates

Polyploidization in the Murine Liver. Hepatology. 2016 Aug;64(2):599-615.

6. Teng KY, Ghoshal K. Role of Noncoding RNAs as Biomarker and Therapeutic Targets

for Liver Fibrosis. Gene Expr. 2015;16(4):155-62.

7. Wang X, He H, Lu Y, Ren W, Teng KY, Chiang CL, Yang Z, Yu B, Hsu S, Jacob ST,

Ghoshal K, Lee LJ. Indole-3-carbinol inhibits tumorigenicity of hepatocellular

carcinoma cells via suppression of microRNA-21 and upregulation of phosphatase and

tensin homolog. Biochim Biophys Acta. 2015 Jan;1853(1):244-53.

8. Ramaswamy B, Lu Y, Teng KY, Nuovo G, Li X, Shapiro CL, Majumder S. Hedgehog

signaling is a novel therapeutic target in tamoxifen-resistant breast cancer aberrantly

activated by PI3K/AKT pathway. Cancer Res. 2012 Oct 1;72(19):5048-59.

9. Peng Z, Shen R, Li YW, Teng KY, Shapiro CL, Lin HJ. Epigenetic repression of

RARRES1 is mediated by methylation of a proximal promoter and a loss of CTCF

binding. PLoS One. 2012;7(5):e36891.

viii

10. Chang SH, Liu CJ, Kuo CH, Chen H, Lin WY, Teng KY, Chang SW, Tsai CH, Tsai FJ,

Huang CY, Tzang BS, Kuo WW. Garlic Oil Alleviates MAPKs- and IL-6-mediated

Diabetes-related Cardiac Hypertrophy in STZ-induced DM Rats. Evidenced Based

Complement Alternate Med. 2011;2011:950150.

Fields of Study

Major Field: Molecular, Cellular and Developmental Biology

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments ...... v

Vita ...... vii

Table of Contents ...... x

List of Tables ...... xiii

List of Figures ...... xiv

List of abbreviations ...... xvii

Chapter 1: Introduction ...... 1

1.1. MicroRNA ...... 1

1.2. MicroRNA-122 ...... 2

1.3. Liver inflammation and miR-122 ...... 3

1.5. Hepatocellular Carcinoma and miR-122 ...... 5

1.6. Short discussion ...... 6

Chapter 2: Blocking CCL2 functions suppresses liver inflammation and hepatocellular carcinoma in a mouse model ...... 18

2.1. Abstract ...... 18

2.2. Introduction ...... 19

2.3. Materials and Methods ...... 20

2.4. Results ...... 24

2.4.1. CCL2 is upregulated in primary human HCC ...... 24

2.4.2. CCL2-CCR2 blockade reduces chronic inflammation and liver injuries in miR-

122 KO mice ...... 25

2.4.3. CCL2 nab therapy inhibits hepatitis by reducing CD11bhighGr1+ inflammatory

myeloid cells accumulation ...... 27

2.4.4. Attenuation of hepatocarcinogenesis in miR-122 KO mice treated with CCL2

nab 28

2.4.5. Oncogenic signaling downstream of IL-6 and TNF-α is blocked in KO tumors

upon CCL2 nab therapy ...... 29

2.4.6. CCL2 nab activates nature killer (NK) cells in the tumor microenvironment to

suppress liver cancer ...... 30

2.5. Discussion ...... 32

Chapter 3. miR-122 is a negative regulator of liver fibrosis ...... 77

3.1. Abstract ...... 77 xi

3.2. Introduction ...... 78

3.3. Method ...... 80

3.4. Results ...... 84

3.4.1. miR-122 is a negative regulator of liver fibrosis ...... 84

3.4.2. Extracellular miR-122 suppresses expression of fibrotic genes in HSCs ...... 85

3.4.3. BCL2 is overexpressed in miR-122 KO mouse and human cirrhotic liver ..... 86

3.4.4. Inhibition of BCL2 functions suppresses liver fibrosis developed in the miR-

122 KO mice ...... 86

3.4.5. Ago-HITS-CLIP identified a miR-122 targetome involved in liver fibrosis .. 87

3.5. Discussion ...... 88

Chapter 4: Conclusion ...... 117

Reference ...... 122

xii

List of Tables

Table 1.1 Common molecular aberrations observed in advanced HCC...... 17

Table 2.1 Primers used for RT-qPCR analysis ...... 72

Table 2.2 Antibody information ...... 73

Table 2.3 CCL2 expression across 4 independent microarrays in HCC patients ...... 74

Table 2.4 Serological analysis of 4 months old KO mice ...... 75

Table 2.5 Histopathological and serological analysis of tumor-bearing miR-122 KO mice treated with CCL2 nab for 8 weeks ...... 76

Table 3.1 Primer used for RT-qPCR ...... 111

Table 3.2 Antibody information ...... 112

Table 3.3 Raw data of RT-qPCR done in the co-cultured LX-2 cell ...... 113

Table 3.4 Pro-fibrotic miR-122 targets identified by Ago-CLIP in miR-122 WT and KO mouse...... 116

xiii

List of Figures

Figure 1.1 Canonical biogenesis of microRNAs...... 8

Figure 1.2 miR-122 is a master regulator of liver physiopathology...... 10

Figure 1.3 miR-122 knockout mouse develops spontaneous hepatitis, steatosis and fibrosis with age...... 11

Figure 1.4 miR-122 KO mouse develops spontaneous HCC with lung metastasis...... 13

Figure 1.5 Inflammation is a double-edged sword in regulating liver homeostasis and pathology...... 14

Figure 1.6 Hepatic architecture under normal and fibrosis status...... 15

Figure 1.7 A globally view of HCC incidence and risk factor...... 16

Figure 2.1 CCL2 is overexpressed in human cirrhotic livers, as well as cirrhotic and noncirrhotic HCCs...... 36

Figure 2.2 CCL2 neutralizing antibody therapy reduces chronic liver inflammation and liver damage in adult miR-122 KO mice...... 37

Figure 2.3 Mice treated with CCL2 neutralizing antibody (nab) for 4 weeks did not show significant changes in body weight...... 40

Figure 2.4 CCR2 inhibitor therapy reduces chronic liver inflammation in miR-122 knockout

(KO) mouse...... 41

xiv

Figure 2.5 CCL2 nab treatment decreases liver CD11bhighGr1+ cell population and IL-6 &

TNF-α expression in adult KO mice...... 42

Figure 2.6 CCL2 nab therapy did not reduces CD11b+ Gr1+ inflammatory cells in the blood.

...... 44

Figure 2.7 Blocking the CCL2-CCR2 axis reduces hepatic macrophages...... 45

Figure 2.8 CCL2 nab does not alter the population of B cell, T cell, or NK cell in liver. .... 46

Figure 2.9 CCR2 inhibitor reduces CD11bhighGr1+ inflammatory myeloid cells in the KO mouse liver...... 51

Figure 2.10 12-month-old male KO mice were assigned to vehicle and CCL2 nab treated group based on serum AFP levels and body weight...... 52

Figure 2.11 Blocking CCL2 function suppresses HCC development in KO mice...... 54

Figure 2.12 CCL2 nab therapy suppresses HCC development in KO mice by inhibiting tumor cell proliferation...... 58

Figure 2.13 CCL2 immunotherapy reduces tumoral p-STAT3, c-MYC, and p65...... 61

Figure 2.14 CCL2 immunotherapy reduces the number of proliferative cells in the adjacent benign tumors...... 64

Figure 2.15 Blocking CCL2 by neutralizing antibody increases tumor cell apoptosis...... 66

Figure 2.16 CCL2 nab activates NK cells in KO mouse livers...... 67

Figure 2.17 CCL2 nab therapy inhibits HCC development in KO mice by modulating tumor microenvironment...... 70

Figure 3.1 The anti-fibrotic role of miR-122 in both mouse and human fibrotic livers...... 92

Figure 3.2 Extracellular miR-122 could be delivered from hepatocytes to HSC ...... 95

Figure 3.3 BCL2 is overexpressed in miR-122 KO livers...... 98

Figure 3.4 Enhanced expression of BCL2 in the CCl4-induced fibrotic livers...... 99

Figure 3.5 BCL2 is overexpressed in the HBV- and HCV-induced cirrhotic livers...... 102

Figure 3.6 Venetoclax inhibits the expression of fibrotic genes both in vitro and in vivo. . 103

Figure 3.7 Argonaute cross linking immune precipitation (Ago-CLIP) defines miR-122 fibrotic targetome in mouse liver...... 106

Figure 3.8 CTGF is a non-canonical miR-122 target...... 108

Figure 4.1 Summary of normal and low miR-122 expressed liver...... 120

xvi

List of abbreviations

122-LM3 miR-122 expressing HCC-LM3 cells

122-PLC miR-122 expressed PLC/PRF5 cells

5’UTR 5’ Untranslated region

AFP Alpha-fetoprotein

AGO Argonaute

Ago HITS-CLIP Argonaute high-throughput sequencing of RNAs pulled down after

UV-crosslinking and immunoprecipitation with Argonaute antibody

ARID AT-rich interaction domain

BCL2 B cell lymphoma 2 c-Cas 7 Cleaved-Casepase 7

CCL2 C-C motif chemokine ligand2

CCl4 Carbon tetrachloride

CCND1 Cyclin D1

CDAA Choline deficient L-amino acid defined

CDKN2A Cyclin-dependent kinase inhibitor 2A

CLL Chronic lymphocytic leukemia c-PARP Cleaved-PARP

Cr-51 Chromium-51

xvii

CSAA Choline supplemented L-amino acid defined

CTGF Connective tissue growth factor

CTNNB1 ß-Catenin

DBILI Direct bilirubin

DGCR8 DiGeorge syndrome chromosomal region 8

ECM Extracellular matrix

ELISA Enzyme-linked immunosorbent assay

EXP5 Exportin 5

FGF Fibroblast growth factor

GGT Gamma-glutamyl transferase

GPC3 Glypecan3

HBsAg Hepatitis B surface antigen

HCC Hepatocellular carcinoma

HCV Hepatitis C Virus

HNF6 Hepatocyte nuclear factor 6

HSC Hepatic stellate cell

HSPG Heparan sulfate proteoglycan

Hy:Pro Hydroxyproline

IHC Immuno-histochemistry

IL-6 Interlukin-6

KEAP1 Kelch like ECH associated protein 1

xviii

KMT Lysine (K)-specific methyltransferase

KO Knockout

LNA Locked nucleic acid

LRP Lipoprotein receptor-related proteins

MAPK MAP kinase miR-122 microRNA-122/miRNA-122 miRNA MicroRNA

MLL Myeloid/lymphoid or mixed-lineage leukaemia (trithorax

homologue, Drosophila)

MRI Magnetic resonance imaging nab Neutralizing antibody

NAFLD Nonalcoholic fatty liver disease

NASH Non-alcoholic steatohepatitis

NFE2L2 Nuclear factor, erythroid 2 like 2

PI3K Phosphoinositide 3-kinase

Pol II Polymerase II

Pol III Polymerase III

Pre-miRNA Precursor microRNA

Pri-miRNA Primary transcript of microRNA

PTEN Phosphatase and tensin homologue

RB1 Retinoblastoma 1

RISC RNA-induced silencing complex

xix

RPS6KA3 Ribosomal protein S6 kinase, 90kDa, polypeptide 3

TBILI Total bilirubin

TERT Telomerase reverse transcriptase

TG Triglyceride

TLR Toll-like receptors

TNFα Tumor necrosis factor alpha

TP53 Tumor Protein P53

TSC Tuberous sclerosis

Vec-LM3 Vector transfected HCCLM3 cells

Vec-PLC Vector transfected PLC/PRF5 cells

VEGFA Vascular endothelial growth factor A

Chapter 1: Introduction

1.1. MicroRNA

MicroRNAs (miRNAs) are short non-coding RNA (~21-22 nucleotides) that regulate genes expression at the post-transcriptional level (1, 2). The majority of mRNA transcripts in mammals are regulated by miRNAs (3). The first miRNA (e.g. lin-4) was discovered in C. elegans through forward genetic screening method (4, 5). Since then, miRNAs have been found in fungi, viruses, plants and animals (1). In humans, most miRNAs are encoded by introns of coding or noncoding transcripts and transcribed by RNA polymerase II/ III (Pol

II/ III) (2). The primary transcripts of miRNA (pri-miRNA) is first processed by the microprocessor complex, which contains nuclear RNase III Drosha and DiGeorge syndrome chromosomal region 8 (DGCR8), to produce precursor miRNA (pre-miRNA), which is then transported to the cytoplasm through exportin 5 (EXP5) in complex with RAN•GTP (Figure

1.1) (6-8). Upon reaching cytoplasm, pre-miRNA is cleaved by Dicer-TRBP complex to form an imperfect dsRNA duplex (miRNA: miRNA*) that contains mature miRNA and its complementary strand (miRNA*) (9-11). The mature miRNA is loaded onto the Argonaute

(AGO) protein and forms an effector complex called RNA-induced silencing complex

(RISC) while most of the miRNA* is degraded (12-14). Once incorporated into the RISC complex, miRNAs base pair to target RNAs and induce translational repression, RNA

1 deadenylation or RNA degradation (15, 16). In the present study, we will mainly focus on one of the liver specific miRNAs, microRNA-122.

1.2. MicroRNA-122

MicroRNA-122 (miR-122) is a liver specific miRNA that accounts for more than 50% of the total liver miRNAs among vertebrates (17-19). miR-122 was first cloned and sequenced from mouse livers in early 2000s (18, 19). Since then, hundreds of studies were conducted to understand the functions of miR-122. miR-122 was shown to regulate lipid metabolism by regulating cholesterol biosynthesis (Figure 1.2) (20-22). Inhibition of miR-122 using antisense-mediated approaches significantly reduces serum cholesterol levels in both mice and non-human primates (20-22). In addition to metabolism, miR-122 forms a positive feedback loop with hepatocyte nuclear factor 6 (HNF6) to drive hepatocyte differentiation

(Figure 1.2) (23). HNF6 transcriptionally increases miR-122 expression, which controls hepatocyte differentiation and enhances the expression of HNF6 (23). Polyploidy, happens in 90% of hepatocytes in mice and 50% in humans, is a unique feature of liver that provides genetic diversity and facilitates adaption to liver injuries (24). The rescue of polyploidy in the predominately diploid miR-122-/- hepatocytes by AAV8-mediated delivery of miR-122 in neonatal mouse suggests miR-122 is necessary and sufficient in regulating liver polyploidization (25). One of the unique role of miR-122 is to facilitate Hepatitis C Virus

(HCV) replication and translation (Figure 1.2). HCV “kidnaps” miR-122 from cells and uses it to stabilized HCV RNA by base-pairing with mature miR-122 at its 5’ untranslated region (5’UTR) (26). With the fact that miR-122 expression is critical for HCV replication, scientists utilized this feature to develop an anti-HCV strategy. Miravirsen, a locked nucleic

2 acid (LNA)-modified antisense oligonucleotide which binds to the 5’UTR of mature miR-

122, has achieved promising results in HCV-infected patients in a phase IIA clinical trial

(NCT01200420) (27, 28). miR-122 has been shown to be associated with other liver diseases such as steatosis (Figure 1.3) (29, 30), liver fibrosis (29-32) and hepatocellular carcinoma (HCC) (Figure 1.4) (29, 30, 33, 34). These associations were demonstrated in the genetic mouse models generated by two independent groups. Mice lacking miR-122 are fertile but progressively develop spontaneous hepatitis, steatosis, fibrosis, and HCC (29, 30), indicating miR-122 is critical in maintaining the overall health of liver.

1.3. Liver inflammation and miR-122

Inflammatory responses are important to protect liver from tissue damage, pathogen invasion, and tumor formation (35). Liver inflammatory mechanisms initiate, progress, and resolve systemically to maintain liver homeostasis; however, local immune responses also contribute to liver pathology. Under certain situations such as metabolic stresses or exposure to microbial products, excessive activation of immune cells might occur and results in liver damage if the immune stimulus persists. Thus, proper control of immune responses is essential to maintain liver health. Uncontrolled inflammation due to chronic liver injuries such as viral infection could be the main driving force for liver pathology (Figure 1.5). miR-122 plays a crucial role in regulating liver inflammation at several levels. First of all, loss of miR-122 produces inflammatory stimulus in mouse hepatocytes due to excessive accumulation of triglyceride (TG) in liver (29), a major source of metabolic stress that promotes Toll-like Receptors (TLR) signaling and inflammasome activation (36). Loss of miR-122 also de-represses the expression of C-C motif chemokine ligand2 (CCL2/MCP1),

3 which has been shown to recruit macrophages and inflammatory myeloid cells (CD11bhigh/

Gr1+) to liver (29, 37). Both macrophages and inflammatory myeloid cells produce pro- inflammatory cytokines including interlukin-6 (IL-6) and tumor necrosis factor α (TNFα) to promote local inflammatory responses (29, 38). As previously mentioned, HCV suppresses miR-122 functions by acting as a sponge to absorb the pool of miR-122 in liver (39). The sponging effect and the HCV infection together activate IL-1β, a well-established cytokine that promotes liver inflammation upon HCV infection (40, 41). A direct evidence of miR-

122 is a suppressor of liver inflammation came from miR-122 KO mouse developed spontaneous liver inflammation with age (29, 30). All these studies underscore the importance of miR-122 in regulating liver inflammation.

1.4. Liver fibrosis and miR-122

Liver fibrosis is a wound-healing response following liver injuries that results in excessive accumulation of extracellular matrix (ECM) (42). Under normal situation, liver fibrosis regresses once liver recovers from injuries. However, if the injury persists for long time, liver fibrosis will progress to the end stage of liver fibrosis (i.e. cirrhosis) (Figure 1.6).

Cirrhosis, characterized by portal hypertension, sinusoidal remodeling, and hepatic endothelial dysfunction, has the highest mortality among non-cancerous diseases in the

United States (43-45).

Hepatic stellate cell (HSC) activation is one of the major events during liver fibrosis (42).

Upon liver injury, HSCs are activated and transformed into myofibroblast-like cells that are capable of synthesizing ECM proteins such as collagens, laminin and fibronectin to cause liver fibrosis (44). Overexpression of miR-122 in the human primary HSC-derived LX-2

4 cells results in reduced cell proliferation, decreased collagen maturation and ECM production (31). In vivo, overexpression of miR-122 using lenti-virus reduced collagen fibrils in the CCl4-treated mice (46), suggesting miR-122 is an anti-fibrotic molecule.

Development of spontaneous fibrosis in miR-122 depleted mice livers further supports this notion (29, 30). Spontaneous liver fibrosis along with the excessive metabolic stress and chronic liver inflammation promote malignant transformation of the damaged hepatocytes during regeneration thereby causing HCC in miR-122 KO mouse liver.

1.5. Hepatocellular Carcinoma and miR-122

Hepatocellular carcinoma (HCC) is the predominant cancer (~90%) in liver, which is the second leading cause of the cancer-related deaths worldwide (47). Approximately 850,000 new cases are identified each year and the incidence is still increasing (47, 48). The major risk factors for HCC includes viral infection (HBV and HCV), alcohol abuse, aflatoxin B1 ingestion, and non-alcoholic steatohepatitis (NASH) (Figure 1.7) (48). Majority of these risk factors are associated with chronic inflammation and fibrosis, which promotes HCC development (49). In addition to the environmental risk factors, HCC is also tightly associated with genetics, epigenetics, and post-transcriptional regulation (50, 51) (Table

1.1). Inactivation of p53, loss of Rb expression, and β-catenin nuclear accumulation are frequently found in HCC tumors (50, 52-54). Tumor suppressors including SOCS-1, GSTP,

APC, E-cadherin, and P15 were shown to be hypermethylated in their promoter region in

HCC (53-55). miR-122, another tumor suppressor, is frequently found to be down-regulated in HCV negative HCC (56-58). Mice depleted of miR-122 spontaneously develop HCC with age with high penetrance (~60%) (29, 30). Furthermore, AAV-mediated delivery of miR-

122 significantly suppresses liver tumor burden in a c-MYC induced mouse HCC model

(29).

Even with the improved medication and advanced technology, it is estimated that about

29,000 deaths caused by HCC will occur this year in the United State (Cancer Facts &

Figures 2017, https://www.cancer.org/). The high mortality rate of HCC might be explained by the shortage of efficient therapy and early detection biomarkers (59). Currently, sorafenib and regorafenib are the only two FDA approved drugs for advanced HCC. Sorafenib increases approximately 2.6 months of overall survival in the advanced HCC while regorafenib extends about 2.8 months in those patients who have been treated with sorafenib

(60, 61). Thus, novel therapeutic strategies are urgently needed to combat this deadly disease. In the next chapter, an anti-HCC strategy will be tested in a pre-clinical HCC mouse model.

1.6. Short discussion

Although the importance of miR-122 in maintaining liver homeostasis is getting clearer with scientists’ efforts, the mechanistic insights of miR-122-regulated networks in maintaining liver homeostasis still remains obscure. A better understanding of the miR-122 regulatory network in liver could possibly facilitate the development of therapeutic agents to combat miR-122-related diseases such as inflammation, fibrosis, and even HCC. In the following chapter, the downstream molecules of miR-122 in regulating liver inflammation, fibrosis, and HCC will be elucidated. Beyond the mechanistic analysis, the possibility to treat liver diseases induced by loss of miR-122 is tested as well. Completion of this study not only

6 uncovers the key players that contribute to liver disease, but also provides proper rationale to develop therapeutic agents which target miR-122 associated liver diseases in clinics.

Figure 1.1 Canonical biogenesis of microRNAs.

The primary transcript of microRNA (pri-miR) is transcribed by RNA polymerase II (Pol II) or RNA polymerase III (Pol III) then processed by DGCR8-Drosha complex to became precursor miRNA (pre-miR). The pre-miR is exported to the cytoplasm by Exportein 5

(EXP5) and Ran:GTP complex and cleaved by Dicer-TRBP complex to form a miR duplex

(miR: miR*). The mature miR will load to Ago2 to induce RNA silencing while the other strand (miR*) is degraded. This figure is reprinted with permission from Dr. Diederichs (8).

Figure 1.2 miR-122 is a master regulator of liver physiopathology. miR-122 has been shown to promote hepatocyte differentiation, cholesterol and fatty acid synthesis, and HCV replication while also suppresses HBV and HCC development. This figure is reprinted with permission from Dr. Zeisel (49).

Figure 1.3 miR-122 knockout mouse develops spontaneous hepatitis, steatosis and fibrosis with age.

(A) H&E staining of control mouse and miR-122 liver specific knockout mouse (LKO) liver. Hepatitis is characterized by the infiltrated immune cells (small dark purple cells).

Steatosis is characterized by the accumulation of lipid droplets (white circular empty spots).

Fibrosis is characterized by the accumulation of extracellular matrix (ECM) (indicated by the yellow arrows). (B) Inflammation, steatosis, and fibrosis scores for the above images generated through blinded evaluation. This figure is modified and reprinted with permission from Dr. Ghoshal (29).

Figure 1.4 miR-122 KO mouse develops spontaneous HCC with lung metastasis.

At approximately 12 months of age, miR-122 KO and LKO mice begin to develop HCC without any external stressor. A portion of KO/LKO tumor bearing mice even develop lung metastasis. The positive signals of AFP shown by IHC confirm that lung tumors were originated from liver. This figure is modified and reprinted with permission from Dr.

Ghoshal (29).

Figure 1.5 Inflammation is a double-edged sword in regulating liver homeostasis and pathology.

Under normal situation, liver inflammation is beneficial to liver homeostasis; however, if the source of inflammation persists, chronic liver inflammation could lead to other liver pathological phenotypes such as liver fibrosis and HCC. This figure is modified and reprinted with permission from Dr. O'Farrelly (35).

Figure 1.6 Hepatic architecture under normal and fibrosis status.

(A) Normal liver. (B) After liver is damaged, inflammatory cells infiltrate into the hepatic parenchyma. The apoptotic hepatocytes and activated kupffer cells release fibrogenic mediators, which activates HSC. HSC proliferates and transforms into myo-fibroblast like cell, producing most of the ECM in liver. Excessive ECM proteins and loss of fenestration from sinusoidal endothelial cells cause increased resistance to blood flow in the hepatic sinusoid, which could lead to cirrhosis. This figure is reprinted with permission from Dr.

Brenner (42).

Figure 1.7 A globally view of HCC incidence and risk factor.

The main risk factors for HCC are HCV infection (Egypt, Mongolia), HBV infection

(China, Mongolia), alcohol intake (Mongolia, United States), non-alcoholic steatohepatitis

(NASH; United States) and aflatoxin B1 ingestion (Sudan). Mongolia has the highest incidence of HCC worldwide. This figure is reprinted with permission from Dr. Gores (47).

Table 1.1 Common molecular aberrations observed in advanced HCC.

This table is reprinted with permission from Dr. Gores (47).

Chapter 2: Blocking CCL2 functions suppresses liver inflammation and hepatocellular

carcinoma in a mouse model

2.1. Abstract

Hepatocellular carcinoma (HCC), a deadly disease, commonly arises in the setting of chronic inflammation. C-C motif chemokine ligand2 (CCL2/MCP1), a chemokine that recruits CCR2-positive immune cells to promote inflammation, is highly upregulated in

HCC patients. Here, we examined the therapeutic efficacy of CCL2-CCR2 axis inhibitors against hepatitis and HCC in the miR-122 knockout (aka KO) mouse model. This mouse model displays upregulation of hepatic CCL2 expression, which correlates with hepatitis that progress to HCC with age. Therapeutic potential of CCL2-CCR2 axis blockade was determined by treating KO mice with a CCL2 neutralizing antibody (nab). This immunotherapy suppressed chronic liver inflammation in these mice by reducing the population of CD11highGr1+ inflammatory myeloid cells, and inhibiting expression of IL-6 and TNF-α in KO livers. Furthermore, treatment of tumor-bearing KO mice with CCL2 nab for 8 weeks significantly reduced liver damage, HCC incidence, and tumor burden.

Phospho-STAT3 (Y705) and c-MYC, the downstream targets of IL-6, as well as NF-κB, the downstream target of TNF-α, were downregulated upon CCL2 inhibition, which correlated with suppression of tumor growth. Additionally, CCL2 nab enhanced hepatic NK cell cytotoxicity and IFN-γ production, which is likely to contribute to the inhibition of 18 tumorigenesis. Collectively, these results demonstrate that CCL2 immunotherapy could be an effective therapeutic approach against inflammatory liver disease and HCC.

2.2. Introduction

Hepatocellular carcinoma (HCC) is the most common liver cancer and the second leading cause of cancer-related death worldwide (62, 63). The incidence of HCC has tripled in the

United States because of the precipitous increase in nonalcoholic fatty liver disease

(NAFLD) and hepatitis C virus (HCV) infection in the past twenty years (51, 64).

Furthermore, sorafenib, one of the two FDA approved drug for advanced HCC extends overall survival by only 2.8 months (65). Recent clinical trials have shown that it is ineffective as an adjuvant therapy after resection or ablation of the tumor (66). Thus, there is an urgent need to develop novel therapeutic strategies for the treatment of HCC.

MicroRNA-122 (miR-122) is the most abundant liver-specific microRNA in vertebrates and loss of miR-122 is associated with metastasis and poor prognosis in HCC patients (67). We previously found that development of spontaneous HCC with age mimicked different stages of tumor progression (e.g. steatohepatitis, fibrosis, primary and metastatic HCC) in miR-122 knockout (KO) mice (29, 30). We also observed that miR-122 depletion in the mouse liver leads to upregulation of chemokine (C-C motif) ligand 2 (CCL2), which recruits

CCR2+CD11bhighGr1+ immune cells to the liver (29). These cells, in turn, produce proinflammatory cytokines, including IL-6 and TNF-α in the liver, resulting in hepatitis and eventually HCC.

CCL2 is known to be involved in the pathogenesis of several diseases characterized by monocytic infiltrates, such as psoriasis, rheumatoid arthritis, and atherosclerosis (68, 69).

CCL2 also promotes local inflammation and macrophage infiltration in the chronically injured liver (37). Very recently, Li et al. demonstrated that a chemical inhibitor of CCR2, the receptor of CCL2, inhibited HCC development by reducing monocytes/macrophage infiltration and M2-macrophage polarization as well as CD8+ T cell activation in the murine xenograft model (70). However, whether targeting the CCL2-CCR2 axis by immunotherapy could be an effective therapeutic approach against chronic inflammation and HCC has not been addressed. In the present study, we addressed this important question using a novel preclinical model, miR-122 knockout mice that develop chronic inflammation driven HCC

(29). Our results clearly showed that targeting CCL2-CCR2 signaling by CCL2 immunotherapy could be an alternative approach in suppressing hepatitis and HCC development. Furthermore, our studies revealed that CCL2 neutralizing antibody (nab) immunotherapy in miR-122 KO mouse model involves suppression of CD11bhighGr1+ inflammatory myeloid cells recruitment and enhancement of NK cell cytotoxicity. These observations underscore the importance of targeting CCL2-CCR2 axis as a potential therapy in a subset of human HCC patients with chronic hepatic inflammation and high CCL2.

2.3. Materials and Methods

Treatment of miR-122 KO mice with CCL2 antibody and CCR2 inhibitor miR-122 KO mice were generated as described (29). Animals were housed in Helicobacter- free facility under a 12/12 hour light/dark cycle. Animals were handled following the guidelines of the Ohio State University Institutional Laboratory Animal Care Committee.

CCL2 neutralizing antibody (anti-mouse CCL2; clone C1142) was generously provided by

Janssen (71). miR-122 KO mice were injected intraperitoneally (ip) with CCL2 nab

(2mg/kg). The control group was injected with vehicle (PBS). To study the function of

CCL2 in hepatitis, 4-month-old KO mice were treating with CCL2 nab twice a week for 4 weeks. To study the role of CCL2 in HCC development, 12-month-old KO mice were injected with CCL2 nab twice a week for 8 weeks.

CCR2 inhibitor (Tocris Bioscience, cat# 2089) was given to KO mice in the drinking water daily (~10mg/kg) for 4 weeks. The amount of CCR2 inhibitor administered to each mouse was estimated based on the daily consumption of water in each cages and mouse body weight. The control group for CCR2 inhibitor was fed with regular water.

Flow cytometric analysis

Murine mononuclear cells from livers were isolated as previously described (29, 72).

Erythrocytes were lysed using RBC lysis buffer (BioLegend, San Diego, CA). Cells isolated from livers were treated with Fc Block antibody (anti CD16/32, BD Biosciences). Cells were stained with mouse-specific immune cell surface markers for 30 min at 4°C. The following anti-mouse antibodies were used at a dilution of 1:200: CD3-APC, CD3-PerCP-

Cy5.5, NK1.1-APC, NK1.1-PE, CD19-FITC, CD69-FITC, CD27-PE-Cy7, CD11b-PE,

CD11b-PerCP-Cy5.5, Gr1 V450 and IFN-γ-FITC (Biolegend, San Diego, CA). For staining of IFN-γ, cells were treated with Cytofix/Cytoperm (BD) following initial cell-surface staining and then performed intracellular staining.

NK cytotoxicity assay

Tumor-derived NK cells were enriched by NK cell isolation kit II (Miltenyi Biotec, San

Diego, CA) and then sorted using a FACSAria II cell sorter (BD Biosciences, San Diego,

CA). Hepa1-6, a mouse hepatoma cell line was generously provided by Dr. Gretchen

Darlington (Baylor College of Medicine). Although we did not authenticate these cells, they exhibited characteristics and gene expression profile of mouse hepatma cells. Hepa cells were labeled with Chromium-51 (Cr-51) and then pre-incubated with C1142 antibody or isotype antibody at a concentration of 10µg/ml for 1h. Then a standard 4-hour Cr-51 release assay (73) was performed to access cytotoxicity of mouse NK cells against Hepa cells.

Enzyme-linked immunosorbent assay (ELISA)

For alpha-fetoprotein (AFP), mouse serum was collected by mandibular punch (before treatment) or cardiac punch (after treatment). KO mouse older than 10 months old would be monitored for their body weight and serum AFP level by every two weeks. AFP level was quantified by DRG® AFP (Alpha Fetoprotein) kit (DRG, cat# EIA-1468).

For IFN-γ, mouse liver mononuclear cells were isolated as previously described (29, 72).

Hepa cells cultured in DMEM containing 10% FBS, were pre-incubated with C1142 (CCL2 nab) at a concentration of 10µg/ml for 1h. Then, 106 mononuclear cells were incubated with the same number of Hepa cells per well in a 96-well V-bottom plate at 37oC for 24h. Culture supernatants were collected for IFN-γ estimation using mouse IFN-γ ELISA Ready-SET-Go kit (eBioscience, cat# 88-7314-88).

Magnetic resonance imaging (MRI)

MRI of liver tumors in miR-122 KO mice was performed as described (74).

Serological, histological and immuno-histochemical (IHC) analysis

Serum was isolated from mice by cardiac puncture after CO2 asphyxiation and cervical dislocation and stored at -80°C. Biochemical analysis of serum enzymes was performed using VetAce (Alfa Wassermann system) as described (29). Macroscopic tumors were counted, dissected and weighed along with benign liver tissues. A fraction of tissues was fixed in 4% paraformaldehyde as described (29) and the rest was snap-frozen for RNA and protein analysis.

For histology, paraffin embedded tissue sections (4µm) on glass slides were stained with

H&E or IHC analysis with Ki-67, AFP, and Glypecan3 (GPC3) specific antibodies as described (29). Inflammation score was generated through blinded evaluation of H&E stained slides (x 100 magnification) as described previously (29). IHC analysis was done as described (29) and quantified by ImageJ (imagej.nih.gov/ij/). Microscopic images were taken by phase contrast microscope (BX41TF, OLYMPUS) and camera (DP71,

OLYMPUS). CellSens Standard 1.13 (OLYMPUS) software was used to capture images.

Antibody information is provided in the supplement.

Reverse transcription - quantitative polymerase chain reaction (RT-qPCR)

Total RNA was extracted from liver tissue using TRIzol (Life Technologies, cat#15596018) followed by DNase I treatment. DNase-treated RNA was reverse-transcribed into complementary DNA (cDNA) using High-capacity cDNA reverse transcription kit (Applied

Biosystem, cat# 4368813). RT-qPCR analysis of each sample performed in triplicate was performed using SYBR Green chemistry. Gene expression was normalized to Gapdh.

Relative expression was calculated by ΔΔCT method. Primer sequences are provided in the

Table 2.1.

Immunoblot analysis

Proteins were extracted from whole liver and tumor tissues by SDS lysis buffer followed by immunoblotting with primary antibodies (29). Signals were developed using Pierce ECL reagent (ThermoFisher, cat# 32106) and quantified by ImageJ (imagej.nih.gov/ij/). Antibody information is provided in the Table 2.2.

Statistical Analysis

All bar diagrams are presented as mean ± standard deviation. Two sample t-tests or ANOVA were used for analysis for comparison of two or more groups. For the GEO data analysis,

CCL2 expression from the quantile-normalized data was compared among tumor tissue types with ANOVA. Holm’s procedure was used to adjust for multiple comparisons (Target alpha level (0.05)/ (n-rank number of pair (by degree of significance) +1)). Fisher's exact test was used to test the difference of tumor incidence and tumor size between the CCL2 nab treated and vehicle treated groups. P-values <0.05 were considered significant and represented as asterisks.

2.4. Results

2.4.1. CCL2 is upregulated in primary human HCC

To verify whether CCL2 plays a role in human HCC development, we first examined its expression across independent microarray datasets (HCC vs. normal liver) downloaded from

Oncomine™ (75). To validate CCL2 expression, we chose The Cancer Genome Atlas

(TCGA), Mas liver, and Guichard liver that contain large patient cohorts (75-77). CCL2

DNA copy number or mRNA expression showed a significant increase in tumor tissues compared to the adjacent benign liver across four datasets (Table 2.3) (75-77). To further

24 evaluate CCL2 expression and disease prognosis, we analyzed HCC RNA microarray data

(n=115, mainly HCV positive) from GEO database (GSE14323) (77). The results showed that CCL2 RNA expression was significantly elevated in HCCs, cirrhotic livers, and cirrhotic HCCs compared to normal livers (Figure 2.1). These data imply that CCL2 might be critical for progression to cirrhosis and HCC. Thus, blocking the CCL2-CCR2 axis seems to be a reasonable therapeutic approach for HCC or even early stages of HCC development such as hepatitis and cirrhosis.

2.4.2. CCL2-CCR2 blockade reduces chronic inflammation and liver injuries in miR-

122 KO mice

CCL2 is a major chemokine that is known to cause various inflammatory diseases in humans (68, 78). Similarly, upregulation of CCL2 in miR-122 knockout (KO) liver correlated with chronic inflammation (29). To determine whether CCL2 is a key player in hepatitis, and blocking CCL2 could inhibit liver inflammation, CCL2 neutralizing antibody

(nab) was administered to 4-month-old miR-122 KO mice by intraperitoneal injection for 4 weeks (Figure 2.2A). CCL2 nab (C1142) displayed certain specificity toward murine CCL2

(71). In addition, CCL2 nab was well tolerated in mice (71). Furthermore, its anti-tumor efficacy has been demonstrated in murine breast cancer metastasis (79), lung (80), brain

(81), and prostate cancer (82, 83) models and no adverse effects were reported.

To evaluate the effects of CCL2 nab in suppressing hepatitis, we treated 4-month-old miR-

122 KO mice with CCL2 nab followed by histopathological and serological analyses. As reported in other mouse models (71, 79-83), CCL2 nab is well tolerated in miR-122 KO mice. There were no obvious changes in the in appearance, activity or body weight (Figure

2.3). Liver histology showed reduced immune cell infiltration after treating with CCL2 nab

(Figure 2.2B). Similarly, CD45 immunohistochemical analysis revealed that CCL2 nab treatment decreased leukocyte accumulation in the liver (Figure 2.2C). Notably, a significant reduction in serum ALT and AST levels corroborated that the liver damage due to hepatitis in KO mice could be reversed by inhibition of CCL2 (Figure 2.2D). Alkaline phosphatase (ALP), a marker of biliary dysfunction, was also reduced by ~50%; however, it was not statistically significant (Figure 2.2D). This result implies that blocking CCL2 may not be able to fully rescue hepatobiliary damage because ALP is a target of miR-122 which is de-repressed in miR-122 depleted livers (29, 30). Serum gamma-glutamyl transferase

(GGT) and bilirubin (total and direct) levels were not significantly altered in CCL2 nab injected mice (Table 2.4).

We have shown that activation of the CCL2-CCR2 axis in miR-122 KO liver correlates with the recruitment of CD11bhighGr1+ cells and hepatic inflammation and injuries (29).

Additionally, CCR2 inhibitors showed anti-inflammatory effects in the animal models with different diseases such as diabetic nephropathy (84), kidney hypertension (85), steatohepatitis (86), renal atrophy (87) models and no adverse effects were reported. We decided to test whether treatment of mice with a CCR2-specific inhibitor could reduce liver inflammation in KO mice. Indeed, KO mice fed water containing a CCR2 inhibitor exhibited reduced hepatic inflammation, as revealed by the dramatic reduction in bridging inflammation (Figure 2.4). Collectively, these data showed that blocking the CCL2-CCR2 axis could effectively inhibit hepatitis in miR-122 KO mice.

2.4.3. CCL2 nab therapy inhibits hepatitis by reducing CD11bhighGr1+ inflammatory

myeloid cells accumulation

It is well established that immune cells play a critical role in liver inflammation (88). Our previous study showed that CD11bhighGr1+ inflammatory myeloid cells excessively accumulated in miR-122 depleted livers (29). Additionally, CD11bhighGr1+ cells, which express CCR2 (29) are known to be recruited to the liver via CCL2-CCR2 interaction (89).

Therefore, we quantitated CD11bhighGr1+ cells in KO livers depleted of CCL2 or CCR2.

Flow cytometric analysis of the immune cells isolated from livers and spleen showed reduced CD11bhighGr1+ cell population in the KO mice treated with CCL2 nab without significant change in their numbers in peripheral blood (Figure 2.5A, 2.5B and Figure 2.6).

Consistent with other studies (37, 70), we also found decreased number of hepatic macrophages in liver by IHC upon blocking the CCL2-CCR2 axis (Figure 2.7). However, no significant alterations in the number of hepatic B, T and Natural killer (NK) cells were observed (Figure 2.8). Additionally, Immune cells isolated from CCR2 inhibitor treated KO mice showed reduced hepatic CD11bhighGr1+ cells (Figure 2.9), further supports the idea that blocking the CCL2-CCR2 axis reduces CD11bhighGr1+ inflammatory myeloid cells in liver.

IL-6 and TNF-α are two pro-inflammatory cytokines that drive hepatitis and HCC in the liver (90). Previously, we have shown that CD11bhighGr1+ inflammatory myeloid cells produce IL-6 and TNF-α in the miR-122 KO liver (29). Therefore, it is expected that hepatic

IL-6 and TNF-α levels would be suppressed if the population of CD11bhighGr1+ cells is decreased in the CCL2 nab treated mice. Indeed, the expressions of both cytokines were

27 reduced in the livers of CCL2 nab treated group (Figure 2.5C). In addition, we also observed a decrease in Il-1β expression upon CCL2 nab treatment (Figure 2.5C). Notably,

CCL2 nab treatment did not altered the expressions of other inflammatory cytokines or chemokines such as Ccl5, Ccl8, Cxcl9, Il-10, Ccl2 or Ccr2 (Figure 2.5C). Taken together, these results suggest that inhibition of the CCL2-CCR2 axis reduces liver inflammation by suppressing CD11bhighGr1+ inflammatory myeloid cells, which leads suppression of the two major pro-inflammatory cytokines, IL-6 and TNF-α.

2.4.4. Attenuation of hepatocarcinogenesis in miR-122 KO mice treated with CCL2

nab

As aforementioned, CCL2 is highly expressed in the tumor tissues of HCC patients (Figure

2.1, Table 2.3). Besides, blocking CCL2 suppresses chronic liver inflammation at early stages of HCC development in the KO mice (Figure 2.2). Therefore, we speculated that impeding CCL2 function would suppress development of spontaneous liver tumors in KO mice. To test this, male KO mice (~12 months old) bearing liver tumors confirmed by the serum alpha-fetoprotein (AFP) levels, were randomly assigned to two treatment groups.

Comparison of the serum AFP level with the tumor size in a large number of KO mice revealed that mice with serum AFP levels >12 IU/ml usually developed tumors (Figure

2.10). This ELISA data was further confirmed by the magnetic resonance imaging (MRI)

(Figure 2.11A). KO mice were injected ip with CCL2 nab (2mg/Kg) or vehicle twice a week for 8 weeks (Figure 2.11A) and analyzed the tumor burden one week after the last injection. We chose male KO mice for this study as these mice, like men, have higher HCC penetrance and tumor burden (29, 30).

Significant phenotypic differences were observed in the liver between the vehicle- and the

CCL2 nab- treated groups after 8 weeks of CCL2 immunotherapy; however, there were no obvious changes in the appearance, activity or body weight between the two groups (Figure

2.10B). While the majority of the PBS-injected mice developed large macroscopic liver tumors, CCL2 nab treated mice mostly developed smaller and fewer tumors (Figure 2.11B,

Table 2.5). Consistent with the previous data (Figure 2.5A), we found hepatic

CD11bhighGr1+ cell population was reduced in the CCL2 nab treated tumor-bearing mice compared to vehicle-treated group (Figure 2.11C). In addition, serum AFP and ALT levels were lower in the CCL2 nab treated mice (Table 2.5). Furthermore, the majority of the

HCCs in the vehicle-treated group were AFP and Glypican 3 (GPC3) positive, and grade 2-3

(intermediate to poorly differentiated) HCCs whereas those in CCL2 inactivated mice were mostly grade 1 (well differentiated) HCCs, AFP and GPC3 negative (Figure 2.11D, 2.11E).

These results suggest that CCL2 blockade is not only anti-inflammatory but also anti- tumorigenic in the liver.

2.4.5. Oncogenic signaling downstream of IL-6 and TNF-α is blocked in KO tumors

upon CCL2 nab therapy

Because CCL2 nab treatment suppressed the level of IL-6 and TNF-α in liver (Figure 2.5C), we hypothesized that the downstream oncogenic signaling pathways of IL-6 and TNF-α namely STAT3 and NF-κB, respectively, would be inhibited in the CCL2 depleted mice.

Indeed, immunoblot analyses showed that phospho-STAT3 (Y705) level was dramatically reduced while total-STAT3 level decreased only slightly in the tumors of CCL2 nab treated mice (Figure 2.12A, 2.12C). Reduced nuclear STAT3 levels in the CCL2 nab treated mice

29 confirmed that CCL2 inhibition suppresses STAT3 functions in liver tumors (Figure 2.12B,

2.12C). Similarly, the level of two major NF-κB subunits, p65, and p50, was reduced in both whole liver lysates and nuclear extracts (Figure 2.12B, 2.12C). The level of c-MYC, a downstream target of STAT3, was also reduced in the CCL2 nab treated mice (Figure

2.12A, 2.12C). Histologically, p65 and c-MYC both showed reduced expression in the transformed hepatocytes while pSTAT3 was decreased in immune cells as well as tumor cells (Figure 2.13). c-MYC is a well-known oncogene that promotes HCC proliferation (91, 92). Therefore, Ki-

67 expression, a marker of cell proliferation, was evaluated in tumor cells by immunohistochemistry. As expected, the number of Ki-67 positive cells were much less in both CCL2 depleted tumor and non-tumor tissues (Figure 2.12D, 2.14). PCNA, another cellular proliferation marker, was also reduced in the CCL2 nab-treated tumor extracts

(Figure 2.12E). Of note, we found cell apoptosis is increased in the CCL2 nab treated group as demonstrated by slight increase in cleaved PARP and cleaved Caspase-7 levels (Figure

2.15). These data suggest that inhibition of HCC growth by targeting CCL2 is, at least in part, through the inhibition of STAT3 and NF-κB signaling and c-MYC expression.

2.4.6. CCL2 nab activates nature killer (NK) cells in the tumor microenvironment to

suppress liver cancer

Cancer immunotherapy has been widely conducted in the clinic over the past several years

(93-95). Since antibody-based immunotherapy is applied in this study, and NK cell is known to be activated by the Fc domain of antibodies (94), we investigated whether enhanced NK cytotoxicity could be a mechanism of tumor suppression in CCL2 nab-treated mice.

Although the total number of NK cells did not change significantly after CCL2 nab immunotherapy (Figure 2.8C, 2.8F), the activity of NK cells increased, as demonstrated by the upregulation of hepatic Ifn-γ expression (Figure 2.16A). To verify the source of the elevated IFN-γ, primary hepatic immune cells isolated from tumor-bearing KO mice were co-cultured with mouse hepatoma (Hepa1-6) cells in the presence of CCL2 nab. The culture supernatants were subjected to IFN-γ ELISA and cells were analyzed by flow cytometry.

Both ELISA and intracellular staining data showed that only NK cells (Figure 2.16B,

2.16C) were the major producers of IFN-γ. Additionally, the cell surface expression of

CD69, a marker for activated NK cells, increased when NK cells were co-cultured with

CCL2 nab treated mouse hepatoma (Hepa) cells (Figure 2.16D), suggesting that NK cells could be activated upon exposure of hepatoma cells to CCL2 nab. To confirm the activated

NK cells could suppress cancer cell growth, NK cell cytotoxicity was assessed by

Chromium-51 (Cr-51) release assay (73, 96). To this end, Hepa cells were first labeled with

Cr-51 followed by incubation with CCL2 nab and subsequently, with tumor-derived NK cells. Enhanced release of Cr-51 from hepa cells is indicative of higher cytotoxicity of NK cells. Indeed, tumor-derived NK cells co-cultured with Hepa cells and CCL2 nab exhibited higher cytotoxicity (Figure 2.16E). This result also suggested that the increased cytotoxicity of NK cells could be triggered independent of other immune cells. Collectively, these results demonstrate that CCL2 nab impedes tumor growth, at least in part, by activating NK cells.

2.5. Discussion

CCL2 is a chemotactic factor to tumor cells and inflammatory macrophage/ monocytes.

Blocking the CCL2-CCR2 axis by CCR2 inhibitor decreases the intrahepatic macrophage infiltration in a diet induced steatohepatitis model (86). CCL2 Spiegelmer, a CCL2 RNA antagonist, suppresses the macrophage infiltration in the carbon tetrachloride (CCl4) and methionine–choline-deficient (MCD) diet-induced liver injury models (37). Recently, Li et al. also demonstrated the importance of CCL2-CCR2 axis in different HCC models (13). In line with the animal models mentioned above, it is clear that CCL2 is crucial for regulating the inflammatory milieu in liver. In the present study, we aimed at elucidating the efficacy of CCL2 immunotherapy against chronic hepatitis and HCC in a murine model. This is a logical extension of our previous study that demonstrated a causal role of CCL2 in the inflamed liver that progressively leads to HCC in miR-122 KO mice (29). Analysis of GEO,

Oncomine™ and Protein Atlas databases showed increased CCL2 expression in patients suffering from HCC (75, 77, 97). These data suggest CCL2 may play a role in liver cancer.

To further clarify the role of the CCL2-CCR2 axis in chronic liver inflammation and tumor development, we used CCL2 nab to block CCL2 signaling. Our data suggest that CCL2 nab inhibits tumor development in miR-122 KO mice through at least two distinct mechanisms: suppressing IL-6 and TNF-α production by reducing the infiltration of CD11bhighGr1+ inflammatory myeloid cells to the liver and enhancing NK cell cytotoxicity (Figure 2.17).

An elegant study demonstrating the correlation of CCL2 protein level with poor survival of

HCC patients was published (70) during the preparation of our manuscript. Li et al. showed that a clinically relevant CCR2 antagonist inhibited murine HCC progression by reducing

M2-type tumor-associated macrophage population and increasing CD8+ cytotoxic T cells in orthotopic and subcutaneous HCC models (70). Their study supports the notion that blocking the CCL2-CCR2 axis has profound effects in modulating liver immune system to suppress HCC progression. While Li et al. found macrophage and CD8+ T cells are regulated by a CCR2 inhibitor in xenograft models, we found CD11bhighGr1+ inflammatory myeloid cells and NK cells could be modulated by CCL2 nab in miR-122 KO mouse livers.

Thus, these studies (ours and Li et al.) demonstrated that targeting CCL2-CCR2 axis by blocking the receptor or the ligand could be effective HCC therapy. In future, it would be interesting to investigate whether combination of both would be more effective.

Furthermore, we used our unique animal model to study the role of CCL2 in liver inflammation and HCC progression in the light of our previous work (29). We demonstrated two different types of immune cells including CD11bhighGr1+ inflammatory myeloid cells and NK cells play key roles in regulating liver inflammation and tumor development after blocking CCL2 function. In addition, CD11bhighGr1+ inflammatory myeloid cells promote tumorigenesis by increasing tumor cell proliferation via IL-6 and TNF-α. Both IL-6 and

TNF-α are well-known pro-inflammatory cytokines that activate oncogenic transcription factor STAT3 to promote HCC development (90, 98). Consistent with the previous findings, our present data reveal that CCL2 nab reduces accumulation of CD11bhighGr1+ cells, which decreases IL-6 and TNF-α levels in the tumor and surrounding liver tissues, and reduces the activation of STAT3 and the expression of c-MYC. Ours and Li et al. data demonstrated that targeting the CCL2-CCR2 axis by blocking the receptor or the ligand could be an effective

HCC therapy. In future, it would be interesting to investigate whether combination of both would be a more effective therapeutic approach.

NK cells are key modulators of liver diseases. Activated NK cells suppress liver disease progression from chronic hepatitis to HCC (99). Based on ours and Li et al. data (70), CCL2 is expressed in liver tumors of both mouse and human origin. The CCL2 nab binds to Fc receptor (CD16) on NK cells (94), and the CCL2 secreted from tumor cells binds to the

CCL2 nab. This can trigger NK cell activation (i.e., enhancement of IFN-γ secretion and cytotoxicity), causing enhanced killing of liver tumor cells. In this study, we observed enhanced cytotoxic capability of NK cells against CCL2 nab-treated hepatoma cells. We also noticed that the level of IFN-γ was higher in the NK cells isolated from livers of tumor- bearing KO mice treated with CCL2 nab, reinforcing the notion that NK cells were activated by the CCL2 nab. Consistent with other literatures that cytotoxic NK cells activate Caspase

3, 7 and PARP (100, 101), we found increased cleaved-Caspase-7 and cleaved-PARP in the

CCL2 nab treated mice (Figure 2.15). Inhibition of STAT3 activity has been shown to sensitize HCC cells to NK cell-mediated cytotoxicity (102). Interestingly, our data show that

STAT3 activity is reduced in the CCL2 nab treated mice tumors (Figure 2.12A-2.12C), suggesting a reciprocal regulation.

Novel therapeutic strategies for HCC are urgently needed since existing therapies could not cure or prolong patients’ survival even by 6 months (65). In this study, we highlight the importance of CCL2 in contributing to chronic liver inflammation and tumorigenesis.

Furthermore, the efficacy of a CCL2 specific neutralizing antibody in the miR-122 mouse model underscores the feasibility of immune-based therapy in inflammatory liver disease

34 and cancer. Our study shows that CCL2 nab immunotherapy elicits two distinct mechanisms that act in concert to inhibit hepatitis and HCC development, and provides rationales for future clinical trials in HCC patients with chronic inflammation.

Figure 2.1 CCL2 is overexpressed in human cirrhotic livers, as well as cirrhotic and noncirrhotic HCCs.

CCL2 RNA expression retrieved from GEO data (www.ncbi.nlm.nih.gov/geo/) with accession_GSE14323 (77) was compared among the four types of liver tissues and statistical analysis was done using ANOVA. Sample size: normal group: n=19; Cirrhosis: n=41;

Cirrhosis HCC: n=17; HCC: n=38.

Figure 2.2 CCL2 neutralizing antibody therapy reduces chronic liver inflammation and liver damage in adult miR-122 KO mice.

(A) The schematic presentation of the CCL2 nab treatment. (B) Liver histology of 4-month- old male KO mice injected IP with CCL2 nab (2mg/kg, n=5) or PBS (vehicle, n=5).

Representative images of H&E stained liver sections (top panel scale bar: 100µm, lower panel: 25µm). Right panel is the quantitation of the inflammation scores generated through blinded evaluation of H&E stained liver sections (x100 magnification, 100µm). The inflammatory area was quantified by imageJ of 3 randomly chosen fields (x 100 magnification, 100um) per animal (n=5). (C) Representative images of CD45 stained liver sections (top panel scale bar: 100µm, lower panel: 25µm). Right panel represents the CD45+ areas quantified by imageJ of 3 randomly chosen fields (x 100 magnification, 100µm) per animal (n=5). (D) Analysis of serum ALT, AST and AFP levels (n=5).

Continued

Figure 2.2 Continued

Figure 2.3 Mice treated with CCL2 neutralizing antibody (nab) for 4 weeks did not show significant changes in body weight.

CCL2 nab or vehicle was given to 4-month-old miR-122 knockout (KO) mice for 4 weeks.

Neither group showed significant changes in body weight during the treatment (n=5).

Figure 2.4 CCR2 inhibitor therapy reduces chronic liver inflammation in miR-122 knockout (KO) mouse.

CCR2 inhibitor (RS 102895) was given to 4-month-old KO mice in the drinking water daily

(10mg/ kg) for 4 weeks. Representative images of H&E stained sections of mouse liver tissues (top panel scale bar: 100µm, lower panel: 25µm). The inflammation score on the right side of this panel was generated through blinded evaluation of histopathology (x100 magnification). The inflamed areas were quantified by imageJ software of 3 randomly chosen fields (x 100 magnification, 100um) per animal (n=5).

Figure 2.5 CCL2 nab treatment decreases liver CD11bhighGr1+ cell population and IL-6

& TNF-α expression in adult KO mice.

(A) Injection of 4 months old KO mice CCL2 nab reduced the hepatic CD11bhighGr1+ population. Right panel is the quantitation of hepatic CD11bhighGr1+ inflammatory myeloid cells (n=5). (B) Reduced CD11bhighGr1+ population in spleen. Right panel is the quantitation of splenic CD11bhighGr1+ inflammatory myeloid cells (n=5). (C) RT-qPCR analysis of inflammatory cytokines and chemokines.

Continued

Figure 2.5 Continued

Figure 2.6 CCL2 nab therapy did not reduces CD11b+ Gr1+ inflammatory cells in the blood.

Flow cytometric analysis of CD11b+ Gr-1+ inflammatory myeloid cells in the vehicle and

CCL2 nab treated mice (n=5).

Figure 2.7 Blocking the CCL2-CCR2 axis reduces hepatic macrophages.

Immunohistochemistry for hepatic macrophages (F4/80) in the vehicle and CCL2 nab treated group (n=5) (scale bar: top 100µm, bottom: 25µm).

Figure 2.8 CCL2 nab does not alter the population of B cell, T cell, or NK cell in liver.

(A) Flow cytometric analysis of CD19+ B cells in the vehicle and CCL2 nab treated mice

(n=5). (B) Flow cytometric analysis of CD3+ T cells in the vehicle and CCL2 nab treated group (n=5). (C) Flow cytometric analysis of NK1.1+ NK cells in the vehicle and CCL2 nab treated group (n=5). (D) Immunohistochemistry for B cell (B220), (E) T cell (CD3), (F) NK cell (NKp46) in the vehicle and CCL2 nab treated group (n=5) (scale bar: top 100µm, bottom: 25µm).

Continued

Figure 2.8 Continued

Continued

Figure 2.8 Continued

Continued

Figure 2.8 Continued

Figure 2.9 CCR2 inhibitor reduces CD11bhighGr1+ inflammatory myeloid cells in the

KO mouse liver.

4-month-old KO mice were treated with CCR2 inhibitor provided in drinking water

(10mg/kg) for 4 weeks. Quantification was demonstrated on the right side of this panel

(n=5).

Figure 2.10 12-month-old male KO mice were assigned to vehicle and CCL2 nab treated group based on serum AFP levels and body weight.

(A) Serum AFP correlated with HCC tumor size (n=16) (103). Serum was collected by mandibular bleeding from 10-month-old KO mice in every other week. Mice with serum

AFP levels >12 IU/ml at 12 months were randomly divided into 2 groups and used for

CCL2 nab and PBS (vehicle) treatment experiments. (B) There were no significant changes in body weights between the two groups during CCL2 nab treatment (control (n=9); CCL2 nab group (n=11)).

Figure 2.11 Blocking CCL2 function suppresses HCC development in KO mice.

(A) The schematic depiction of the immunotherapy protocol and the representative magnetic resonance images (MRI) of liver tumors (axial view, T1 weighted) for 12-month-old miR-

122 KO mice. (B) Five representative images of the livers harvested from the vehicle (n=9) and CCL2 nab treated mice (n=11). (C) Flow cytometric analysis of hepatic CD11bhighGr1+ immune cells in the vehicle and CCL2 nab treated mice. (D) Representative images of liver histology, AFP and GPC3 immunohistochemical images of the vehicle and (E) nab treated groups (scale bar: top: 250µm, middle: 100µm, bottom: 25µm).

Continued

Figure 2.11 Continued

Continued

Figure 2.11 Continued

Figure 2.12 CCL2 nab therapy suppresses HCC development in KO mice by inhibiting tumor cell proliferation.

(A) Immunoblotting of the key downstream proteins of IL-6 and TNF-α in KO tumors. (B)

Immunoblotting of NK-κB subunit and STAT3 in the tumor nuclear extracts. (C) Respective protein levels were normalized to those of GAPDH or KU86. Quantitation was done by

ImageJ. (D) Representative images of Ki-67 staining (scale bar: 100µm, inset: 25µm).

Quantitation was done by ImageJ of 3 randomly chosen fields (x 200 magnification, 50µm) per animal (n=5). (E) Immunoblotting of PCNA in liver tumors.

Continued

Figure 2.12 Continued

Figure 2.13 CCL2 immunotherapy reduces tumoral p-STAT3, c-MYC, and p65.

Immunohistochemistry for (A) pSTAT3, (B) c-MYC, (C) p65 in the vehicle and CCL2 nab treated group (n=5) (scale bar: top 100µm, bottom: 25µm).

Continued

Figure 2.13 Continued

Figure 2.14 CCL2 immunotherapy reduces the number of proliferative cells in the adjacent benign tumors.

Histological and immunohistochemical staining in the non-tumor area of vehicle and CCL2 nab treated group (n=5) (scale bar: left panel: 100µm, right panel: 25µm).

Figure 2.15 Blocking CCL2 by neutralizing antibody increases tumor cell apoptosis.

(A) Immunoblotting of Cleaved-Casepase 7(c-Cas 7) and Cleaved-PARP (c-PARP) in the vehicle and CCL2 nab treated mice. (B) Cleaved-PARP and cleaved-Cas 7 protein levels were normalized to GAPDH. Quantification was done by ImageJ (n=4 in each group).

Figure 2.16 CCL2 nab activates NK cells in KO mouse livers.

(A) RT-qPCR analysis of hepatic Ifn-γ in CCL2 nab treated mice (n=5 in each group). (B)

IFN-γ quantification by ELISA in the supernatants of cultured Hepa and hepatic immune cells (n=3 in each group). (C) Intracellular flow cytometric analysis of IFN-γ in the co- cultured immune cells (n=3 in each group). (D) Flow cytometric analysis of the activated

NK cell surface marker, CD69 (n=3 in each group). (E) Cytotoxicity of hepatic NK cells towards CCL2 nab treated Hepa cells (n=3 in each group).

Continued

Figure 2.16 Continued

Figure 2.17 CCL2 nab therapy inhibits HCC development in KO mice by modulating tumor microenvironment.

Loss of miR-122 induces CCL2 expression in liver, which, in turn, recruits CCR2 positive

CD11bhighGr1+ inflammatory myeloid cells. These inflammatory myeloid cells activate IL-

6-Stat3-c-Myc axis and TNF-α-NF-κB axis, respectively, to promote liver inflammation and tumorigenesis in a later stage. Blocking CCL2 activity reduces CD11bhighGr1+ inflammatory myeloid cells recruitment and downstream signaling pathways to inhibit hepatitis and hepatocarcinogenesis. Depletion of CCL2 by CCL2 nab also increases NK cell activation in the tumor microenvironment and enhances their cytotoxicity towards tumor cells. Thus,

CCL2 nab exhibits both anti-inflammatory and anti-tumorigenic functions.

Table 2.1 Primers used for RT-qPCR analysis

Gapdh-RT-F 5'-TCCTGCACCACCAACTGCTTAG-3'

Gapdh-RT-R 5'-TGCTTCACCACCTTCTTGATGTC-3'

IL6-RT-F 5′-AGACTTCACAGAGGATACCACTCCC-3′

IL6-RT-R 5′-TCTCATTTCCACGATTTCCCAG-3′

TNF-a-RT-F 5′-ACCGTCAGCCGATTTGCTATC-3′

TNF-a-RT-R 5′-TCAGAGTAAAGGGGTCAGAGTGGG-3′

IFNg-RT-F 5'- TAGCCAAGACTGTGATTGCGG-3'

IFNg-RT-R 5'- AGACATCTCCTCCCATCAGCAG-3'

Ccl5-RT-F 5’-GTGCCAACCCAGAGAAGAAGTG-3’

Ccl5-RT-R 5'- ATGCCCATTTTCCCAGGACC -3'

Ccl8-RT-F 5'- TGCTTCTTTGCCTGCTGCTC -3'

Ccl8-RT-R 5'- ATACCCTGCTTGGTCTGGAAAAC -3'

IL-1b-RT-F 5’- CCTGTCCTGTGTAATGAAAGACGG -3’

IL-1b-RT-R 5’- TGCTCTGCTTGTGAGGTGCTGATG -3’

Table 2.2 Antibody information

Immunohistochemistry

CD45 (Santa Cruz, cat# SC-1121)

AFP (Santa Cruz, cat#sc-15375)

GPC3 (Santa Cruz cat# SC-11395)

c-MYC (Santa Cruz, cat# SC-764)

Ki-67 (Abcam, cat# ab15580)

pSTAT3 Y705 (Cell Signaling Technology, cat# CS-9131)

Immunoblotting

p-STAT3 Y705 (Cell Signaling Technology, cat# CS-9131)

t-STAT3 (Santa Cruz, cat# SC-482)

c-MYC (Cell Signaling Technology, cat# CS-5605)

p65 (Santa Cruz, cat# SC-109X)

p50 (Santa Cruz, cat# SC-114X)

PCNA (eBioscience, cat# 14-6748-81)

GAPDH (EMD Millipore, cat# MAB374)

KU86 (Santa Cruz, cat# SC-1484)

Table 2.3 CCL2 expression across 4 independent microarrays in HCC patients

Cancer vs. Normal Array type P-value Fold Change Sample Size

Mas Liver RNA 0.012 3.324 n=115

Guichard Liver cohort 2 DNA 0.036 2.025 n=52

TCGA Liver DNA 0.003 2.063 n=212

Guichard Liver cohort 1 DNA 0.026 2.153 n=185

CCL2 expression in human clinical samples (Cancer VS. Normal). Data provided by

Oncomine (www.oncomine.com, January 2016, Thermo Fisher Scientific, Ann Arbor, MI).

TCGA database (TCGA Research Network: http://cancergenome.nih.gov) contain >200

HCC specimens of different etiology (HBV, HCV, alcohol, and NAFLD). Mas liver, is a

RNA array which contains >100 HCC samples (mostly HCV positive) (77). The Guichard liver DNA array consisting of 237 HCCs of various risk factors including HBV, HCV, alcohol, and NAFLD, was analyzed by exosome sequencing and SNP array (76).

Table 2.4 Serological analysis of 4 months old KO mice

Vehicle CCL2 nab p-Value

GGT (U/L) 3.8± 2.2 3± 2.1 0.55

DBILI (mg/dL) 0.3± 0.1 0.24± 0.2 0.517

TBILI (mg/dL) 0.35± 0.1 0.3± 0.1 0.514

Serum was collected from mice by cardiac puncture. Gamma glutamyl transferase (GGT), direct bilirubin (DBILI), and total bilirubin (TBILI) levels were analyzed. Statistical analysis was done by student’s 2-tailed t test.

Table 2.5 Histopathological and serological analysis of tumor-bearing miR-122 KO mice treated with CCL2 nab for 8 weeks

Vehicle Ccl2 nab P-value

HCC incidence 8/9 5/11 0.07

Tumor number 2.67± 1.22 0.73± 0.79 0.0004

Sum of tumor diameter (cm) 1.71± 1.24 0.57± 0.65 0.01

Serum ALT (U/L) 151.89± 68.06 73.57± 28.5 0.027

Serum AFP (IU/ml) 14.07± 3.51 9.95± 0.61 0.001

All measurements are presented as mean ± SD. P-values were calculated using student’s 2- tailed t-test (for tumor number, sum of tumor diameter, tumor weight, ALT, and AFP) or

Fisher's exact test (for HCC incidence).

Chapter 3. miR-122 is a negative regulator of liver fibrosis

3.1. Abstract microRNA 122 (miR-122), a small non-coding RNA that regulates liver homeostasis, has been identified as a liver injury marker in viral infected, alcohol induced, and non-alcoholic steatohepatitis (NASH) induced liver fibrosis. However, why low expression of miR-122 is associated with liver fibrosis and whether loss of miR-122 in hepatocytes would affect hepatic stellate cells (HSC), the main extracellular matrix (ECM) protein producer during liver fibrosis, are still not clear. To address these questions, a series of bioinformatic and molecular experiments were performed in the human liver derived HSC cell line (LX-2) and mice liver fibrosis models. Co-cultured miR-122 expressed HCC cells and LX-2 using transwell plates demonstrated that extracellular miR-122 could be delivered from hepatocytes to HSCs to suppress HSC proliferation and expression of fibrotic markers. In addition, analyzing published datasets of cirrhosis patients from GEO and RNA-seq of wildtype (WT) and miR-122 knockout (KO) mice revealed BCL2 could be an important pro-fibrotic factor. BCL2 was found to be overexpressed in the HBV- and HCV-infected cirrhotic livers in human patients. In addition, BCL2 is consistently upregulated in the miR-

122 KO livers as well as CCl4-induced fibrotic livers. Venetoclax, an FDA approved selective BCL2 inhibitor recently approved for chronic lymphocytic leukemia (CLL), showed promising effects in suppressing LX-2 proliferation and the expression of fibrotic

77 genes. Treating miR-122 KO mice with Venetoclax for two weeks significantly reduced collagen levels and expression of fibrotic genes in the liver. Moreover, re-analyzing the data generated by high-throughput sequencing of RNAs isolated by crosslinking and Argonaute immunoprecipitation (Ago-HITS-CLIP) identified several fibrotic genes that are directly associated with miR-122 and dysregulated in the miR-122 KO livers. Collectively, our data not only revealed the regulatory mechanisms and networks of miR-122 and its downstream molecules in suppressing liver fibrosis but also demonstrated the therapeutic potential of

Venetoclax in the treatment of liver fibrosis.

3.2. Introduction

MicroRNAs are a small non-coding RNA regulates gene expression (1, 2). microRNA-122

(miR-122), a liver specific and abundant microRNA among vertebrates, has been shown to play a critical role in lipid metabolism (21, 22, 29), cell differentiation (104), hepatic polyploidization (25), and HCV replication (39). miR-122 was shown to be associated with various liver diseases as well. Lower expression of miR-122 was consistently found in HCV negative liver cancers and correlated with metastasis in hepatocellular carcinoma (HCC) patients (34, 105, 106). The role of miR-122 in mediating liver pathology is probably best demonstrated by the genetic knockout mouse models. Mouse liver lacking of miR-122 develops spontaneous hepatitis, steatosis, fibrosis, and eventually HCC (29, 30), suggesting the importance and complexity of miR-122 in regulating liver homeostasis.

Liver fibrosis is the process attributed to excessive accumulation of extracellular matrix

(ECM) proteins due to repeated wound-healing responses, which occurs in most types of chronic liver diseases (42, 107). It is widely accepted that the hepatic stellate cells (HSCs)

78 are the main collagen-producing cells during liver fibrosis (42, 108, 109). When the liver is injured, damaged hepatocytes release TGFβ1 and ROS while Kupffer cells release TNF-α and TGFβ1 to activate quiescent HSCs into myofibroblast-like cells (42, 110-113). In addition to HSCs, immune cells in the liver also contribute to fibrosis by secreting IL-6 and

TNF-α to the microenvironments upon liver injury (114). If the damages persist for a long period of time, the space of dead hepatic cells will be filled up with ECM, which increase the stiffness of the liver and leads to cirrhosis (45, 115).

One such a pro-fibrotic gene that promotes HSC activation is B cell lymphoma 2 (BCL2).

BCL2 was the first discovered as an apoptotic regulator localized to the outer membrane of mitochondria (116-118). The role of BCL-2 is to promote cell survival by inhibiting the action of pro-apoptotic proteins such as BAX and BAK (119). BCL2 and other pro-survival

BCL2 family members (e.g. BCL-XL and MCL-1) inhibit BAX- and BAK-induced permeabilization of mitochondria membrane and release of cytochrome C and ROS (120).

Not only does BCL2 control cell survival, but also function as an oncogene in different kinds of cancer (118). Overexpression of BCL2 in the chronic lymphocytic leukemia (CLL) and acute lymphocytic leukemia (ALL) cancer cells allow these cancer cells to be resistant to chemotherapy (121). Similar to cancer cells, overexpression of BCL2 in HSC increases the resistance to apoptotic stimulus such as Fas activation and exposure to TNF-α (122).

Mouse liver overexpressing BCL2 resulted in cholestatic liver fibrosis (123), indicating

BCL2 is a pro-fibrogenic factor in liver.

In this study, we aimed at understanding the relationship between miR-122 and liver fibrosis. We found extracellular miR-122 could be delivered from hepatocytes to HSCs,

79 which inhibits HSC proliferation and fibrotic gene expression. In addition, we identified a pro-fibrotic gene that is dysregulated in both human and mouse fibrotic livers. Bioinformatic analysis using published datasets deposited in GEO revealed BCL2 is overexpressed in both

HBV- and HCV- induced cirrhotic livers in human patients. Consistent with the expression profile in human, BCL2 was also found to be upregulated in the miR-122 KO livers and in

CCl4-iuduced liver fibrosis model in wild type (C57bl/6) mice. Furthermore, Venetoclax, an

FDA approved BCL2 inhibitor for CLL, suppressed HSC proliferation and the expression of fibrotic genes in LX-2 cells. Venetoclax also reduced the collagen level and expression of fibrotic genes in the mice liver depleted of miR-122, suggesting targeting BCL2 could be an effective approach in inhibiting liver fibrosis. In the end, we re-analyzed our recent published data generated through Ago-HITS-CLIP (124) and found a list of fibrotic genes that are directly associated with miR-122 and dysregulated in the mice liver depleted of miR-122, providing a global view of miR-122 mediated suppression of liver fibrosis.

3.3. Method

Dataset analysis

All the dataset analysis was done using R software (version 3.4.0). miR-122 expression data in cirrhotic patients (risk factors: HBV and HCV) were queried from ArrayExpress

(https://www.ebi.ac.uk/arrayexpress/) (Accession number # E-TABM-866) (125, 126) hosted by The European Bioinformatics Institute (EMBL-EBI). Expression profiles of miR-

122 KO mouse were downloaded from GEO (https://www.ncbi.nlm.nih.gov/geo/) using

GEOquery package from Bioconductor (https://bioconductor.org/). GSE20610 (microarray of WT and miR-122 LKO mice) (29) and GSE97060 (RNA-seq of WT and miR-122 KO

80 mice) (124) were used to identify potential downstream fibrotic genes regulated by miR-

122. BCL2 expression data was downloaded from GEO. BCL2 mRNA expression was evaluated from the studies of GSE84044 (HBV-induced cirrhosis) (127) and GSE14323

(HCV-induced cirrhosis) (77). Identification of miR-122 direct targets involved in liver fibrosis was performed using published dataset from High-throughput sequencing of RNAs isolated by crosslinking and Argonaute immunoprecipitation (Ago-HITS-CLIP) of wild type and miR-122 KO livers (GSE97061) (124).

Mouse treatment miR-122 knockout (KO) mouse was generated as described (29). All animals were housed in temperature-controlled and Helicobacter-free facility under a 12/12-hour light/dark cycle.

Animals studies were conducted under the guidelines of the Ohio State University

Institutional Laboratory Animal Care Committee.

Carbon tetrachloride (CCl4) (Sigma-Aldrich, # 289116) was dissolved in corn oil and administered to both wild type C57bl/6 mouse and miR-122 KO mice by i.p. injection. WT and miR-122 KO mice (~4 months old, 25 g) were treated with vehicle (corn oil) or CCl4

(0.25ml/ Kg) twice a week for 4 weeks. Mice were harvested one day after the last injection.

Venetoclax (ABT-199) was purchased from PharmaBlock (USA) (cat# PBLJ6036).

Venetoclax was dissolved in Propylene Glycan (Sigma-Aldrich # P4347), PEG400 (Sigma-

Aldrich # 91893), ethanol (Decon Labs # 2705HC) following the ratio of 60:30:10 (v/v/v), respectively. Venetoclax (100mg/kg) was administered to miR-122 KO mice that develop developed hepatic fibrosis ~8-month of age (29) by oral gavage for 14 consecutive days.

Proliferation assay

LX-2 cells were a kind gift from Dr. Scott Friedman (Icahn School of Medicine). Cell proliferation was evaluated by CellTiter-Glo® Luminescent Cell Viability Assay kit

(Promega, #G7572) following manufacturer’s protocol.

Histological analysis

Mice livers were fixed in 4% paraformaldehyde for 24 hours then switched to PBS before embedding into paraffin. Paraffin embedded tissue sections (4mm) on glass slides were used for H&E staining, Sirius Red staining or immunohistochemistry (IHC) as previously described (128).

Co-culture of HCC and LX-2 cells

HCC cell line HCCLM3 and PLC/PRF5 was engineered to overexpress miR-122 using lenti-virus as described (129). Co-cultures were performed in the transwell plates (6.5 mm,

3µm pore polycarbonate membrane insert, Costar# 3415). Briefly, HCC cells (105 cells) were seeded at the top transwells while LX-2 cells (2 x 105 cells) were seeded at the bottom wells. Cells were cultured in the DMEM supplemented with 10% exosome free FBS (SBI #

EXO-FBS-50A-1). After 72 hours of incubation, culture supernatants were collected and exosomes were purified using ExoQuick TC (SBI # EXOTC10A-1) kit following manufacture’s guidelines. Exosomes and LX-2 cells were used for RNA isolation as described below.

Reverse-transcription polymerase chain reaction (RT-qPCR)

Total RNA was isolated by using TRIzol (Life Technologies, # 15596018) and de- contaminated of DNA with DNase I. DNA-free RNA was reverse-transcribed into complementary DNA (cDNA) by High-capacity cDNA reverse transcription kit (Applied

Biosystem, # 4368813). RT-qPCR analysis of each sample was performed in triplicate using

SYBR Green chemistry. Relative expression was calculated by ΔΔCT method with ACTB as the normalizer. Primer sequences are provided in the Table 3.1. For detecting mature miR-122, Taqman assay was used (Applied Biosystems, #TM: 002245). Mature miR-122 expression was calculated by ΔΔCT method using RNU6B (Applied Biosystems, #TM:

001093) as the normalizer.

Immunoblot analysis

Proteins were extracted from whole cells or liver tissues in RIPA buffer followed by immunoblotting with primary antibodies as described (128). X-ray films were developed using Pierce ECL reagent (ThermoFisher, cat# 32106). Near-infrared (NIR) fluorescence signals were developed by Odyssey® CLx Imaging System (LICOR). Primary antibody information is provided in Table 3.2.

Target validation using luciferase assay

CTGF 5’UTR harbors miR-122 seed region (CACTCCA) was cloned into the 5’ end of luciferase (hRluc) of psiCHECK-2 plasmid using Nhe I restriction enzyme. The mutant

CTGF 5’UTR was created by mutating the miR-122 seed region with the complementary sequences (CACTCCA GTGAGGT). Luciferase was performed in the Hepa 1-6 cells transfected with 25nM of scram RNA (negative control) or miR-122 mimic RNA (miR-122) using Lipofectamine 3000 (Thermo Fisher, #L3000015). Luciferase activity was measured using Dual-Luciferase® Reporter Assay System (Promega # E1960) following manufacture’s protocol.

Statistical analysis

Statistical analysis for datasets was performed with R (3.4.0). Expression counts of each sample per gene were log2 transformed. Significance was determined using Wilcox.test() function in R. All Bar diagrams in this study were shown as mean ± standard deviation. Two sample t-tests was used for two group comparisons. For more than two group comparisons,

ANOVA was performed. P-values <0.05 was considered significant and represented with asterisks.

3.4. Results

3.4.1. miR-122 is a negative regulator of liver fibrosis

Studies from different groups including ours have reported that mouse depleted of miR-122 developed spontaneous liver fibrosis (29), suggesting miR-122 is an anti-fibrotic microRNA in the liver. To confirm that this association was common between mouse and human, the expression profiles of miR-122 in the normal and HCC adjacent benign cirrhosis tissue was evaluated by accessing the microRNA microarray data (E-TABM-866) downloaded from

The European Bioinformatics Institute (EMBL-EBI) (126). miR-122 expression was evaluated in total 111 samples (21 normal liver; 90 cirrhosis). Significant downregulation of miR-122 was found in the cirrhosis patient group (Figure 3.1A). This data along with the phenotype of miR-122 KO mice (29) suggest that miR-122 might serve as an anti-fibrotic molecule. To confirm the anti-fibrotic function of miR-122, LX-2 cells (immortalized human hepatic stellate cell line) (130) were transfected with either scrambled negative control RNA or miR-122 mimic. After 72 hours, miR-122 transfected LX-2 cells showed reduced proliferation compared to the control group (Figure 3.1B). This data correlated with reduced expression of fibrotic genes in the miR-122 transfected groups. miR-122 suppressed

84 the expression of two well-established fibrosis genes (COL1A1 and ACTA2) (Figure 3.1C).

In addition to the in vitro data, mouse lacking hepatic miR-122 was found to be very sensitive to fibrosis stimulus. miR-122 KO mouse treated with low dose of CCl4 (0.25 ml/Kg) showed drastic liver fibrosis compare to the control group (Figure 3.1D, 3.1E), indicating the importance of miR-122 in suppressing liver fibrosis.

3.4.2. Extracellular miR-122 suppresses expression of fibrotic genes in HSCs

Hepatocyte-generated miR-122 has been found in circulation carried by extracellular vehicles e.g. exosomes (131, 132). In addition, it has been reported that miR-122 could be transferred from miR-122-abundant cells to low expressing cells (133). To explore how hepatocyte-made miR-122 could regulate hepatic stellate cell (HSC), we co-cultured the

HCC cell lines (PLC/PRF5 and HCC-LM3) overexpressing miR-122 with LX-2 using a transwell system. (Figure 3.2A). LX-2 co-cultured with miR-122 overexpressed-HCC cells

(miR-122-HCC) showed increased expression of mature miR-122 but not the primary transcripts (pri-miR-122 not detectable, Table 3.3), suggesting that the increased miR-122 levels in HSCs are exogenous. The primer used for measuring pri-miR-122 were used in other studies (34, 134) and confirmed by the UCSC In-Silico PCR program

(https://genome.ucsc.edu/cgi-bin/hgPcr). The raw Ct values were listed in Table 3.3. These data suggested that miR-122 could be delivered from hepatocytes to HSCs. Furthermore, the expressions of COL1A1 and ACTA2 mRNAs were much lower in LX-2 cells co-cultured with miR-122-HCC cells (Figure 3.2B), indicating the increased expression of miR-122 in

HSCs is functional in suppressing HSC activation. Moreover, we purified exosomes from co-cultured medium and found exosomal miR-122 level was much higher in the miR-122

85 expressed cells compared to the control cells (Figure 3.2C). This result suggested that the increased level of miR-122 in HSC might be, at least in part, delivered through exosomes.

3.4.3. BCL2 is overexpressed in miR-122 KO mice and human cirrhotic livers

To understand the molecular mechanisms of miR-122 mediated suppression of liver fibrosis, we re-analyzed the microarray and RNA-seq data in the wild type (WT) and miR-122 KO mice (29, 124). We specifically focus on pro-fibrotic genes that were found to be upregulated in the KO mice livers. BCL2, a pro-fibrotic gene shown in several studies (122,

123, 135) was significantly increased in the miR-122 KO mice liver compared to WT mice at both RNA and protein levels (Figure 3.3A-3.3D). Consistent with miR-122 KO mice,

BCL2 level was enhanced in another well-established liver fibrosis model induced by CCl4 treatment in the C57Bl/6 mice (Figure 3.4A, 3.4B). Besides, BCL2 expression was found to be upregulated in both HBV- and HCV-infected livers of cirrhotic patients (Figure 3.5A,

3.5B) (77, 127). Collectively, data from mice fibrosis models and human cirrhotic livers suggests upregulated BCL2 might play an important role in promoting liver fibrosis.

3.4.4. Inhibition of BCL2 activity using a small molecule, Venetoclax, suppresses liver

fibrosis developed in the miR-122 KO mice

In order to study the role of BCL2 in contributing to liver fibrosis, an FDA approved BCL2 selective inhibitor (Venetoclax, ABT-199) originally developed for treatment of chronic lymphocytic leukemia (CLL) patients with 17p deletion (136), was used in our study. The efficacy of Venetoclax was first tested in the LX-2 cells. We found Venetoclax inhibited

50% of cell proliferation at ~10 µM concentration in LX-2 cells (Figure 3.6A). In addition,

COL1A1 and ACTA2 expressions were decreased upon Venetoclax exposure (Figure

3.6B). Given that BCL2 is upregulated in both miR-122 KO mice and CCl4-induced fibrosis models, and Venetoclax reduces fibrotic gene expression in the LX-2 cells, it is very likely that Venetoclax might also suppress liver fibrosis in vivo. We, therefore, treated miR-122

KO mice (~32 weeks old) with Venetoclax (100mg/kg) or vehicle for two weeks. The expressions of the fibrotic gene (ACTA2) was down regulated in the Venetoclax treated group compared to the control group (Figure 3.6C). Sirius red staining showed reduced collagen level in the Venetoclax treated group (Figure 3.6D). Collectively, both in vitro and in vivo data with Venetoclax suggests that BCL2 is a critical gene for promoting liver fibrosis. Blocking BCL2 activity using BCL2 selective inhibitor (Venetoclax) is an effective therapeutic approach in suppressing liver fibrosis.

3.4.5. Ago-HITS-CLIP identified a miR-122 targetome involved in liver fibrosis

In addition to RNA-seq and microarray analysis which led us to find BCL2, we also used an unbiased and transcriptome-wide method to identify direct interactions between miR-122 and fibrotic targets. High-throughput sequencing of RNAs isolated by crosslinking and

Argonaute immunoprecipitation (Ago-HITS-CLIP) is known for its ability to biochemically determine the precise sequences of the miRNA-mRNA interactions in vivo (39, 137). We revisited the miR-122 target list identified by Ago-HITS-CLIP performed in our earlier study (Figure 3.7A) (124). Several pro-fibrotic transcripts, e.g. Ctgf, Tgfb1 and Pdgfb were found to harbor the miR-122 dependent peaks in the wild type but not in miR-122 KO livers, indicating miR-122 binds to these mRNAs (Figure 3.7B, Table 3.4).

Among these, we found CTGF harbors one miR-122 binding site (5’-CACTCCA-3’) at its

5’ UTR. This is a rare binding event as most of the miRNA and mRNA interactions occur at

87 the 3’UTRs or coding DNA sequence (CDS) of the targets (124). Due to this unique binding and the well-characterized pro-fibrotic feature of CTGF, we decided to investigate the relationship between CTGF and miR-122. We found CTGF expression was upregulated in the 5 weeks old miR-122 KO mouse liver by analyzing microarray and RNA-seq data

(Figure 3.8A). CTGF protein level was found to be consistently upregulated in the miR-122

KO liver compared to the WT liver (Figure 3.8B), suggesting miR-122 could regulate

CTGF expression. This idea was further supported by the higher CTGF level in miR-122-/- primary hepatocytes (Fig. 3.8C). An inverse correlation between miR-122 and CTGF was found by transfecting miR-122 mimic in the KO hepatocytes or anti-miR-122 in the WT hepatocytes (Figure 3.8C- 3.8E). Finally, we validated miR-122 binding at the 5’UTR of

CTGF mRNA by the luciferase reporter assay (Figure 3.8F). A luciferase reporter containing mouse Ctgf 5’UTR that harbors a miR-122 seed region (CACTCCA) showed reduced luciferase activity while no effects was found in the mutant reporter that harbors a mutant miR-122 seed region (GTGAGGT) in the presence of miR-122. These results confirmed that miR-122 regulates CTGF expression through direct binding.

3.5. Discussion

In our earlier study, we showed that mice depleted of miR-122 in livers developed spontaneous liver fibrosis (29). Although few studies have demonstrated that de-repressed miR-122 downstream targets (e.g. KLF6, P4HA1) could promote liver fibrosis (30, 31), the mechanisms of how hepatocyte-produced miR-122 could regulate HSC and whether these miR-122 downstream molecules could serve as a therapeutic target for liver fibrosis are still not clear. In the present study, we aimed at understanding the mechanistic role of miR-122

88 in blocking fibrosis and identifying potential therapeutic targets for liver fibrosis. We analyzed published datasets from cirrhosis patients and the data from various mouse models to demonstrate the anti-fibrotic functions of miR-122. To elucidate the direct connections between miR-122 and HSC, we used a transwell system to co-culture miR-122-expressed

HCC cells with hepatic stellate cells (LX-2). Similar to other microRNAs that could be transferred between hepatocytes and HSCs (138), our data showed miR-122 could be delivered from HCC cells to HSC through extracellular mechanisms. This notion is further supported by the enrichment of miR-122 in exosomes purified from the co-cultured media of miR-122 expressed HCC cells and LX-2 cells. The increased level of mature miR-122 in

HSCs, but not the primary miR-122 transcripts, suggests the source of miR-122 is exogenous (from HCC cells). Furthermore, we showed the increased miR-122 level inversely correlated with the reduced expression of fibrosis genes, suggesting the increased level of miR-122 is functional. Consistent with other study (31), our data showed miR-122 could be delivered from hepatic cells to HSCs to inhibit cell proliferation and expression of fibrosis genes.

To look for the downstream molecules that are responsible for the liver fibrosis developed in miR-122 KO mice, we analyzed the microarray and RNA-seq data in the WT and miR-122

KO mice. We found a pro-fibrotic gene (i.e. BCL2) is highly expressed in the KO livers in both analyses (122, 123, 135). The increased level of BCL2 is further confirmed by the western blotting of the WT and KO liver lysates. Consistent with the KO mice data, BCL2 level was significantly elevated in another fibrosis model induced by CCl4. In addition to the data from mice fibrosis models, BCL2 was found to be highly expressed in both HBV-

89 and HCV-infected cirrhotic liver. Collectively, our data strongly suggest BCL2 is a critical factor for liver fibrosis. Besides liver fibrosis, BCL2 is known to play a critical role in leukemia (118, 139, 140). The anti-apoptotic functions of BCL2 help tumor cells overcome the apoptosis stimulus such as radiation. Many studies, including solid tumors, have been exploring the possibility to inhibit tumor growth by targeting BCL2 (118, 119). The first generation of BCL2 inhibitor: Navitoclax (ABT-263), had profound effects in suppressing leukemia cell growth (141). However, navitoclax also targets BCL-XL, another BCL2 family member. These off-target effects caused thrombocytopenia in patients with relapsed or refractory chronic lymphocytic leukemia (CLL) (142, 143). The second generation BCL2 selective inhibitor: Venetoclax (ABT-199) did not have this issue due to its high specificity toward BCL2 (144, 145). Moreover, Venetoclax is approved by FDA for therapy of CLL patients with a chromosomal deletion at 17p. Thus, it is reasonable to use Venetoclax as a tool to verify the mechanistic role and test therapeutic value of targeting miR-122 - BCL2 axis for liver fibrosis. In the in vitro system, we found Venetoclax suppressed cell proliferation and fibrogenic gene expression in the human liver derived HSC cell line (LX-

2). Venetoclax also suppressed liver fibrosis developed in the miR-122 KO mice demonstrated by the reduced expression of fibrosis markers and reduced collagen stained by

Sirius red. These data along with the dysregulation of BCL2 in both cirrhosis patients and mouse fibrosis models demonstrated its therapeutic potential for treating liver fibrosis patients.

In addition to BCL2, we also identified several pro-fibrotic targets (e.g. Ctgf, Pdgfrβ,

Tgfβr1…etc.) that are significantly increased in KO livers and directly interacted with miR-

122 using Ago-HITS-LCIP (Table 3.4). To validate the results of Ago-HITS-CLIP analysis, we further evaluated one of the targets and confirmed Ctgf as a direct target of miR-122 using luciferase assay. We chose Ctgf because of its non-canonical binding with miR-122 at the 5’UTR and its well-established role in contributing liver fibrosis. With the target list generated by Ago-HITS-CLIP, we are able to explain why loss of miR-122 in liver would have such a huge impact in promoting liver fibrosis. In future, we would like to explore some of the targets listed in Table 3.4 for their functions in mediating liver fibrosis.

Meanwhile, we would like to continuously explore the regulatory mechanisms between miR-122 and BCL2, as Ago-HITS-CLIP did not identify the direct interactions between these two.

Drug re-purposing or use of existing drugs for new diseases, has gained its popularity due to the low cost and time-saving benefits in clinics. In the current study, we applied an FDA approved drug originally designed for CLL patients to treat liver fibrosis. We selected BCL2 as a therapeutic target based on the expression profiles in two mice models and in human cirrhotic livers from published datasets deposited at GEO and EMBL-EBI. Our analyses clearly demonstrated that BCL2 are upregulated in both human and mouse fibrotic livers.

Inhibition of BCL2 functions using Venetoclax suppressed human HSC derived cells (LX-2) from transforming into pro-fibrogenic status, and inhibited liver fibrosis developed in the miR-122 KO mice. With the mechanistic and phenotypic data generated in the current study, it is possible that blocking BCL2 using Venetoclax might be able to serve as a therapeutic approach for fibrosis patients with high BCL2 expression.

Figure 3.1 The anti-fibrotic role of miR-122 in both mouse and human livers.

(A) miR-122 expression in normal liver (n=21) and non-tumor cirrhotic liver tissues (n=90).

(B) Proliferation of LX-2 cells transfected with miR-122 or scrambled RNA was measured

72hr post-transfection using CellTiter-GLO assay. (C) Expression of fibrotic genes. (D)

Hydroxyproline (Hy:Pro) level in the CCl4 treated WT and miR-122 KO mouse liver. (E)

Histology of the CCl4 treated WT and miR-122 KO mouse liver (x100 magnification,

100µm).

Continued

Figure 3.1 Continued

Figure 3.2 Extracellular miR-122 could be delivered from hepatocytes to HSC

(A) A scheme of co-culture experiment between miR-122 overexpressed HCC cells and LX-

2. (B) RT-qPCR data in the LX-2 cells co-cultured with Vector transfected PLC/PRF5 (Vec-

PLC), miR-122 expressed PLC/PRF5 (122-PLC), Vector transfected HCCLM3 (Vec-LM3), and miR-122 expressed HCC-LM3 (122-LM3) (n=3 in each group). Raw data is provided in

Table 3.3. (C) RT-qPCR data for the exosomal miR-122 purified from co-cultured media

(showed in Figure 3.4A) (n=3 in each group).

Figure 3.2 Continued

Figure 3.3 BCL2 is overexpressed in miR-122 KO livers.

(A) Microarray analysis of BCL2 in wild type (WT, n=5) and miR-122 KO mouse (n=5).

(B) Microarray analysis of BCL2 in wild type (WT, n=5) and miR-122 KO mouse (n=4).

(C) BCL2 mRNA expression measured by qRT-PCR in WT (n=6) and miR-122 KO mouse liver (n=6). (D) BCL2 protein level in WT (n=3) and miR-122 KO mouse liver (n=3). (E)

Quantification of BCL2 levels from Figure 3.3D.

Figure 3.3 Continued

Figure 3.4 Enhanced expression of BCL2 in the CCl4-induced fibrotic livers.

(A) Histology of corn oil (vehicle) and CCl4 in corn oil (0.25ml/kg) treated mice livers

(n=3). (B) Immunoblot analysis of BCL2 in the vehicle (corn oil) and CCl4 injected mice

(n=3).

Continued 100

Figure 3.4 Continued

101

Figure 3.5 BCL2 is overexpressed in the HBV- and HCV-induced cirrhotic livers.

(A) BCL2 expression in HBV-induced cirrhotic liver tissues. Normal group: both inflammation and fibrosis score = 0 (n=26); Cirrhosis group: fibrosis score >=1 (n=81)

(Data retrieved from GSE84044, X represents median in the boxplot) (B) BCL2 expression in normal liver (n=19) and HCV-induced cirrhotic liver tissues (n=41) (Data retrieved from

GSE14323, X represents median in the boxplot).

102

Figure 3.6 Venetoclax inhibits the expression of fibrotic genes both in vitro and in vivo.

(A) Venetoclax suppresses LX-2 growth in a dose-dependent manner (n=3). (B) Fibrosis genes (ACTA2, COL1A1) expression was measured by RT-qPCR in the LX-2 cells (n=3).

122 KO mice (~32 weeks old) treated with vehicle or Venetoclax (n=5). (D)

Histopathological analysis of the liver section treated with vehicle or Venetoclax (n=5)

(scale bar: 250 µm).

Continued

103

Figure 3.6 Continued

Continued

104

Figure 3.6 Continued

105

Figure 3.7 Argonaute cross-linking immune precipitation (Ago-CLIP) identified fibrotic targetome of miR-122 in mouse liver.

(A) A scheme of Ago-CLIP. (B) Examples of miR-122 dependent binding peaks in the miR-

122 WT and KO mouse (n=4).

Continued

106

Figure 3.7 Continued

107

Figure 3.8 CTGF is a 5’UTR-specific miR-122 target.

(A) Microarray (n=5) and RNA-seq (n=4) of Ctgf expression in the WT and KO mice livers

(5 weeks old). (B) CTGF protein level in the WT and LKO mouse liver (n=4). (C) CTGF protein level in the hepatocytes of WT and LKO mice. (D) CTGF protein level negatively correlated with miR-122 in the transfected mouse hepatocytes. (E) Reciprocal expression of

CTGF and miR-122 in the hepatocytes. (F) Reduced luciferase activity in the miR-122 and wild type CTGF 5’UTR (CACTCCA)co-transfected luciferase reporter but not in the CTGF

5’UTR mutant confirms the miR-122 binding at the CTGF 5’UTR. NC-S: Negative control for miR-122 sense; 122-S: miR-122 sense; NC-AS: Negative control for miR-122 antisense; 122-AS: miR-122 anti-sense.

Continued

108

Figure 3.8 Continued

Continued

109

Figure 3.8 Continued

110

Table 3.1 Primer used for RT-qPCR

ACTA2 RT-F 5'- ATGCTCCCAGGGCTGTTTTC -3

ACTA2 RT-R 5'- TCCCAGTTGGTGATGATGCC -3'

COL1A1 RT-F 5'- GCAAGAATGGAGATGATGGGG -3'

COL1A1 RT-R 5'- AAACCACTGAAGCCTCGGTGTC -3'

ACTB RT-F 5'- CTGGCACCACACCTTCTACAATG -3'

ACTB RT-R 5'- TAGCACAGCCTGGATAGCAACG -3'

CTGF RT-F 5'- CACCAGTGTGAAGACATACAGGGC -3'

CTGF RT-R 5'- TCGGGGCATTTGAACTCCAC -3' pri-miR122 RT-F 5′ GCTCTTCCCATTGCTCAAGATG 3' pri-miR122 RT-R 5′ GTATGTAACAACAGCATGTG 3′

111

Table 3.2 Antibody information

Immunoblotting

CTGF Santa Cruz Biotech, #SC-14939

BCL2 Santa Cruz Biotech, #SC-7382

GAPDH Santa Cruz Biotech, #SC-365062

Immunohistochemistry

α-SMA DAKO, #M085129-2

112

Table 3.3 Raw data of RT-qPCR done in the co-cultured LX-2 cell

Sample Gene Ct value Mean Ct Stdev

Vec-PLC++LX2 COL1A1 27.63 27.39 0.208

Vec-PLC++LX2 COL1A1 27.28 27.39 0.208

Vec-PLC++LX2 COL1A1 27.26 27.39 0.208

122-PLC++LX2 COL1A1 27.87 27.74 0.151

122-PLC++LX2 COL1A1 27.58 27.74 0.151

122-PLC++LX2 COL1A1 27.76 27.74 0.151

Vec-LM3++LX2 COL1A1 26.93 26.96 0.038

Vec-LM3++LX2 COL1A1 27 26.96 0.038

Vec-LM3++LX2 COL1A1 26.95 26.96 0.038

122-LM3++LX2 COL1A1 26.9 26.92 0.023

122-LM3++LX2 COL1A1 26.91 26.92 0.023

122-LM3++LX2 COL1A1 26.94 26.92 0.023

Vec-PLC++LX2 ACTA2 29.34 29.18 0.292

Vec-PLC++LX2 ACTA2 29.35 29.18 0.292

Vec-PLC++LX2 ACTA2 28.84 29.18 0.292

122-PLC++LX2 ACTA2 29.93 29.45 0.484

122-PLC++LX2 ACTA2 28.96 29.45 0.484

122-PLC++LX2 ACTA2 29.45 29.45 0.484 Continued

113

Vec-LM3++LX2 ACTA2 29.24 29.13 0.106

Vec-LM3++LX2 ACTA2 29.03 29.13 0.106

Vec-LM3++LX2 ACTA2 29.13 29.13 0.106

122-LM3++LX2 ACTA2 29.34 29.04 0.29

122-LM3++LX2 ACTA2 29.02 29.04 0.29

122-LM3++LX2 ACTA2 28.76 29.04 0.29

Vec-PLC++LX2 ACTB 22.61 22.48 0.13

Vec-PLC++LX2 ACTB 22.35 22.48 0.13

Vec-PLC++LX2 ACTB 22.48 22.48 0.13

122-PLC++LX2 ACTB 22.02 21.79 0.31

122-PLC++LX2 ACTB 21.44 21.79 0.31

122-PLC++LX2 ACTB 21.91 21.79 0.31

Vec-LM3++LX2 ACTB 21.91 21.85 0.194

Vec-LM3++LX2 ACTB 21.63 21.85 0.194

Vec-LM3++LX2 ACTB 22.01 21.85 0.194

122-LM3++LX2 ACTB 21.54 21.3 0.255

122-LM3++LX2 ACTB 21.03 21.3 0.255

122-LM3++LX2 ACTB 21.35 21.3 0.255

Vec-PLC++LX2 pri-miR122 0 0 N/A

122-PLC++LX2 pri-miR122 0 0 N/A Continued

114

122-PLC++LX2 pri-miR122 0 0 N/A

Vec-LM3++LX2 pri-miR122 0 0 N/A

122-LM3++LX2 pri-miR122 0 0 N/A

Vec-PLC++LX2: LX-2 co-cultured with vector transfected PLC/PRF5

122-PLC++LX2: LX-2 co-cultured with miR-122 transfected PLC/PRF5

Vec-LM3++LX2: LX-2 co-cultured with vector transfected HCCLM3

122-LM3++LX2: LX-2 co-cultured with miR-122 transfected HCCLM3

115

Table 3.4 Pro-fibrotic miR-122 targets identified by Ago-CLIP in miR-122 WT and

KO mouse.

Gene Annotation Seed Type p-value WT VS KO expression (log2)

Ctgf 5'UTR 7A1 0.016988 -1.518624491

Pdgfrb CDS 8mer 3.75E-21 -0.96498004

Col5a1 CDS 8mer 3.26E-20 -0.952192613

P4ha1 3'UTR 8mer 0.001085 -3.232382621

Mmp2 3'UTR 7A1 1.02E-24 -1.307276327

Klf6 3'UTR 7m8 5.68E-47 -1.190961957

Col4a1 3'UTR 6mer 7.08E-24 -1.102863702

Vcam1 3'UTR 7A1 4.68E-06 -0.979919528

Ngfr 3'UTR 7m8 7.06E-12 -0.954727937

Myh9 3'UTR 6mer 1.67E-16 -0.796204606

Tnfrsf1b 3'UTR 6mer 0.002705 -0.766674249

Tgfbr1 3'UTR 7A1 3.83E-06 -0.332025716

Tgfbr1 3'UTR 6mer 3.83E-06 -0.332025716

116

Chapter 4: Concluding remarks miR-122 is a liver specific microRNA that maintains liver homeostasis in many aspects such as lipid metabolism, cell differentiation, polyploidy and viral infections (25, 49). In the current study, we explored the role of miR-122 in suppressing liver inflammation, fibrosis, and HCC using combination of bioinformatic analysis, pathological evaluation and mouse genetics. Our study provides a comprehensive and systematic analysis in understanding the regulatory mechanisms of miR-122-mediated liver pathology. We answered few interesting questions that were unclear in the field: 1. Is de-repression of CCL2 a cause of liver inflammation in the miR-122 KO mouse? 2. Is it possible to inhibit liver inflammation, fibrosis, and even hepatocarcinogenesis caused by loss of miR-122 through targeting miR-

122 downstream targets? 3. What are the downstream molecules of miR-122 that promote liver fibrosis in the miR-122 KO mouse? 4. How does hepatocyte-produced miR-122 regulate HSC during fibrosis? Our data directly addressed the questions mentioned above with well-designed experiments conducted in both human and mouse models.

In chapter 2, we confirmed the role of miR-122 in regulating liver inflammation is, at least in part, through the CCL2-CCR2 axis. CCL2 is a both directly and indirectly regulated by miR-122 (29). Blocking the CCL2-CCR2 axis reduces the accumulation of pro- inflammatory and pro-tumorigenic CD11b+/Gr1+ cells in liver (146, 147). With reduced

CD11b+/Gr1+ cell population in the liver, the level of TNF-α and IL-6 are decreased

117 significantly in the CCL2 neutralizing antibody treated group. Therefore, their corresponding downstream pathways are inhibited in the liver as well, which leads to reduced tumorigenesis. In fact, it would be more conclusive if we could have shown

CD11b+/Gr1+ cells are indeed the main player in both inflammation and tumorigenesis by depleting CD11b+/Gr1+ cells in the CCL2 neutralizing antibody treated mouse. However, there are no appropriate approaches that could specifically target CD11b+/Gr1+ cells. Thus, it is beyond our capability to address this question at present. Even so, our data still clearly demonstrated the therapeutic potential of CCL2 neutralizing antibody in suppressing liver inflammation and tumor development.

The association between miR-122 and liver fibrosis has been suggested by various studies

(30, 31, 148, 149). However, only a few studies addressed the mechanistic role of miR-122 in liver fibrosis (30, 31). In chapter 3, the anti-fibrotic role of miR-122 is strongly suggested by a series of molecular and bioinformatic analysis. Using a co-culture system, we found miR-122 could be delivered from hepatocytes to HSCs extracellularly. To our knowledge, our study is the first report explaining the mechanism of how hepatocyte-produced miR-122 could regulate liver fibrosis and HSCs. The RT-qPCR data generated from purified exosomes in the co-culture media suggests the presence of miR-122 in the HSCs could be the results of exosomal delivery. In addition, we tried to suppress liver fibrosis by targeting the downstream molecules of miR-122 that are dysregulated in the miR-122 KO mice livers as well as cirrhotic patients. BCL2 was found to be upregulated in the HBV- and HCV- induced cirrhotic livers as well as two different mouse models that develop liver fibrosis.

We re-purposed a BCL2 selective inhibitor (Venetoclax) originally designed for CLL

118 patients for liver fibrosis in mice models. Venetoclax significantly reduces collagen levels in the miR-122 KO livers and fibrotic gene expressions in both human HSCs and miR-122 KO mice livers, suggesting the therapeutic potential of Venetoclax for the treatment of liver fibrosis patients.

In summary, our study highlights the importance of miR-122 in regulating liver homeostasis

(Figure 4.1). We linked miR-122 and miR-122 downstream genes to the pathological phenotype developed in miR-122 KO mice using CCL2-specific neutralizing antibody and a

BCL2 selective inhibitor (Venetoclax). We chose CCL2 and BCL2 as therapeutic targets not only based on the mouse studies but also taking into consideration their expression in human patients. Besides the molecular mechanisms, our study also demonstrated the therapeutic potentials that could possibly translate our research from the bench to the bedside. With an ultimate goal to develop effective therapy for HCC and HCC-associated liver pathology, this study demonstrated the potential of treating liver inflammation and HCC using CCL2 neutralizing antibody while suppressing liver fibrosis using Venetoclax.

119

Figure 4.1 Summary of normal and low miR-122 expressed liver.

Loss of miR-122 de-represses several pro-inflammatory and pro-fibrotic molecules in liver.

In one hand, the increased level of CCL2 causes excessive accumulation of CD11b+/Gr1+ inflammatory myeloid cells, which releases TNF-α and IL-6 to promote inflammation and tumorigenesis. On the other hand, the reduced level of exosomal miR-122 in HSC also leads to HSC activation due to the increased expression of pro-fibrotic genes such as BCL2 and

CTGF.

Continued

120

Figure 4.1 Continued

121

Reference

1. Ha M, Kim VN. Regulation of microRNA biogenesis. Nature reviews Molecular cell biology. 2014;15:509-24.

2. He L, Hannon GJ. MicroRNAs: small RNAs with a big role in gene regulation.

Nature reviews Genetics. 2004;5:522-31.

3. Friedman RC, Farh KK, Burge CB, Bartel DP. Most mammalian mRNAs are conserved targets of microRNAs. Genome research. 2009;19:92-105.

4. Ambros V. A hierarchy of regulatory genes controls a larva-to-adult developmental switch in C. elegans. Cell. 1989;57:49-57.

5. Chalfie M, Horvitz HR, Sulston JE. Mutations that lead to reiterations in the cell lineages of C. elegans. Cell. 1981;24:59-69.

6. Lee Y, Ahn C, Han J, Choi H, Kim J, Yim J, et al. The nuclear RNase III Drosha initiates microRNA processing. Nature. 2003;425:415-9.

7. Lund E, Guttinger S, Calado A, Dahlberg JE, Kutay U. Nuclear export of microRNA precursors. Science. 2004;303:95-8.

8. Winter J, Jung S, Keller S, Gregory RI, Diederichs S. Many roads to maturity: microRNA biogenesis pathways and their regulation. Nature cell biology. 2009;11:228-34.

9. Grishok A, Pasquinelli AE, Conte D, Li N, Parrish S, Ha I, et al. Genes and mechanisms related to RNA interference regulate expression of the small temporal RNAs 122 that control C. elegans developmental timing. Cell. 2001;106:23-34.

10. Hutvagner G, McLachlan J, Pasquinelli AE, Balint E, Tuschl T, Zamore PD. A cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA. Science. 2001;293:834-8.

11. Ketting RF, Fischer SE, Bernstein E, Sijen T, Hannon GJ, Plasterk RH. Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C. elegans. Genes & development. 2001;15:2654-9.

12. Hammond SM, Boettcher S, Caudy AA, Kobayashi R, Hannon GJ. Argonaute2, a link between genetic and biochemical analyses of RNAi. Science. 2001;293:1146-50.

13. Mourelatos Z, Dostie J, Paushkin S, Sharma A, Charroux B, Abel L, et al. miRNPs: a novel class of ribonucleoproteins containing numerous microRNAs. Genes & development.

2002;16:720-8.

14. Schwarz DS, Hutvagner G, Du T, Xu Z, Aronin N, Zamore PD. Asymmetry in the assembly of the RNAi enzyme complex. Cell. 2003;115:199-208.

15. Ameres SL, Zamore PD. Diversifying microRNA sequence and function. Nature reviews Molecular cell biology. 2013;14:475-88.

16. Krol J, Loedige I, Filipowicz W. The widespread regulation of microRNA biogenesis, function and decay. Nature reviews Genetics. 2010;11:597-610.

17. Chang J, Nicolas E, Marks D, Sander C, Lerro A, Buendia MA, et al. miR-122, a mammalian liver-specific microRNA, is processed from hcr mRNA and may downregulate the high affinity cationic amino acid transporter CAT-1. RNA biology. 2004;1:106-13.

18. Girard M, Jacquemin E, Munnich A, Lyonnet S, Henrion-Caude A. miR-122, a

123 paradigm for the role of microRNAs in the liver. Journal of hepatology. 2008;48:648-56.

19. Lagos-Quintana M, Rauhut R, Yalcin A, Meyer J, Lendeckel W, Tuschl T.

Identification of tissue-specific microRNAs from mouse. Current biology : CB.

2002;12:735-9.

20. Elmen J, Lindow M, Schutz S, Lawrence M, Petri A, Obad S, et al. LNA-mediated microRNA silencing in non-human primates. Nature. 2008;452:896-9.

21. Esau C, Davis S, Murray SF, Yu XX, Pandey SK, Pear M, et al. miR-122 regulation of lipid metabolism revealed by in vivo antisense targeting. Cell metabolism. 2006;3:87-98.

22. Krutzfeldt J, Rajewsky N, Braich R, Rajeev KG, Tuschl T, Manoharan M, et al.

Silencing of microRNAs in vivo with 'antagomirs'. Nature. 2005;438:685-9.

23. Laudadio I, Manfroid I, Achouri Y, Schmidt D, Wilson MD, Cordi S, et al. A feedback loop between the liver-enriched transcription factor network and miR-122 controls hepatocyte differentiation. Gastroenterology. 2012;142:119-29.

24. Duncan AW, Taylor MH, Hickey RD, Hanlon Newell AE, Lenzi ML, Olson SB, et al. The ploidy conveyor of mature hepatocytes as a source of genetic variation. Nature.

2010;467:707-10.

25. Hsu SH, Delgado ER, Otero PA, Teng KY, Kutay H, Meehan KM, et al. MicroRNA-

122 regulates polyploidization in the murine liver. Hepatology. 2016;64:599-615.

26. Shimakami T, Yamane D, Welsch C, Hensley L, Jangra RK, Lemon SM. Base pairing between hepatitis C virus RNA and microRNA 122 3' of its seed sequence is essential for genome stabilization and production of infectious virus. Journal of virology.

2012;86:7372-83.

124

27. Janssen HL, Reesink HW, Lawitz EJ, Zeuzem S, Rodriguez-Torres M, Patel K, et al.

Treatment of HCV infection by targeting microRNA. The New England journal of medicine.

2013;368:1685-94.

28. Ottosen S, Parsley TB, Yang L, Zeh K, van Doorn LJ, van der Veer E, et al. In vitro antiviral activity and preclinical and clinical resistance profile of miravirsen, a novel anti- hepatitis C virus therapeutic targeting the human factor miR-122. Antimicrobial agents and chemotherapy. 2015;59:599-608.

29. Hsu SH, Wang B, Kota J, Yu J, Costinean S, Kutay H, et al. Essential metabolic, anti- inflammatory, and anti-tumorigenic functions of miR-122 in liver. The Journal of clinical investigation. 2012;122:2871-83.

30. Tsai WC, Hsu SD, Hsu CS, Lai TC, Chen SJ, Shen R, et al. MicroRNA-122 plays a critical role in liver homeostasis and hepatocarcinogenesis. The Journal of clinical investigation. 2012;122:2884-97.

31. Li J, Ghazwani M, Zhang Y, Lu J, Li J, Fan J, et al. miR-122 regulates collagen production via targeting hepatic stellate cells and suppressing P4HA1 expression. Journal of hepatology. 2013;58:522-8.

32. Trebicka J, Anadol E, Elfimova N, Strack I, Roggendorf M, Viazov S, et al. Hepatic and serum levels of miR-122 after chronic HCV-induced fibrosis. Journal of hepatology.

2013;58:234-9.

33. Budhu A, Jia HL, Forgues M, Liu CG, Goldstein D, Lam A, et al. Identification of metastasis-related microRNAs in hepatocellular carcinoma. Hepatology. 2008;47:897-907.

34. Tsai WC, Hsu PW, Lai TC, Chau GY, Lin CW, Chen CM, et al. MicroRNA-122, a

125 tumor suppressor microRNA that regulates intrahepatic metastasis of hepatocellular carcinoma. Hepatology. 2009;49:1571-82.

35. Robinson MW, Harmon C, O'Farrelly C. Liver immunology and its role in inflammation and homeostasis. Cellular & molecular immunology. 2016;13:267-76.

36. Tall AR, Yvan-Charvet L. Cholesterol, inflammation and innate immunity. Nature reviews Immunology. 2015;15:104-16.

37. Baeck C, Wehr A, Karlmark KR, Heymann F, Vucur M, Gassler N, et al.

Pharmacological inhibition of the chemokine CCL2 (MCP-1) diminishes liver macrophage infiltration and steatohepatitis in chronic hepatic injury. Gut. 2012;61:416-26.

38. Arango Duque G, Descoteaux A. Macrophage cytokines: involvement in immunity and infectious diseases. Frontiers in immunology. 2014;5:491.

39. Luna JM, Scheel TK, Danino T, Shaw KS, Mele A, Fak JJ, et al. Hepatitis C virus

RNA functionally sequesters miR-122. Cell. 2015;160:1099-110.

40. Burdette D, Haskett A, Presser L, McRae S, Iqbal J, Waris G. Hepatitis C virus activates interleukin-1beta via caspase-1-inflammasome complex. The Journal of general virology. 2012;93:235-46.

41. Negash AA, Ramos HJ, Crochet N, Lau DT, Doehle B, Papic N, et al. IL-1beta production through the NLRP3 inflammasome by hepatic macrophages links hepatitis C virus infection with liver inflammation and disease. PLoS pathogens. 2013;9:e1003330.

42. Bataller R, Brenner DA. Liver fibrosis. The Journal of clinical investigation.

2005;115:209-18.

43. El-Serag HB, Rudolph KL. Hepatocellular carcinoma: epidemiology and molecular

126 carcinogenesis. Gastroenterology. 2007;132:2557-76.

44. Hernandez-Gea V, Friedman SL. Pathogenesis of liver fibrosis. Annual review of pathology. 2011;6:425-56.

45. Tsochatzis EA, Bosch J, Burroughs AK. Liver cirrhosis. Lancet. 2014;383:1749-61.

46. Zeng C, Wang YL, Xie C, Sang Y, Li TJ, Zhang M, et al. Identification of a novel

TGF-beta-miR-122-fibronectin 1/serum response factor signaling cascade and its implication in hepatic fibrogenesis. Oncotarget. 2015;6:12224-33.

47. Llovet JM, Zucman-Rossi J, Pikarsky E, Sangro B, Schwartz M, Sherman M, et al.

Hepatocellular carcinoma. Nature reviews Disease primers. 2016;2:16018.

48. Balogh J, Victor D, 3rd, Asham EH, Burroughs SG, Boktour M, Saharia A, et al.

Hepatocellular carcinoma: a review. Journal of hepatocellular carcinoma. 2016;3:41-53.

49. Bandiera S, Pfeffer S, Baumert TF, Zeisel MB. miR-122--a key factor and therapeutic target in liver disease. Journal of hepatology. 2015;62:448-57.

50. Buendia MA. Genetics of hepatocellular carcinoma. Seminars in cancer biology.

2000;10:185-200.

51. Farazi PA, DePinho RA. Hepatocellular carcinoma pathogenesis: from genes to environment. Nature reviews Cancer. 2006;6:674-87.

52. Teufel A, Staib F, Kanzler S, Weinmann A, Schulze-Bergkamen H, Galle PR.

Genetics of hepatocellular carcinoma. World journal of gastroenterology. 2007;13:2271-82.

53. Yeo W, Wong N, Wong WL, Lai PB, Zhong S, Johnson PJ. High frequency of promoter hypermethylation of RASSF1A in tumor and plasma of patients with hepatocellular carcinoma. Liver international : official journal of the International

127

Association for the Study of the Liver. 2005;25:266-72.

54. Yu J, Zhang HY, Ma ZZ, Lu W, Wang YF, Zhu JD. Methylation profiling of twenty four genes and the concordant methylation behaviours of nineteen genes that may contribute to hepatocellular carcinogenesis. Cell research. 2003;13:319-33.

55. Yang B, Guo M, Herman JG, Clark DP. Aberrant promoter methylation profiles of tumor suppressor genes in hepatocellular carcinoma. The American journal of pathology.

2003;163:1101-7.

56. Kutay H, Bai S, Datta J, Motiwala T, Pogribny I, Frankel W, et al. Downregulation of miR-122 in the rodent and human hepatocellular carcinomas. Journal of cellular biochemistry. 2006;99:671-8.

57. Spaniel C, Honda M, Selitsky SR, Yamane D, Shimakami T, Kaneko S, et al. microRNA-122 abundance in hepatocellular carcinoma and non-tumor liver tissue from

Japanese patients with persistent HCV versus HBV infection. PloS one. 2013;8:e76867.

58. Wang S, Qiu L, Yan X, Jin W, Wang Y, Chen L, et al. Loss of microRNA 122 expression in patients with hepatitis B enhances hepatitis B virus replication through cyclin

G(1) -modulated P53 activity. Hepatology. 2012;55:730-41.

59. Tsuchiya N, Sawada Y, Endo I, Saito K, Uemura Y, Nakatsura T. Biomarkers for the early diagnosis of hepatocellular carcinoma. World journal of gastroenterology.

2015;21:10573-83.

60. Bruix J, Qin S, Merle P, Granito A, Huang YH, Bodoky G, et al. Regorafenib for patients with hepatocellular carcinoma who progressed on sorafenib treatment (RESORCE): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2017;389:56-66.

128

61. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. The New England journal of medicine. 2008;359:378-

90.

62. Llovet JM, Villanueva A, Lachenmayer A, Finn RS. Advances in targeted therapies for hepatocellular carcinoma in the genomic era. Nature reviews Clinical oncology.

2015;12:408-24.

63. Marquardt JU, Andersen JB, Thorgeirsson SS. Functional and genetic deconstruction of the cellular origin in liver cancer. Nature reviews Cancer. 2015;15:653-67.

64. El-Serag HB. Hepatocellular carcinoma. The New England journal of medicine.

2011;365:1118-27.

65. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. The New England journal of medicine. 2008;359:378-

90.

66. Bruix J, Takayama T, Mazzaferro V, Chau GY, Yang J, Kudo M, et al. Adjuvant sorafenib for hepatocellular carcinoma after resection or ablation (STORM): a phase 3, randomised, double-blind, placebo-controlled trial. The Lancet Oncology. 2015;16:1344-54.

67. Thakral S, Ghoshal K. miR-122 is a unique molecule with great potential in diagnosis, prognosis of liver disease, and therapy both as miRNA mimic and antimir.

Current gene therapy. 2015;15:142-50.

68. Antonelli A, Fallahi P, Delle Sedie A, Ferrari SM, Maccheroni M, Bombardieri S, et al. High values of Th1 (CXCL10) and Th2 (CCL2) chemokines in patients with psoriatic arthtritis. Clinical and experimental rheumatology. 2009;27:22-7.

129

69. Xia M, Sui Z. Recent developments in CCR2 antagonists. Expert opinion on therapeutic patents. 2009;19:295-303.

70. Li X, Yao W, Yuan Y, Chen P, Li B, Li J, et al. Targeting of tumour-infiltrating macrophages via CCL2/CCR2 signalling as a therapeutic strategy against hepatocellular carcinoma. Gut. 2015.

71. Tsui P, Das A, Whitaker B, Tornetta M, Stowell N, Kesavan P, et al. Generation, characterization and biological activity of CCL2 (MCP-1/JE) and CCL12 (MCP-5) specific antibodies. Human antibodies. 2007;16:117-25.

72. Yu J, Mitsui T, Wei M, Mao H, Butchar JP, Shah MV, et al. NKp46 identifies an

NKT cell subset susceptible to leukemic transformation in mouse and human. The Journal of clinical investigation. 2011;121:1456-70.

73. Brunner KT, Mauel J, Cerottini JC, Chapuis B. Quantitative assay of the lytic action of immune lymphoid cells on 51-Cr-labelled allogeneic target cells in vitro; inhibition by isoantibody and by drugs. Immunology. 1968;14:181-96.

74. Majumder S, Roy S, Kaffenberger T, Wang B, Costinean S, Frankel W, et al. Loss of metallothionein predisposes mice to diethylnitrosamine-induced hepatocarcinogenesis by activating NF-kappaB target genes. Cancer research. 2010;70:10265-76.

75. Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Varambally R, Yu J, Briggs BB, et al. Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia (New York, NY). 2007;9:166-80.

76. Guichard C, Amaddeo G, Imbeaud S, Ladeiro Y, Pelletier L, Maad IB, et al.

Integrated analysis of somatic mutations and focal copy-number changes identifies key

130 genes and pathways in hepatocellular carcinoma. Nature genetics. 2012;44:694-8.

77. Mas VR, Maluf DG, Archer KJ, Yanek K, Kong X, Kulik L, et al. Genes involved in viral carcinogenesis and tumor initiation in hepatitis C virus-induced hepatocellular carcinoma. Molecular medicine. 2009;15:85-94.

78. Stankovic A, Slavic V, Stamenkovic B, Kamenov B, Bojanovic M, Mitrovic DR.

Serum and synovial fluid concentrations of CCL2 (MCP-1) chemokine in patients suffering rheumatoid arthritis and osteoarthritis reflect disease activity. Bratislavske lekarske listy.

2009;110:641-6.

79. Qian BZ, Li J, Zhang H, Kitamura T, Zhang J, Campion LR, et al. CCL2 recruits inflammatory monocytes to facilitate breast-tumour metastasis. Nature. 2011;475:222-5.

80. Fridlender ZG, Kapoor V, Buchlis G, Cheng G, Sun J, Wang LC, et al. Monocyte chemoattractant protein-1 blockade inhibits lung cancer tumor growth by altering macrophage phenotype and activating CD8+ cells. American journal of respiratory cell and molecular biology. 2011;44:230-7.

81. Zhu X, Fujita M, Snyder LA, Okada H. Systemic delivery of neutralizing antibody targeting CCL2 for glioma therapy. Journal of neuro-oncology. 2011;104:83-92.

82. Loberg RD, Ying C, Craig M, Yan L, Snyder LA, Pienta KJ. CCL2 as an important mediator of prostate cancer growth in vivo through the regulation of macrophage infiltration.

Neoplasia (New York, NY). 2007;9:556-62.

83. Loberg RD, Ying C, Craig M, Day LL, Sargent E, Neeley C, et al. Targeting CCL2 with systemic delivery of neutralizing antibodies induces prostate cancer tumor regression in vivo. Cancer research. 2007;67:9417-24.

131

84. Seok SJ, Lee ES, Kim GT, Hyun M, Lee JH, Chen S, et al. Blockade of CCL2/CCR2 signalling ameliorates diabetic nephropathy in db/db mice. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association -

European Renal Association. 2013;28:1700-10.

85. Elmarakby AA, Quigley JE, Olearczyk JJ, Sridhar A, Cook AK, Inscho EW, et al.

Chemokine receptor 2b inhibition provides renal protection in angiotensin II - salt hypertension. Hypertension. 2007;50:1069-76.

86. Miura K, Yang L, van Rooijen N, Ohnishi H, Seki E. Hepatic recruitment of macrophages promotes nonalcoholic steatohepatitis through CCR2. American journal of physiology Gastrointestinal and liver physiology. 2012;302:G1310-21.

87. Kashyap S, Warner GM, Hartono SP, Boyilla R, Knudsen BE, Zubair AS, et al.

Blockade of CCR2 reduces macrophage influx and development of chronic renal damage in murine renovascular hypertension. American journal of physiology Renal physiology.

2016;310:F372-84.

88. Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell.

2010;140:883-99.

89. Zhao L, Lim SY, Gordon-Weeks AN, Tapmeier TT, Im JH, Cao Y, et al. Recruitment of a myeloid cell subset (CD11b/Gr1 mid) via CCL2/CCR2 promotes the development of colorectal cancer liver metastasis. Hepatology (Baltimore, Md). 2013;57:829-39.

90. Park EJ, Lee JH, Yu GY, He G, Ali SR, Holzer RG, et al. Dietary and genetic obesity promote liver inflammation and tumorigenesis by enhancing IL-6 and TNF expression. Cell.

2010;140:197-208.

132

91. Hui L, Zatloukal K, Scheuch H, Stepniak E, Wagner EF. Proliferation of human

HCC cells and chemically induced mouse liver cancers requires JNK1-dependent p21 downregulation. The Journal of clinical investigation. 2008;118:3943-53.

92. Kawate S, Fukusato T, Ohwada S, Watanuki A, Morishita Y. Amplification of c-myc in hepatocellular carcinoma: correlation with clinicopathologic features, proliferative activity and p53 overexpression. Oncology. 1999;57:157-63.

93. June CH. Adoptive T cell therapy for cancer in the clinic. The Journal of clinical investigation. 2007;117:1466-76.

94. Weiner GJ. Building better monoclonal antibody-based therapeutics. Nature reviews

Cancer. 2015;15:361-70.

95. Prieto J, Melero I, Sangro B. Immunological landscape and immunotherapy of hepatocellular carcinoma. Nature reviews Gastroenterology & hepatology. 2015;12:681-700.

96. Trotta R, Dal Col J, Yu J, Ciarlariello D, Thomas B, Zhang X, et al. TGF-beta utilizes

SMAD3 to inhibit CD16-mediated IFN-gamma production and antibody-dependent cellular cytotoxicity in human NK cells. Journal of immunology (Baltimore, Md : 1950).

2008;181:3784-92.

97. Uhlen M, Fagerberg L, Hallstrom BM, Lindskog C, Oksvold P, Mardinoglu A, et al.

Proteomics. Tissue-based map of the human proteome. Science (New York, NY).

2015;347:1260419.

98. Toffanin S, Friedman SL, Llovet JM. Obesity, inflammatory signaling, and hepatocellular carcinoma-an enlarging link. Cancer cell. 2010;17:115-7.

99. Tian Z, Chen Y, Gao B. Natural killer cells in liver disease. Hepatology (Baltimore,

133

Md). 2013;57:1654-62.

100. Saini RV, Wilson C, Finn MW, Wang T, Krensky AM, Clayberger C. Granulysin delivered by cytotoxic cells damages endoplasmic reticulum and activates caspase-7 in target cells. Journal of immunology (Baltimore, Md : 1950). 2011;186:3497-504.

101. Warren HS, Smyth MJ. NK cells and apoptosis. Immunology and cell biology.

1999;77:64-75.

102. Sui Q, Zhang J, Sun X, Zhang C, Han Q, Tian Z. NK cells are the crucial antitumor mediators when STAT3-mediated immunosuppression is blocked in hepatocellular carcinoma. Journal of immunology (Baltimore, Md : 1950). 2014;193:2016-23.

103. Ong HT, Federspiel MJ, Guo CM, Ooi LL, Russell SJ, Peng KW, et al. Systemically delivered measles virus-infected mesenchymal stem cells can evade host immunity to inhibit liver cancer growth. Journal of hepatology. 2013;59:999-1006.

104. Kim N, Kim H, Jung I, Kim Y, Kim D, Han YM. Expression profiles of miRNAs in human embryonic stem cells during hepatocyte differentiation. Hepatology research : the official journal of the Japan Society of Hepatology. 2011;41:170-83.

105. Coulouarn C, Factor VM, Andersen JB, Durkin ME, Thorgeirsson SS. Loss of miR-

122 expression in liver cancer correlates with suppression of the hepatic phenotype and gain of metastatic properties. Oncogene. 2009;28:3526-36.

106. Bai S, Nasser MW, Wang B, Hsu SH, Datta J, Kutay H, et al. MicroRNA-122 inhibits tumorigenic properties of hepatocellular carcinoma cells and sensitizes these cells to sorafenib. The Journal of biological chemistry. 2009;284:32015-27.

107. Friedman SL. Liver fibrosis -- from bench to bedside. Journal of hepatology. 2003;38

134

Suppl 1:S38-53.

108. Friedman SL, Roll FJ, Boyles J, Bissell DM. Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. Proceedings of the National Academy of

Sciences of the United States of America. 1985;82:8681-5.

109. Gabele E, Brenner DA, Rippe RA. Liver fibrosis: signals leading to the amplification of the fibrogenic hepatic stellate cell. Frontiers in bioscience : a journal and virtual library.

2003;8:d69-77.

110. Higuchi H, Gores GJ. Mechanisms of liver injury: an overview. Current molecular medicine. 2003;3:483-90.

111. Canbay A, Friedman S, Gores GJ. Apoptosis: the nexus of liver injury and fibrosis.

Hepatology. 2004;39:273-8.

112. Naito M, Hasegawa G, Ebe Y, Yamamoto T. Differentiation and function of Kupffer cells. Medical electron microscopy : official journal of the Clinical Electron Microscopy

Society of Japan. 2004;37:16-28.

113. Thurman RG. II. Alcoholic liver injury involves activation of Kupffer cells by endotoxin. The American journal of physiology. 1998;275:G605-11.

114. Casini A, Ceni E, Salzano R, Biondi P, Parola M, Galli A, et al. Neutrophil-derived superoxide anion induces lipid peroxidation and stimulates collagen synthesis in human hepatic stellate cells: role of nitric oxide. Hepatology. 1997;25:361-7.

115. Schuppan D, Afdhal NH. Liver cirrhosis. Lancet. 2008;371:838-51.

116. Tsujimoto Y, Gorham J, Cossman J, Jaffe E, Croce CM. The t(14;18) chromosome translocations involved in B-cell neoplasms result from mistakes in VDJ joining. Science.

135

1985;229:1390-3.

117. Monaghan P, Robertson D, Amos TA, Dyer MJ, Mason DY, Greaves MF.

Ultrastructural localization of bcl-2 protein. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society. 1992;40:1819-25.

118. Delbridge AR, Grabow S, Strasser A, Vaux DL. Thirty years of BCL-2: translating cell death discoveries into novel cancer therapies. Nature reviews Cancer. 2016;16:99-109.

119. Czabotar PE, Lessene G, Strasser A, Adams JM. Control of apoptosis by the BCL-2 protein family: implications for physiology and therapy. Nature reviews Molecular cell biology. 2014;15:49-63.

120. Kluck RM, Bossy-Wetzel E, Green DR, Newmeyer DD. The release of cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis. Science.

1997;275:1132-6.

121. Letai AG. Diagnosing and exploiting cancer's addiction to blocks in apoptosis.

Nature reviews Cancer. 2008;8:121-32.

122. Novo E, Marra F, Zamara E, Valfre di Bonzo L, Monitillo L, Cannito S, et al.

Overexpression of Bcl-2 by activated human hepatic stellate cells: resistance to apoptosis as a mechanism of progressive hepatic fibrogenesis in humans. Gut. 2006;55:1174-82.

123. Zhang Y, Liu C, Barbier O, Smalling R, Tsuchiya H, Lee S, et al. Bcl2 is a critical regulator of bile acid homeostasis by dictating Shp and lncRNA H19 function. Scientific reports. 2016;6:20559.

124. Luna JM, Barajas JM, Teng KY, Sun HL, Moore MJ, Rice CM, et al. Argonaute

CLIP Defines a Deregulated miR-122-Bound Transcriptome that Correlates with Patient

136

Survival in Human Liver Cancer. Molecular cell. 2017;67:400-10 e7.

125. Kolesnikov N, Hastings E, Keays M, Melnichuk O, Tang YA, Williams E, et al.

ArrayExpress update--simplifying data submissions. Nucleic acids research.

2015;43:D1113-6.

126. Pineau P, Volinia S, McJunkin K, Marchio A, Battiston C, Terris B, et al. miR-221 overexpression contributes to liver tumorigenesis. Proceedings of the National Academy of

Sciences of the United States of America. 2010;107:264-9.

127. Wang M, Gong Q, Zhang J, Chen L, Zhang Z, Lu L, et al. Characterization of gene expression profiles in HBV-related liver fibrosis patients and identification of ITGBL1 as a key regulator of fibrogenesis. Scientific reports. 2017;7:43446.

128. Teng KY, Han J, Zhang X, Hsu SH, He S, Wani NA, et al. Blocking the CCL2-CCR2

Axis Using CCL2-Neutralizing Antibody Is an Effective Therapy for Hepatocellular Cancer in a Mouse Model. Molecular cancer therapeutics. 2017;16:312-22.

129. Sun HL, Cui R, Zhou J, Teng KY, Hsiao YH, Nakanishi K, et al. ERK Activation

Globally Downregulates miRNAs through Phosphorylating Exportin-5. Cancer cell.

2016;30:723-36.

130. Xu L, Hui AY, Albanis E, Arthur MJ, O'Byrne SM, Blaner WS, et al. Human hepatic stellate cell lines, LX-1 and LX-2: new tools for analysis of hepatic fibrosis. Gut.

2005;54:142-51.

131. Povero D, Eguchi A, Li H, Johnson CD, Papouchado BG, Wree A, et al. Circulating extracellular vesicles with specific proteome and liver microRNAs are potential biomarkers for liver injury in experimental fatty liver disease. PloS one. 2014;9:e113651.

137

132. Bala S, Petrasek J, Mundkur S, Catalano D, Levin I, Ward J, et al. Circulating microRNAs in exosomes indicate hepatocyte injury and inflammation in alcoholic, drug- induced, and inflammatory liver diseases. Hepatology. 2012;56:1946-57.

133. Lou G, Song X, Yang F, Wu S, Wang J, Chen Z, et al. Exosomes derived from miR-

122-modified adipose tissue-derived MSCs increase chemosensitivity of hepatocellular carcinoma. Journal of hematology & oncology. 2015;8:122.

134. Young DD, Connelly CM, Grohmann C, Deiters A. Small molecule modifiers of microRNA miR-122 function for the treatment of hepatitis C virus infection and hepatocellular carcinoma. Journal of the American Chemical Society. 2010;132:7976-81.

135. Gonzalez-Puertos VY, Hernandez-Perez E, Nuno-Lambarri N, Ventura-Gallegos JL,

Lopez-Diazguerrero NE, Robles-Diaz G, et al. Bcl-2 overexpression in hepatic stellate cell line CFSC-2G, induces a pro-fibrotic state. Journal of gastroenterology and hepatology.

2010;25:1306-14.

136. Stilgenbauer S, Eichhorst B, Schetelig J, Coutre S, Seymour JF, Munir T, et al.

Venetoclax in relapsed or refractory chronic lymphocytic leukaemia with 17p deletion: a multicentre, open-label, phase 2 study. The Lancet Oncology. 2016;17:768-78.

137. Chi SW, Zang JB, Mele A, Darnell RB. Argonaute HITS-CLIP decodes microRNA- mRNA interaction maps. Nature. 2009;460:479-86.

138. Chen L, Chen R, Kemper S, Charrier A, Brigstock DR. Suppression of fibrogenic signaling in hepatic stellate cells by Twist1-dependent microRNA-214 expression: Role of exosomes in horizontal transfer of Twist1. American journal of physiology Gastrointestinal and liver physiology. 2015;309:G491-9.

138

139. Yip KW, Reed JC. Bcl-2 family proteins and cancer. Oncogene. 2008;27:6398-406.

140. Reed JC. Bcl-2-family proteins and hematologic malignancies: history and future prospects. Blood. 2008;111:3322-30.

141. Tse C, Shoemaker AR, Adickes J, Anderson MG, Chen J, Jin S, et al. ABT-263: a potent and orally bioavailable Bcl-2 family inhibitor. Cancer research. 2008;68:3421-8.

142. Mason KD, Carpinelli MR, Fletcher JI, Collinge JE, Hilton AA, Ellis S, et al.

Programmed anuclear cell death delimits platelet life span. Cell. 2007;128:1173-86.

143. Roberts AW, Seymour JF, Brown JR, Wierda WG, Kipps TJ, Khaw SL, et al.

Substantial susceptibility of chronic lymphocytic leukemia to BCL2 inhibition: results of a phase I study of navitoclax in patients with relapsed or refractory disease. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2012;30:488-96.

144. Souers AJ, Leverson JD, Boghaert ER, Ackler SL, Catron ND, Chen J, et al. ABT-

199, a potent and selective BCL-2 inhibitor, achieves antitumor activity while sparing platelets. Nature medicine. 2013;19:202-8.

145. Roberts AW, Davids MS, Pagel JM, Kahl BS, Puvvada SD, Gerecitano JF, et al.

Targeting BCL2 with Venetoclax in Relapsed Chronic Lymphocytic Leukemia. The New

England journal of medicine. 2016;374:311-22.

146. Kapanadze T, Medina-Echeverz J, Gamrekelashvili J, Weiss JM, Wiltrout RH,

Kapoor V, et al. Tumor-induced CD11b(+) Gr-1(+) myeloid-derived suppressor cells exacerbate immune-mediated hepatitis in mice in a CD40-dependent manner. European journal of immunology. 2015;45:1148-58.

147. Ilkovitch D, Lopez DM. The liver is a site for tumor-induced myeloid-derived

139 suppressor cell accumulation and immunosuppression. Cancer research. 2009;69:5514-21.

148. Waidmann O, Koberle V, Brunner F, Zeuzem S, Piiper A, Kronenberger B. Serum microRNA-122 predicts survival in patients with liver cirrhosis. PloS one. 2012;7:e45652.

149. Cermelli S, Ruggieri A, Marrero JA, Ioannou GN, Beretta L. Circulating microRNAs in patients with chronic hepatitis C and non-alcoholic fatty liver disease. PloS one.

2011;6:e23937.

140