UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Pathogenesis and reversal of liver fibrosis: Effects of genes and environment

Sokolović, A.

Link to publication

Citation for published version (APA):Sokolovi, A. (2013). Pathogenesis and reversal of liver fibrosis: Effects of genes and environment.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 05 Sep 2020

Pathogenesis and reversal of liver fibrosis

– effects of genes and environment

PATHOGENESIS AND REVERSAL OF LIVER FIBROSIS

– EFFECTS OF GENES AND ENVIRONMENT

Author:

Aleksandar Sokolović/PhD Thesis/Faculty of Medicine, University of Amsterdam, 2013

Cover and illustrations:

Tomislav Sokolović

Printed by:

Ipskamp Drukkers BV

ISBN: 978-94-6191-623-5

Copyright © 2013 Aleksandar Sokolović

The research described in this thesis was performed at Tytgat Institute for Liver and Intestinal

Research Academic Medical Center, University of Amsterdam, The Netherlands.

The work was supported by Dutch MLDS grant CG 102001.

The study presented in Chapter 4 of the thesis was partly performed at Division of Hepatology and

Gene Therapy, CIMA University of Navarra, and CIBERehd, Pamplona, Spain. The research

described in Chapter 5 was done in collaboration with Department of Medical Biochemistry,

Academic Medical Center, University of Amsterdam, The Netherlands.

Printing of this thesis was financially supported by The Netherlands Federation for Hepatology, and

Tytgat Institute for Liver and Intestinal Research Academic Medical Center, University of

Amsterdam, The Netherlands.

PATHOGENESIS AND REVERSAL OF LIVER FIBROSIS

– EFFECTS OF GENES AND ENVIRONMENT

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus prof.dr. D.C. van den Boom

ten overstaan van een door het college voor promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op donderdag 28 februari 2013, te 10:00 uur

door

Aleksandar Sokolović

geboren te Leskovac, Serviё

Promotiecommissie:

Promotor Prof.dr. R.P.J. Oude Elferink

Co-promotor Dr. P.J. Bosma

Overige leden: Prof.dr.em. W.H. Lamers

Prof.dr. U.H.W. Beuers

Prof.dr. J. Prieto

Prof.dr. A.J. Moshage

Prof.dr. K. Poelstra

Prof.dr. J. Rothuizen

Faculteit der Geneeskunde

za mamu, tatu, Ducu, Milku i Miu

TABLE OF CONTENTS

Chapter 1 Intro: Scope of the thesis 8

Introduction

Outline of the thesis

Chapter 2 IGFBP5 enhances survival of LX2 human hepatic stellate cells 30

Fibrogenesis & Tissue Repair 2010, 3:3

Chapter 3 Unexpected reduction of murine liver fibrosis by IGFBP5 46

Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease;

2012 Jun;1822(6):996-1003.

Chapter 4 IGF1 enhances bile-duct proliferation and fibrosis in Abcb4-/- mice 64

accepted for publication in:

Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 2013

Chapter 5 Fasting reduces liver fibrosis in mouse model 80

for chronic cholangiopathies

under revision

Chapter 6 Closure: Summarizing discussion and conclusions 96

Summaries

Abbreviations

Personalia Acknowledgements 116

Curriculum vitae

List of publications

c h a p t e r 1

intro:

scope of the thesis

introduction

outline of the thesis

chapter 1

10

SCOPE OF THE THESIS

Hepatic fibrosis results from repeated liver tissue damage, caused by chronic viral hepatitis, parasitic

diseases, inborn errors of metabolism, toxic damage and non-alcoholic fatty liver disease (NAFLD).

Fibrosis is a generic wound healing response to chronic insults, regardless of their mechanism. In the

liver, resulting progressive accumulation of fibrillar extracellular matrix (ECM) is associated with

ECM degradation and remodeling that can result either in restoration of normal hepatic structure

and function, or in cirrhosis, conversion of normal liver architecture into structurally abnormal

nodules.

Efforts to elucidate the cellular and molecular basis of fibrogenesis have generated a

comprehensive picture of progression and regression of liver fibrosis, essential for developing

antifibrotic therapies. Due to the dynamic evolving nature of the disease, it seems that interventions

even as late as in cirrhosis could improve clinical outcomes. Though translation of these findings into

diagnostic tools and treatments does not seem too far away, there are still no effective antifibrotics.

There is also a requirement for better markers of the stage and activity of fibrosis, predicting if

patients will progress to cirrhosis or reverse.

The objective of this thesis was therefore to make use of multiple complementary

experimental model systems, molecular techniques and metabolic challenges, to study liver fibrosis,

with the final aim to create new opportunities for antifibrotic therapies.

intro

11

INTRODUCTION

Fibrotic diseases account for up to 45% of deaths in the developed world (1). They are the most

common non-neoplastic cause of death among hepatobiliary and digestive diseases in the USA and

Europe, even though only 25-30% of patients affected by chronic liver diseases are likely to develop

significant fibrosis and cirrhosis, with an inevitably poor prognosis, where 1-year mortality ranges

from 1% in early cirrhosis to 57% in decompensated disease (2). The main causes of liver fibrosis in

developed countries are chronic viral hepatitis, alcohol and drug abuse. The association of

fibrosis/cirrhosis with primary liver cancer further increases the relative mortality rate (3, 4). The

worldwide burden is equally compelling, with chronic liver diseases affecting hundreds of millions of

individuals (5). In addition, the explosive growth in obesity worldwide has led to a dramatic increase

in the prevalence of non-alcoholic steatohepatitis (NASH) (6), which also confers a risk of

hepatocellular carcinoma (HCC), (fastest rising cancer incidence of any neoplasm in the USA and

Western Europe) once cirrhosis develops (7, 8). The large number of people affected worldwide and

the lack of effective anti-fibrotic treatment necessitates further studies of the pathophysiology in

order to deliver novel therapies.

HEPATIC FIBROGENESIS

Liver fibrosis results from perpetuation of the normal wound healing response. Repeated cycles of

hepatocyte injury and repair, accompanied with inflammation and excessive deposition of ECM

proteins, result in an abnormal continuation of fibrogenesis. Its progression depends on the etiology

of liver disease and is influenced by environmental and genetic factors (9-11).

During hepatic fibrogenesis, significant changes in quality, quantity and distribution of ECM

components occur in the periportal and perisinusoidal space, increasing the content of collagens and

non-collagenous components (12).

In normal liver, the subendothelial space of Disse that separates the epithelium (hepatocytes)

from the sinusoidal endothelium, contains both an interstitial (high-density) and a basement

membrane–like, perisinusoidal (low density) ECM (Figure 1a). During fibrosis, perisinusoidal ECM is

replaced by an interstitial (scar-type) matrix with bundles of collagen fibrils and an electron-dense

basement membrane (13). This is tied in with a loss of hepatocyte microvilli and a disappearance of

endothelial fenestrations (Figure 1b). The dominant event in fibrogenic cascade is activation of

hepatic stellate cells (HSCs) (14), remarkably adaptable mesenchymal cells that are vital to

hepatocellular function and the liver’s response to injury. These cells resident in the space of Disse

are the major storage of vitamin A in the body. They represent one-third of the nonparenchymal

cells, and ∼15% of the total number of resident cells in normal liver (Figure 1a). HSCs are primary

effector cells that play a pivotal role in activating the immune response through secretion of

cytokines and chemokines and interacting with immune cells, contributing thereby to angiogenesis

and the regulation of oxidant stress, and orchestrating the deposition of ECM in normal and fibrotic

liver (15). In the injured liver, numbers of activated hepatic stellate cells/myofibroblasts

(HSCs/MFs) dramatically increase at sites of infection, inflammation and injury, promoting the

chapter 1

12

deposition of scar-tissue (Figure 1b). These cells are generated locally, by transdifferentiation of

resident perisinusoidal HSCs and periportal fibroblasts (16, 17), or via the recruitment and

differentiation of bone-marrow-derived mesenchymal stem cells (18, 19). Activated HSCs secrete

large amount of interstitial collagen, accompanied by matrix metalloproteinases (MMPs), which

induce three-dimensional changes in the ECM. They degrade the normal basement membrane

matrix, whilst having a mild effect on the neo-matrix rich in fibrillar collagen. On the other hand, the

production of tissue inhibitor of metalloproteinase (TIMPs) by HSCs also increases, promoting

ECM accumulation by inhibiting matrix degradation, improving survival of activated HSCs and

preventing their clearance, all delaying the regression of liver fibrosis (20-22).

The changes in ECM impair the flow of metabolites between the hepatocytes and the

perfusing plasma, exposing hepatocytes to hypoxia (particularly in pericentral zones). In unopposed

fibrosis this causes distortion of hepatic architecture, formation of septae (broad bands that divide

the parenchyma into regenerative nodules), increase in intrahepatic vascular resistance, angiogenesis

and sinusoidal remodeling, inducing portal hypertension (23). This ultimately leads to

pathophysiological damage and cirrhosis as the final stage (24) . Advanced fibrotic changes are

strongly associated with hepatocellular carcinoma, which further reduces chances of survival (25).

Potential reversibility of hepatic fibrosis varies depending on the duration of injury, the

degree of angiogenesis, and the composition, spatial distribution, and cellularity of scar. Cirrhotic

areas resistant to degradation are fairly acellular and rich in elastin and highly cross-linked collagen,

suggesting that ECM cross-linking might represent a “point of no return” in fibrosis (22). Fibrotic

patients have a good short-term prognosis, as it usually requires several months to years of ongoing

insult to reach chirrosis (26). Cirrhosis, however, as present evidence indicates, is not completely

reversible.

SIGNALS FOR PATHOGENESIS IN LIVER FIBROSIS

Similar to fibrotic disorders in other organs, chronic activation of the wound-healing process is the

mechanism behind hepatic fibrogenesis. Upon hepatic injury, a fibrogenic response is provoked by

activation of key signals in resident and infiltrating liver cells – reactive oxygen species (ROS),

hypoxia, inflammation and apoptosis. As a consequence, ECM proteins accumulate in the liver due

to imbalance between the fibrogenesis and fibrinolysis.

Oxidative stress in chronic liver diseases (CLD) results from an increased generation of

ROS (and other reactive intermediates) and from a decreased efficiency of antioxidant defenses. This

stress contributes to excessive tissue remodeling and fibrogenesis by stimulating the production of

profibrogenic mediators by Kupffer cells and other resident and circulating inflammatory cells. It

directly affects the behavior of HSCs/MFs by induction of critical profibrogenic genes, like

procollagen type 1, monocyte chemoattractant protein 1, and tissue inhibitor of metalloproteinase 1

(TIMP1). This is likely enabled by the activation of transcription factors, such as c-jun N-terminal

kinases (JNKs), transcription factor AP1 and nuclear factor κB (NFκB) (27).

Hypoxia is a critical, early fibrogenic stimulus, caused by capillarisation of sinusoids,

intrahepatic shunting, vasoconstriction, compression and thrombosis (28). It impairs mitochondrial

function, produces oxidative stress, induces vascular endothelial cell growth factor (VEGF) and its

intro

13

receptors, and stimulates synthesis of collagen 1 in HSCs (29). Hypoxia also potentiates expression

of transforming growth factor, beta 1 (TGFβ1) (30), contributing to both autocrine and paracrine

loops that drive angiogenesis, chemotaxis and fibrogenesis.

Figure 1: Hepatic liver cells and the hepatic sinusoid in normal and injured liver

Inflammation initiates and regulates fibrosis by the release of soluble mediators and the

secretion of products that remodel/degrade the normal basement membrane matrix within the liver

(31). Inflammatory cells, resident (e.g. NK cells and macrophages) or recruited (e.g. T and B cells),

are involved in pathogen elimination, cell removal (e.g. hepatocytes damage during anti-viral immune

reaction), recruitment and activation of MFs, and spontaneous recovery of fibrosis (32, 33).

chapter 1

14

Apoptosis, a common feature of chronic liver disease, results in the generation of apoptotic

bodies, which phagocytotic clearance creates pro-inflammatory and fibrogenic stimuli. After

engulfing the apoptotic bodies, Kupffer cells secrete death ligands and tumor necrosis factor alpha

(TNFα). Similarly, the engulfment of apoptotic bodies by HSCs triggers a profibrogenic response

with production of oxidative radicals, and upregulation of expression of TGFβ1 and collagen 1 (34-

37).

HEPATIC STELLATE CELLS AND FIBROGENESIS

In response to liver injury, HSC undergo rapid activation, functional and morphological changes.

Activation, which occurs in two phases, initiation and perpetuation, is, if liver injury has subsided,

followed by resolution (Figure 2) (38). Initiation of HSCs activation comprises early changes in

gene expression and phenotype, which make them responsive to cytokines and other stimuli released

from affected cells, like damaged hepatocytes, bile duct cells, Kupffer cells and inflammatory cells

(39). Perpetuation encompasses cellular series of events that amplify the activated phenotype

through enhanced growth factor expression and responsiveness. It results from autocrine and

paracrine stimulation, and accelerated ECM remodeling.

In response to acute and chronic liver injury, HSCs proliferation is promoted by strongly

increased mitogens. They include platelet-derived growth factor (PDGF), VEGFA, thrombin and its

receptor, epidermal growth factor (EGF), transforming growth factor α (TGFα), keratinocyte growth

factor and fibroblast growth factors 2 (FGF2) and its receptors (FGFR1) (11). The most potent

fibrogenic factor for HSCs is TGFβ1 (40, 41). Activated TGFβ1 released from the local ECM binds

to an increased number of TGFβ receptors on HSCs. In response, signaling via the SMAD pathway

enhances proliferation and survival of activated HSCs, increases collagen synthesis, upregulates

TIMPs, and decreases MMPs expression (42). HSCs migrate and accumulate in zones of

injury/repair (43) secreting large amounts of cytokines (amplifying the inflammatory and fibrogenic

tissue responses), and MMPs (accelerating the replacement of normal with "scar" matrix). Activated

HSCs accumulated in the collagenous bands impede portal blood flow by constricting individual

sinusoids and entire organ (44, 45).

Resolution of fibrosis refers to the fate of activated HSCs when the primary insult is

withdrawn or attenuated, and refers to pathways that cause apoptosis, senescence, or quiescence of

HSCs (Figure 2) (46-48).

ANTIFIBROTIC THERAPIES

Current treatments of cirrhosis include withdrawal of the causative agent and treatment of

complications, but the only curative treatment for advanced cirrhosis is liver transplantation. Its

clinical applicability is, however, limited by shortage of donors, by the commitment of recipients to

life-long toxic immunosuppression, and by poor condition of the potential recipient.

intro

15

Figure 2: The role of hepatic stellate cells in recurrent liver injury

Since the progress of fibrosis is basically similar in all liver diseases regardless of the etiology,

the development of antifibrotic therapies should benefit all patients with fibrosing liver injury. The

liver's ability to regenerate and its first-pass metabolism allow lower doses of orally administered

compounds to achieve a therapeutic response, minimizing systemic distribution and non-hepatic side

effects, which makes liver fibrosis an attractive target for therapy. The progress in clinical research

and basic science and improved diagnostic tools promise to halt the progression of liver fibrosis

and/or promote its reversion. Potential therapeutic approaches for liver fibrosis include: treatment

of the primary disease, suppression of hepatic inflammation, inhibition of HSCs activation,

promotion of matrix degradation, stimulation of HSCs apoptosis, and targeted antifibrotic therapies

(Figure 3). These approaches will be addressed individually.

Treatment of the primary disease. Clearing the primary cause of liver disease is still the

most effective intervention in the treatment of liver fibrosis. Removal of excess iron or copper in

haemochromatosis or Wilson’s disease, respectively, abstinence in alcoholic liver disease, anti-

helminthic therapy in parasitic infections, clearance of HBV or HCV in chronic viral hepatitis, and

biliary decompression in bile duct obstruction, or, most recently, weight loss in patients with NASH,

lead to reduced fibrogenesis (49-53) This approach can be highly effective when applied early,

resulting in a near-normal restitution of hepatic histology.

chapter 1

16

Figure 3: Targets for antifibrotic therapies.

Suppression of hepatic inflammation. In sustained liver injury, persistent inflammation

perpetuates the fibrogenic HSCs phenotype and leads to fibrosis. Therefore, a number of anti-

inflammatory agents have been evaluated in vitro and in vivo. Treatment with corticosteroids in

patients with severe autoimmune hepatitis leads to hepatic fibrosis only in a minority of patients (54).

Blocking interleukin 1 by expressing the interleukin-1 receptor antagonist, also reduced liver damage

and pro-inflammatory cytokine levels in an in vivo model of ischaemia-reperfusion injury (55).

Recombinant interleukin 10 reduces inflammation and fibrosis in patients with chronic hepatitis C,

but results in an increased viral load (56). Local expression of interferon alpha (IFNα) improved liver

fibrosis in a dimethylnitrosamine-induced model of cirrhosis (57). Finally, the beneficial effects of

ursodeoxycholic acid in the treatment of primary biliary cirrhosis (PBC) are thought to be in part via

anti-inflammatory mechanisms (58).

Inhibition of HSCs activation. HSCs activation, proliferation and fibrogenesis are pivotal

steps in the development of liver fibrosis, making the mechanisms that mediate them attractive

targets for therapeutic intervention. Oxidative stress, which stimulates HSCs activation, is one such

target. Several therapeutic agents have been tested in animal models to counteract the activation of

HSCs by oxidative stress, like vitamin E and C (59), and S-adenosyl- L- methionine (60) but the

effects were minor. A number of other agents have been shown to inhibit the HSC activation. They

include interferon gamma (IFNγ) (61) and ligands for peroxisomal proliferator activated receptor

(PPAR) gamma.(62). TGFβ antagonists are extensively tested, since neutralizing of this potent

cytokine would have the dual effect: inhibition of matrix production and acceleration of its

degradation. Adenoviral expression of Tgfβ1 antisense mRNA, and gene transfer of Smad7, which

blocks TGFβ intracellular signaling, also had positive effect. Gene transfer of soluble TGFβR2 that

intro

17

inhibits liver fibrogenesis in rats is shown to attenuate apoptosis, injury, and fibrosis in bleomycin-

induced lung fibrosis in mice. Blocking TGFβ activity by synthetic peptide reduces fibrosis in mice in

carbon tetrachloride (CCL4)-induced liver fibrosis, and bleomycin-induced skin and lung fibrosis (41,

63). Although inhibitors of TGFβ1 are effective in short-term animal models, they are not suitable

for long term therapy because of the significant role of TGFβ1 in homeostasis and repair (47).

Promotion of matrix degradation. Remodeling of ECM occurs through the action of

MMPs that efficiently cleave collagens and other matrix components, and the TIMPs, critical

determinants of fibrosis reversal, which inhibit their activity. During the regression, decreased TIMP-

1 levels are associated with clearance of activated HSCs through apoptosis (64). Fibrinolysis and

matrix rearrangement can be induced by a number of agents, as shown by studies in animal models

of liver fibrosis. For example, direct adenoviral administration of Mmp1 mRNA attenuated

established liver fibrosis in thioacetamide treated rats, while administration of an anti-TIMP1

antibody decreased HSCs activation and attenuated fibrosis in a CCl4 rat model (65, 66).

Downregulation of TIMPs by the reproductive hormone relaxin also inhibits liver fibrosis in vitro

(67). Furthermore, adenoviral delivery of urokinase-type plasminogen activator (uPA), which initiates

the matrix proteolysis and upregulates hepatocyte growth factor (HGF) (68), results in enhanced

collagenase activity, fibrosis regression and hepatocyte regeneration (69). MMPs gene delivery in

advanced liver fibrosis shows that even transient shifts in the imbalance between MMPs and TIMPs

are sufficient for the resolution - if the underlying causative stimuli have been successfully removed

(70). Dietary supplementation with zinc, an essential co-factor for MMP activity, has also been

shown to promote collagen degradation in CCl4-injured rats (71). Collagen synthesis is also sensitive

to changes in food intake, and malnutrition may have profound effects on its production (72).

Stimulation of HSCs apoptosis. Apoptosis is a key event in the spontaneous recovery

from liver fibrosis (73) and can be induced by two general mechanisms (34). The first, ”extrinsic”

pathway, works via stimulation of specific cell surface death receptors, which trigger an intracellular

cascade of caspases, ensuing cellular apoptosis. Second, ”intrinsic” pathway, operates at the level of

the mitochondria. Stimuli like toxins, UV radiation, and reactive oxygen species stimulate the

proapoptotic factors in the mitochondrial membrane, resulting in leakage of cytochrome C into the

cytosol, caspase activation, and apoptosis. In addition, upon activation of HSCs, the expression of a

cell surface death receptors CD95 (APO-1/Fas) and CD95L (APO-1-/Fas-ligand) increases, which

may explain the increased apoptosis seen upon their activation in vitro and in vivo (74). Also, activation

of TNFα receptor, and low-affinity nerve growth factor (NGF) receptor (p75) increases HSCs

apoptosis in vitro (75). Transformed HSCs also express peripheral benzodiazepine receptors, which

render the cells sensitive to peripheral benzodiazepine receptor-ligand-induced apoptosis (76).

NFκB, on the other hand, protects HSCs from apoptosis, and its inhibition with gliotoxin accelerates

recovery from drug-induced liver fibrosis in rats (77), proving that HSCs pro-apoptotic agents could

abrogate liver fibrosis. The approaches mentioned in this section may sound attractive for in vitro

work, but the selectivity may be too low to use this approach in vivo.

Targeted antifibrotic therapy. Wound healing is an important universal process, hence

liver-specific targeting of antifibrotic therapies minimizes the potential harmful effects on other

tissues. Furthermore the extensive first pass metabolism makes liver an attractive target for directed

therapy with orally administered drugs. Still, hepatic delivery needs to be optimized via cell-specific

chapter 1

18

targeting (of hepatocytes, Kupffer cells, or HSCs). It has been shown, for instance, that

administration of recombinant insulin-like growth factor 1 (rIGF1) or HSCs-targeted induction of its

expression using a viral vector in CCl4-treated mice and rats, significantly reduced liver fibrosis (78-

80). Treatment of cirrhotic rats with rIGF1 promotes weight gain, nitrogen retention, and intestinal

absorption of nutrients and was able to exert hepatoprotective effect (78, 81). Concerns regarding

the safety of gene therapy with regard to genotoxicity and effects on the immune system remain.

However, with improved delivery techniques and the conditional expression systems, the anti-

fibrotic trials in humans are no longer a matter of (distant) future.

IGF AXIS IN LIVER FIBROSIS

Insulin-like growth factor 1 (IGF1) is a potent cytoprotective and anabolic hormone, synthesized

mainly in the liver. It circulates bound to a set of binding proteins (IGFBPs) that prolong its half-life

and, by modulating its interaction with the IGF1 receptor (IGF1R), control its biological activity

(82). Detachment from IGFBPs releases free IGF1, which can bind to and activate the IGF1R.

Interaction with IGF1R leads to activation of mitogen activated protein (MAP) kinase and

phosphatidylinositol 3-kinases (PI3) cascades that regulate genes involved in cell survival, growth,

and differentiation (83). In advanced liver fibrosis, the IGF1 axis is severely impaired mostly due to a

reduced number of healthy, IGF1 producing hepatocytes. The decrease in IGF1 signalling can

disturb systemic metabolism (e.g. in liver cirrhosis), and provide a pro-fibrotic environment, since

the progression of liver fibrosis could be delayed by IGF1 administration (79, 84).

As mentioned, the action of IGF1 depends on six IGFBPs of multifunctional nature, which

can act both via IGF-dependent and independent mechanisms. Among them, insulin-like growth

factor binding protein 5 (IGFBP5) that binds to IGFs with high affinity (85) is strongly induced in

the fibrotic livers of Abcb4-/- mice (86), suggesting a potential role in the pathogenesis of chronic

cholangiopathy. Its expression is also high in skin and lung fibrosis (SSc and IPF; (87, 88)). IGFBP5

induces collagen and fibronectin production in fibroblasts, and their transdifferentiation to

myofibroblast in vitro and in vivo (89, 90). In IGFBP5-induced dermal fibrosis, an increase in the

number of fibroblasts expressing proliferating cell nuclear antigen PCNA is accompanied by

increased vimentin and α-SMA expression (suggesting an IGFBP5-induced myofibroblast

differentiation). This leads to an excessive production of ECM and the development of fibrosis.

IGFBP5 seems to play an important role in other aspects of fibrosis, such as epithelial injury.

IGFBP5 expression is dramatically increased in the involuting mammary gland and prostate during

apoptosis, and in the brain after ischaemia (91, 92). IGFBP5 induces epithelial–mesenchymal

transition in pulmonary epithelial cells in vitro (90). Senescence, a process in which cells lose the

ability to proliferate, is also associated with increased IGFBP5 expression in human dermal

fibroblasts and endothelial cells (93, 94). A knockdown of IGFBP5 partially reverses, whereas its

induction causes a premature senescence in human umbilical-vein endothelial cells (95).

intro

19

LIVER FIBROSIS AND DIET

As the prevalence of obesity is rapidly increasing, the prevalence of NASH is also on the rise .

NASH has been recognized as a major cause of liver fibrosis (96), ranging from steatosis to cirrhosis,

eventually leading to HCC. NASH is a component of metabolic syndrome and type 2 diabetes,

characterized by obesity, dyslipidemia and insulin resistance (97). Insulin resistance is a prominent

mechanism that links steatosis and fibrogenesis (98), hence adipokines that play a role in insulin

resistance but also steatosis and inflammation are considered to be involved in liver fibrosis.

Fibrogenic action of insulin itself works via the release of TGFβ1, which stimulates type 1 collagen

production in HSC (99). Leptin promotes stellate cell mitogenesis and cell survival, two seminal

events that promote liver fibrosis (100). In contrast, administration of recombinant adiponectin and

its agonists reduced steatosis, attenuated inflammation and the elevated levels of serum alanine

aminotransferase (ALT) in (non)alcoholic fatty liver diseases in mice, indicating potential clinical

application (101, 102).

Treatment options in obesity-linked liver fibrosis are limited so lifestyle changes like

introducing exercise and healthy diet are advised. Weight loss in morbidly obese patients is associated

with a reduction in the prevalence of hepatic fibrosis (103). In a mouse model of metabolic

syndrome, stimulation of weight loss by blocking the endocannabinoid receptor CB1 prevents

progression of fibrosis (104). Still, beneficial effects of insulin sensitizers on progression of liver

fibrosis need to be documented in therapeutic trials.

MODELS OF LIVER FIBROSIS

Although fibrosis occurs as a part of normal wound healing, excessive (dysregulated) deposition of

ECM leads to severe organ dysfunction and is a feature of a variety of diseases. Due to its insidious

onset, fibrosis tends to go undetected in its early stages. This is, in part, why these diseases remain so

poorly understood. To improve our understanding of development and reversal of liver fibrosis, a

number of experimental models have been established.

The first are in vitro models - purified primary cells from normal or experimentally injured

livers, or stable cell lines representative of specific hepatic cell types. These models enhance the

understanding of the molecular pathogenesis of liver fibrosis (15), and permit high-throughput

testing and improvement of potential antifibrotic agents, thanks to low cost and high reproducibility.

However, they cannot recapitulate the in vivo situation that involves the complex interplay of resident

and incoming cells in a dynamic microenvironment.

The alternative are the ex vivo models - human hepatic tissue obtained from biopsies,

resections or explants. This type of models is vital for validating observations made in animal and

cell culture models (105). However, ethical reasons prevent multiple liver biopsies in humans,

resulting in limited amount of data, usually at the advanced end of the disease.

In vivo experimental animal models of fibrosis are used to address the limitations of the

other two types, and to facilitate the study of a highly dynamic process (106, 107). In spite their

questionable relevance to human disease, they allow consecutive sampling of tissue in the volumes

required for detailed studies of cellular and molecular pathogenesis. These models provide a chance

chapter 1

20

to examine the early stages of fibrosis and changes in signaling pathways, chemokines, and cytokines.

The two most frequently used models of experimental fibrosis are repetitive toxic damage by carbon

tetrachloride (CCl4) and bile duct ligation (BDL) (107). CCl4 induces zone III necrosis, hepatocyte

apoptosis and, with iterative dosing, can be used to generate bridging hepatic fibrosis, early cirrhosis

and advanced micronodular cirrhosis (4, 8 and 12 weeks of twice weekly dosing, respectively) (73,

108). BDL stimulates the proliferation of biliary epithelial cells and oval cells (hepatocyte

progenitors), resulting in proliferating bile ductules and an accompanying portal inflammation and

fibrosis. Fibrosis can also be established using CCl4 or BDL in specific transgenic mice (over-

expressing pro- or anti-fibrotic factors) and knock-out mice (in which the expression of a pro or anti

fibrotic factor is abolished) (79, 109). This allows addressing the molecular mechanisms and/or

importance of these factors in hepatic functions and diseases.

In vitro and in vivo models used in this thesis will be described in more detail in the coming

paragraphs. The LX2 cell and MFs were used to study the mechanism of action of IGFBP5 in

(activated) HSCs. LX2 human HSCs lines are generated by spontaneous immortalisation in low

serum culture conditions. They retain key features of HSCs, like expression of α-SMA, vimentin, and

glial fibrillary acid protein. Similarly to primary HSCs, LX2 cells express key receptors regulating liver

fibrosis, like platelet derived growth factor receptor β (PDGFβ-R), as well as proteins involved in

matrix remodelling; MMP2, TIMP1 and 2 (110).

Abcb4-/- mice used in this thesis are a well characterized non-surgical mouse model for

chronic cholangiopathies. These mice deficient in the canalicular phospholipid flippase are created

on FVB background (FVB.129P2-Abcb4tm1Bor; (111)). They are characterized by a lack of Mdr2 P-

glycoprotein (multidrug resistance transporter 2, Abcb4), a phosphatidylcholine transporter. This

results in absent phospholipid secretion into bile and spontaneous bile duct injury, with macroscopic

and microscopic features closely resembling human sclerosing cholangitis (PSC) (112). These animals

are an animal model for human MDR3 deficiency (progressive familial intrahepatic cholestasis type

3) which progresses to liver cirrhosis (113). The hepatic pathogenesis in Abcb4-/- mice include

disruption of tight junctions and basement membranes of bile ducts, and bile leakage to the portal

tract. This triggers a multistep process that upregulates fibrogenesis and downregulates fibrinolysis,

leading to the formation of periportal biliary fibrosis (114), which makes these mice an attractive

model for liver fibrosis. As a consequence of chronic inflammation and progressing fibrosis, Abcb4-/-

mice develop HCC within 12 to 15 months of age (115), and, at later age, they serve as a model for

studying HCC. The effect of genetic and environmental factors on the development of liver fibrosis

was studied in this thesis by exposing these mice either to liver specific overexpression of genes

(belonging to IGF axis), or to metabolic challenges (like food deprivation).

intro

21

OUTLINE OF THE THESIS

The aim of the research described in this thesis was to identify novel approaches and therapeutic

targets for the treatment of liver fibrosis. The initial experiments were based on our previous finding

that IGFBP5 was strongly upregulated upon HSC activation and transdifferentiation into MFs in vitro

and in vivo (86). In Chapter 2, we therefore investigated the role of IGFBP5 in HSCs using gain- and

loss-of-function approaches. We used the human LX2 cell line (110) that recapitulates many features

of the activated HSCs phenotype. To assess its role in more advanced stages of fibrosis, the response

to IGFBP5 was also dissected in human primary liver MFs. To clarify the role of IGFBP5 in the

pathogenesis of liver fibrosis caused by chronic cholangiopathies (Chapter 3), we overexpressed this

gene in hepatocytes of Abcb4-/- mice using an adeno-associated viral (AAV) vector large amounts of

IGFBP5 were delivered to the hepatic tissue including HSCs. To substantiate the findings in the

Abcb4-/- model and relate them to the physiological role of hepatic IGFBP5 expression, a systems

genetics approach was applied to a mouse genetic reference population, using an open source

database. Since the biliary epithelium expresses a receptor for IGF1 (116), and insulin-like growth

factor 1 (IGF1) protects cholangiocytes against cholestatic injury in vitro (117), overexpression of

IGF1 could be beneficial in cholestatic liver diseases. Therefore, in Chapter 4 we aimed to study if a

prolonged increase of hepatic IGF1 expression in a model for chronic cholangiopathy, the Abcb4-/-

mouse, was beneficial. Using a rat IGF1 gene behind an α-Sma promoter the expression of this

growth factor was enhanced in activated HSCs.

Intermittent fasting and caloric restriction have beneficial effect on health status of

laboratory animals (118). In NASH and NAFLD patients, histological improvements of liver

pathology positively correlate with the size of weight loss (119-121), but the mechanism is not fully

understood. Weight loss has beneficial effects on fibrosis induced by nutrient overload in men and

mice (103, 104). We therefore postulated that the complex adaptive response to nutrient deprivation

would also benefit the fibrotic pathology in cholestatic liver injury. In Chapter 5 we therefore tested

the hypothesis that food deprivation could positively influence pathology in fibrotic Abcb4-/- mouse

liver. Finally, in Chapter 6 the work described in this thesis is discussed, and future perspectives are

given.

chapter 1

22

References

1. Mehal WZ, Iredale J, Friedman SL. Scraping fibrosis: expressway to the core of fibrosis. Nat Med

2011 May;17(5):552-553.

2. Tsochatzis EA, Bosch J, Burroughs AK. A new therapeutic paradigm for patients with cirrhosis.

Hepatology 2012 Jun 22.

3. Befeler AS, Di Bisceglie AM. Hepatocellular carcinoma: diagnosis and treatment. Gastroenterology

2002 May;122(6):1609-1619.

4. Bruix J, Boix L, Sala M, Llovet JM. Focus on hepatocellular carcinoma. Cancer Cell 2004

Mar;5(3):215-219.

5. Lim YS, Kim WR. The global impact of hepatic fibrosis and end-stage liver disease. Clin Liver Dis

2008 Nov;12(4):733-46, vii.

6. Cheung O, Sanyal AJ. Recent advances in nonalcoholic fatty liver disease. Curr Opin Gastroenterol

2009 May;25(3):230-237.

7. Bugianesi E. Non-alcoholic steatohepatitis and cancer. Clin Liver Dis 2007 Feb;11(1):191-1xi.

8. Hashizume H, Sato K, Takagi H, Hirokawa T, Kojima A, Sohara N, et al. Primary liver cancers with

nonalcoholic steatohepatitis. Eur J Gastroenterol Hepatol 2007 Oct;19(10):827-834.

9. Bataller R, North KE, Brenner DA. Genetic polymorphisms and the progression of liver fibrosis: a

critical appraisal. Hepatology 2003 Mar;37(3):493-503.

10. Friedman SL. Liver fibrosis -- from bench to bedside. J Hepatol 2003;38 Suppl 1:S38-S53.

11. Lee UE, Friedman SL. Mechanisms of hepatic fibrogenesis. Best Pract Res Clin Gastroenterol 2011

Apr;25(2):195-206.

12. Schuppan D. Structure of the extracellular matrix in normal and fibrotic liver: collagens and

glycoproteins. Semin Liver Dis 1990 Feb;10(1):1-10.

13. Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a modulator of hepatic fibrogenesis.

Semin Liver Dis 2001 Aug;21(3):351-372.

14. Friedman SL. Molecular regulation of hepatic fibrosis, an integrated cellular response to tissue injury.

J Biol Chem 2000 Jan 28;275(4):2247-2250.

15. Friedman SL. Hepatic stellate cells: protean, multifunctional, and enigmatic cells of the liver. Physiol

Rev 2008 Jan;88(1):125-172.

16. Iwaisako K, Brenner DA, Kisseleva T. What's new in liver fibrosis? The origin of myofibroblasts in

liver fibrosis. J Gastroenterol Hepatol 2012 Mar;27 Suppl 2:65-68.

17. Parola M, Marra F, Pinzani M. Myofibroblast - like cells and liver fibrogenesis: Emerging concepts in

a rapidly moving scenario. Mol Aspects Med 2008 Feb;29(1-2):58-66.

18. Forbes SJ, Russo FP, Rey V, Burra P, Rugge M, Wright NA, et al. A significant proportion of

myofibroblasts are of bone marrow origin in human liver fibrosis. Gastroenterology 2004

Apr;126(4):955-963.

19. Russo FP, Alison MR, Bigger BW, Amofah E, Florou A, Amin F, et al. The bone marrow

functionally contributes to liver fibrosis. Gastroenterology 2006 May;130(6):1807-1821.

20. Han YP. Matrix metalloproteinases, the pros and cons, in liver fibrosis. J Gastroenterol Hepatol 2006

Oct;21 Suppl 3:S88-S91.

21. Mormone E, George J, Nieto N. Molecular pathogenesis of hepatic fibrosis and current therapeutic

approaches. Chem Biol Interact 2011 Sep 30;193(3):225-231.

22. Ramachandran P, Iredale JP. Liver fibrosis: a bidirectional model of fibrogenesis and resolution. QJM

2012 May 29.

intro

23

23. Thabut D, Shah V. Intrahepatic angiogenesis and sinusoidal remodeling in chronic liver disease: new

targets for the treatment of portal hypertension? J Hepatol 2010 Nov;53(5):976-980.

24. Schuppan D, Afdhal NH. Liver cirrhosis. Lancet 2008 Mar 8;371(9615):838-851.

25. Zhang DY, Friedman SL. Fibrosis-dependent mechanisms of hepatocarcinogenesis. Hepatology 2012

Mar 1.

26. Fattovich G, Giustina G, Degos F, Tremolada F, Diodati G, Almasio P, et al. Morbidity and mortality

in compensated cirrhosis type C: a retrospective follow-up study of 384 patients. Gastroenterology

1997 Feb;112(2):463-472.

27. Urtasun R, Conde dlR, Nieto N. Oxidative and nitrosative stress and fibrogenic response. Clin Liver

Dis 2008 Nov;12(4):769-90, viii.

28. Rosmorduc O, Housset C. Hypoxia: a link between fibrogenesis, angiogenesis, and carcinogenesis in

liver disease. Semin Liver Dis 2010 Aug;30(3):258-270.

29. Corpechot C, Barbu V, Wendum D, Kinnman N, Rey C, Poupon R, et al. Hypoxia-induced VEGF

and collagen I expressions are associated with angiogenesis and fibrogenesis in experimental cirrhosis.

Hepatology 2002 May;35(5):1010-1021.

30. Jeong WI, Do SH, Yun HS, Song BJ, Kim SJ, Kwak WJ, et al. Hypoxia potentiates transforming

growth factor-beta expression of hepatocyte during the cirrhotic condition in rat liver. Liver Int 2004

Dec;24(6):658-668.

31. Mehal WZ, Friedman SL. The role of inflammation and immunity in the pathogenesis of liver

fibrosis. In: Gershwin ME, Veirling JM, Manns MP, eds. Liver Immunology.Totowa: New Jersey:

Humana Press, 2007. 99-109.

32. Henderson NC, Iredale JP. Liver fibrosis: cellular mechanisms of progression and resolution. Clin Sci

(Lond) 2007 Mar;112(5):265-280.

33. Mehal WZ. Activation-induced cell death of hepatic stellate cells by the innate immune system.

Gastroenterology 2006 Feb;130(2):600-603.

34. Canbay A, Friedman S, Gores GJ. Apoptosis: the nexus of liver injury and fibrosis. Hepatology 2004

Feb;39(2):273-278.

35. Zhan SS, Jiang JX, Wu J, Halsted C, Friedman SL, Zern MA, et al. Phagocytosis of apoptotic bodies

by hepatic stellate cells induces NADPH oxidase and is associated with liver fibrosis in vivo.

Hepatology 2006 Mar;43(3):435-443.

36. Canbay A, Taimr P, Torok N, Higuchi H, Friedman S, Gores GJ. Apoptotic body engulfment by a

human stellate cell line is profibrogenic. Lab Invest 2003 May;83(5):655-663.

37. Chakraborty JB, Oakley F, Walsh MJ. Mechanisms and biomarkers of apoptosis in liver disease and

fibrosis. Int J Hepatol 2012;2012:648915.

38. Friedman SL. Mechanisms of disease: Mechanisms of hepatic fibrosis and therapeutic implications.

Nat Clin Pract Gastroenterol Hepatol 2004 Dec;1(2):98-105.

39. Friedman SL, Bansal MB. Reversal of hepatic fibrosis -- fact or fantasy? Hepatology 2006 Feb;43(2

Suppl 1):S82-S88.

40. Dooley S, Ten DP. TGF-beta in progression of liver disease. Cell Tissue Res 2012 Jan;347(1):245-

256.

41. Hayashi H, Sakai T. Biological Significance of Local TGF-beta Activation in Liver Diseases. Front

Physiol 2012;3:12.

42. Yoshida K, Matsuzaki K. Differential Regulation of TGF-beta/Smad Signaling in Hepatic Stellate

Cells between Acute and Chronic Liver Injuries. Front Physiol 2012;3:53.

43. Hashmi AZ, Hakim W, Kruglov EA, Watanabe A, Watkins W, Dranoff JA, et al. Adenosine inhibits

cytosolic calcium signals and chemotaxis in hepatic stellate cells. Am J Physiol Gastrointest Liver

Physiol 2007 Jan;292(1):G395-G401.

chapter 1

24

44. Reynaert H, Thompson MG, Thomas T, Geerts A. Hepatic stellate cells: role in microcirculation and

pathophysiology of portal hypertension. Gut 2002 Apr;50(4):571-581.

45. Rockey DC. Vascular mediators in the injured liver. Hepatology 2003 Jan;37(1):4-12.

46. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005 Feb;115(2):209-218.

47. Kisseleva T, Brenner DA. Hepatic stellate cells and the reversal of fibrosis. J Gastroenterol Hepatol

2006 Oct;21 Suppl 3:S84-S87.

48. Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, Miething C, et al. Senescence of activated

stellate cells limits liver fibrosis. Cell 2008 Aug 22;134(4):657-667.

49. Dixon JB, Bhathal PS, Hughes NR, O'Brien PE. Nonalcoholic fatty liver disease: Improvement in

liver histological analysis with weight loss. Hepatology 2004 Jun;39(6):1647-1654.

50. Falize L, Guillygomarc'h A, Perrin M, Laine F, Guyader D, Brissot P, et al. Reversibility of hepatic

fibrosis in treated genetic hemochromatosis: a study of 36 cases. Hepatology 2006 Aug;44(2):472-477.

51. Hammel P, Couvelard A, O'toole D, Ratouis A, Sauvanet A, Flejou JF, et al. Regression of liver

fibrosis after biliary drainage in patients with chronic pancreatitis and stenosis of the common bile

duct. N Engl J Med 2001 Feb 8;344(6):418-423.

52. Klein S, Mittendorfer B, Eagon JC, Patterson B, Grant L, Feirt N, et al. Gastric bypass surgery

improves metabolic and hepatic abnormalities associated with nonalcoholic fatty liver disease.

Gastroenterology 2006 May;130(6):1564-1572.

53. Roberts S, Gordon A, McLean C, Pedersen J, Bowden S, Thomson K, et al. Effect of sustained viral

response on hepatic venous pressure gradient in hepatitis C-related cirrhosis. Clin Gastroenterol

Hepatol 2007 Aug;5(8):932-937.

54. Czaja AJ, Carpenter HA. Progressive fibrosis during corticosteroid therapy of autoimmune hepatitis.

Hepatology 2004 Jun;39(6):1631-1638.

55. Harada H, Wakabayashi G, Takayanagi A, Shimazu M, Matsumoto K, Obara H, et al. Transfer of the

interleukin-1 receptor antagonist gene into rat liver abrogates hepatic ischemia-reperfusion injury.

Transplantation 2002 Nov 27;74(10):1434-1441.

56. Nelson DR, Tu Z, Soldevila-Pico C, Abdelmalek M, Zhu H, Xu YL, et al. Long-term interleukin 10

therapy in chronic hepatitis C patients has a proviral and anti-inflammatory effect. Hepatology 2003

Oct;38(4):859-868.

57. Suzuki K, Aoki K, Ohnami S, Yoshida K, Kazui T, Kato N, et al. Adenovirus-mediated gene transfer

of interferon alpha improves dimethylnitrosamine-induced liver cirrhosis in rat model. Gene Ther

2003 May;10(9):765-773.

58. Pares A, Caballeria L, Rodes J. Excellent long-term survival in patients with primary biliary cirrhosis

and biochemical response to ursodeoxycholic Acid. Gastroenterology 2006 Mar;130(3):715-720.

59. Soylu AR, Aydogdu N, Basaran UN, Altaner S, Tarcin O, Gedik N, et al. Antioxidants vitamin E and

C attenuate hepatic fibrosis in biliary-obstructed rats. World J Gastroenterol 2006 Nov

14;12(42):6835-6841.

60. Yang H, Ramani K, Xia M, Ko KS, Li TW, Oh P, et al. Dysregulation of glutathione synthesis during

cholestasis in mice: molecular mechanisms and therapeutic implications. Hepatology 2009

Jun;49(6):1982-1991.

61. Li Q, Yan Z, Li F, Lu W, Wang J, Guo C. The improving effects on hepatic fibrosis of interferon-

gamma liposomes targeted to hepatic stellate cells. Nanotechnology 2012 Jun 15;23(26):265101.

62. Rockey DC. Current and future anti-fibrotic therapies for chronic liver disease. Clin Liver Dis 2008

Nov;12(4):939-62, xi.

63. Hawinkels LJ, Ten DP. Exploring anti-TGF-beta therapies in cancer and fibrosis. Growth Factors

2011 Aug;29(4):140-152.

intro

25

64. Hemmann S, Graf J, Roderfeld M, Roeb E. Expression of MMPs and TIMPs in liver fibrosis - a

systematic review with special emphasis on anti-fibrotic strategies. J Hepatol 2007 May;46(5):955-975.

65. Iimuro Y, Nishio T, Morimoto T, Nitta T, Stefanovic B, Choi SK, et al. Delivery of matrix

metalloproteinase-1 attenuates established liver fibrosis in the rat. Gastroenterology 2003

Feb;124(2):445-458.

66. Parsons CJ, Bradford BU, Pan CQ, Cheung E, Schauer M, Knorr A, et al. Antifibrotic effects of a

tissue inhibitor of metalloproteinase-1 antibody on established liver fibrosis in rats. Hepatology 2004

Nov;40(5):1106-1115.

67. Williams EJ, Benyon RC, Trim N, Hadwin R, Grove BH, Arthur MJ, et al. Relaxin inhibits effective

collagen deposition by cultured hepatic stellate cells and decreases rat liver fibrosis in vivo. Gut 2001

Oct;49(4):577-583.

68. Narmada BC, Chia SM, Tucker-Kellogg L, Yu H. HGF regulates the activation of TGF-beta1 in rat

hepatocytes and hepatic stellate cells. J Cell Physiol 2012 Jun 20.

69. Salgado S, Garcia J, Vera J, Siller F, Bueno M, Miranda A, et al. Liver cirrhosis is reverted by

urokinase-type plasminogen activator gene therapy. Mol Ther 2000 Dec;2(6):545-551.

70. Iimuro Y, Brenner DA. Matrix metalloproteinase gene delivery for liver fibrosis. Pharm Res 2008

Feb;25(2):249-258.

71. Gimenez A, Pares A, Alie S, Camps J, Deulofeu R, Caballeria J, et al. Fibrogenic and collagenolytic

activity in carbon-tetrachloride-injured rats: beneficial effects of zinc administration. J Hepatol 1994

Sep;21(3):292-298.

72. Spanheimer R, Zlatev T, Umpierrez G, DiGirolamo M. Collagen production in fasted and food-

restricted rats: response to duration and severity of food deprivation. J Nutr 1991 Apr;121(4):518-524.

73. Iredale JP, Benyon RC, Pickering J, McCullen M, Northrop M, Pawley S, et al. Mechanisms of

spontaneous resolution of rat liver fibrosis. Hepatic stellate cell apoptosis and reduced hepatic

expression of metalloproteinase inhibitors. J Clin Invest 1998 Aug 1;102(3):538-549.

74. Saile B, Knittel T, Matthes N, Schott P, Ramadori G. CD95/CD95L-mediated apoptosis of the

hepatic stellate cell. A mechanism terminating uncontrolled hepatic stellate cell proliferation during

hepatic tissue repair. Am J Pathol 1997 Nov;151(5):1265-1272.

75. Trim N, Morgan S, Evans M, Issa R, Fine D, Afford S, et al. Hepatic stellate cells express the low

affinity nerve growth factor receptor p75 and undergo apoptosis in response to nerve growth factor

stimulation. Am J Pathol 2000 Apr;156(4):1235-1243.

76. Fischer R, Schmitt M, Bode JG, Haussinger D. Expression of the peripheral-type benzodiazepine

receptor and apoptosis induction in hepatic stellate cells. Gastroenterology 2001 Apr;120(5):1212-

1226.

77. Wright MC, Issa R, Smart DE, Trim N, Murray GI, Primrose JN, et al. Gliotoxin stimulates the

apoptosis of human and rat hepatic stellate cells and enhances the resolution of liver fibrosis in rats.

Gastroenterology 2001 Sep;121(3):685-698.

78. Castilla-Cortazar I, Garcia M, Muguerza B, Quiroga J, Perez R, Santidrian S, et al. Hepatoprotective

effects of insulin-like growth factor I in rats with carbon tetrachloride-induced cirrhosis.

Gastroenterology 1997 Nov;113(5):1682-1691.

79. Sanz S, Pucilowska JB, Liu S, Rodriguez-Ortigosa CM, Lund PK, Brenner DA, et al. Expression of

insulin-like growth factor I by activated hepatic stellate cells reduces fibrogenesis and enhances

regeneration after liver injury. Gut 2005 Jan;54(1):134-141.

80. Sobrevals L, Rodriguez C, Romero-Trevejo JL, Gondi G, Monreal I, Paneda A, et al. Insulin-like

growth factor I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to

cirrhosis reversion in rats. Hepatology 2010 Mar;51(3):912-921.

chapter 1

26

81. Castilla-Cortazar I, Prieto J, Urdaneta E, Pascual M, Nunez M, Zudaire E, et al. Impaired intestinal

sugar transport in cirrhotic rats: correction by low doses of insulin-like growth factor I.

Gastroenterology 1997 Oct;113(4):1180-1187.

82. Mohan S, Baylink DJ. IGF-binding proteins are multifunctional and act via IGF-dependent and -

independent mechanisms. J Endocrinol 2002 Oct;175(1):19-31.

83. Riedemann J, Macaulay VM. IGF1R signalling and its inhibition. Endocr Relat Cancer 2006 Dec;13

Suppl 1:S33-S43.

84. Conchillo M, de Knegt RJ, Payeras M, Quiroga J, Sangro B, Herrero JI, et al. Insulin-like growth

factor I (IGF-I) replacement therapy increases albumin concentration in liver cirrhosis: results of a

pilot randomized controlled clinical trial. J Hepatol 2005 Oct;43(4):630-636.

85. Beattie J, Allan GJ, Lochrie JD, Flint DJ. Insulin-like growth factor-binding protein-5 (IGFBP-5): a

critical member of the IGF axis. Biochem J 2006 Apr 1;395(1):1-19.

86. Boers W, Aarrass S, Linthorst C, Pinzani M, Elferink RO, Bosma P. Transcriptional profiling reveals

novel markers of liver fibrogenesis: gremlin and insulin-like growth factor-binding proteins. J Biol

Chem 2006 Jun 16;281(24):16289-16295.

87. Feghali CA, Wright TM. Identification of multiple, differentially expressed messenger RNAs in

dermal fibroblasts from patients with systemic sclerosis. Arthritis Rheum 1999 Jul;42(7):1451-1457.

88. Pilewski JM, Liu L, Henry AC, Knauer AV, Feghali-Bostwick CA. Insulin-like growth factor binding

proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular

matrix deposition. Am J Pathol 2005 Feb;166(2):399-407.

89. Yasuoka H, Jukic DM, Zhou Z, Choi AM, Feghali-Bostwick CA. Insulin-like growth factor binding

protein 5 induces skin fibrosis: A novel murine model for dermal fibrosis. Arthritis Rheum 2006

Sep;54(9):3001-3010.

90. Yasuoka H, Zhou Z, Pilewski JM, Oury TD, Choi AM, Feghali-Bostwick CA. Insulin-like growth

factor-binding protein-5 induces pulmonary fibrosis and triggers mononuclear cellular infiltration.

Am J Pathol 2006 Nov;169(5):1633-1642.

91. Flint DJ, Tonner E, Allan GJ. Insulin-like growth factor binding proteins: IGF-dependent and -

independent effects in the mammary gland. J Mammary Gland Biol Neoplasia 2000 Jan;5(1):65-73.

92. O'Donnell SL, Frederick TJ, Krady JK, Vannucci SJ, Wood TL. IGF-I and microglia/macrophage

proliferation in the ischemic mouse brain. Glia 2002 Jul;39(1):85-97.

93. Hampel B, Fortschegger K, Ressler S, Chang MW, Unterluggauer H, Breitwieser A, et al. Increased

expression of extracellular proteins as a hallmark of human endothelial cell in vitro senescence. Exp

Gerontol 2006 May;41(5):474-481.

94. Yoon IK, Kim HK, Kim YK, Song IH, Kim W, Kim S, et al. Exploration of replicative senescence-

associated genes in human dermal fibroblasts by cDNA microarray technology. Exp Gerontol 2004

Sep;39(9):1369-1378.

95. Kim KS, Seu YB, Baek SH, Kim MJ, Kim KJ, Kim JH, et al. Induction of cellular senescence by

insulin-like growth factor binding protein-5 through a p53-dependent mechanism. Mol Biol Cell 2007

Nov;18(11):4543-4552.

96. Brunt EM. Nonalcoholic steatohepatitis. Semin Liver Dis 2004 Feb;24(1):3-20.

97. Anstee QM, Goldin RD. Mouse models in non-alcoholic fatty liver disease and steatohepatitis

research. Int J Exp Pathol 2006 Feb;87(1):1-16.

98. Fartoux L, Poujol-Robert A, Guechot J, Wendum D, Poupon R, Serfaty L. Insulin resistance is a

cause of steatosis and fibrosis progression in chronic hepatitis C. Gut 2005 Jul;54(7):1003-1008.

99. Svegliati-Baroni G, Ridolfi F, Di SA, Casini A, Marucci L, Gaggiotti G, et al. Insulin and insulin-like

growth factor-1 stimulate proliferation and type I collagen accumulation by human hepatic stellate

cells: differential effects on signal transduction pathways. Hepatology 1999 Jun;29(6):1743-1751.

intro

27

100. Saxena NK, Titus MA, Ding X, Floyd J, Srinivasan S, Sitaraman SV, et al. Leptin as a novel

profibrogenic cytokine in hepatic stellate cells: mitogenesis and inhibition of apoptosis mediated by

extracellular regulated kinase (Erk) and Akt phosphorylation. FASEB J 2004 Oct;18(13):1612-1614.

101. Kamada Y, Tamura S, Kiso S, Matsumoto H, Saji Y, Yoshida Y, et al. Enhanced carbon tetrachloride-

induced liver fibrosis in mice lacking adiponectin. Gastroenterology 2003 Dec;125(6):1796-1807.

102. Xu A, Wang Y, Keshaw H, Xu LY, Lam KS, Cooper GJ. The fat-derived hormone adiponectin

alleviates alcoholic and nonalcoholic fatty liver diseases in mice. J Clin Invest 2003 Jul;112(1):91-100.

103. Moretto M, Kupski C, da Silva VD, Padoin AV, Mottin CC. Effect of Bariatric Surgery on Liver

Fibrosis. Obes Surg 2011 Nov 23.

104. Deleve LD, Wang X, Kanel GC, Atkinson RD, McCuskey RS. Prevention of hepatic fibrosis in a

murine model of metabolic syndrome with nonalcoholic steatohepatitis. Am J Pathol 2008

Oct;173(4):993-1001.

105. Lagaye S, Shen H, Saunier B, Nascimbeni M, Gaston J, Bourdoncle P, et al. Efficient replication of

primary or culture hepatitis C virus isolates in human liver slices: A relevant ex vivo model of liver

infection. Hepatology 2012 Sep;56(3):861-872.

106. Hayashi H, Sakai T. Animal models for the study of liver fibrosis: new insights from knockout mouse

models. Am J Physiol Gastrointest Liver Physiol 2011 May;300(5):G729-G738.

107. Constandinou C, Henderson N, Iredale JP. Modeling liver fibrosis in rodents. Methods Mol Med

2005;117:237-250.

108. Issa R, Zhou X, Constandinou CM, Fallowfield J, Millward-Sadler H, Gaca MD, et al. Spontaneous

recovery from micronodular cirrhosis: evidence for incomplete resolution associated with matrix

cross-linking. Gastroenterology 2004 Jun;126(7):1795-1808.

109. Cadoret A, Rey C, Wendum D, Elriz K, Tronche F, Holzenberger M, et al. IGF-1R contributes to

stress-induced hepatocellular damage in experimental cholestasis. Am J Pathol 2009 Aug;175(2):627-

635.

110. 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 Jan;54(1):142-151.

111. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice

with homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of

nonsuppurative inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994

Nov;145(5):1237-1245.

112. Smit JJ, Schinkel AH, Oude Elferink RP, Groen AK, Wagenaar E, van DL, et al. Homozygous

disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from

bile and to liver disease. Cell 1993 Nov 5;75(3):451-462.

113. Jacquemin E. Role of multidrug resistance 3 deficiency in pediatric and adult liver disease: one gene

for three diseases. Semin Liver Dis 2001 Nov;21(4):551-562.

114. Fickert P, Fuchsbichler A, Wagner M, Zollner G, Kaser A, Tilg H, et al. Regurgitation of bile acids

from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology

2004 Jul;127(1):261-274.

115. Katzenellenbogen M, Mizrahi L, Pappo O, Klopstock N, Olam D, Barash H, et al. Molecular

mechanisms of the chemopreventive effect on hepatocellular carcinoma development in Mdr2

knockout mice. Mol Cancer Ther 2007 Apr;6(4):1283-1291.

116. Alvaro D, Metalli VD, Alpini G, Onori P, Franchitto A, Barbaro B, et al. The intrahepatic biliary

epithelium is a target of the growth hormone/insulin-like growth factor 1 axis. J Hepatol 2005

Nov;43(5):875-883.

chapter 1

28

117. Gatto M, Drudi-Metalli V, Torrice A, Alpini G, Cantafora A, Blotta I, et al. Insulin-like growth

factor-1 isoforms in rat hepatocytes and cholangiocytes and their involvement in protection against

cholestatic injury. Lab Invest 2008 Sep;88(9):986-994.

118. Anson RM, Guo Z, de CR, Iyun T, Rios M, Hagepanos A, et al. Intermittent fasting dissociates

beneficial effects of dietary restriction on glucose metabolism and neuronal resistance to injury from

calorie intake. Proc Natl Acad Sci U S A 2003 May 13;100(10):6216-6220.

119. Harrison SA, Day CP. Benefits of lifestyle modification in NAFLD. Gut 2007 Dec;56(12):1760-1769.

120. Harrison SA, Fecht W, Brunt EM, Neuschwander-Tetri BA. Orlistat for overweight subjects with

nonalcoholic steatohepatitis: A randomized, prospective trial. Hepatology 2009 Jan;49(1):80-86.

121. Promrat K, Kleiner DE, Niemeier HM, Jackvony E, Kearns M, Wands JR, et al. Randomized

controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis. Hepatology 2010

Jan;51(1):121-129.

intro

29

c h a p t e r 2

igfbp5 enhances survival

of lx2 human hepatic stellate cells

chapter 2

32

ABSTRACT

Background: Expression of insulin-like growth factor binding protein 5 (IGFBP5) is strongly

induced upon activation of hepatic stellate cells and their transdifferentiation into myofibroblasts in

vitro. This was confirmed in vivo in an animal model of liver fibrosis. Since IGFBP5 has been shown

to promote fibrosis in other tissues, the aim of this study was to investigate its role in the progression

of liver fibrosis.

Methods: The effect of IGFBP5 was studied in LX2 cells, a model for partially activated hepatic

stellate cells, and in human primary liver myofibroblasts. IGFBP5 signalling was modulated by the

addition of recombinant protein, by lentiviral overexpression, and by siRNA mediated silencing.

Furthermore, the addition of IGF1 and silencing of the IGF1R was used to investigate the role of

the IGF-axis in IGFBP5 mediated effects.

Results: IGFBP5 enhanced the survival of LX2 cells and myofibroblasts via a >50% suppression of

apoptosis. This effect of IGFBP5 was not modulated by the addition of IGF1, nor by silencing of

the IGF1R. Additionally, IGFBP5 was able to enhance the expression of established pro-fibrotic

markers, such as collagen Iα1, TIMP1 and MMP1.

Conclusion: IGFBP5 enhances the survival of (partially) activated hepatic stellate cells and

myofibroblasts by lowering apoptosis via an IGF1-independent mechanism, and enhances the

expression of profibrotic genes. Its lowered expression may, therefore, reduce the progression of

liver fibrosis.

Fibrogenesis & Tissue Repair 2010, 3:3

Authors

Aleksandar Sokolović1, Milka Sokolović2, Willem Boers1, Ronald P.J. Oude Elferink1 and Piter J.

Bosma1

Affiliations

1Tytgat Institute for Liver and Intestinal Research,

2Department of Medical Biochemistry,

Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands

IGFBP5 in LX2

33

INTRODUCTION

The extensive accumulation of extracellular matrix (ECM) produced by activated hepatic stellate cells

(HSCs), which normally reside in the space of Disse as the major vitamin A storage site, is a hallmark

of liver fibrosis (1, 2). Liver damage induces HSCs activation and, upon repeated and/or chronic

injury, they transdifferentiate into myofibroblast-like cells (1, 3). These cells migrate to the damaged

regions of the liver (4-6) where they play a central role in the pathogenic deposition of ECM (7, 8).

In order to identify novel therapeutic targets, we used gene expression profiling at different

stages of the pathogenic transdifferentiation of HSCs (9). One of the factors found to be upregulated

upon HSCs activation and further enhanced upon transdifferentiation into myofibroblasts was

IGFBP5 (insulin-like growth factor binding protein 5). This strong induction of IGFBP5 expression

was confirmed during the development of liver fibrosis in the Mdr2-/- mice, a well established animal

model of liver fibrosis (10). Expression of IGFBP5 in HSCs has been reported to be enhanced by

insulin-like growth factor 1 (IGF1) via a post-translational mechanism, while its novel synthesis was

decreased by TGFβ1 (transforming growth factor beta 1) (11).

IGFBP5 is a member the IGFBPs that bind IGF1 (12). IGF1 is mainly synthesized by the

liver and gets secreted into the circulation bound to IGFBPs, which prolong its half-life and, by

modulating its interaction with the IGF1 receptor (IGF1R), control its biological activity (12-14). In

advanced liver fibrosis, the IGF1 axis is severely impaired mostly due to a reduced number of healthy

IGF1 producing hepatocytes (15). The decrease in IGF1 signalling seems to provide a pro-fibrotic

environment, since the progression of liver fibrosis could be delayed by IGF1 administration (16,

17). As IGFBP5 impairs the binding of IGF1 to the cell-surface receptor IGF1R (18), its increased

expression in activated HSCs and myofibroblasts may reduce IGF1 signalling and, thus, promote

liver fibrosis. In contrast, in another study IGF1 has been reported to exert pro-fibrotic activity (19).

In that case the inhibition of IGF1 signalling by IGFBP5 would impair the pathogenesis of liver

fibrosis.

In lung and skin, IGFBP5 has also been shown to induce fibrosis upon epithelial injury (20-

22). Induction of IGFBP5 expression initiated the activation and transdifferentiation of resident

fibroblasts into myofibroblasts, causing increased ECM production and deposition in these tissues.

Moreover, it seemed to cause cellular senescence and epithelial-mesenchymal transition (23).

The aim of this study was to investigate the role of IGFBP5 in liver fibrosis by employing

both gain and loss of function approaches. We focused on the effect of IGFBP5 on HSCs, using the

human LX2 cell line (24), which recapitulates many features of the activated HSCs phenotype. In

addition, to see if IGFBP5 could play a role in more advanced stages of fibrosis, we analysed its

effect on human primary liver myofibroblasts.

MATERIALS AND METHODS

Cell culturing

LX2 cells (kindly provided by Prof. dr. S. Friedman) and human myofibroblasts (obtained as

described (9)) were cultured in Dulbecco’s modified Eagle’s medium (Lonza, Verviers, Belgium),

supplemented with 10% fetal calf serum (FCS), 1 mmol/l L-glutamine, 100 IU/ml penicillin and

chapter 2

34

streptomycin. Human recombinant IGFBP5 and IGF1 (rIGFBP5 and rIGF1; Gro-Pep, Reutlingen,

Germany) were added in concentration 0.1 ng/μl (25) and 1ng/μl, respectively, at 24 and 45 h of cell

culturing. The cells were used for additional assays 3 h after the last protein additions.

In order to induce apoptosis, the cells were serum deprived for 48 h. For drug induced

apoptosis, the cells grown in 10% FCS were incubated with 0.5 μM gliotoxin (Alexis Biochemicals,

Lausen, Switzerland) for 3 h. This dose induced caspase activity sixfold 3h after gliotoxin

administration (data not shown).

Lentiviral transduction

A lentiviral vector encoding human IGFBP5 behind the constitutive PGK promoter was generated

using the self-inactivating cppt-PGKiresGFP-PRE vector (26). Sequencing was performed to

exclude mutations.

siRNA mediated silencing

IGFBP5, IGF1R and negative control small interfering RNAs (siRNAs) were purchased from

Invitrogen (Breda, Netherlands). siRNA was introduced to the cells using Lipofectamine 2000

(Invitrogen, Breda, Netherland) in Optimem (Gibco, Invitrogen, Paisley, UK), according to the

manufacturer’s instructions. For the downstream applications siRNA transfected cells were grown in

10% FCS and used 48 h after transfection. When apoptosis was induced by serum starvation, the

cells were first grown o/n (16 h) in serum containing medium, which was then replaced for 48 h by

serum depleted medium.

Cell viability test

Cell Proliferation Reagent WST-1 (Roche Applied Science, Mannheim, Germany) was added (1/10th

of the culture volume) to cells grown in 96-well plates. Absorbance was measured at 450 nm

(reference at 630 nm) immediately and after 1h of incubation at 37°C.

Bromo-2’-deoxy-uridine (BrdU) assay (Roche applied Science, Penzberg, Germany) was

performed according to the manufacturer’s protocol. The cells grown in 96 well plates were

incubated with BrdU during the last 16 h of culturing.

Apoptosis assay

The Apo-One homogeneous Caspase 3/7 assay (Promega, Medison, WI, USA) was performed

according to the manufacturer’s protocol. The Apo-One Caspase 3/7 reagent was added in a 1:1

ratio with medium and fluorescence was measured using Novostar plate reader at 521 nm.

Reverse transcription and quantitative polymerase chain reaction (qPCR)

Total RNA was isolated using Trizol (Invitrogen, Breda, The Netherlands). RNA concentration was

determined spectrophotometrically and the integrity was checked by gel electrophoreses. cDNA was

synthesized using an oligo-dT primer and Superscript III (Invitrogen, Breda, The Netherlands).

qPCR was performed on a LightCycler using FastStart DNA Master Plus SYBR Green I (Roche,

Mannheim, Germany). The relative expression levels were calculated using the LinReg program (27)

IGFBP5 in LX2

35

and for each sample normalized by the expression of housekeeping gene 36B4 (acidic ribosomal

phosphoprotein P0). Primer sequences are shown in Table 1.

Western blot analysis

Cell extracts were analysed by Western blotting as described (28). We used antisera against IGFBP5

(1:250; Abcam, Cambridge, UK), IGF1R (1:200; Santa Cruz Biotechnology, CA, USA), poly (ADP-

ribose) polymerase (PARP; 1:100; Promega, Madison, USA) and β-actine (1:3000; Sigma, St Louis,

USA). Goat anti-mouse (1:2500; Bio-Rad Laboratories, Veenendaal, The Netherlands) and goat anti-

rabbit IgG horseradish peroxidase conjugated (1:5000; Santa Cruz Biotechnology, CA, USA) were

used as secondary antibodies. Chemiluminescence was quantified on the Lumi-Imager F1 using

CDP-Star (Roche, Mannheim, Germany). Amido-black staining and β-actine immuno-blot verified

similar protein loading.

Statistical analysis

All cell culturing experiments were performed in quadruplo, and were repeated at least three times. In

order to assess the significance of the data, ANOVA analysis and Student’s t test were employed.

The error bars in the figures represent the standard deviation. Significance threshold was set to

P<0.05 (denoted by the asterisks in all the Figures).

RESULTS

IGFBP5 expression in LX2 cells

In human HSCs in vitro, IGFBP5 expression is enhanced twofold upon activation and 200-fold upon

transdifferentiation into myofibroblasts (9). LX2 cells also express IGFBP5 (Table 2). The 30%

higher expression in these cells compared to resting HSCs confirms reports demonstrating that it is a

model for partially activated HSCs (24).

Table 2. Relative gene expression of IGF1 and IGFBP5 in LX2 cells, activated HSCs and myofibroblasts given

as fold changes of basal expression in freshly isolated human HSCs.

LX2 cells HSCs Activated HSCs Myofibroblasts

IGFBP5 1.3 1 2.3 236

IGF1 0.7 1 1.2 0.3

In order to investigate a potential role of IGFBP5 in liver fibrosis, we developed methods to

modulate its expression in LX2 cells. The endogenous expression of IGFBP5 mRNA was doubled

using lentiviral transduction (Figure 1A). This resulted in a 2.5-fold increase in protein levels (Figure

1B). Transfection with IGFBP5 siRNA reduced IGFBP5 mRNA 80% and led to complete absence

of detectable IGFBP5 signal on a Western blot (Figure 1C and D, respectively)

chapter 2

36

Figure 1. Modulation of IGFBP5 expression in LX2 cells. A) IGFBP5 mRNA levels in control cells (LX2) and in

cells with lentiviral overexpression of IGFBP5 (LX2(BP5)). RNA levels normalized to 36B4 are shown as a % of control

cells. B) IGFBP5 protein levels in control (LX2) and transduced (LX2(BP5)) cells as demonstrated on a Western blot

and C) as quantified using the LumiImager. D) Silencing of IGFBP5 expression. Proteins were isolated from the cells

and medium 48 h after small interfering RNA (sRNA) transfection. The presence of IGFBP5 protein was visualized by

Western blotting in LX2 and LX2(BP5) cells transfected with control (siCtrl) and IGFBP5 siRNA (siBP5) and in medium

obtained from transfected LX2(BP5). E) Cells were transfected with small interfering (si) RNA for IGFBP5 (siBP5) or

with a control siRNA (siCtrl). RNA was isolated 48 h later and IGFBP5 mRNA levels were quantified and given as a

percentage of control cells. The results are given as a % of the control cells. The error bars represent standard deviations,

with asterisks denoting significant difference (P<0.05).

IGFBP5 enhances survival of LX2 cells

Several studies have shown that IGFBP5 can affect cell survival (29-32). We therefore compared the

viability of control LX2 cells with cells over-expressing IGFBP5. When grown with 10% FCS, cell

growth did not differ (data not shown). However, upon culturing in serum free medium for 48 h, a

WST assay (for cell viability) indicated 100% higher viability of the transduced cells (Figure 2A). The

similar 60% increase in viability seen upon addition of recombinant IGFBP5 (rIGFBP5) confirmed

the effect of IGFBP5 on LX2 survival (Figure 2B). In serum free medium, the siRNA mediated

reduction of IGFBP5 expression resulted in a >40% decrease in cell viability. This reduction in

viability was prevented by the addition of rIGFBP5 to the silenced cells (Figure 2C).

In order to investigate if IGFBP5 affected proliferation, we studied its effect on the BrdU

incorporation. In the cells grown in serum depleted medium the addition of 0.1 ng/µl rIGFBP5 at 24

h and 45 h of culturing did not effect the incorporation of BrdU (Figure 2D). This indicates that

IGFBP5 does not enhance LX2 cell proliferation.

IGFBP5 in LX2

37

Figure 2. IGFBP5 increases survival of LX2 cells. A) The viability of IGFBP5 overexpressing (LX2(BP5)) and

control cells (LX2) cultured in serum-free medium for 48 h was measured using WST. (B) The viability of LX2 cells

cultured in serum-free medium for 48 h was determined using WST. At 24 and 45 h, recombinant IGFBP5 was added

(rBP5). C) LX2 cells were transfected with control (siCtrl) or IGFBP5 siRNA (siBP5). At 16 h after transfection the

medium was replaced by serum-free medium and 48 h later the viability of the cells was measured using WST. rIGFBP5

was added at 24 and 45 h of serum-free culturing (siBP5+rBP5). D) BrdU incorporation was used to determine the

proliferation of LX2 cells cultured in serum-free medium for 48 h. rIGFBP5 was added at 24 and 45 h (rBP5). BrdU was

added at t=32 h. The results are shown as % of the control cells.

Impact of IGFBP5 on apoptosis in LX2 cells

In order to investigate if the effect of IGFBP5 on viability was caused by decreased apoptosis, we

determined the caspase activity using a Caspase 3/7 assay. The addition of rIGFBP5 to LX2 cells

cultured in serum free medium did reduce the caspase activity by 40% (Figure 3A). In order to

substantiate this protective effect, we also tested it in a model of drug induced apoptosis by gliotoxin

administration (33). The addition of rIGFBP5 3 h prior to the addition of 0.5 μM of gliotoxin

resulted in a significant (40%) lowering of caspase activity (Figure 3B). In order to investigate if

lowering of IGFBP5 would increase apoptosis, we studied the effect of siRNA mediated silencing on

its expression. Compared to control-siRNA transfected cells, the caspase activity in silenced cells was

increased by 30%. This increase was seen when apoptosis was induced both by serum starvation and

gliotoxin. Administration of rIGFBP5 reduced caspase activity in silenced cells to a level significantly

lover than that in control cells (Figure 3C and D, respectively). The increase in apoptosis in silenced

cells was further confirmed by Western blot analysis, which demonstrated a 150% increase of poly

(ADP-ribose) polymerase (PARP) protein expression in silenced compared to control cells (Figure

3E).

Altogether, these data indicate that IGFBP5 enhances the survival of this model for partially

activated HSCs by arresting apoptosis.

chapter 2

38

Figure 3. IGFBP5 impact on apoptosis in LX2 cells. A) LX2 cells were cultured in serum-free medium for 48 h.

Recombinant IGFBP5 (rBP5) was added at 24h and 45 h and, at 48 h, a Caspase 3/7 assay was performed. B) Cells

grown in 10% FCS were treated with rBP5 as above, with the addition of 0.5 μM gliotoxin after 45 h. C) LX2 cells were

transfected with control (siCtrl) or IGFBP5 siRNA (siBP5). 16 h after transfection, the medium was replaced by serum-

free medium and cells were cultured for additional 48 h before performing the caspase assay. To siIGFBP5 transfected

cells, rBP5 was added at 24 h and 45 h (siBP5+rBP5). D) To cells treated with si- and rBP5 as in C, 0.5 µM gliotoxin was

added 3 h before performing the caspase assay. All results are shown as a % of the control cells. E) Apoptosis was

induced in LX2 cells transfected with siCtrl or siBP5 by the addition of 0.5 uM gliotoxin at 45 h. Proteins were isolated at

48 h and Western blotting was performed for PARP detection (left panel). Quantification is shown in the right panel.

IGFBP5 also enhances survival of human myofibroblasts

Activated HSCs transdifferentiate into myofibroblasts in advanced stages of liver fibrosis. This

transdifferentiation results in a further induction of IGFBP5 expression (9). Given the effects on

LX2 cells, IGFBP5 may also affect the survival of liver myofibroblasts. In order to investigate this,

we performed siRNA mediated silencing of IGFBP5 expression in primary human myofibroblasts

(Figure 4A). Forty-eight hours after transfection, IGFBP5 mRNA levels in primary myofibroblasts

were decreased by 60%. Although the silencing was not as effective as that in LX2 cells, it did cause

a small, but significant, decrease in survival, which could be prevented by the addition of both rIGF1

and rIGFBP5 (Figure 4B). As in LX2 cells, this decrease in survival was due to an increase in caspase

activity that could be avoided by the addition of rIGF1 and rIGFBP5 (Figure 4C). Thus, IGFBP5

also increases the survival of human liver myofibroblasts by reducing apoptosis.

Involvement of BCL2

In order to investigate the mechanism involved in the anti-apoptotic effect of IGFBP5 we studied its

effect on the expression of several pro- and anti-apoptotic genes. The enhanced expression of

IGFBP5 in transduced cells resulted in a significant induction of B-cell CLL/lymphoma 2 (BCL2)

mRNA levels (Figure 5). Also, the addition of rIGFBP5 to IGFBP5 silenced LX2 cells resulted in a

clear increase in BCL2 expression. The lack of effect of IGFBP5 silencing on BCL2 expression is

unexpected and seems to suggest that, in presence of low levels of IGFBP5, other factors do

determine (basal) BCL2 expression. BCL2 mRNA induction by high IGFBP5 levels seems to suggest

IGFBP5 in LX2

39

that it could play a role in the anti-apoptotic effect of IGFBP5, but additional studies such as BCL2

silencing are needed in order to establish this possibility.

Figure 4. In human myofibroblasts IGFBP5 silencing decreases survival and promotes apoptosis. A) Primary

human myofibroblasts were transfected with siRNA for IGFBP5 (siBP5) or with a control siRNA (siCtrl). At 16 h after

transfection the medium was replaced by serum-free medium, and cultured for an additional 48 h. Then RNA was

isolated and IGFBP5 mRNA level determined by qPCR. B) Recombinant IGFBP5 (rBP5), IGF1 (rIGF1), or both

(rIGF1+ rBP5), where added at 24 h and 45 h, after replacing the medium, to siBP5 transfected cells. At 48 h the WST

assay was performed. C) To cells treated as above but grown in 10% FCS, 0,5 µM gliotoxin was added at 45 h. Caspase

3/7 assay was performed at 48 h. Results are given as a % of control cells.

Figure 5. IGFBP5 influences BCL2 expression. LX2 cells were

transfected with control (siCtrl) or IGFBP5 siRNA (siBP5). 16 h after

transfection the medium was replaced by serum-free medium.

Recombinant IGFBP5 was added at 24 h and 45 h, after replacing the

medium, to siIGFBP5 transfected cells (siBP5+rBP5). After 48 h RNA

was isolated from control (siCtrl), silenced (siBP5), silenced treated with

rIGFBP5 (siBP5+rBP5) and LX2 cells overexpression IGFBP5

(LX2(BP5)) and used for qPCRin order to determine BCL2 mRNA. The

BCL2 mRNA levels were given as % of control cells.

Increased cell survival by IGFBP5 seems independent of IGF1

IGFBP5 binds IGF1 with high affinity. Therefore, its protective role in LX2 cells and myofibroblasts

could be due to modulation of the pro-survival effects of IGF1 signalling (18, 34). In order to

investigate if IGFBP5 exerts its action via IGF1, we also determined the effect of IGF1 on LX2 cells

viability. As with IGFBP5, the addition of IGF1 induced a 70% increase in survival of LX2 cells

(Figure 6A), and lowering of apoptosis by 40% (Figure 6B). In order to investigate if IGF1 effect is

modulated by IGFBP5, we studied the effect of the two factors added together. The response in

viability and apoptosis were similar when IGF1 and IGFBP5 were added, alone or in combination

(Figure 6A and B). However, in contrast to IGFBP5, IGF1 was able to enhance proliferation of LX2

cells for about 40% (Figure 6C), indicating that the two factors exert different effects on LX2 cells.

chapter 2

40

In order to establish a possible IGF1-independent effect of IGFBP5, we chose to silence

IGF1R, the receptor that mediates IGF1 signalling. Transfection of LX2 cells with siRNA for

IGF1R resulted in a 90% drop of IGF1R protein expression after 48 h (Figure 6D). This large

decrease effectively inhibited the pro-survival effect of IGF1, but not that of IGFBP5 (Figure 6E).

The possibility of IGF1-independent action of IGFBP5 was further confirmed by the addition of

rIGFBP5 to IGF1R-silenced LX2 cells, which caused a decrease in caspase activity, even to the level

below that in the control cells. The addition of IGF1, on the other hand, was not able to overcome

the effect of IGF1R silencing and its effect on caspase activity was lacking (Figure 6F). Thus, in

contrast to IGF1, the effect of IGFBP5 on LX2 cells survival is not mediated by the IGF1R,

suggesting that the effect it exerts on partly activated stellate cells is not by modulation of IGF1

signalling.

Figure 6. Cell survival by IGFBP5 is not affected by IGF1 or mediated by IGF1R. A) LX2 cells were transfected

with IGFBP5 siRNA (siBP5). 16 h after transfection the medium was replaced by serum-free medium for 48 h and then

the viability was determined by WST assay. Recombinant IGFBP5 (rBP5), IGF1 (rIGF1), or both (rIGF1+ rBP5), were

added at 24 h and 45 h. B) To cells treated as above, but cultured in 10% FCS, 0.5 μM gliotoxin was added at 45 h

followed by a Caspase 3/7 assay 3 h later. C) LX2 cells were cultured in serum-free medium for 48 h. rIGF1 was added

in different concentrations at 24 h and 45 h, BrdU was added at 32 h. Cells were lysed 16 h later (at 48 h) and assayed for

BrdU incorporation. D) LX2 cells were transfected with control (siCtrl) or siRNA for IGF1R (siIGF1R). 16 h after

transfection the medium was replaced by serum-free medium and cells were cultured for additional 48 h, and then lysed

and used for Western blot analysis in order to detect IGF1R. E) Cells were transfected and cultured as in D. rIGF1 or

rIGFBP5 (rBP5) was added at 24 h and 45 h after replacing medium. After 48 h a WST assay was performed. F) 0.5 µM

gliotoxin was added at 45 h, after replacing the medium. Caspase activity was measured 3 h later. All results are given as a

% of control cells.

IGFBP5 and fibrotic markers

In tissues such as lung and skin, IGFBP5 was reported to enhance the expression of established pro-

fibrotic markers, such as αSMA, collagen and fibronectin. As changed expression of these markers

could further establish the (pro-fibrotic) role of IGFBP5 in liver fibrosis, we examined its influence

on their expression in LX2 cells. As shown in Figure 7, overexpression of IGFBP5 induced the

mRNA level of collagen, type I, alpha 1 (COL1A1) by 70%, tissue inhibitor of metalloproteinase

(TIMP) metallopeptidase inhibitor 1 (TIMP1) by 70% and matrix metallopeptidase 1 (MMP1) by

100% in LX2 cells (Figure 7A-C). Silencing of IGFBP5 reduced the expression of these genes, which

IGFBP5 in LX2

41

could be recuperated by the addition of recombinant IGFBP5. This indicates that IGFBP5 may not

only enhance fibrosis by promoting the survival of activated HSCs and myofibroblasts, but also by

stimulating the expression of pro-fibrotic genes in these cells.

Figure 7. IGFBP5 influences the expression of genes involved in fibrogenesis. LX2 cells were transfected with

small interfering control (siCtrl) or IGFBP5 siRNA (siBP5). Sixteen hours after transfection, medium was replaced by

serum-free medium. Recombinant IGFBP5 (rIGFBP5) was added at 24 h and 45 h, after replacing the medium, to

siIGFBP5 transfected cells (siBP5+rBP5). After 48 h of culturing on serum-free medium RNA was isolated from

control, transfected and transfected treated with rIGFBP5. For comparison, RNA was isolated from LX2 cells

overexpressing IGFBP5 cultured in serum-free medium (LX2(BP5)). A) Collagen 1α1, B) TIMP1 and C) MMP1 mRNA

levels were measured with quantitative polymerase chain reaction, normalized to 36B4 and given as a percentage of

control cells.

DISCUSSION

We have previously shown that IGFBP5 expression was strongly upregulated during HSCSs

transdifferentiation in vitro and during development of liver fibrosis in vivo (9). The aim of this study

was to scrutinize the role of IGFBP5 in (partially) activated and transdifferentiated hepatic stellate

cells.

IGFBP5 expression is increased in fibrotic lung, skin and liver (21). The effects of IGFBP5

depend on cell type and tissue. For instance, in the process of mammary gland involution, IGFBP5

promotes apoptosis of epithelial cells in vivo and in vitro (35). In contrast, an anti-apoptotic role of

IGFBP5 has been reported, for instance, in myogenesis, in breast cancer cells grown in vitro, and in

gingival epithelial cells (36-38).

In liver fibrosis, the level of apoptosis is reduced in activated HSCs that play a pivotal role in

the initiation and perpetuation of this pathological process (39). Since IGFBP5 is induced in these

cells upon activation and can promote survival, it may play a role in increasing the numbers of these

activated cells seen in fibrotic liver. In order to study the effects of IGFBP5 in vitro, we used primary

human myofibroblasts and LX2 cells, an established human model cell line for partially activated

HSCs (40). The expression of IGFBP5 in LX2 cells was 30% higher than in quiescent normal HSCs.

For comparison, in cultured (that is, partially activated) HSCs, IGFBP5 expression was doubled,

while in hepatic myofibroblasts expression went up more than 200-fold. We used lentiviral

transduction of LX2 cells to enhance the IGFBP5 expression to the levels seen in activated HSCs.

Upon serum depletion, both IGFBP5 over-expression and the addition of rIGFBP5 promoted the

survival of LX2 cells by lowering the apoptosis. This suggests that the increased IGFBP5 expression

chapter 2

42

could play a role in the reduction of apoptosis seen in activated HSCs in the fibrotic liver. In

accordance, lowered IGFBP5 expression by RNA silencing decreased the viability of LX2 cells.

IGFBP5 silencing also decreased survival in human primary liver myofibroblasts. In both cell types

this was due to increased apoptosis, demonstrating that IGFBP5 functions as an anti-apoptotic, pro-

survival factor in these two pro-fibrotic cell types. Two recent studies reported a high expression of

IGFBP5 in hepatocellular carcinoma (HCC) and intra-hepatic cholangiocarcinoma (CC), suggesting

that it has a similar role in vivo by promoting the survival of cancer cells [48, 49]. We demonstrated

that IGFBP5 expression strongly increased during development of liver fibrosis in Mdr2 -/- mice (9).

This model of liver fibrosis not only spontaneously develops portal fibrosis, but is also prone to

HCC formation (41). The presence of IGFBP5 in this animal model and in patients with HCC or CC

suggests that, in addition to a role in the pathogenesis of liver fibrosis, IGFBP5 may also contribute

to tumour formation. Lowering IGFBP5 expression may, therefore, not only impair the

development of liver fibrosis but also reduce the risk of tumour formation.

The role of IGFBP5 in IGF1 signalling is well established. IGFBP5 binds IGF1 with high

affinity and protects it from rapid degradation (18) but, at the same time, prevents induction of the

IGF1R and thereby inhibits pro-survival signalling by IGF1. The observed effects of IGFBP5 may

therefore be due to its role in IGF1 signalling – that is to its disturbance of the IGF1 axis observed

in liver fibrosis (17). Our in vitro studies show, however, that, in contrast to IGFBP5, IGF1 increases

proliferation of LX2 cells. Moreover, no effect of IGF1 on IGFBP5 action, nor of IGFBP5 on IGF1

signalling, was detected. Therefore, it seems that the two factors exert their effect on LX2 cells via

different routes. In line with this, silencing of IGF1R resulted in loss of the pro-survival effects of

IGF1 but did not interfere with the pro-survival effects of IGFBP5. This strongly supports the

notion that IGFBP5 promotes survival of activated HSCs by lowering apoptosis in an IGF1-

independent manner. Several mechanisms, such as activation of TGFβ1, modulation of the

coagulation cascade and integrin activation may be involved in the IGF1-independent effect of

IGFBP5 (42). In addition, the existence of an IGFBP5 receptor has been proposed to explain the

IGF1-independent actions of IGFBP5 (43, 44).

The finding that IGFBP5 promoted survival without inducing proliferation suggests that it

may induce cellular senescence. Overexpression of IGFBP5 is shown to induce senescence in

HUVEC cells by induction of the tumour suppressor p53 (25). The induction of senescence plays a

promoting role in the development of pulmonary fibrosis (23). On the other hand, senescence in

activated HSCs was associated with impediment of liver fibrosis (45), leaving the role of IGFBP5 in

the development of senescence in liver fibrogenesis opened for further inquiries. Additionally,

adenovirally mediated overexpression of IGFBP5 induced the expression pro-fibrotic genes and

ECM deposition in lung and skin (20, 21). We showed that IGFBP5 also enhanced the expression of

genes directly involved in liver fibrogenesis such as collagen1α1, TIMP1 and MMP1 in LX2 cells. Its

effect on these pro-fibrotic genes in a model for activated HSCs indicates that IGFBP5 may also

have a direct effect on the ECM deposition in liver fibrosis. Increase of both TIMP1 and MMP1

expression, and thus exertion of both fibrotic and matrix-degrading effects, could be seen in the light

of a necessity to establish a balance between matrix deposition and restructuring during the process

of fibrogenesis. However, since our data only demonstrate changes on the mRNA level, it seems

possible that the changes in the expression of these opposing factors may not have occurred on the

protein level due to posttranscriptional and posttranslational modifications.

IGFBP5 in LX2

43

In conclusion, our data show that IGFBP5 improves the survival of (partially) activated

HSCs and liver myofibroblasts by lowering the level of apoptosis via an IGF1-independent

mechanism. In addition, IGFBP5 increases the expression of genes involved in ECM deposition.

Accordingly, lowering expression of this apparently pro-fibrotic factor may impair the progression of

liver fibrosis.

References

1. Friedman SL. Liver fibrosis -- from bench to bedside. J Hepatol 2003;38 Suppl 1:S38-S53.

2. Blomhoff R, Wake K. Perisinusoidal stellate cells of the liver: important roles in retinol metabolism

and fibrosis. FASEB J 1991 Mar 1;5(3):271-277.

3. Wynn TA. Cellular and molecular mechanisms of fibrosis. J Pathol 2008 Jan;214(2):199-210.

4. Ikeda K, Wakahara T, Wang YQ, Kadoya H, Kawada N, Kaneda K. In vitro migratory potential of

rat quiescent hepatic stellate cells and its augmentation by cell activation. Hepatology 1999

Jun;29(6):1760-1767.

5. Kinnman N, Hultcrantz R, Barbu V, Rey C, Wendum D, Poupon R, et al. PDGF-mediated

chemoattraction of hepatic stellate cells by bile duct segments in cholestatic liver injury. Lab Invest

2000 May;80(5):697-707.

6. Marra F, Romanelli RG, Giannini C, Failli P, Pastacaldi S, Arrighi MC, et al. Monocyte chemotactic

protein-1 as a chemoattractant for human hepatic stellate cells. Hepatology 1999 Jan;29(1):140-148.

7. Gressner AM. Liver fibrosis: perspectives in pathobiochemical research and clinical outlook. Eur J

Clin Chem Clin Biochem 1991 May;29(5):293-311.

8. Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a modulator of hepatic fibrogenesis.

Semin Liver Dis 2001 Aug;21(3):351-372.

9. Boers W, Aarrass S, Linthorst C, Pinzani M, Elferink RO, Bosma P. Transcriptional profiling reveals

novel markers of liver fibrogenesis: gremlin and insulin-like growth factor-binding proteins. J Biol

Chem 2006 Jun 16;281(24):16289-16295.

10. van Nieuwkerk CM, Elferink RP, Groen AK, Ottenhoff R, Tytgat GN, Dingemans KP, et al. Effects

of Ursodeoxycholate and cholate feeding on liver disease in FVB mice with a disrupted mdr2 P-

glycoprotein gene. Gastroenterology 1996 Jul;111(1):165-171.

11. Gentilini A, Feliers D, Pinzani M, Woodruff K, Abboud S. Characterization and regulation of insulin-

like growth factor binding proteins in human hepatic stellate cells. J Cell Physiol 1998 Feb;174(2):240-

250.

12. Schneider MR, Wolf E, Hoeflich A, Lahm H. IGF-binding protein-5: flexible player in the IGF

system and effector on its own. J Endocrinol 2002 Mar;172(3):423-440.

13. Jones JI, Clemmons DR. Insulin-like growth factors and their binding proteins: biological actions.

Endocr Rev 1995 Feb;16(1):3-34.

14. Twigg SM, Baxter RC. Insulin-like growth factor (IGF)-binding protein 5 forms an alternative ternary

complex with IGFs and the acid-labile subunit. J Biol Chem 1998 Mar 13;273(11):6074-6079.

15. Scharf JG, Schmitz F, Frystyk J, Skjaerbaek C, Moesus H, Blum WF, et al. Insulin-like growth factor-I

serum concentrations and patterns of insulin-like growth factor binding proteins in patients with

chronic liver disease. J Hepatol 1996 Nov;25(5):689-699.

16. Conchillo M, de Knegt RJ, Payeras M, Quiroga J, Sangro B, Herrero JI, et al. Insulin-like growth

factor I (IGF-I) replacement therapy increases albumin concentration in liver cirrhosis: results of a

pilot randomized controlled clinical trial. J Hepatol 2005 Oct;43(4):630-636.

chapter 2

44

17. Sanz S, Pucilowska JB, Liu S, Rodriguez-Ortigosa CM, Lund PK, Brenner DA, et al. Expression of

insulin-like growth factor I by activated hepatic stellate cells reduces fibrogenesis and enhances

regeneration after liver injury. Gut 2005 Jan;54(1):134-141.

18. Beattie J, Allan GJ, Lochrie JD, Flint DJ. Insulin-like growth factor-binding protein-5 (IGFBP-5): a

critical member of the IGF axis. Biochem J 2006 Apr 1;395(1):1-19.

19. Allen JT, Spiteri MA. Growth factors in idiopathic pulmonary fibrosis: relative roles. Respir Res

2002;3:13.

20. Yasuoka H, Jukic DM, Zhou Z, Choi AM, Feghali-Bostwick CA. Insulin-like growth factor binding

protein 5 induces skin fibrosis: A novel murine model for dermal fibrosis. Arthritis Rheum 2006

Sep;54(9):3001-3010.

21. Yasuoka H, Zhou Z, Pilewski JM, Oury TD, Choi AM, Feghali-Bostwick CA. Insulin-like growth

factor-binding protein-5 induces pulmonary fibrosis and triggers mononuclear cellular infiltration.

Am J Pathol 2006 Nov;169(5):1633-1642.

22. Pilewski JM, Liu L, Henry AC, Knauer AV, Feghali-Bostwick CA. Insulin-like growth factor binding

proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular

matrix deposition. Am J Pathol 2005 Feb;166(2):399-407.

23. Allan GJ, Beattie J, Flint DJ. Epithelial injury induces an innate repair mechanism linked to cellular

senescence and fibrosis involving IGF-binding protein-5. J Endocrinol 2008 Nov;199(2):155-164.

24. 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 Jan;54(1):142-151.

25. Kim KS, Seu YB, Baek SH, Kim MJ, Kim KJ, Kim JH, et al. Induction of cellular senescence by

insulin-like growth factor binding protein-5 through a p53-dependent mechanism. Mol Biol Cell 2007

Nov;18(11):4543-4552.

26. Seppen J, Rijnberg M, Cooreman MP, Oude Elferink RP. Lentiviral vectors for efficient transduction

of isolated primary quiescent hepatocytes. J Hepatol 2002 Apr;36(4):459-465.

27. Ramakers C, Ruijter JM, Deprez RHL, Moorman AFM. Assumption-free analysis of quantitative real-

time polymerase chain reaction (PCR) data. Neuroscience Letters 2003 Mar 13;339(1):62-66.

28. Keijzer R, Blommaart PJ, Labruyere WT, Vermeulen JL, Doulabi BZ, Bakker O, et al. Expression of

thyroid hormone receptors A and B in developing rat tissues; evidence for extensive

posttranscriptional regulation. J Mol Endocrinol 2007 May;38(5):523-535.

29. Johnson SK, Haun RS. Insulin-like growth factor binding protein-5 influences pancreatic cancer cell

growth. World J Gastroenterol 2009 Jul 21;15(27):3355-3366.

30. Meadows KA, Holly JM, Stewart CE. Tumor necrosis factor-alpha-induced apoptosis is associated

with suppression of insulin-like growth factor binding protein-5 secretion in differentiating murine

skeletal myoblasts. J Cell Physiol 2000 Jun;183(3):330-337.

31. Miyake H, Nelson C, Rennie PS, Gleave ME. Overexpression of insulin-like growth factor binding

protein-5 helps accelerate progression to androgen-independence in the human prostate LNCaP

tumor model through activation of phosphatidylinositol 3'-kinase pathway. Endocrinology 2000

Jun;141(6):2257-2265.

32. Perks CM, McCaig C, Holly JM. Differential insulin-like growth factor (IGF)-independent

interactions of IGF binding protein-3 and IGF binding protein-5 on apoptosis in human breast

cancer cells. Involvement of the mitochondria. J Cell Biochem 2000 Oct 20;80(2):248-258.

33. Hagens WI, Olinga P, Meijer DK, Groothuis GM, Beljaars L, Poelstra K. Gliotoxin non-selectively

induces apoptosis in fibrotic and normal livers. Liver Int 2006 Mar;26(2):232-239.

34. Butt AJ, Firth SM, Baxter RC. The IGF axis and programmed cell death. Immunol Cell Biol 1999

Jun;77(3):256-262.

IGFBP5 in LX2

45

35. Flint DJ, Boutinaud M, Tonner E, Wilde CJ, Hurley W, Accorsi PA, et al. Insulin-like growth factor

binding proteins initiate cell death and extracellular matrix remodeling in the mammary gland.

Domest Anim Endocrinol 2005 Aug;29(2):274-282.

36. Cobb LJ, Salih DA, Gonzalez I, Tripathi G, Carter EJ, Lovett F, et al. Partitioning of IGFBP-5

actions in myogenesis: IGF-independent anti-apoptotic function. J Cell Sci 2004 Apr 1;117(Pt

9):1737-1746.

37. Perks CM, Bowen S, Gill ZP, Newcomb PV, Holly JM. Differential IGF-independent effects of

insulin-like growth factor binding proteins (1-6) on apoptosis of breast epithelial cells. J Cell Biochem

1999 Dec 15;75(4):652-664.

38. McCaig C, Perks CM, Holly JM. Intrinsic actions of IGFBP-3 and IGFBP-5 on Hs578T breast cancer

epithelial cells: inhibition or accentuation of attachment and survival is dependent upon the presence

of fibronectin. J Cell Sci 2002 Nov 15;115(Pt 22):4293-4303.

39. Iredale JP. Hepatic stellate cell behavior during resolution of liver injury. Semin Liver Dis 2001

Aug;21(3):427-436.

40. Wang XM, Yu DM, McCaughan GW, Gorrell MD. Fibroblast activation protein increases apoptosis,

cell adhesion, and migration by the LX-2 human stellate cell line. Hepatology 2005 Oct;42(4):935-945.

41. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice

with homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of

nonsuppurative inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994

Nov;145(5):1237-1245.

42. Andress DL. Comparison studies of IGFBP-5 binding to osteoblasts and osteoblast-derived

extracellular matrix. Prog Growth Factor Res 1995;6(2-4):337-344.

43. Andress DL. Heparin modulates the binding of insulin-like growth factor (IGF) binding protein-5 to

a membrane protein in osteoblastic cells. J Biol Chem 1995 Nov 24;270(47):28289-28296.

44. Andress DL. Insulin-like growth factor-binding protein-5 (IGFBP-5) stimulates phosphorylation of

the IGFBP-5 receptor. Am J Physiol 1998 Apr;274(4 Pt 1):E744-E750.

45. Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, Miething C, et al. Senescence of activated

stellate cells limits liver fibrosis. Cell 2008 Aug 22;134(4):657-667.

c h a p t e r 3

unexpected reduction of

murine liver fibrosis by igfbp5

chapter 3

48

ABSTRACT

Background: The ATP-binding cassette, sub-family B member 4 knock-out mouse (Abcb4-/-) is a

relevant model for chronic cholangiopathy in man. Due to the lack of this P-glycoprotein in the

canalicular membrane of hepatocytes, the secretion of phospholipids into bile is absent, resulting in

increased bile toxicity. Expression of insulin like growth factor binding protein 5 (Igfbp5) increases in

time in the livers of these mice. It is unclear whether this induction is a consequence of or plays a

role in the progression of liver pathology. Aim of this study was therefore to investigate the effect of

IGFBP5 induction on the progression of liver fibrosis caused by chronic cholangiopathy.

Methods: IGFBP5 and, as a control, green fluorescent protein were overexpressed in the

hepatocytes of Abcb4-/- mice, using an adeno-associated viral vector (AAV). Progression of liver

fibrosis was studied 3, 6, and 12 weeks after vector injection by analyzing serum parameters, collagen

deposition, expression of pro-fibrotic genes, inflammation and oxidative stress.

Results: A single administration of the AAV vectors provided prolonged expression of IGFBP5 and

GFP in the livers of Abcb4-/-mice. Compared to GFP control, fractional liver weight, ECM

deposition and amount of activated HSCs significantly decreased in IGFBP5 overexpressing mice

even 12 weeks after treatment. This effect was not due to a change in bile composition, but driven by

reduced inflammation, oxidative stress, and proliferation.

Conclusion: Overexpression of IGFBP5 seems to have a protective effect on liver pathology in this

model for chronic cholangiopathy.

Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease

Volume 1822, Issue 6, June 2012, Pages 996–1003

Authors

Aleksandar Sokolović1, Paula S. Montenegro-Miranda1, Dirk Rudi de Waart1, Radha M.N. Cappai1,

Suzanne Duijst1, Milka Sokolović2 and Piter J. Bosma1

Affiliations

1 Tytgat Institute for Liver and Intestinal Research, 2 Department of Medical Biochemistry,

Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands

Acknowledgments

We are indebted to Profs. M. Kotb and R.W. Moyer for permission to use their unpublished

phenotypic data deposited in the GeneNetwork database. We thank Prof. Dr. R.P.J. Oude Elferink

for constructive discussions, Dr. J. Ruijter for help with quantification in Scion Image, and J. K.

Hiralall, L. ten Bloemendaal and C. Kunne for technical assistance.

IGFBP5 in Abcb4-/-

49

INTRODUCTION

Abcb4-/- (ATP-binding cassette, subfamily b, member 4) mice are the model for progressive familial

intrahepatic cholestasis type 3. In humans, ABCB4 deficiency results in spontaneous development of

biliary fibrosis shortly after birth, due to the lack of biliary phospholipids essential for neutralizing

the toxic effects of the high concentrations of bile acids (1). Since murine bile acids are less toxic, the

phenotype is milder in mice than in humans, which also qualifies Abcb4-/- mice as a model for other

forms of progressive cholangiopathies, such as progressive sclerosing cholangitis (2, 3). Recurrent

chronic injury in this mouse model, caused by leakage of toxic bile into the portal area leads to

progressive accumulation of extracellular matrix (ECM) components, mainly produced by activated

hepatic stellate cells (HSCs) in an attempt to counter hepatic damage (4-6). This results in an

imbalance between fibrogenesis and fibrolysis, leading to liver fibrosis, architectural distortion,

cirrhosis and ultimately liver failure (7, 8).

Insulin like growth factor binding protein 5 (Igfbp5) is strongly induced in the livers of Abcb4-

/- mice suggesting a potential role in the pathogenesis of chronic cholangiopathy (9). This seems to

be supported by its high expression in patients with cholangiocarcinoma (10). We have shown

previously that overexpression of IGFBP5 increases the survival of partially activated HSCs and

myofibroblasts (MF) (11). This pointed to its profibrotic role in liver, as also reported for skin and

lung where IGFBP5 overexpression stimulates the progression of fibrosis (12, 13).

To clarify the role of IGFBP5 in the pathogenesis of chronic cholangiopathies, we overexpressed it

in the Abcb4-/- mice, using an adeno-associated viral (AAV) vector. This in vivo AAV delivery that

efficiently provides expression in hepatocytes, should result in the supply of large amounts this

excretory protein to the intracellular space and to the HSCs (14).To substantiate the experimental

findings in this model, and to relate them to the physiological role of hepatic IGFBP5 expression, a

systems genetics approach was applied to a mouse genetic reference population, using the open

source GeneNetwork database. The results have pointed to a beneficial role of IGFBP5 in liver

fibrosis caused by chronic cholangiopathy in Abcb4-/- mouse model.

MATERIALS AND METHODS

scAAV vectors

scAAV8-LP1-Igfbp5 and scAAV8-LP1-eGFP were generated by replacing the factor IX cDNA from

scAAV-LP1-IX by the cDNA of mouse Igfbp5 or eGFP, using the EcoR1 and Bspm1 sites (Figure

1A) (15). Correctness of constructs was checked by sequencing. scAAV8 pseudotyped vector was

produced and titrated as described (16).

Animals

The animal studies were reviewed and approved by the AMC committee for animal care and use.

Three months after birth, a single dose of 5x1012 genomic copies (gc)/kg body weight of scAAV8-

LP1- Igfbp5 or scAAV8-LP1-eGFP was injected in the tail vein of male FVB and Abcb4-/- mice (FVB

background). Animals were sacrificed 3 (n=5), 6 (n=7) and 12 (n=7) weeks thereafter. In FVB

controls (n=5) the effect was studied only after 3 weeks. Five animals per treatment group were used

at 3 weeks, and 7 animals per group at weeks 6 and 12.

chapter 3

50

Tissue collection and biochemical analyses

Upon sacrificing, blood was collected, livers were snap-frozen in liquid nitrogen and stored at -80oC,

or fixed in 4% formaldehyde in PBS. Total liver collagen was determined by measuring

hydroxyproline content as described (17). GSH/GSSG ratio in liver and bile was determined (18).

ASAT, ALAT and alkaline phosphatase (ALP) concentrations in blood were determined using

standard clinical chemistry methods.

RNA isolation, reverse transcription and quantitative PCR

These analyses were performed and analyzed as previously described (11). We used 5–7 animals for

qRT-PCR, depending on the time point—5 animals per group at 3 weeks, and 7 animals per group at

6 and 12 weeks.

Western blot analysis

They were performed as described (11), using the following antibodies: IGFBP5 (Abcam,

Cambridge, UK), GFP, PCNA and p21 (Santa Cruz biotechnology, CA, USA). Goat anti-mouse

(Bio-Rad Laboratories, Veenendaal, The Netherlands) and goat anti-rabbit IgG horseradish

peroxidase conjugated (Santa Cruz Biotechnology, CA, USA) were used as secondary antibodies.

Chemiluminescence was quantified by Lumi-Imager F1, using CDP-Star (Roche, Mannheim,

Germany) and β-actin as a reference. For detection of IGFBP5 in serum, the same amount of (10x

diluted) serum was loaded for each sample.

Histology and immunohistochemistry

Formaldehyde-fixed liver sections were stained with Sirius red (morphometrical connective tissue

assessment), or immunohistochemically, as described (19). Antibodies directed against PCNA (Santa

Cruz biotechnology, CA, USA) and α-SMA (Sigma, Zwijndrecht, The Netherlands) were visualized

with goat anti-mouse or goat anti-rabbit IgG coupled to horse-radish peroxidase (Sigma,

Zwijndrecht, The Netherlands), or alkaline phosphatase, respectively. Anti-α-F4/80 (AbDSerotec,

Dusseldorf, Germany) and anti-MAC-1 (R&D systems, Abingdon UK) were both visualized using

rabbit-anti-rat-biotin as secondary antibody (DAKO, Glostrup, Denmark).

Correlation analyses and QTL interval mapping in GeneNetwork

These analyses were performed as described (20). Analysis of Igfbp5 expression in liver of the BxD

genetic reference population. The BxD mouse genetic reference population (GRP) is a set of

recombinant inbred (RI) mouse strains, derived by crossing F2 mice obtained from an intercross of

C57BL/6J (B6) and DBA/2J (D2) mice, and inbreeding the resulting progeny for at least 20

generations to generate the BxD GRP (21, 22). RI panels, comprised of a number of lines with

limited intra- and large inter-line phenotypic variation, are widely used to determine genotype-

phenotype associations with quantitative trait locus (QTL) mapping techniques, or eQTL mapping,

when mRNA levels are used as phenotypes (23-27). The relationships between the phenotypes and

genotypes are calculated with a likelihood ratio statistic (LRS), a measure of the probability that a

given genetic marker explains the variation in the phenotype. This genetic reference panel contains

data for hundreds of phenotypic traits, including clinical and molecular ones (i.e. microarray

expression studies in multiple tissues) acquired over 25-years. The data are publically available at

GeneNetwork (www.genenetwork.org). The web-based software further allows extraction of sets of

IGFBP5 in Abcb4-/-

51

transcripts or clinical phenotypes that tightly co-vary with a gene of interest across genetically

different populations. Patterns of co-variations can be studied in multiple experimental crosses and

multiple tissues.

For this study we used data from BxD published phenotypes database

(www.genenetwork.com), and gene expression data from:

liver (http://www.genenetwork.org/dbdoc/LV_G_0106_B.html),

kidney (http://www.genenetwork.org/dbdoc/MA_M2_0806_R.html)

brain (http://www.genenetwork.org/dbdoc/BR_M2_1106_R.html).

Statistical analysis

ANOVA analysis and Student’s t-test were used. The error bars represent the standard deviation, and

significance threshold was set to P<0.05.

RESULTS

Figure 1: Overexpression of IGFBP5 in liver using AAV. A) Time course and genome structure of scAAV-LP1-Igfbp5 and scAAV-LP1-GFP. B) Western blot of GFP (left panel) and IGFBP5 (30kD; right panel) present in liver homogenates obtained 3, 6 and 12 weeks after injection, and of IGFBP5 in plasma after 3 weeks (right panel). In both panels, the GFP-overexpressing animals are on the left (“g”) and IGFBP5-overexpressing mice (“i”) on the right side. For detection, the Western blots in the left panel were incubated with anti-GFP, while those in the right panel were incubated with an anti-IGFBP5 antibody. Every lane represents an individual animal. In total, 4–6 animals of each treatment group were analyzed. C–D) Relative gene expression of Igfbp5 in Abcb4−/− mice 3, 6 and 12 weeks after administration of scAAV-LP1-Igfbp5 (right panel) or scAAV-LP1-GFP as a control (left panel) (n=5-7). Asterisks denote significance (p<0.05) in comparison with control treatment.

IGFBP5 overexpression reduces hepatocyte damage and proliferation in Abcb4-/- mice

To investigate the effect of IGFBP5 on the liver pathogenesis in Abcb4-/-mice, its expression

was increased in hepatocytes using the liver specific scAAV8-LP1-IGFBP5 vector and scAAV8-LP1-

GFP as a control (Figure 1A). qPCR and Western blot confirmed the hepatic overexpression of both

chapter 3

52

transgenes 3, 6 and 12 weeks after injection (Figure 1B). During this period, the expression declined,

but was still significantly increased for Igfbp5 at 12 weeks (~15-fold) compared to mock transduction

(Figure 1C). Because of the ongoing hepatocyte proliferation in the Abcb4-/- mice this loss of

episomal AAV vectors was expected (28, 29). The lower expression of GFP seen at 3 weeks after

injection while the vector dose was equal, suggested a more rapid decline of GFP expression during

this period, which could be due to a higher hepatocyte proliferation rate in these control mice

(Figure 1D).

Therefore, the expression of proliferation markers PCNA and Ki67 was determined (Figure

2A-C). The 2-3 fold lower PCNA expression level after three weeks, and the lower expression level

of Ki67 at both later time points, confirmed that proliferation was reduced in mice overexpressing

IGFBP5.

Figure 2: IGFBP5 overexpression decreases liver cell proliferation in Abcb4-/- mice. A) Representative PCNA

immunostaining in liver at 3 weeks after injection of scAAV-LP1-GFP (control) or scAAV-LP1-Igfbp5 (IGFBP5). B)

Quantification of PCNA expressing nuclei in control (GFP) or IGFBP5 overexpressing mice per mm2 at 3 weeks after

treatment. C) mRNA level of Ki67 in liver after 3, 6 and 12 weeks determined by qPCR expressed as % of the control.

Asterisks denote significance (p<0.05).

This may be due to enhanced expression of p21, involved in repair of DNA damage by

arresting cell cycle and proliferation. Both p21 mRNA and protein expression level were significantly

upregulated in Igfbp5 injected animals (Figure 3A-B). This was accompanied by an increased presence

of senescence-associated β-galactosidase (a widely used biomarker for senescent and aging cells),

especially in fibrotic regions (Figure 3C). Lower ALP (alkaline phosphatase) and ASAT (aspartate

transaminase) levels in these mice 6, 9 and 12 weeks after injection, suggested that IGFBP5 alleviates

hepatocyte damage (Figure 3D–E).

IGFBP5 in Abcb4-/-

53

Figure 3: IGFBP5 overexpression reduces hepatocyte damage in Abcb4-/- mice. A) Gene expression of p21 in liver after 3, 6 and 12 weeks after vector administration. B) p21 level 12 weeks after vector administration as detected by Western blot, and quantified as percentage of the control (GFP). C) Representative SA-β-gal staining of control (GFP) and IGFBP5-treated animals 12 weeks after vector injection. D–E) Plasma concentrations of ALP and ASAT in GFP- and IGFBP5-treated Abcb4−/−mice 1, 3, 6, 9, and 12 weeks upon vector administration.

Overexpression of IGFBP5 reduces liver fibrosis in Abcb4-/-mice

A protective role of IGFBP5, as suggested by decreased hepatocyte proliferation and lowered liver

enzymes in serum, may also slow down the progression of fibrosis. The 20% lowered fractional liver

weight, the reduced presence of ECM shown by hydroxyproline content (Figure 4A-B) and Sirius red

staining (Figure 4C) all confirmed that IGFBP5 overexpression impairs the progression of liver

fibrosis in this model.

Figure 4: IGFBP5 overexpression

reduces liver fibrosis in Abcb4-/- mice.

A) Average fractional liver weight of

control and IGFBP5 overexpressing mice 6

and 12 weeks after vector administration.

B) Hydroxyproline amount in liver 6 and

12 weeks after viral treatment. C)

Representative Sirius red staining in control

(GFP) and IGFBP5 overexpressing livers

12 weeks after experiment started.

chapter 3

54

Figure 5: IGFBP5 overexpression decreases expression of fibrotic markers in Abcb4-/- mice. A) Representative α-SMA staining in control (GFP; left and middle panel; 5X and 20X magnified respectively) and IGFBP5 overexpressing livers (right panel) 12 weeks after vector administration. B) Number of α-SMA expressing cells per mm2 in liver 12 weeks after vector administration as determined by a custom written counting macro in Scion Image (Scion Corp, Frederick, MD). C–F) Relative mRNA level for procollagen, collagens 1, 3, 4 and Timp1, after 3, 6 and 12 weeks. All results are given as a percentage of the control. Asterisks denote significance (p < 0.05).

This was supported by 70% reduction of α-SMA (marker for activated HSCs/MFs)

expressing HSCs/MFs 12 weeks after vector administration (Figure 5A-B), and, as a consequence,

reduced expression of procollagen, collagens 1, 3, 4 and Timp1 (Figure 5C-G).

Since the pathogenesis in the Abcb4-/- mouse results from toxic bile, an effect of IGFBP5 on

bile composition could provide an alternative explanation of its protective role in this model. The

absence of significant differences in bile composition 12 weeks after vector administration however

renders this explanation highly unlikely (Table 1).

Table 1: Bile salt composition of GFP control and IGFBP5-treated Abcb4-/- mice and FVB.

FVB Abcb4-/- ctrl Abcb4-/- BP5

tauro β-muricholate 223.52 ± 67.90 277.96 ± 142.05 253.98 ± 172.18 taurocholate 94.97 ± 22.84 145.85 ± 68.17 180.53 ± 162.50 tauro α-muricholate 1.92 ± 0.84 6.73 ± 4.86 9.33 ± 7.36 tauroursodeoxycholate 0.35 ± 0.14 0.74 ± 0.36 1.18 ± 0.89 taurochenodeoxycholate 1.56 ± 0.59 4.36 ± 1.66 5.55 ± 3.14 ∑ (bile salts) 322.32 ± 85.73 435.65 ± 209.62 450.58 ± 334.74

IGFBP5 overexpression reduces inflammation and oxidative stress in the liver of Abcb4-/-

mice

IGFBP5 has been reported to impair the influx of proinflammatory cells (30). Influx of these cells

into the liver will result in the release of inflammatory mediators that enhance fibrosis and cause

oxidative stress (31). Inhibition of this influx by IGFBP5 may therefore explain the protective effect

observed in this model. To investigate this, markers for inflammatory cells were studied. Presence of

F4/80, a marker for infiltrating macrophages and liver-resident Kupffer cells, was significantly lower

IGFBP5 in Abcb4-/-

55

6 and 12 weeks (mRNA) and 12 weeks (protein) after vector administration (Figure 6A-C).

Expression of Mpo, a marker for infiltrating neutrophils, was reduced at all three time points (Figure

6B). Furthermore, the presence of MAC1-expressing activated macrophages was clearly reduced in

IGFBP5 overexpressing animals (Figure 6D). In addition, the expression of Mcp1, which plays a role

in the recruitment of monocytes to sites of injury and infection, was reduced 3 and 6 weeks upon

IGFBP5 administration (Figure 6E) Finally, Tgfβ1 mRNA level was also reduced in Abcb4-/- mice

overexpressing IGFBP5 (Figure 6F).

Figure 6: IGFBP5 reduces expression of inflammatory markers. A, B, E and F) Relative mRNA levels of F4/80,

Mpo, Mcp1 and Tgfβ1, respectively, in livers of control (GFP) and IGFBP5 overexpressing mice, 3, 6 and 12 weeks

upon vector administration. All results are given as a percentage of the control (GFP). Asterisks denote significance (P <

0.05). C and D) Representative pictures of immunostainings for F4/80 and MAC1, respectively, 12 weeks after vector

administration.

The reduced influx caused by IGFBP5 overexpression will also lower the release of

proinflammatory cytokines in liver and as such may explain the lower the oxidative stress in these

chapter 3

56

livers as indicated by the significant increased ratio of glutathione (GSH) and its oxidized form

glutathione-disulfide (GSSG; Figure 7A). In support, mRNA expression level of heme oxygenase 1

was significantly upregulated upon injection of IGFBP5 (Figure 7B)

Figure 7: IGFBP5 overexpression lower the oxidative stress. A) The GSH/GSSG ratio determined at 6 weeks (liver) and 12 weeks (liver and bile). B) Relative mRNA level for HO-1 at 6 and 12 weeks. All results are given as a percentage of the control. Asterisks denote significance (P < 0.05). C) Absence of anti-GFP antibodies in plasma of GFP- and IGFBP5-expressing mice demonstrated by an ELISA.

Specific expression of GFP in hepatocytes did not induce an immune response, as shown by

the lack of anti-GFP antibodies in serum of control mice and the comparable degree of liver fibrosis

in non-treated and GFP-treated Abcb4-/-mice (Figure 7C). This confirms previous findings of our

group and others that hepatocyte specific expression provided by an AAV vector does not result in

an immune response towards the transgene (16).

Altogether these data support an anti-inflammatory effect of IGFBP5 in the liver of Abcb4-/- mice.

Co-variation of Igfbp5 expression with phenotypic traits in a mouse genetic reference

population supports its anti-inflammatory role

To gain insight into the physiological role of Igfbp5 expression in adult liver, its co-variation with

published phenotypic traits in the BxD genetic reference population (GRP) was analyzed. This

experimental model that incorporates the natural range of genetic variation is obtained by 20

generation inbreeding in the F2 mice from an intercross between C57BL/6J and DBA/2J strains

(www.genenetwork.org) (24, 26).The analysis revealed a significant negative correlation between liver

Igfbp5 expression and plasma concentrations of ALAT and ASAT (Figure 8A and B) (32). These

relationships underline a hepato-protective role of IGFBP5. The negative correlation between Timp1

and Igfbp5 expression in the BxD GRP livers (5) provides further support for a possible physiological

role of IGFBP5 in ECM remodeling (Figure 8C). In the reference population, the expression of pro-

fibrotic cytokine Tgfβ1 positively correlated with that of F4/80 (Figure 8D), which corroborated the

reduction of inflammatory response by IGFBP5.

Furthermore, the correlation of IGFBP5 expression in liver of BxD GRP with clinical

phenotypes was calculated. A number of phenotypes were related to immune function

(Supplementary file 2 in a published manuscript), which was consistent with the anti-inflammatory

effect of IGFBP5 in the liver of Abcb4-/- mice. Igfbp5 expression in the livers of BxD mice negatively

IGFBP5 in Abcb4-/-

57

correlated with inflammation in spleen, bone marrow and with a general inflammatory response, the

exception being a positive correlation with TBC infection in the lung (Figure 8E).

Figure 8: Igfbp5 expression in liver correlates with suppressed immune response in the BxD reference population. Correlation between Igfbp5 mRNA in liver and ASAT A) and ALAT B) serum levels (n=7) as found in (GeneNetwork database). C) Correlations between Igfbp5 and Timp1 mRNA levels liver of BxD recombinant inbred strains (as determined in GeneNetwork database). D) Expression of Tgfβ in liver correlates with that of F4/80 in BxD genetic reference population. E) A correlation network between Igfbp5 expression level (blue rectangle) in livers of BxD mice and immune-linked phenotypic features measured in the same animals. The phenotypic features (depicted in green rectangles) are as follows: “IrradSpleen” (10407)- Antigenic activity of irradiated BXD spleen cells for Thy-1+CD3+CD4+CD8- T-cell clone E4.3 [Thymidine uptake cpm]; (33); “CpvInDMCs” (12678) - Infectious disease, immune function: Cowpox virus measure of disease onset measured by the day of maximum clinical score over two weeks after 10^6 pfu intranasal inoculation (Rice AD, Moyer RW, 2010, unpublished data); “ProgCellMob” (10206)- Immune function: Granulocyte colony stimulating factor (G-CSF) induced progenitor cell mobilization from bone marrow to blood [PB-CFC/ul]; (34); “BCGPulmGranInflm” (10695) - Immune function: Lung inflammatory response, pulmonary granulomatous inflammation induced by BCG [units] (35).

Tissue specific regulation of Igfbp5 expression

Given the unanticipated effect of IGFBP5 on liver fibrosis, we addressed the tissue-specificity of its

expression. Notably, a direct analysis of variation in Igfbp5 mRNA levels in different organs - liver,

kidney, (whole) brain, hippocampus and cartilage - did not reveal any correlation (not shown),

underscoring organ specific regulation of Igfbp5 expression. The correlations between Igfbp5

expression in different organs and phenotypic features in the BxD GRP were distinct for liver,

kidney and brain, (as published in Supplementary file 2). Unsurprisingly, the variation of Igfbp5

expression in other two organs did not correlate with ALAT and ASAT levels in serum.

The systems genetic approach was further employed to identify the quantitative trait loci

(QTLs) that specifically modulate the expression of Igfbp5 in mouse liver. No cisQTLs were

determined on Chr.1, where Igfbp5 resides (Figure 9A). A putative trans-regulatory region that

modulates Igfbp5 expression in liver was identified on Chr.7. The QTL peak (135-142Mb) comprised

chapter 3

58

83 positional candidates that may modulate hepatic Igfbp5 expression (Figure 9B, as published in

Supplementary table 1).

Figure 9: Interval mapping, a WebQTL analysis of upstream controllers of liver Igfbp5 expression. A) The X-axis represents 19 mouse autosomes and the X chromosome. The small purple triangle marks the location of the Igfbp5 gene. The thick blue line summarizes the strength of association between variation in Igfbp5 expression and the two genotypes of all markers and the intervals between markers. The height of the peak (primary Y-axis) provides the likelihood ratio statistic (LRS; which can be read as a chi-square-like statistic). The additive coefficient (green line, secondary Y-axis) indicates that DBA/2J alleles increase trait values. In contrast, a negative additive coefficient (red line) indicates that C57BL/6J alleles increase trait values. The yellow histogram bars summarize the results of a whole-genome bootstrap of the trait performed 1000 times, supporting thereby the robustness of the results. The horizontal dashed lines at 10.5 and 17.3 are the LRS values associated with the suggestive and significant genome-wide probabilities that were established by permutations of phenotypes across genotypes. B) Physical map of variation in Igfbp5 expression in liver on distal Chr 7 (a blow up of the whole-genome map in panel A). A physical scale in megabases (Mb) is shown on the X-axis. The small irregular colored blocks and marks at the top of the map indicate the locations of genes superimposed on the physical map. The orange hash marks along the X-axis represent the number of single nucleotide polymorphisms that distinguish the two parental strains (C57BL/6J and DBA/2J) from each other.

To narrow them down to a biologically relevant subset, genes were scrutinized by analyzing

their co-expression, by QTL mapping (to establish if they were cisQTLs per se), by examining the

number of SNPs, and by studying the literature. Among the four most promising candidates (Tacc2,

Nkx1-2, Zranb1 and Ctbp) Nkx1-2 appeared the most likely one, since it was a cisQTL, it was

modulated by the same transQTL on Chr.14 that possibly also regulates the expression of Igfbp5, and

IGFBP5 in Abcb4-/-

59

its expression positively correlated with that of Igfbp5 (Figure 10A-C). The almost linear relationship

between Nkx1-2 expression in the liver and the survival coefficient in infectious diseases suggests its

potential important role in orchestrating inflammatory responses, possibly including Igfbp5

expression (Figure 10D; (Kotb M., 2010, unpublished data)). Although this analysis reveals Nkx1 as an

interesting candidate, the role of other candidate genes cannot be excluded.

Figure 10: NKX1-2 is a likely modulator of Igfbp5 expression in liver. A) Interval mapping for Nkx1-2. (For explanation see Supplemtary figure 6.) B) QTL heatmap for Chrs. 1, 7 and 14 for Igfbp5 (left column) and Nkx1-2 (right column). The yellow triangles indicate the location of the two genes (Igfbp5 on Chr.1 and Nkx1-2 on Chr.7). The intensity of color indicates the likelihood ratio statistic (LRS) values associated with the suggestive and significant genome-wide probabilities established by permutations of phenotypes across genotypes for both of the parental strains. C) Correlation of Igfbp5 and Nkx1-2 expression in BxD GRP. D) Correlation between the expression level of Nkx1-2 and the coefficient of survival in infectious diseases (M. Kotb, 2010, unpublished data).

DISCUSSION

The aim of this study was to investigate the role of IGFBP5 in the liver pathogenesis in the Abcb4-/-

mice, a model for chronic cholangiopathy. Prolonged overexpression of IGFBP5 in hepatocytes

decreased their proliferation, reduced inflammation and oxidative stress, and lowered number of

activated HSCs, all leading to reduced development of liver fibrosis.

IGFBP5 has cell- and tissue-type specific effects. It induces apoptosis in the involuting

mammary gland (36) and inhibits cell proliferation in the involuting prostate gland (37), but supports

the survival of skeletal myoblast (38). To clarify the differential role and expression of Igfbp5 in liver

and other organs, its steady-state expression was analyzed in the BxD genetic reference population.

chapter 3

60

The correlations with clinical features not only indicated an organ specific regulation of IGFBP5, but

also a role that differed between organs. This seems to explain why IGFBP5 was found a profibrotic

factor in lung and skin (12, 13), while the current study revealed its protective effect in liver fibrosis.

The latter is underscored by the negative correlation between Igfbp5 expression and serum

parameters of liver damage in IGFBP5 overexpressing Abcb4-/- mice, and in the BxD genetic

reference population. The systems genetic approach made it clear that the relation between these

liver enzyme levels and IGFBP5 goes beyond the issue of damage in the Abcb4-/- model. In addition,

a protective effect of IGFBP5 on hepatocytes in vivo is consistent with its prosurvival effect seen in

activated HSCs in vitro (11). Reduced liver fibrosis after 12 weeks of IGFBP5 expression, as

demonstrated by considerably reduced portal bridging compared to that seen in control animals at

the time of vector administration and compared to GFP-expressing controls (Figure 4C), indicated

involvement of IGFBP5 not only in impediment of ongoing fibrogenesis, but in reduction of the

existing scar tissue.

The most likely mechanism of the anti-fibrotic effect of IGFBP5 expression in the Abcb4-/-

mice is a reduction of inflammation. The portal inflammation, caused by leakage of toxic bile is one

of the most prominent pathological features in this model for chronic cholangiopathy (3).

Overexpression of IGFBP5 lowers the influx of inflammatory cells into this area, possibly by

impairing their migration (39). This would also lower the production of inflammatory cytokines,

known to stimulate collagen synthesis and to enhance hepatocyte damage. The negative correlation

of Igfbp5 expression with inflammation in the BxD genetic reference population further supports

such an anti-inflammatory role. In addition, the systems genetics approach indicated Nkx1-2 as the

putative modulator of liver Igfbp5 expression, possibly in response to an inflammatory assault, which

would then explain the Igfbp5 upregulation in Abcb4-/-mice upon cholate feeding (9). Damaged

hepatocytes and activated inflammatory cells release reactive oxygen species (ROS) and other

reactive intermediates (40, 41), which enhance the expression of profibrotic genes in human HSCs

and MFs (42-44). The reduced oxidative stress in IGFBP5 overexpressing livers is in agreement with

the observed decrease in inflammation, and may additionally alleviate liver fibrosis. However, the

lower influx may also reduce the immune surveillance in these livers and as such may provide

protection of cancer cells. Such a mechanism could explain the high IGFBP5 expression reported in

patients with cholangiocarcinoma (10).

IGFBP5, however, may have a more direct effect. The increased amount of p21 in IGFBP5

overexpressing livers may be instrumental for the reduced proliferation. p21 regulates PCNA binding

to damaged DNA (45), reducing its nuclear presence as seen in this study (Figure 2A-B). In addition,

the increased presence of p21 may play a role in senescence (46), particularly that in the fibrotic

regions of IGFBP5 treated livers. Changes in gene expression profiles of senescent HSCs point to

the cell cycle standstill, decreased synthesis of ECM, increased release of ECM remodeling enzymes

(47). In this case, the increased expression of IGFBP5 seems aimed at reducing the pathology in this

model of chronic cholangiopathy.

In conclusion, this study reveals that expression of IGFBP5 in the Abcb4-/-mice reduces

inflammation, oxidative stress, ECM deposition and hepatocyte proliferation, thereby evidently

reducing liver fibrosis. Furthermore, the systems genetics approach reveals correlation between

IGFBP5 in Abcb4-/-

61

endogenous hepatic Igfbp5 expression and decrease of inflammation in a reference population,

underscoring its role in ameliorating pathology in the model for chronic cholangiopathy.

References

1. Jacquemin E, de Vree JM, Cresteil D, Sokal EM, Sturm E, Dumont M, et al. The wide spectrum of multidrug resistance 3 deficiency: from neonatal cholestasis to cirrhosis of adulthood. Gastroenterology 2001 May;120(6):1448-1458.

2. Katzenellenbogen M, Pappo O, Barash H, Klopstock N, Mizrahi L, Olam D, et al. Multiple adaptive mechanisms to chronic liver disease revealed at early stages of liver carcinogenesis in the Mdr2-knockout mice. Cancer Res 2006 Apr 15;66(8):4001-4010.

3. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice with homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of nonsuppurative inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994 Nov;145(5):1237-1245.

4. Benyon RC, Arthur MJ. Extracellular matrix degradation and the role of hepatic stellate cells. Semin Liver Dis 2001 Aug;21(3):373-384.

5. Gatti D, Maki A, Chesler EJ, Kirova R, Kosyk O, Lu L, et al. Genome-level analysis of genetic regulation of liver gene expression networks. Hepatology 2007 Aug;46(2):548-557.

6. Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a modulator of hepatic fibrogenesis. Semin Liver Dis 2001 Aug;21(3):351-372.

7. Friedman SL. Mechanisms of disease: Mechanisms of hepatic fibrosis and therapeutic implications. Nat Clin Pract Gastroenterol Hepatol 2004 Dec;1(2):98-105.

8. Pinzani M, Rombouts K. Liver fibrosis: from the bench to clinical targets. Dig Liver Dis 2004 Apr;36(4):231-242.

9. Boers W, Aarrass S, Linthorst C, Pinzani M, Elferink RO, Bosma P. Transcriptional profiling reveals novel markers of liver fibrogenesis: gremlin and insulin-like growth factor-binding proteins. J Biol Chem 2006 Jun 16;281(24):16289-16295.

10. Nishino R, Honda M, Yamashita T, Takatori H, Minato H, Zen Y, et al. Identification of novel candidate tumour marker genes for intrahepatic cholangiocarcinoma. J Hepatol 2008 Aug;49(2):207-216.

11. Sokolovic A, Sokolovic M, Boers W, Elferink RP, Bosma PJ. Insulin-like growth factor binding protein 5 enhances survival of LX2 human hepatic stellate cells. Fibrogenesis Tissue Repair 2010;3:3.

12. Pilewski JM, Liu L, Henry AC, Knauer AV, Feghali-Bostwick CA. Insulin-like growth factor binding proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular matrix deposition. Am J Pathol 2005 Feb;166(2):399-407.

13. Yasuoka H, Jukic DM, Zhou Z, Choi AM, Feghali-Bostwick CA. Insulin-like growth factor binding protein 5 induces skin fibrosis: A novel murine model for dermal fibrosis. Arthritis Rheum 2006 Sep;54(9):3001-3010.

14. Clemmons DR. Insulin-like growth factor binding proteins and their role in controlling IGF actions. Cytokine Growth Factor Rev 1997 Mar;8(1):45-62.

15. Nathwani AC, Gray JT, Ng CY, Zhou J, Spence Y, Waddington SN, et al. Self-complementary adeno-associated virus vectors containing a novel liver-specific human factor IX expression cassette enable highly efficient transduction of murine and nonhuman primate liver. Blood 2006 Apr 1;107(7):2653-2661.

16. Seppen J, Bakker C, de JB, Kunne C, van den OK, Vandenberghe K, et al. Adeno-associated virus vector serotypes mediate sustained correction of bilirubin UDP glucuronosyltransferase deficiency in rats. Mol Ther 2006 Jun;13(6):1085-1092.

17. van WR, de Waart DR, Akman S, Krediet RT. Assessment of peritoneal fibrosis by conventional light microscopy and hydroxyproline measurements. Perit Dial Int 2004 May;24(3):290-292.

18. Tietze F. Enzymic method for quantitative determination of nanogram amounts of total and oxidized glutathione: applications to mammalian blood and other tissues. Anal Biochem 1969 Mar;27(3):502-522.

19. Sokolovic M, Wehkamp D, Sokolovic A, Vermeulen J, Gilhuijs-Pederson LA, van Haaften RI, et al. Fasting induces a biphasic adaptive metabolic response in murine small intestine. BMC Genomics 2007;8:361.

20. Geisert EE, Lu L, Freeman-Anderson NE, Templeton JP, Nassr M, Wang X, et al. Gene expression in the mouse eye: an online resource for genetics using 103 strains of mice. Mol Vis 2009;15:1730-1763.

21. Peirce JL, Lu L, Gu J, Silver LM, Williams RW. A new set of BXD recombinant inbred lines from advanced intercross populations in mice. BMC Genet 2004 Apr 29;5:7.

22. Taylor BA, Wnek C, Kotlus BS, Roemer N, MacTaggart T, Phillips SJ. Genotyping new BXD recombinant inbred mouse strains and comparison of BXD and consensus maps. Mamm Genome 1999 Apr;10(4):335-348.

chapter 3

62

23. Abiola O, Angel JM, Avner P, Bachmanov AA, Belknap JK, Bennett B, et al. The nature and identification of quantitative trait loci: a community's view. Nat Rev Genet 2003 Nov;4(11):911-916.

24. Argmann CA, Chambon P, Auwerx J. Mouse phenogenomics: the fast track to "systems metabolism". Cell Metab 2005 Dec;2(6):349-360.

25. Chesler EJ, Lu L, Wang J, Williams RW, Manly KF. WebQTL: rapid exploratory analysis of gene expression and genetic networks for brain and behavior. Nat Neurosci 2004 May;7(5):485-486.

26. Chesler EJ, Lu L, Shou S, Qu Y, Gu J, Wang J, et al. Complex trait analysis of gene expression uncovers polygenic and pleiotropic networks that modulate nervous system function. Nat Genet 2005 Mar;37(3):233-242.

27. Farrall M. Quantitative genetic variation: a post-modern view. Hum Mol Genet 2004 Apr 1;13 Spec No 1:R1-R7.

28. Grimm D, Pandey K, Nakai H, Storm TA, Kay MA. Liver transduction with recombinant adeno-associated virus is primarily restricted by capsid serotype not vector genotype. J Virol 2006 Jan;80(1):426-439.

29. van Nieuwkerk CM, Elferink RP, Groen AK, Ottenhoff R, Tytgat GN, Dingemans KP, et al. Effects of Ursodeoxycholate and cholate feeding on liver disease in FVB mice with a disrupted mdr2 P-glycoprotein gene. Gastroenterology 1996 Jul;111(1):165-171.

30. Kershenobich SD, Weissbrod AB. Liver fibrosis and inflammation. A review. Ann Hepatol 2003 Oct;2(4):159-163.

31. Lieber CS. Alcoholic fatty liver: its pathogenesis and mechanism of progression to inflammation and fibrosis. Alcohol 2004 Aug;34(1):9-19.

32. Koutnikova H, Laakso M, Lu L, Combe R, Paananen J, Kuulasmaa T, et al. Identification of the UBP1 locus as a critical blood pressure determinant using a combination of mouse and human genetics. PLoS Genet 2009 Aug;5(8):e1000591.

33. Miconnet I, Huchet R, Bonardelle D, Motta R, Canon C, Garay-Rojas E, et al. Graft-versus-host mortality induced by noncytolytic CD4+ T cell clones specific for non-H-2 antigens. J Immunol 1990 Oct 1;145(7):2123-2131.

34. Hasegawa M, Baldwin TM, Metcalf D, Foote SJ. Progenitor cell mobilization by granulocyte colony-stimulating factor controlled by loci on chromosomes 2 and 11. Blood 2000 Mar 1;95(5):1872-1874.

35. Sternick JL, Schrier DJ, Moore VL. Genetic control of BCG-induced granulomatous inflammation in mice. Exp Lung Res 1983 Nov;5(3):217-228.

36. Tonner E, Barber MC, Travers MT, Logan A, Flint DJ. Hormonal control of insulin-like growth factor-binding protein-5 production in the involuting mammary gland of the rat. Endocrinology 1997 Dec;138(12):5101-5107.

37. Thomas LN, Cohen P, Douglas RC, Lazier C, Rittmaster RS. Insulin-like growth factor binding protein 5 is associated with involution of the ventral prostate in castrated and finasteride-treated rats. Prostate 1998 Jun 1;35(4):273-278.

38. Meadows KA, Holly JM, Stewart CE. Tumor necrosis factor-alpha-induced apoptosis is associated with suppression of insulin-like growth factor binding protein-5 secretion in differentiating murine skeletal myoblasts. J Cell Physiol 2000 Jun;183(3):330-337.

39. Novo E, Busletta C, di Bonzo LV, Povero D, Paternostro C, Mareschi K, et al. Intracellular Reactive Oxygen Species Are Required For Directional Migration Of Resident And Bone Marrow-Derived Hepatic Pro-Fibrogenic Cells. J Hepatol 2010 Oct 31.

40. Greenwel P, Dominguez-Rosales JA, Mavi G, Rivas-Estilla AM, Rojkind M. Hydrogen peroxide: a link between acetaldehyde-elicited alpha1(I) collagen gene up-regulation and oxidative stress in mouse hepatic stellate cells. Hepatology 2000 Jan;31(1):109-116.

41. Nieto N. Oxidative-stress and IL-6 mediate the fibrogenic effects of [corrected] Kupffer cells on stellate cells. Hepatology 2006 Dec;44(6):1487-1501.

42. Bataller R, Sancho-Bru P, Gines P, Brenner DA. Liver fibrogenesis: a new role for the renin-angiotensin system. Antioxid Redox Signal 2005 Sep;7(9-10):1346-1355.

43. Novo E, Marra F, Zamara E, Valfre di BL, Caligiuri A, Cannito S, et al. Dose dependent and divergent effects of superoxide anion on cell death, proliferation, and migration of activated human hepatic stellate cells. Gut 2006 Jan;55(1):90-97.

44. Zamara E, Novo E, Marra F, Gentilini A, Romanelli RG, Caligiuri A, et al. 4-Hydroxynonenal as a selective pro-fibrogenic stimulus for activated human hepatic stellate cells. J Hepatol 2004 Jan;40(1):60-68.

45. Cazzalini O, Perucca P, Savio M, Necchi D, Bianchi L, Stivala LA, et al. Interaction of p21(CDKN1A) with PCNA regulates the histone acetyltransferase activity of p300 in nucleotide excision repair. Nucleic Acids Res 2008 Mar;36(5):1713-1722.

46. Zhao B, Benson EK, Qiao R, Wang X, Kim S, Manfredi JJ, et al. Cellular senescence and organismal ageing in the absence of p21(CIP1/WAF1) in ku80(-/-) mice. EMBO Rep 2009 Jan;10(1):71-78.

47. Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, Miething C, et al. Senescence of activated stellate cells limits liver fibrosis. Cell 2008 Aug 22;134(4):657-667.

IGFBP5 in Abcb4-/-

63

c h a p t e r 4

igf1 enhances

bile-duct proliferation and fibrosis

in abcb4-/- mice

chapter 4

66

ABSTRACT

Background and Aims: Adamant progression of chronic cholangiopathies towards cirrhosis and

limited therapeutic options leave a liver transplantation the only effective treatment. Insulin-like

growth factor 1 (IGF1) effectively blocks fibrosis in acute models of liver damage in mice, and a

phase I clinical trial suggested an improved liver function. IGF1 targets the biliary epithelium, but its

potential benefit in chronic cholangiopathies has not been studied.

Methods: To investigate the possible therapeutic effect of increased IGF1 expression, we crossed

Abcb4-/- mice (a model for chronic cholangiopathy), with transgenic animals that overexpress IGF1.

The effect on disease progression was studied in the resulting IGF1-overexpressing Abcb4-/- mice,

and compared to that of Abcb4-/- littermates. The specificity of this effect was further studied in an

acute model of fibrosis.

Results: The overexpression of IGF1 in transgenic Abcb4-/- mice resulted in stimulation of

fibrogenic processes - as shown by increased expression of Tgfß, and collagens 1, 3 and 4, and

confirmed by Sirius red staining and hydroxyproline measurements. Excessive extracellular matrix

deposition was favoured by raise in Timp1 and Timp2, while a reduction of tPA expression indicated

lower tissue remodelling. These effects were accompanied by an increase in expression of

inflammation markers like Tnfα, and higher presence of infiltrating macrophages. Finally, increased

number of Ck19-expressing cells indicated proliferation of biliary epithelium.

Conclusion: In contrast to liver fibrosis associated with hepatocellular damage, IGF1

overexpression does not inhibit liver fibrogenesis in chronic cholangiopathy.

accepted for publication in Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease

Authors

Aleksandar Sokolović1, Carlos M Rodriguez-Ortigosa2, Lysbeth ten Bloemendaal1, Ronald P.J. Oude

Elferink1, Jesús Prieto2, Piter J Bosma1

Affiliations

1Tytgat Institute for Liver and Intestinal Research, Academic Medical Center, University of

Amsterdam, Amsterdam, The Netherlands

2Division of Hepatology and Gene Therapy, CIMA University of Navarra, and CIBERehd,

Pamplona, Spain

Acknowledgements:

We thank Nerea Juanarena and Sara Arcelus for their excellent technical assistance, and dr. Dirk

Rudi de Waart for hydroxyplroline quantification. We would also like to thank dr. Milka Sokolović

for her assistance and advice during preparation of the manuscript.

IGF1 in Abcb4-/-

67

INTRODUCTION

Primary sclerosing cholangitis and primary biliary cirrhosis are chronic diseases of the bile duct

characterized by progressive inflammation that leads to the development of biliary fibrosis. There is

no effective therapy to halt the disease progression and most patients progress to cirrhosis and liver

failure (1, 2). The lack of sufficient donor organs and surgical contraindications render development

of therapeutic interventions that stop progression of liver fibrosis urgently needed.

Insulin-like growth factor 1 (IGF1) regulates the proliferation and differentiation of many

cell types (3). The liver is responsible for the abundant presence of IGF1 in the circulation (4) where

it is present attached to binding proteins (IGFBPs) that prolong its half-life and regulate its biological

activity (3, 5). Once released from IGFBPs, IGF1 can bind and activate the receptor (IGF1R), and

by activation of mitogen activated protein (MAP) kinase and PI3 kinase induce genes involved in cell

survival, growth, and differentiation (6). Liver fibrosis reduces the IGF1 expression in the liver (7, 8).

Administration of recombinant IGF1 or enhancing IGF1 expression using a viral vector significantly

reduced liver fibrosis in rats with CCl4-induced liver cirrhosis (9, 10). In the follow-up phase I-II

clinical trials, IGF1 administration increased the serum albumin level, suggesting an improved liver

function in patients suffering from this devastating disease (11).

The potential use of IGF1 in chronic cholangiopathies has not been studied. Since the biliary

epithelium expresses IGF1R (12), and IGF1 protects cholangiocytes against cholestatic injury in vitro

(13), overexpression of IGF1 could be beneficial in these diseases. Therefore the aim of this study

was to investigate if a prolonged increase of hepatic IGF1 expression in a model for chronic

cholangiopathy, the Abcb4-/- mouse, was beneficial.

The Abcb4-/- mouse is a model for primary familial cholestasis type 3 (14). ABCB4 is a

flippase essential for the excretion of phospholipids into bile, needed to neutralize the harmful

effects of bile acids. Bile lacking phospholipids is toxic and causes hepatocyte and cholangiocyte

injury (15). Since murine bile acids are less toxic, the damage in the Abcb4-/- is smaller compared to

that in PFIC type 3 patients, making this mouse also a model for other chronic cholangiopathies,

more specifically, primary sclerosing cholangitis (PSC) (16).To study the long term effect of IGF1 on

chronic cholangiopathy we crossbred the Abcb4-/- with the transgenic SMP8-IGF1 mouse. This

transgenic mouse expresses the rat IGF1 behind an α-SMA promoter (17). The increased α-SMA

expression in Abcb4-/- livers (15) should make this cross-bred model well suited to study long-term

effects of IGF1 on PSC.

MATERIALS AND METHODS

Animals.

This study was approved by the Ethical Committee for Animal Research of the University of

Navarra with the number: 008/08.

Abcb4-/- IGF1 model

Homozygous Abcb4-/- mice (FVB background; (18)) were cross-bred with transgenic SMP8-IGF1

mice, with a germline integration of a transgene composed of the murine α-Sma promoter linked to a

chapter 4

68

full length rat Igf1 cDNA (17). The resulting heterozygotes for Abcb4 were back-crossed with Abcb4-

/- mice to obtain Abcb4-/- SMP8-Igf1+/- transgenic mice (TG) and their nontransgenic (NT)

littermates as a control. After weaning, the mice WT (FVB; n=4), Abcb4-/- (n=7) and Abcb4-/-

SMP8-Igf1+/- (n=7) were fed a 0.03% cholate diet to augment fibrosis, and were sacrificed after 3, 6

and 12 weeks (WT control for 3 weeks).

Bile duct ligation (BDL) model

SMP8-IGF1 (n=5) and WT (FVB; n=5) mice were bile duct ligated. After a 1 cm abdominal midline

incision, the common bile duct was isolated, double ligated close to the liver hilus (immediately

below the bifurcation) with a silk suture, and transected between ligatures. In sham operated mice

(SMP8-IGF1 and WT; n=5), the same operation was performed, but neither ligatures nor bile duct

section were carried out. All the mice used in the protocol were 12 week old at the beginning of the

experiment. The mice were sacrificed a week after BDL.

Tissue collection and biochemical analyses

Upon sacrificing, blood was collected, livers were snap-frozen in liquid nitrogen and stored at -80oC,

or fixed in 4% formaldehyde in PBS. Total liver collagen was determined by measuring

hydroxyproline content as described (19) Serum transaminases (ALT and AST), alkaline phosphatase

and bilirubin were determined in retro-orbital collected blood, using standard clinical chemistry

methods in a Cobas C311 autoanalyzer (Roche Diagnostics, Mannheim, Germany).

Reverse transcription, quantitative PCR and data analyses

The qPCR analysis was performed and the data analyzed as previously described (20). Shortly, total

liver RNA was extracted from snap-frozen livers using TRIzol reagent (Invitrogen, Breda, The

Netherlands) following the manufacturer’s instructions. 4M LiCl was used to specifically precipitate

RNA. All RNA isolates had A260/A280 ratio around 2, as assessed by NanoDrop ND-1000

spectrophotometer. The RNA integrity was confirmed by denaturing agarose gel electrophoresis.

The absence of DNA contamination was established by including RT- controls in cDNA synthesis

reactions. Gene specific primers (for Igf1, Hgf, aSma, Tgfb, Col1, Col3, Col4, Timp1, Timp2, tPa,

Tnfa, Ki67 and Ck19) were designed using Primer3Plus program and their sequences are available

upon request. qPCR was performed using a Light Cyler 480 (Roche) as described (21). The

expression level for each of the studied genes in liver was calculated using the LinRegPCR software

(22, 23), thereby avoiding the assumption of equal amplification efficiency between the reactions.

Gene expression was normalized by dividing the expression levels of tested genes by those of a

reference gene (36B4), and calculating the average of normalized values per condition. 36B4 was

used only after having validated it against three reference genes in our previous studies (24). In BDL

model, qPCR was performed as previously described (25). To assess the significance of the data,

ANOVA analysis and Student’s t test were employed. The error bars in the Figures represent the

standard deviation. Significance threshold was set to P<0.05 (denoted by the asterisks in the Figures

and Tables).

Histology and immunohistochemistry

Formaldehyde-fixed liver sections were stained with Sirius red (morphometrical connective tissue

assessment), or immunohistochemically, as described (26). Liver collagen content was quantified by

IGF1 in Abcb4-/-

69

Sirius red staining, using an Axioplan 2 Imaging microscope (Zeiss, Jena, Germany); random

acquisition of images was performed with Metamorph software (Molecular Devices, Sunnyvale, CA,

USA), and scoring of areas was carried out with Matlab software and the imaging processing libraries

of DIPlib (MathWorks, Natick, MA, USA). Ck19 (Eurogentec, Koln, Germany) was detected using

GAR-AP (DAKO, Glostrup, Denmark) as a secondary antibody. F4/80 (AbD Serotec, Dusseldorf,

Germany) was visualized using rabbit anti Rat-Ig (DAKO, Glostrup, Denmark) as a secondary

antibody.

Double staining for Ck19 and Ki67 was performed on frozen sections, fixed for 5 minutes in ice

cold acetone and air dried for 30 minutes. Slides were rinsed in PBS, blocked for 30 minutes in

TengT and incubated o/n at 4oC with the primary anti-Ki67 Ab (Abcam, Cambridge, UK; 1:500).

Washing was followed by 30 minutes incubation at RT with the secondary Alexa Fluor488 goat anti

rabbit IgG (Invitrogen, Paisley, UK; 1:1000). Sections were then rinsed and incubated with the

primary anti-Ck19 Ab (Eurogentec, Koln, Germany; 1:1000) for 1 hour at RT, washed again, and

incubated with Alexa Fluor594 goat anti rabbit IgG (Invitrogen, Paisley, UK; 1:1000) for 30 minutes

at RT. After washing, the slides were covered with vectashield + DAPI and visualized by confocal

microscopy. Quantification of Ki67-positive cholangiocytes after 12 weeks of cholate diet was

perfomed in 5 animals per phenotype, 5 images per mouse. The results are expressed as a percentage

of Ki67-positive in the total number of cholangiocytes.

In BDL model, bromodeoxyuridine (BrdU) method was used for counting proliferating

cholangiocytes as previously described (27). Shortly, formalin-fixed and paraffin-embedded liver

sections were stained with mouse monoclonal anti-BrdU antibody (1:1000 GE Healthcare,

Barcelona, Spain), at 4°C, overnight. Secondary peroxidase-labeled goat anti-mouse antibody (Dako

Envision System, Carpinteria, CA) was incubated at room temperature for 30 min. Two persons

unaware of the study groups scored BrdU positive cholangiocytes in SMP8-IGF1 BDL (n=5) and

WT BDL (n=5), analyzing 6 panels per animal.

RESULTS

Overexpression of IGF1 in Abcb4-/- mice

To study the long-term effect of IGF1 on the pathology of chronic cholangiopathy, Abcb4-/- mice

were crossbred with a transgenic mouse expressing rat Igf1 behind an α-Sma promoter (17). This

restricts rat Igf1 hepatic expression to the α-SMA expressing, activated hepatic stellate cells (HSC)

(27). To aggravate the pathology, after weaning the animals were given a cholate-supplemented diet

(0.03%) (15, 18). The expression of rat Igf1 was confirmed by qPCR in livers of the transgenic (TG)

Abcb4-/-IGF1 mice (Fig.1A). Since rat Igf1 expression in TG mice is controlled by the α-Sma

promoter, its stable expression reflects the α-Sma expression, that remained comparable at all three

time points (Fig.1B). IGF1 induces HGF expression in HSC (28). The significant increase of Hgf

expression at 6 and 12 weeks in TG animals compared to non-transgenic (NT) Abcb4-/- littermates

confirmed that IGF1 signaling was indeed enhanced (Fig.1C).

chapter 4

70

Fig.1: Functional expression of IGF-1 in livers of Abcb4-/- mice. After weaning, WT, Abcb4-/- and Abcb4-/- IGF1 mice - WT (n=4; for 3 weeks), Abcb4-/- (n=7) and Abcb4-/- IGF1 (n=7) - were fed a 0.03% cholate diet. (A-C) mRNA levels of transgenic Igf1, α-Sma and Hgf in livers of WT, Abcb4-/- and Abcb4-/- IGF1 mice fed a cholate supplemented diet for 3, 6 and 12 weeks. D) Average fractional liver weight of Abcb4-/- and Abcb4-/- IGF1 mice, after 3, 6 and 12 weeks of cholate diet. Asterisks denote significance (P<0.05).

Overexpression of IGF1 induced liver fibrosis in Abcb4-/- mice

Liver function tests were performed to investigate if IGF1 could protect this mouse model against

the toxicity of bile salts, thereby reducing chronic bile duct proliferation. Serum levels of ALT, AST,

ALP and bilirubin significantly increased in Abcb4-/-IGF1 and Abcb4-/- mice compared to WT,

indicating liver damage in both models (Table 1). The initially high ALP serum levels, a result of

osteoblast activity in growing bones (29), decreased to normal after weaning (66-260 U/L). After 12

weeks of cholate diet, however, ALP and bilirubin concentration significantly increased in the IGF1

transgenics compared to their Abcb4-/- littermates, suggesting cholestasis and possibly an enhanced

progression of liver fibrosis (Table 1).

Table 1: The influence of IGF1 overexpression on liver function shown by clinical parameters in plasma.

ALT (U/L)

AST (U/L)

ALP (U/L)

tBil (mg/dl)

TP (g/dl)

Chol (mg/dl)

TG (mg/dl)

WT

3 weeks 76.1±16.9 126.9±29.5 188.2±36.6 0.08±0.01 5.2±0.2 183.5±9.9 83.50±17.4

Abcb4-/-

3 weeks 355.4±125.8 502.5±186.9 1992.6±1257.6 0.2±0.2 5.2±0.3 107.2±25.9 137.3±24.7

6 weeks 185.3±95.4 238.7±74.8 448.8±319.5 0.2±0.1 5.3±0.8 103.6±28.7 144.9±12.9

12 weeks 356.6±220.3 591.1±460.4 285±136.4 0.3±0.3 5.9±0.5 80.1±16.6 89.14±34.2

Abcb4-/-IGF1

3 weeks 320.7±86.2 502.2±231.2 1020.5±657.4 0.2±0.1 5.9±0.5 112.9±15.9 116.6±29.1

6 weeks 238.5±127.7 314.6±152.9 533.4±323.2 0.2±0.1 5.4±1.4 101.5±29.8 132.1±56.7

12 weeks 388.8±191.2 600.9±322.7 1712.9±1095.5* 1.9±1.6* 5.0±1.2 103.3±24.6 128.3±42.9

The table shows mean values and standard deviation of serum parameters measured in plasma of WT (n=4), Abcb4 -/- (n=7) and Abcb4-/-IGF1 (n=7) mice. The differences between Abcb4-/- and Abcb4-/-IGF1 mice were tested by Student’s t-test, with asterisks denoting significant difference (P<0.05). The values for WT are shown as a reference. ALT (alanine transaminase), AST (aspartate transaminase), ALP (alkaline phosphatase), tBil (total bilirubin), TP (total protein), Chol (cholesterol), Tg (triglycerides).

After 3 weeks of cholate diet, the fractional liver weight in Abcb4-/- mice was significantly

higher than in Abcb4-/- IGF1 group. At 6 weeks this difference had disappeared, while at 12 weeks

there was even a tendency toward an increased liver/body weight ratio in Abcb4-/- IGF1 mice

(Fig.1D). Sirius red staining and hydroxyproline measurement revealed a significant increase in the

IGF1 in Abcb4-/-

71

amount of connective tissue in the Abcb4-/- IGF1 group, indicating increased liver fibrosis in

transgenic mice after 12 weeks of cholate diet (Fig.2A-C). The increased expression of the pro-

fibrotic transforming growth factor beta 1 (Tgfβ1), which stimulates trans-differentiation of HSC into

myofibroblasts and enhances TIMP and collagen production (30), may also support the increased

fibrosis in IGF1 expression Abcb4-/- mice at twelve weeks (Fig.3A). Also the mRNA levels of Col1, 3

and 4 at 12 weeks were significantly increased in these mice (Fig.3B-D), providing additional support

that IGF1 enhanced the progression of liver fibrosis in the Abcb4-/- mouse model.

Fig.2: IGF1 overexpression promotes fibrosis Abcb4-/- livers. (A) Sirius red staining of liver of WT, Abcb4-/- mice and transgenic Abcb4-/-IGF1 mice with representative pictures of large bile duct structures 12 weeks of cholate diet. (B) Percentage of fibrotic area estimated by image analysis of Sirius red stainings in Abcb4-/- and Abcb4-/-IGF1 mice after 12 weeks of cholate diet. Y axe represents the ratio between collagen area and total tissue area. (C) Hydroxyproline amount at 3, 6 and 12 weeks of cholate diet in livers of Abcb4-/- and Abcb4-/-IGF1 mice. Asterisks denote significance (P<0.05).

In addition to enhanced synthesis, the increased ECM deposition in mice overexpressing

IGF1 may also in part be due to reduced matrix remodeling. Inhibition of matrix metalloproteinases

(MMP) by their tissue inhibitors (TIMPs) plays an important role in this process (31, 32). The

significantly increased expression of Timp1 and 2 after 12 weeks suggested that tissue remodeling was

reduced in the Abcb4-/- IGF1 mice (Fig.3E, F). Additional support for this was provided by the

decreased expression of tPa, an activator of plasminogen that plays an essential role in the activation

of MMP needed for tissue remodeling (33).

chapter 4

72

Fig.3: IGF1 overexpression hinders matrix remodeling in Abcb4-/- livers. (A-G) Relative mRNA level of Tgfβ1, Col1, 3, 4, tPA, Timp1, 2 and Tnfα in livers of WT, Abcb4-/-IGF1 and Abcb4-/- mice after 3, 6 and 12 weeks of cholate diet. Asterisks denote significant difference (P<0.05).

Overexpression of IGF1 enhances inflammation in Abcb4-/- liver

Inflammation due to regurgitation of bile acids plays an important role in the development of

pathology in the Abcb4-/- mice. The increased pathology in Abcb4-/- IGF1 animals was accompanied

by an increased inflammation, as shown by the enhanced expression of Tnfa (Fig.4A).

Immunostaining for F4/80, a marker for infiltrating macrophages and liver-resident Kupffer cells,

was higher at 12 weeks in Abcb4-/- IGF1 animals, which suggested an increased influx of

macrophages compared with non-transgenic controls (Fig.4B).

Fig.4: IGF1 overexpression induces expression of inflammatory markers in Abcb4-/- mice. Relative mRNA level of Tnfα in livers of WT, Abcb4-/-IGF1 and Abcb4-/- mice after 3, 6 and 12 weeks of cholate diet. Asterisks denote significant difference (P<0.05). (B) The representative pictures of F4/80 immunostaining in non-transgenic (left panel) and Abcb4-/-IGF1 mice (right panel) after 12 weeks of cholate diet are shown. (C) Representative CK19 immunostaining of the same livers.

IGF1 in Abcb4-/-

73

Overexpression of IGF1 increases cholangiocyte proliferation in Abcb4-/- mice

Since bile duct proliferation is one of the major characteristics of the pathology in Abcb4-/- mice, and

cholangiocytes are target of IGF1 signaling (12), the increased pathology could be due to an

enhanced cholangiocyte proliferation in response to the IGF1 overexpression. Therefore, the level

of, the cholangiocyte marker Ck19 was determined in these livers. A significant increase in mRNA

expression level and an increased presence of Ck19-expressing cells detected by

immunohistochemistry confirmed an enhanced presence of cholangiocytes in the transgenic animals

(Fig.5C and Fig.4C).

Subsequently, the expression of the proliferation marker Ki67 was determined to investigate if this

effect was caused by a higher proliferation. Although a higher Ki67 level was observed, the

difference did not reach significance (Fig.5B). This lack of significance could be due to the fact that

the proliferation was not limited to cholangiocytes, hence the proliferation of biliary epithelium was

determinated by fluorescent cell labelling and co-localisation for Ck19 and Ki67. This indicated an

ongoing proliferation of Ck19 positive cells in the transgenic animals (Fig.5A). Quantification of

Ki67-positive cholangiocytes with respect to the total number of cholangiocytes demonstrated a

significant increase in bile duct proliferation in IGF1 overexpressing animals after 12 weeks of

cholate diet (Fig.5D). This enhanced proliferation of biliary epithelium could have resulted in the

formation of numerous large bile duct structures seen in the IGF1 overexpressing mice (Fig.2A).

Table 2: Plasma liver function parameters, hepatic mRNA expression levels, and percentage of BrdU-positive cholangiocyte in WT and SMP8-IGF1 mice in response to BDL

liver function tests WT ctrl WT BDL SMP8-IGF1 ctrl SMP8-IGF1 BDL

AST 133.30 ± 59.90 730.70 ± 186* 113.30 ± 47.72 507.20 ± 158.70*

ALT 77 ± 48.08 777 ± 262*# 73.67 ± 38.84 386.20 ± 105.40*

ALP 83 ±10.17 1543 ± 332.90* 112.30 ± 14.47 1333 ± 523.60*

bilirubin 0.08 ± 0.01 10.07 ± 0.65* 0.10 ± 0.02 11.23 ± 2.09*

mRNA expression

transgen Igf1 no expression no expression 2.28 ±0.63 5.03 ± 2.13*

αSma 1.56 ± 0.69 23.55 ± 7.69*# 2.53 ± 0.33 11.65 ± 5.44*

Ck19 1.22 ± 0.54 7.04 ± 3.09* 1.47 ± 0.83 10.63 ± 4.70*

Tnfα 6.83 ± 3.63 214.35 ± 88.56* 9.22 ± 7.08 115.68 ± 56.46*

Col1 0.73 ± 0.82 1.18 ± 0.79 0.31 ± 0.19 0.78 ± 0.17*

Tgfβ 1.34 ± 0.89 1.73 ± 1.07 1.35 ± 0.01 1.34 ± 0.61

Asbt 19.16 ± 13.64 25.01 ± 15.68 20.89 ±4.72 49.07 ± 20.57*

BrdU+ cholangiocytes (%) 11.69 ± 4.30 12.70 ± 5.89

The table shows mean values and standard deviations of liver function parameters and hepatic mRNA expression levels (relative to housekeeping 36B4) in WT and SMP8-IGF1 mice and one week after BDL treatment and in sham-operated controls (n=5 per group). Percentage of BrdU-positive cholangiocytes was assessed in WT and SMP8-IGF1 mice one week after BDL. Asterisks denote significant difference between sham-operated animals and BDL counterparts; hashtags indicate significant difference between BDL groups (P<0.05).

chapter 4

74

Fig.5: IGF1 overexpression in Abcb4-/- mice enhances proliferation of cholangiocytes: (A) Representative pictures

of double staining for Ck19, Ki67 in Abcb4-/- (left panels) and Abcb4-/-IGF1 (right panels) after 12 weeks of cholate diet.

(B-C) Relative gene expression of Ki67 and Ck19 in control, Abcb4-/- and Abcb4-/-IGF1 mice. (D) Percent of Ki67-

positive cholangiocytes with respect to the total number of cholangiocytes, as detected by immunohistochemistry.

The number of a-Sma expressing activated HSC increased, and was accompanied by an increased scar

formation reflected by higher Col1 levels, by increased cholangiocyte proliferation demonstrated by

higher Ck19 levels, and more inflammation based on Tnfα expression.

Overexpression of IGF1 in BDL model

Though Abcb4-/- is a relevant model for chronic cholangiopathy, it does not reproduce all features of

acute models of cholestasis, e.g. CCl4- and BDL-induced. To compare the effect of IGF1 in Abcb4-/-

mice with that in an acute model, we used BDL to induced acute cholestasis in SMP8-IGF1 mice.

Sirius red staining of livers in BDL-treated SMP8-IGF1 and WT animals demonstrated fibrosis in

peribiliary areas (data not shown). Upon BDL, the expression of IGF1 was significantly higher in

transgenic SMP8-IGF1 mice in comparison to sham-operated animals indicating transgenic

upregulation by cholestatic injury. Changes in gene expression profiles in BDL treated animals and

IGF1 in Abcb4-/-

75

liver function test (ALT, AST, ALP and bilirubin) confirmed the development of liver fibrosis in

both SMP8-IGF1 and WT mice (Table 2).

The gene expression data one week after BDL did not reveal an increased development of

fibrosis in SMP8-IGF1 mice compared to WT mice. In fact, the only significant difference was a

decreased expression of α-Sma in transgenic compared to WT (Table 2), while that of Tnfα, Tgfβ, and

Col1 did not differ between the models. In contrast to the Abcb4-/- model, IGF1 overexpression did

not cause an increased Ck19 level in BDL-treated mice (Table 2). This absence of an effect on

proliferation was confirmed by the similar percentage of BrdU-positive cholangiocytes (Table 2). In

conclusion, in contrast to the chronic model for bile duct damage, increased IGF expression did not

affect pathology in the acute BDL model, but did not display a protective effect either.

DISCUSSION

Prolonged overexpression of IGF1 increased fibrosis in the Abcb4-/- mouse, a model for primary

sclerosing cholangitis. The enhanced IGF1 signaling resulted in stimulated cholangiocyte

proliferation, one of the main pathological features of the liver damage in these mice.

The expression of transgenic Igf1, controlled by the α-Sma promoter, was stable during the

entire experiment. It was previously reported that in Abcb4-/- mice expression of α-Sma (also of pro-

fibrotic Tgfβ and Pdgfr) reaches maximum at 4 weeks of age and then gradually decreases (34). This

was due to the generation of fibrous septa that shield the healthy liver from the toxic bile, thereby

limiting the increase in pro-fibrotic factors. In our study, the expression of pro-fibrotic factors in the

Abcb4-/- mice remained high, resulting in a hydroxyproline level about 2-fold higher than previously

reported. The difference between the studies is most likely due to the cholate diet, which enhances

the pathology in this model, resulting in prolonged high expression of pro-fibrotic factors (15).

IGF1 induces the expression of hepato-protective HGF in HSC (28). Confirming our

previous observation (27), Hgf mRNA levels also increased in the IGF1 overexpressing Abcb4-/- mice

compared to the non-transgenic litter-mates. The extent of damage in this model is much less severe

than in the previously used acute CCl4 model (100-fold lower levels of liver enzymes in serum; (9,

10)). The increase of AST and ALT in serum of the Abcb4-/- mice was comparable to the data

reported before for this model (15), and indicated an ongoing liver damage. The elevation of serum

ALP and bilirubin after 12 weeks of cholate diet points to an increase in progression of bile duct

damage in mice that overexpress IGF1. This worsening of pathology was subsequently confirmed by

the increased level of hydroxyproline and collagen, indicating an enhanced deposition of ECM. Thus,

in contrast to our previous study in animals with CCl4- induced parenchymal damage, in the present

work IGF1 overexpression enhanced the damage and increased portal fibrosis in the cholangiopathic

livers of Abcb4-/- mice.

Fibrogenesis depends on tissue remodeling, composed of strictly controlled processes. A

subtle balance of MMPs (mainly involved in the ECM degradation) and TIMPs (their endogenous

inhibitors) modulates liver fibrogenesis (35). MMPs are regulated by uPA/tPA/plasmin proteases,

which directly contribute to degradation of collagens, and so play a pivotal role in the maintenance of

ECM and tissue homeostasis (36-38). IGF1 overexpression increased the expression of Timp1, which

could reduce MMP activity and ECM degradation in Abcb4-/-IGF1 animals. Moreover, lower tPA

chapter 4

76

expression points to a less active tissue remodeling resulting in a more prominent fibrosis in IGF1

overexpressing mice, which may further promote liver fibrosis.

Several studies have shown that bile-duct epithelium cells, in contrat to hepatocytes, express

the IGF1R and directly respond to IGF1. IGF1 enhances proliferation and survival of

cholangiocytes via IRS1/2 and ERK and PI3-kinase pathways (12). The proliferative effect of IGF1

has also been reported for smooth muscle cells in IGF1-overexpressing mice (17). Activation of the

IGF1 system improves the survival of cholangiocytes in PBC patients suffering from chronic bile

duct damage (39). In the acute CCl4 model, IGF1 did not reduce the initial damage, but caused a

more rapid DNA synthesis, resulting in liver regeneration (27). Depletion of IGF1R reduces ductular

and fibrogenic responses and increased cholestasis tolerance in BDL mice (40). The damage in the

Abcb4-/- mouse is chronic, less severe, and localized to the portal areas (16). This results in a

persistent increase in IGF1 signaling in the portal area, enhancing cholangiocyte proliferation (12). In

addition, IGF1 enhances the proliferation of liver cystic epithelium and reduction of its expression

reduces the size of liver cysts in a mouse model (41). A similar action may explain the formation of

extremely large, but ill-developed bile-duct structures observed in those mice, leading to the

formation of strictures. Presence of strictures may explain the cholestasis that, according to serum

bilirubin and ALP level, developed eventually in our transgenic mice. Alternatively, the increased

proliferation of bile ducts could promote the cholehepatic shunt (42), thereby increasing the amount

of bile salts reaching the hepatocytes.

Submitting SMP8-IGF1 and WT mice to BDL revealed an absence of protective effect of

overexpression of IGF1 also in this model. However, no difference in bile duct proliferation was

seen between the WT and the SMP8-IGF1 mice. BDL did induce a-Sma mRNA expression in both

models, suggesting that one week was enough to enhance IGF1 signaling. Nevertheless, 1 week BDL

protocol seemed insufficient to induce the extent of liver damage needed to reveal an effect of

enhanced IGF1 signaling.

Bile duct proliferation increased in IGF1 overexpressing Abcb4-/- and showed a tendency to

increase in the SMP8-IGF1 mice submitted to BDL based on Ck-19 expression and BrdU staining.

Increased Asbt expression in SMP8-IGF1 BDL mice suggests an increased cholehepatic shunt of bile

salts, probably in an attempt to reduce the presence of damaging bile salts in biliary ducts. Likewise,

Abcb4-/- mice showed an increased expression of Asbt (43). The augmented expression of IGF1 in

Abcb4-/-IGF1 mice could result in an even higher increase of Asbt that would enhance the excessive

return of bile salts to the hepatocyte. Thus although these responses are aimed at the protection and

repair of the bile duct epithelium, the additional increase of IGF1 to supra-normal levels induced

extensive proliferation resulting in more bile duct abnormalities.

A difference in pathogenesis between the models may explain the discrepancy of the effect of

IGF1 seen between Abcb4-/- mice and other mouse models of liver fibrosis (e.g. BDL- and CCl4-

induced). Ligation of the common bile duct results in an acute interstitial (biliary) fibrosis, associated

with massive proliferation of bile ducts that is only rarely observed in man. On the other hand,

progression of fibrosis in Abcb4-/- mice is spontaneous due to cholangiocyte proliferation and

massive up-regulation of profibrogenic genes, and bears more resemblance to human biliary fibrosis

(34).

IGF1 in Abcb4-/-

77

Our data demonstrate that rise in IGF1 expression in Abcb4-/- mouse, model for primary

sclerosing cholangitis, worsens the pathology by increasing cholangiocyte proliferation leading to the

formation of bile duct strictures, comparable to those in patients (44). This indicates that, in contrast

to chronic parenchymal live damage, IGF1 administration does not seem an option for treating

fibrosis caused by chronic cholangiopathies, and should be taken into account in the design of IGF1

replacement therapy for cirrhotic patients.

References

1. Selmi C, Zuin M, Gershwin ME. The unfinished business of primary biliary cirrhosis. J Hepatol 2008 Sep;49(3):451-460.

2. Weismuller TJ, Wedemeyer J, Kubicka S, Strassburg CP, Manns MP. The challenges in primary sclerosing cholangitis--aetiopathogenesis, autoimmunity, management and malignancy. J Hepatol 2008;48 Suppl 1:S38-S57.

3. Cohick WS, Clemmons DR. The insulin-like growth factors. Annu Rev Physiol 1993;55:131-153. 4. Sjogren K, Liu JL, Blad K, Skrtic S, Vidal O, Wallenius V, et al. Liver-derived insulin-like growth factor I (IGF-

I) is the principal source of IGF-I in blood but is not required for postnatal body growth in mice. Proc Natl Acad Sci U S A 1999 Jun 8;96(12):7088-7092.

5. Mohan S, Baylink DJ. IGF-binding proteins are multifunctional and act via IGF-dependent and -independent mechanisms. J Endocrinol 2002 Oct;175(1):19-31.

6. Riedemann J, Macaulay VM. IGF1R signalling and its inhibition. Endocr Relat Cancer 2006 Dec;13 Suppl 1:S33-S43.

7. Assy N, Pruzansky Y, Gaitini D, Shen OZ, Hochberg Z, Baruch Y. Growth hormone-stimulated IGF-1 generation in cirrhosis reflects hepatocellular dysfunction. J Hepatol 2008 Jul;49(1):34-42.

8. Caregaro L, Alberino F, Angeli P, Gatta A. Insulin-like growth factor 1 (IGF-1) in liver cirrhosis: a marker of hepatocellular dysfunction? J Hepatol 1998 Aug;29(2):342.

9. Castilla-Cortazar I, Garcia M, Muguerza B, Quiroga J, Perez R, Santidrian S, et al. Hepatoprotective effects of insulin-like growth factor I in rats with carbon tetrachloride-induced cirrhosis. Gastroenterology 1997 Nov;113(5):1682-1691.

10. Sobrevals L, Rodriguez C, Romero-Trevejo JL, Gondi G, Monreal I, Paneda A, et al. Insulin-like growth factor I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to cirrhosis reversion in rats. Hepatology 2010 Mar;51(3):912-921.

11. Conchillo M, de Knegt RJ, Payeras M, Quiroga J, Sangro B, Herrero JI, et al. Insulin-like growth factor I (IGF-I) replacement therapy increases albumin concentration in liver cirrhosis: results of a pilot randomized controlled clinical trial. J Hepatol 2005 Oct;43(4):630-636.

12. Alvaro D, Metalli VD, Alpini G, Onori P, Franchitto A, Barbaro B, et al. The intrahepatic biliary epithelium is a target of the growth hormone/insulin-like growth factor 1 axis. J Hepatol 2005 Nov;43(5):875-883.

13. Gatto M, Drudi-Metalli V, Torrice A, Alpini G, Cantafora A, Blotta I, et al. Insulin-like growth factor-1 isoforms in rat hepatocytes and cholangiocytes and their involvement in protection against cholestatic injury. Lab Invest 2008 Sep;88(9):986-994.

14. Oude Elferink RP, Paulusma CC. Function and pathophysiological importance of ABCB4 (MDR3 P-glycoprotein). Pflugers Arch 2007 Feb;453(5):601-610.

15. van Nieuwerk CM, Groen AK, Ottenhoff R, van WM, van den Bergh Weerman MA, Tytgat GN, et al. The role of bile salt composition in liver pathology of mdr2 (-/-) mice: differences between males and females. J Hepatol 1997 Jan;26(1):138-145.

16. Fickert P, Fuchsbichler A, Wagner M, Zollner G, Kaser A, Tilg H, et al. Regurgitation of bile acids from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology 2004 Jul;127(1):261-274.

17. Wang J, Niu W, Nikiforov Y, Naito S, Chernausek S, Witte D, et al. Targeted overexpression of IGF-I evokes distinct patterns of organ remodeling in smooth muscle cell tissue beds of transgenic mice. J Clin Invest 1997 Sep 15;100(6):1425-1439.

18. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice with homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of nonsuppurative inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994 Nov;145(5):1237-1245.

19. van WR, de Waart DR, Akman S, Krediet RT. Assessment of peritoneal fibrosis by conventional light microscopy and hydroxyproline measurements. Perit Dial Int 2004 May;24(3):290-292.

chapter 4

78

20. Sokolovic A, Sokolovic M, Boers W, Elferink RP, Bosma PJ. Insulin-like growth factor binding protein 5 enhances survival of LX2 human hepatic stellate cells. Fibrogenesis Tissue Repair 2010;3:3.

21. Lekanne Deprez RH, Fijnvandraat AC, Ruijter JM, Moorman AF. Sensitivity and accuracy of quantitative real-time polymerase chain reaction using SYBR green I depends on cDNA synthesis conditions. Anal Biochem 2002 Aug 1;307(1):63-69.

22. Ramakers C, Ruijter JM, Deprez RHL, Moorman AFM. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neuroscience Letters 2003 Mar 13;339(1):62-66.

23. Ruijter JM, Ramakers C, Hoogaars WM, Karlen Y, Bakker O, van den Hoff MJ, et al. Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res 2009 Apr;37(6):e45.

24. Sokolovic A, Montenegro-Miranda PS, de Waart DR, Cappai RM, Duijst S, Sokolovic M, et al. Overexpression of insulin like growth factor binding protein 5 reduces liver fibrosis in chronic cholangiopathy. Biochim Biophys Acta 2012 Jun;1822(6):996-1003.

25. Latasa MU, Gil-Puig C, Fernandez-Barrena MG, Rodriguez-Ortigosa CM, Banales JM, Urtasun R, et al. Oral methylthioadenosine administration attenuates fibrosis and chronic liver disease progression in Mdr2-/- mice. PLoS One 2010;5(12):e15690.

26. Sokolovic M, Wehkamp D, Sokolovic A, Vermeulen J, Gilhuijs-Pederson LA, van Haaften RI, et al. Fasting induces a biphasic adaptive metabolic response in murine small intestine. BMC Genomics 2007;8:361.

27. Sanz S, Pucilowska JB, Liu S, Rodriguez-Ortigosa CM, Lund PK, Brenner DA, et al. Expression of insulin-like growth factor I by activated hepatic stellate cells reduces fibrogenesis and enhances regeneration after liver injury. Gut 2005 Jan;54(1):134-141.

28. Skrtic S, Wallenius V, Ekberg S, Brenzel A, Gressner AM, Jansson JO. Hepatocyte-stimulated expression of hepatocyte growth factor (HGF) in cultured rat hepatic stellate cells. J Hepatol 1999 Jan;30(1):115-124.

29. Dimai HP, Linkhart TA, Linkhart SG, Donahue LR, Beamer WG, Rosen CJ, et al. Alkaline phosphatase levels and osteoprogenitor cell numbers suggest bone formation may contribute to peak bone density differences between two inbred strains of mice. Bone 1998 Mar;22(3):211-216.

30. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005 Feb;115(2):209-218. 31. Iredale JP, Benyon RC, Arthur MJ, Ferris WF, Alcolado R, Winwood PJ, et al. Tissue inhibitor of

metalloproteinase-1 messenger RNA expression is enhanced relative to interstitial collagenase messenger RNA in experimental liver injury and fibrosis. Hepatology 1996 Jul;24(1):176-184.

32. Murphy FR, Issa R, Zhou X, Ratnarajah S, Nagase H, Arthur MJ, et al. Inhibition of apoptosis of activated hepatic stellate cells by tissue inhibitor of metalloproteinase-1 is mediated via effects on matrix metalloproteinase inhibition: implications for reversibility of liver fibrosis. J Biol Chem 2002 Mar 29;277(13):11069-11076.

33. Hsiao Y, Zou T, Ling CC, Hu H, Tao XM, Song HY. Disruption of tissue-type plasminogen activator gene in mice aggravated liver fibrosis. J Gastroenterol Hepatol 2008 Jul;23(7 Pt 2):e258-e264.

34. Popov Y, Patsenker E, Fickert P, Trauner M, Schuppan D. Mdr2 (Abcb4)-/- mice spontaneously develop severe biliary fibrosis via massive dysregulation of pro- and antifibrogenic genes. J Hepatol 2005 Dec;43(6):1045-1054.

35. Hemmann S, Graf J, Roderfeld M, Roeb E. Expression of MMPs and TIMPs in liver fibrosis - a systematic review with special emphasis on anti-fibrotic strategies. J Hepatol 2007 May;46(5):955-975.

36. Li YY, McTiernan CF, Feldman AM. Interplay of matrix metalloproteinases, tissue inhibitors of metalloproteinases and their regulators in cardiac matrix remodeling. Cardiovasc Res 2000 May;46(2):214-224.

37. Nagase H, Woessner JF, Jr. Matrix metalloproteinases. J Biol Chem 1999 Jul 30;274(31):21491-21494. 38. Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function,

and biochemistry. Circ Res 2003 May 2;92(8):827-839. 39. Onori P, Alvaro D, Floreani AR, Mancino MG, Franchitto A, Guido M, et al. Activation of the IGF1 system

characterizes cholangiocyte survival during progression of primary biliary cirrhosis. J Histochem Cytochem 2007 Apr;55(4):327-334.

40. Cadoret A, Rey C, Wendum D, Elriz K, Tronche F, Holzenberger M, et al. IGF-1R contributes to stress-induced hepatocellular damage in experimental cholestasis. Am J Pathol 2009 Aug;175(2):627-635.

41. Spirli C, Okolicsanyi S, Fiorotto R, Fabris L, Cadamuro M, Lecchi S, et al. Mammalian target of rapamycin regulates vascular endothelial growth factor-dependent liver cyst growth in polycystin-2-defective mice. Hepatology 2010 May;51(5):1778-1788.

42. Hulzebos CV, Voshol PJ, Wolters H, Kruit JK, Ottenhof R, Groen AK, et al. Bile duct proliferation associated with bile salt-induced hypercholeresis in Mdr2 P-glycoprotein-deficient mice. Liver Int 2005 Jun;25(3):604-612.

43. Hulzebos CV, Voshol PJ, Wolters H, Kruit JK, Ottenhof R, Groen AK, et al. Bile duct proliferation associated with bile salt-induced hypercholeresis in Mdr2 P-glycoprotein-deficient mice. Liver Int 2005 Jun;25(3):604-612.

44. Trauner M, Fickert P, Wagner M. MDR3 (ABCB4) defects: a paradigm for the genetics of adult cholestatic syndromes. Semin Liver Dis 2007 Feb;27(1):77-98.

IGF1 in Abcb4-/-

79

c h a p t e r 5

fasting reduces liver fibrosis

in mouse model

for chronic cholangyopathies

chapter 5

82

ABSTRACT

Background: Chronic cholangiopathies often lead to fibrosis, as a result of a perpetuated wound

healing response, characterized by increased inflammation and excessive deposition of proteins of

the extracellular matrix. Our previous studies have shown that food deprivation suppresses the

immune response, which led us to postulate its beneficial effects on pathology when liver fibrosis

driven by portal inflammation.

Methods: We investigated the consequences of fasting on liver fibrosis in Abcb4-/- mice that

spontaneously develop it due to a lack of phospholipids in bile. The effect of up to 48 hours of food

deprivation was studied by gene expression profiling, (immuno)histochemistry, and biochemical

assessments of biliary output, and hepatic and plasma lipid composition.

Results: Bile composition in Abcb4-/- mice remained unchanged with fasting, in contrast to

increased biliary output in the wild type counterparts. Markers of inflammation and cell proliferation

dramatically decreased in livers of Abcb4-/- mice already after 12 hours of fasting. Reduced presence

of activated hepatic stellate cells and actively increased tissue remodeling further propelled a decrease

in parenchymal fibrosis in fasting.

Conclusion:This study is the first to show that food deprivation positively influences liver pathology

in a fibrotic mouse model, opening a door for new strategies to improve liver regeneration in chronic

disease.

under revision

Authors

Aleksandar Sokolović1, Cindy P.A.A. van Roomen2, Roelof Ottenhoff2, Saskia Scheij2, Johan K.

Hiralall1, Nike Claessen3, Jan Aten3, Ronald P.J. Oude Elferink1, Albert K. Groen4 and Milka

Sokolović2

Affiliations

1Tytgat Institute for Liver and Intestinal Research, 2Department of Medical Biochemistry, 3Pathology

Department, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands;

4Department of Paediatrics, Center for Liver, Digestive and Metabolic Diseases, University Medical

Centre, Groningen, The Netherlands

Acknowledgements

The authors would like to thank dr. D.R. de Waart for hydroxyproline measurement.

fibrosis in fasting

83

INTRODUCTION

Chronic cholangiopathies, like primary sclerosing cholangitis (PSC), are caused by progressive

inflammation and scarring of the bile ducts in the liver (1). The inflammation impedes the flow of

bile to the gut, which can lead to liver fibrosis (and cirrhosis as its ultimate stage), the final common

phase of all chronic liver diseases. In general, liver injury is followed by a wound healing response,

characterized by compensatory proliferation, inflammation and deposition of extracellular matrix

(ECM) (2). Initially, fibrosis is reversible - normal architecture is restored by fibrolysis, and ECM

producing cells are removed by apoptosis (3, 4). Recurrent chronic injury, however, results in an

imbalance between fibrogenesis and fibrolysis, leading to scar formation, architectural distortion,

cirrhosis and liver failure (5, 6). Liver transplantation is the only effective treatment, but surgical

contraindications and the lack of donors urge for interventions that halt disease progression.

Starvation provokes a massive response of the body and an interorgan integration of the

activities to prevent an irreversible loss of resources (7). When the reserves are low, there is necessity

for a trade-off between the risk of starvation and disease (8, 9). It is not surprising therefore that the

suppression of the immune response is also one of the major transcriptomics adaptations to food

deprivation (10, 11). Given inflammation as a driving force in the pathogenesis of biliary fibrosis

(12), we postulated that starvation-induced immune suppression could alleviate the pathology in

sclerosing cholangitis. To assess the effect of fasting on liver fibrosis, we used the Abcb4-/- mice in

which the lack of ABCB4 (ATP-binding cassette, subfamily b, member 4) leads to an absence of

phosphatidylcholine in the bile, making it toxic. (13). Bile salts accumulate in the intrahepatic biliary

system disrupting tight junctions and basement membranes of bile ducts, and lead to non-

suppurative inflammatory cholangitis, periductal fibrosis shortly after birth, and ductular proliferation

(14). Results of our study of fibrotic Abcb4-/- mouse livers, indeed, confirmed the hypothesis that

food deprivation influenced pathology, and revealed an ameliorating effect on fibrosis of as short as

12 hours of fasting.

MATERIALS AND METHODS

Animals

We used 3 months old male Abcb4-/- (FVB background), and WT FVB mice as a control (Charles

River, Maastricht, The Netherlands). The animals were kept separately in regular cages, and in grid

bottom cages 24 h prior to and during fasting, to prevent coprophagia and intake of bedding. The

study was approved by the AMC Animal Experiments Committee.

Plasma and tissue collection and analytical procedures

The gallbladders in anesthetized mice (Hypnorm and Diazepam, 1 ml/kg and 10 mg/kg respectively)

were cannulated; bile was collected for 15 minutes and stored at –20°C. Blood sample was collected

by cardiac puncture and plasma was stored at -20°C. The liver was quickly excised and parts were

snap-frozen in liquid nitrogen and stored at -80°C, or fixed in 4% buffered formalin and embedded

in paraffin. Total liver collagen was determined by measuring hydroxyproline content (15). Plasma

concentrations of total cholesterol, triglycerides, FFA and choline-containing PL were determined

using the equivalent colourimetric enzymatic kits (Wako Chemicals GmbH, Neuss, Germany and

chapter 5

84

bioMérieux Benelux, Boxtel, The Netherlands). Biliary BS and PL content were analyzed as

described (16). Biliary cholesterol was measured fluorescently (17). Tissue lipids were extracted using

a chloroform–methanol-based (2:1 by volume) method (18); cholesterol and triglycerides were

measured using the CHOD-PAP (Ecoline 25 cholesterol, Merck, Darmstadt, Germany) and GPO-

PAP method (Ecoline S+, DiaSys Diagnostic Systems, Holzheim, Germany), and PL was measured

with an enzymatic colourimetric assay (Biolabo, Maizy, France). To correct the obtained lipid values

for the amount of tissue, the protein content of the liver was measured using the BCA method

(Pierce, Perbio Science Nederland BV, Etten-Leur, The Netherlands).

Histology and immunohistochemistry

Four animals per time point and five sections per animal were analyzed. Paraffin sections were

dewaxed and stained with hematoxylin and eosin (HE) for general histology, or with 0.2% picro-

sirius red (PSR) to detect fibrillar collagen. Immunohistology was performed as described (19).

Sections were incubated with rat IgG2b anti-mouse F4/80 monoclonal antibody (AbD Serotec,

Oxford, UK) and Ki67 (Dako, Glostrup, Denmark). Primary antibodies were detected using

appropriate secondary horseradish peroxidase-conjugated polyclonal goat IgG antibodies, directed

against mouse IgG2a, or rat IgG2a (Southern Biotech, Birmingham, AL). Bound peroxidase activity

was visualized using H2O2 and 3,3’-diaminobenzidine as chromogen. Sections were counterstained

with hematoxylin.

Reverse transcription, quantitative PCR and data analyses

Total RNA was extracted from frozen livers with TRIzol reagent (Invitrogen, Breda, The

Netherlands), and its quality assessed with RNA 6000 Nano LabChip® Kit in an Agilent 2100

bioanalyzer (Agilent Technologies, Palo Alto, USA). qPCR was performed as described (20). The

average of three house-keeping genes (36b4, Hprt and cyclophilin b) was used to normalize the

expression of the studied genes. mRNA concentration was calculated using the LinRegPCR program

(21, 22).

Western blot analysis

was performed as described (20), using antibodies against MMP2 and MMP9 (goat anti-mouse; Bio-

Rad Laboratories, Veenendaal, The Netherlands), with goat anti-rabbit IgG horseradish peroxidase

conjugated (Santa Cruz Biotechnology) as secondary antibody. Chemiluminescence was quantified by

Lumi-Imager F1 using CDP-Star (Roche, Mannheim, Germany), using β-actine as a reference.

Statistical analysis

ANOVA and two-tailed Student’s t-test were used to determine statistical significance. P<0.05 is

denoted by asterisks in all figures, with error bars representing the standard deviation.

RESULTS

Fasting reduces ECM deposition in the livers of Abcb4-/- mice

To investigate the effects of fasting on the fibrotic livers in Abcb4-/- mice, we assessed the

connective tissue by picro-Sirius red staining (Fig.1A). Parenchymal fibrosis was clearly reduced

already after 12h of fasting, causing a decrease of inter-portal bridging.

fibrosis in fasting

85

Figure 1: Food deprivation ameliorates parenchymal fibrosis in the livers of Abcb4-/- mice. A) Representative Picro-sirius red immunostaining of liver sections of wild-type (WT) and Abcb4-/- mice, fed or starved for 12, 24 and 48h. Scale bar represents 100 µm. B) Hydroxyproline concentration (expressed in μg/mg wet tissue) in Abcb4-/- and WT livers, 0, 12, 24h and 48h after starvation started. C-I) Relative mRNA expression level of collagens 1, 3, 4 and fibronectin 1 (components of ECM), αSma and vimentin (markers of activated HSC), and Gfap (marker of quiescent HSC), in livers at 0, 12, 24 and 48h of fasting. mRNa expression is normalized by 3 house-keeping genes and shown as fold change compared with WT control at 0h. Asterisks denote significance compared to fed condition in the corresponding strain (P<0.05), and error bars represent SD.

Total liver collagen, responsible for the structural integrity of the ECM and used as a marker for liver fibrogenesis (23), was determined by measuring hydroxyproline content. In Abcb4-/- mice it decreased 40, 70 and 80% at 12, 24 and 48h of fasting, respectively, and was not affected in WT animals (Fig.1B).mRNA expression of collagens 1 and 3 was lower in fasting in both Abcb4-/- and WT mice (Fig.1C-D). Expression of Col4 and fibronectin decreased at 12 and 24h (Fig.1E-F). These data strongly suggest a decrease in matrix deposition in response to fasting in Abcb4-/- mice.

Hepatic stellate cells (HSC) play a central role in ECM remodeling by production of MMPs

and TIMPs (24, 25), and by deposition of fibrillar collagens type 1 and 3 (25). The impact of fasting

was analyzed by measuring the markers for activated (α-Sma and vimentin) and quiescent HSC (Gfap).

While quiescent were more abundant in the WT (Fig.1I), activated HSC were dominating in fibrotic

Abcb4-/- livers (Fig.1G-H). A significant decrease in expression levels was observed for all three

markers in Abcb4-/- mice, regardless of the duration of fasting (Fig.1F-I). In WT mice, these markers

were downregulated after 12 and 24h of fasting. In support, the expression of fibronectin, which

decreases starvation-induced apoptosis in rat HSC (26), was downregulated.

chapter 5

86

Clearly different response at 48h compared to previous time points does not come as a

surprise (27), since such a long stretch in mice (comparable to several weeks in humans) provokes a

complex survival response. Although we mainly focus on the effect of short and moderate fasting,

we provide the later time point to complete the picture of the response in fibrotic livers.

Figure 2: Starvation causes an

active matrix remodeling in

Abcb4-/- livers. A-D) Gene

expressions of Plat, Plau, Timp1 and

Tgfβ1 at 0, 12, 24h and 48h of fasting

in Abcb4-/- and WT mice. mRNa

expression is normalized by 3 house-

keeping genes and shown as fold

change compared with WT control at

0h. Asterisks denote significance

compared to fed condition in the

corresponding strain (P<0.05), and

error bars represent SD.

Fibrotic Abcb4-/- livers respond to fasting by matrix remodeling

To assess the mechanism behind the reduced amount of ECM, we studied the components of the

matrix remodeling machinery.

The expression levels of Plau and Plat, which activate MMPs (28), increased in prolonged

fasting in Abcb4-/- mice (80% Plat and 30% Plau; Fig.2A-B), after an initial decrease in short fasting.

In WT animals, food deprivation did not affect their gene expression. Timp1, which inhibits most

MMP activity (29) and promotes the survival of HSC (30), was significantly downregulated by fasting

in fibrotic livers at all time points (Fig.2C). Tgfβ1, which promotes hepatic fibrosis by prompting

HSC differentiation into myofibroblasts, was 2-fold higher in fed fibrotic than in WT livers (Fig.2D).

Compared to fed Abcb4-/-, Tgfβ1 expression was 75, 70 and 50% lower at three fasting time points,

respectively. Matrix remodeling in fasting, therefore, seems to be regulated by reduced expression of

profibrotic cytokine Tgfβ1 and thereby reduced number of activated HSC/MFs.

Hepatic cell turnover is strongly reduced by fasting

Abcb4-/- mice have an intrinsically high hepatocyte proliferation (31). We therefore analyzed the

effect of fasting on proliferation markers Pcna and Ki67. They decreased in both Abcb4-/- and WT

mice (Fig.3B-C; Pcna returned to control values in prolonged fasting). Immunostaining (Fig.3A)

confirmed that the fibrotic livers of Abcb4-/- mice contained cycling hepatocytes and Ki67-positive

cells in the portal tract. Ki67-positive cells were not found in fasted Abcb4-/- livers at any time point,

or in any WT livers.

fibrosis in fasting

87

Figure 3: Fasting reduces originally high cell turnover in fibrotic livers of Abcb4-/- mice. A) Immunostaining for

Ki67 in the liver sections of fed and 12h fasted WT and Abcb4-/- mice. Scale bar represents 100 µm. B-D) Gene

expression of Ki67, Pcna and p21 upon 0, 12, 24 and 48h of food deprivation. mRNa expression is normalized by 3

house-keeping genes and shown as fold change compared with WT control at 0h. Asterisks denote significance

compared to fed condition in the corresponding strain (P<0.05), and error bars represent SD.

Fasting, therefore, was a strong enough stimulus to reduce proliferation in Abcb4-/- livers.

This was apparently supported by the increased cell-cycle arrest, since the expression of p21,

involved in repair of DNA damage, was strongly upregulated in Abcb4-/- mice (even 700% at 48h;

Fig.3D).

The frequency of apoptotic changes (Councilman bodies) did not change dramatically in

Abcb4-/- mice upon fasting, especially in acinar zones 2 (Fig.4A).

Similarly, mRNA concentrations of proapoptotic markers Bax and Apaf barely changed in Abcb4-/-

mice (Fig.4B-C), while in the WT they significantly decreased with fasting (40-60%). Expression of

antiapoptotic Bcl-xl was unaffected by fasting in fibrotic livers, though it has strongly increased in

prolonged fasting in WT mice (Fig.4D). This suggests a possible different mechanism behind

apoptosis between the highly proliferative livers of fed Abcb4-/-mice, and those affected by fasting

Biliary output and hepatic and plasma lipid composition in fasted Abcb4-/- mice

In contrast to the earlier reported increased biliary output in fasted WT mice (32), bile flow in

Abcb4-/- mice tended to decrease after 12h (40%, P<0.06; Table 1). As expected, PL were barely

measurable in Abcb4-/- animals. Cholesterol concentration was 2.7-6.5 fold lower during fasting

compared to WT, while none of the measured biliary secretion rates – BS, cholesterol or PL- was

affected by fasting, indicating that the Abcb4-/- mice had lost the adaptive biliary response to fasting

seen in WT animals.

chapter 5

88

Figure 4: Apoptosis in fasted fibrotic livers. A) Hematoxylin-eosin staining of liver sections of fed and fasted WT and

Abcb4-/- mice. Scale bar represents 100 µm. B-D) Relative mRNA expression level of Bax, Apaf and Bcl-xl, after 0, 12,

24h and 48h of fasting. mRNa expression is normalized by 3 house-keeping genes and shown as fold change compared

with WT control at 0h. Asterisks denote significance compared to fed condition in the corresponding strain (P<0.05),

and error bars represent SD.

Cholesterol concentration in the liver of Abcb4-/- mice increased during fasting (Table 1),

hepatic triglyceride concentration was strongly increased, while that of PL did not change. Hepatic

TG and cholesterol concentrations were somewhat lower in fed Abcb4-/- mice than in their WT

counterparts, but fasting altered them so that both lipids were present in higher concentration in

Abcb4-/- mice.

To assess the lipid status in the systemic circulation of fasted Abcb4-/- mice, we measured the

plasma concentrations of cholesterol, TG, FFA and PL (Table 1). Similarly to WT, plasma

cholesterol concentration increased after 12 and 24h of fasting, TG concentration strongly decreased

in longer fasting, PL concentrations remained stable, while FFA concentration increased (50 and

25% after 12 and 24h). Initially, plasma cholesterol level in fed Abcb4-/- mice was two-fold lower

than in the WT, and remained such also in fasting.

fibrosis in fasting

89

Table 1: Biliary output and hepatic and plasma lipid composition in up to 48 hours fasted Abcb4-/- and wild type mice. The values are given as mean ± S.D. Asterisks denote significance (P<0.05)

Fasting reduces hepatic inflammation in Abcb4-/- mice

Bile duct proliferation and portal inflammation (13) are the most prominent histopathological

features of Abcb4-/- mice and are caused by secretion of toxic bile (33, 34). To assess the effect of

fasting on portal inflammation, we analyzed the presence of inflammatory cells in livers of Abcb4-/-

and WT mice.

Markers for activated monocytes and macrophages (Cd11b; Cd11c), neutrophils (Mpo), circulating monocites (Mcp1), and liver-resident Kupffer cells (F4/80), as well as inflammatory marker Tnfα, were clearly (and expectedly) higher in fibrotic than in control livers (Fig.5A-E; Mpo and Mcp1 not shown). Inflammation seemed attenuated in fasted fibrotic livers by significantly lowered expression of general markers F4/80, Cd11b and Cd11c. In fibrotic livers of fed Abcb4-/- mice, immunostaining for F4/80 showed areas with abundant periportal inflammation, with large numbers of F4/80 positive macrophages and relatively few F4/80 positive Kupffer cells in the adjacent lobules (Fig.5A, top right panel). More Kupffer cells were present in the areas where periportal inflammation was less abundant. A striking shift in localization of F4/80 macrophages was observed at 12 and 24h of fasting (Fig.5A, two mid-right panels). The periportal tracts contained far less F4/80 macrophages, whereas their number was increased in acinar zone 2. Upon 24 and 48h of fasting the number of macrophages was further reduced, with adherent F4/80 macrophages now often present in central veins (Fig.5A, two bottom-right panels). In WT mice, the relative number of F4/80 positive Kupffer cells decreased only after 48h of starvation (Fig.5A, left panels). A decrease in expression of proinflammatory marker Irf5 that promotes inflammatory macrophage polarization

Abcb4-/- WT

0 h 12 h 24 h 48 h 0 h 12 h 24 h 48 h

biliary lipids

bile flow (µl/min.100g)

9.5 ± 1.64 5.6 ± 0.290.06 6.6 ± 1.00 7.4 ± 1.14 3.7 ± 0.9 6.5 ± 0.4* 6.2 ± 0.8* 5.1 ± 0.6

bile salts (nmol/min.100g)

272 ± 87 197 ± 30 260 ± 50 398 ± 85 168 ± 36 267 ± 27* 228 ± 7 367 ± 107*

cholesterol

(nmol/min.100g) 0.5 ± 0.16 0.3 ± 0.04 0.5 ± 0.15 0.8 ± 0.14 1.3 ± 0.2 2.2 ± 0.2* 2.1 ± 0.1 3.0 ± 0.5*

phospholipids (nmol/min.100g)

0.16 ± 0.12 0.01 ± 0.01 0.08 ± 0.04 0.22 ± 0.04 3.0 ± 0.7 8.0 ± 1.3* 6.5 ± 0.4* 14.4 ± 2.6*

hepatic lipids

triglycerides

(nmol/mg) 5.7 ± 0.8 15.2 ± 0.8* 17.6 ± 1.2* 12.8 ± 2.4* 7.9 ± 1.1 11.7 ± 1.6* 13.5 ± 2.3* 12.2 ± 1.0*

cholesterol (nmol/mg)

8.1 ± 0.2 11.9 ± 0.5* 12.7 ± 1.0* 13.2 ± 0.8* 10.6 ± 0.3 10.5 ± 0.4 9.5 ± 0.4 17.4 ± 1.1*

phospholipids (nmol/mg)

19.2 ± 1.2 21.2 ± 0.8 23.0 ± 1.2 21.8 ± 1.2 20.6 ± 0.6 19.2 ± 0.6 21.5 ± 0.6 21.0 ± 0.5

plasma lipids

cholesterol

(mmol/L) 2.2 ± 0.03 2.5 ± 0.1* 2.5 ± 0.1* 2.2 ± 0.2 3.8 ± 0.2 4.3 ± 0.1* 4.5 ± 0.1* 4.9 ± 0.2*

triglicerides (mmol/L)

1.7 ± 0.1 1.8 ± 0.1 1.1 ± 0.07* 0.55 ± 0.04* 1.4 ± 0.2 0.9± 0.1* 0.6 ± 0.1* 0.8 ± 0.1*

free fatty acids (mmol/L)

0.53 ± 0.02 0.77 ± 0.05* 0.66 ± 0.04* 0.64 ± 0.11 1.3 ± 0.3 1.0 ± 0.1 1.0 ± 0.1 1.4 ± 0.3

phospholipids (mmol/L)

2.0 ± 0.03 2.1 ± 0.1 2.0 ± 0.1 1.8 ± 0.1 3.6 ± 0.2 3.5 ± 0.1 3.6 ± 0.1 3.8 ± 0.2

chapter 5

90

(35), and an increase of Ym1, a marker for alternatively activated macrophages, indicate a shift towards alternatively activated macrophages in fasted fibrotic livers.

Figure 5: Fasting affects inflammatory status and macrophage localization in fibrotic livers. A) Representative Immunostainings for F4/80 in the livers of fed and fasted WT and Abcb4-/- mice. B-G) Relative mRNA level of F4/80, Tnfα, Cd11b, Cd11c, Irf5 and Ym1 upon 0, 12, 24 and 48h of fasting. mRNa expression is normalized by 3 house-keeping genes and shown as fold change compared with WT control at 0h. Asterisks denote significance compared to fed condition in the corresponding strain (P<0.05), and error bars represent SD.

DISCUSSION

This study reveals that fasting causes a reduction of liver fibrosis in a mouse model for biliary

cholangiopathies. Rapid adaptive response to food deprivation in Abcb4-/- mice (already after 12h)

lowered inflammation, decreased hepatocyte proliferation, lowered the number of activated

HSC/MFs, decreased production of ECM components, and increased expression of genes involved

in tissue remodeling, leading ultimately to amelioration of liver fibrosis.

Collagen, responsible for the structural integrity of the ECM, was strongly affected by

fasting. Within 48 hours, total collagen synthesis in livers of Abcb4-/- mice (measured by

hydroxyproline) decreased to 20%, similarly to the response found in the articular cartilage of fasted

guinea pigs (36). This corresponds with a notion that collagen production is sensitive to changes in

food intake, and that malnutrition may have profound effects on its production (37). Amount and

the types of produced collagen are regulated by the nutritional state of the animal or by the disease

processes - either directly as in wound healing, inflammation, and fibrosis, or indirectly, as in

starvation and diabetes. In advanced starvation, collagen is one of the two major sources of energy

fibrosis in fasting

91

(proteins) in the body (38). The regulation of collagen mass in fasted fibrotic livers occurred,

however, already in early stages of starvation, before the other energy sources were depleted. This

points to an active ECM remodeling, rather than to an ad hoc degradation for mere provision of

amino acids. The most prominent drop was noticed in expression of fibril-forming collagens 1 and 3

and vimentin, predominantly synthesized by activated HSC (39), suggesting that the process of liver

remodeling was specifically induced by food deprivation.

Chronic liver diseases are characterized by aberrant matrix deposition, calling for our

attention to the role of ECM in resolution of liver fibrosis. Tissue remodeling is regulated by MMPs,

involved in the ECM degradation, and TIMPs, their endogenous inhibitors. Their subtle balance

maintains liver fibrogenesis. Tissue homeostasis is further regulated by proteolytic activity of the

PLAU/PLAT/plasmin, responsible for the maintenance of the physiologic levels of ECM (40).

PLAU promotes ECM degradation through activation of MMPs (MMP-2, -3 and -9; (41, 42),

increases the differentiation of hepatic stem cells, and HGF-dependent regeneration of hepatocytes

(43). PLAT protects ECM proteins from proteolytic degradation and helps expedite wound healing

(44). In fed Abcb4-/- mice, hepatic Plau and Plat expression levels were much higher than in their WT

counterparts (3- and 8-fold, respectively). Short and moderate fasting initially reduced the expression

of both, but prolonged fasting posed a demand for higher mRNA concentrations, indicating an

ongoing balancing action between fibrolysis and fibrogenesis. Together with strongly (6-fold)

reduced Timp1 expression, our data indicate that fibrotic livers responded to food deprivation by an

active ECM remodeling. Matrix remodeling and in fasting was further supported by reduced

expression Mmp13. Expression of this collagenase mainly produced by Kupffer cells is high in

collagen 1a1 r/r mouse model that fails to recover from fibrotic liver injury, due to resistance to

MMP degradation of ECM (45). However, given the contradictory findings by other groups (46), the

role of Mmp13 in (resolution of) liver fibrosis remains controversial.

Apart from ECM synthesis and degradation, tissue remodeling comprises cell proliferation,

death, and migration. In general, hepatocyte proliferation is low under normal conditions, except to

compensate for a loss of cells (47) and it gets further reduced when the energy supply is low (48).

Caloric restriction induces apoptosis and decreases hepatocyte proliferation in a murine strain with a

high incidence of spontaneous liver tumors (49). We now show that, as a part of tissue remodeling,

fibrotic livers of Abcb4-/- mice respond to fasting mainly by a decrease in their pathologically high

proliferation. Furthermore, consistently with previously shown increased apoptosis in HSCs in

response to nutrient deprivation (26), this study shows a decrease in number of activated and

quiescent stellate cells, most likely in response to decreased production of fibronectin and TIMP1,

which both prevent HSCs apoptosis (26, 30).

The impressive decrease in inflammation in Abcb4-/- mice, the macrophage migration from the

periportal tract towards the central vein, and the shift from classically to alternatively activated

macrophages, seem a likely driving force for improved pathology (50). Overall, altered

concentrations of cytokines, metabolic and hormonal trophic factors induced by fasting could

decrease inflammation in fibrotic livers. Notably, profibrotic cytokine leptin that increases TIMP1

expression (51) is strongly reduced by fasting (52). Fasting increases circulating corticosterone (53,

54), decreasing thereby the production of cytokines and interleukins (55, 56), with an

immunosuppressive effects in humans (57, 58) and mice (59, 60). In Abcb4-/- mice, persistent hepatic

chapter 5

92

inflammation triggers profibrotic signaling via activation of cytokines, promoting the formation of

MFs, which in turn synthetize elevated amounts of ECM proteins (61, 62). TGF-β1, the most potent

fibrogenic cytokine, prompts HSC differentiation into myofibroblasts, by enhancing expression of

TIMPs (that block ECM degradation), and by directly stimulating synthesis of interstitial fibrillar

collagens (12). Liver fibrosis in Abcb4-/- mice was attenuated in fasting, most likely driven by

downregulation of Tgfβ1, which reduced number of activated HSC. Downregulation of a number of

proinflammatory markers, a.o. Tnfα and Irf5 (a transcriptional activator of pro-inflammatory

cytokines and chemokines (35)), clearly points to fasting-induced suppression of immune response.

The change in number, but also localization of resident and infiltrated macrophages, may be causal in

resolution of the parenchymal fibrosis (e.g. by altered production of MMPs (46, 63), especially given

the pivotal but divergent roles of macrophages in matrix remodelling - favoring ECM accumulation

during ongoing injury, and enhancing matrix degradation during recovery (64). In the starved

animals, the high energy demand of the immune response (e.g. for production of acute phase

response proteins) leads to a massive change in hepatic transcriptome directed towards its

suppression (11, 50). The decrease in inflammation in Abcb4-/- mice, is therefore compatible with

our previous notion that the suppression of immune response (subserving energy preservation) is

one of the highlights of the overall body’s fasting response (50).

In conclusion, this study demonstrates that fasting leads to alleviation of biliary fibrosis by

decreased inflammation and by actively increased matrix remodeling. Fasting in the fibrotic Abcb4-/-

model may help improve understanding of the mechanisms of resolution, and inform strategies to

improve liver regeneration in chronic disease.

References

1. Hirschfield GM, Heathcote EJ, Gershwin ME. Pathogenesis of cholestatic liver disease and therapeutic

approaches. Gastroenterology 2010 Nov;139(5):1481-1496.

2. Friedman SL. Liver fibrosis -- from bench to bedside. J Hepatol 2003;38 Suppl 1:S38-S53.

3. Benyon RC, Arthur MJ. Extracellular matrix degradation and the role of hepatic stellate cells. Semin Liver Dis

2001 Aug;21(3):373-384.

4. Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a modulator of hepatic fibrogenesis. Semin

Liver Dis 2001 Aug;21(3):351-372.

5. Friedman SL. Mechanisms of disease: Mechanisms of hepatic fibrosis and therapeutic implications. Nat Clin

Pract Gastroenterol Hepatol 2004 Dec;1(2):98-105.

6. Pinzani M, Rombouts K. Liver fibrosis: from the bench to clinical targets. Dig Liver Dis 2004 Apr;36(4):231-

242.

7. Diamond J. The double puzzle of diabetes. Nature 2003 Jun 5;423(6940):599-602.

8. Hanssen SA, Hasselquist D, Folstad I, Erikstad KE. Costs of immunity: immune responsiveness reduces

survival in a vertebrate. Proc Biol Sci 2004 May 7;271(1542):925-930.

9. Houston AI, McNamara JM, Barta Z, Klasing KC. The effect of energy reserves and food availability on

optimal immune defence. Proc Biol Sci 2007 Nov 22;274(1627):2835-2842.

10. Hakvoort TB, Moerland PD, Frijters R, Sokolovic A, Labruyere WT, Vermeulen JL, et al. Interorgan

coordination of the murine adaptive response to fasting. J Biol Chem 2011 May 6;286(18):16332-16343.

11. Martin SA, Douglas A, Houlihan DF, Secombes CJ. Starvation alters the liver transcriptome of the innate

immune response in Atlantic salmon (Salmo salar). BMC Genomics 2010;11:418.

12. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005 Feb;115(2):209-218.

fibrosis in fasting

93

13. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice with

homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of nonsuppurative

inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994 Nov;145(5):1237-1245.

14. Fickert P, Fuchsbichler A, Wagner M, Zollner G, Kaser A, Tilg H, et al. Regurgitation of bile acids from leaky

bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology 2004 Jul;127(1):261-

274.

15. van Westrhenen R., de Waart DR, Akman S, Krediet RT. Assessment of peritoneal fibrosis by conventional

light microscopy and hydroxyproline measurements. Perit Dial Int 2004 May;24(3):290-292.

16. Frijters CM, Ottenhoff R, van Wijland MJ, Van NC, Groen AK, Oude Elferink RP. Influence of bile salts on

hepatic mdr2 P-glycoprotein expression. Adv Enzyme Regul 1996;36:351-363.

17. Carey MC, Small DM. The physical chemistry of cholesterol solubility in bile. Relationship to gallstone

formation and dissolution in man. J Clin Invest 1978 Apr;61(4):998-1026.

18. FOLCH J, LEES M, SLOANE STANLEY GH. A simple method for the isolation and purification of total

lipides from animal tissues. J Biol Chem 1957 May;226(1):497-509.

19. Bijl N, Sokolovic M, Vrins C, Langeveld M, Moerland PD, Ottenhoff R, et al. Modulation of glycosphingolipid

metabolism significantly improves hepatic insulin sensitivity and reverses hepatic steatosis in mice. Hepatology

2009 Nov;50(5):1431-1441.

20. Sokolovic A, Sokolovic M, Boers W, Elferink RP, Bosma PJ. Insulin-like growth factor binding protein 5

enhances survival of LX2 human hepatic stellate cells. Fibrogenesis Tissue Repair 2010;3:3.

21. Ruijter JM, Ramakers C, Hoogaars WM, Karlen Y, Bakker O, van den Hoff MJ, et al. Amplification efficiency:

linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res 2009 Apr;37(6):e45.

22. Ruijter JM, Pfaffl MW, Zhao S, Spiess AN, Boggy G, Blom J, et al. Evaluation of qPCR curve analysis methods

for reliable biomarker discovery: Bias, resolution, precision, and implications. Methods 2012 Sep 3.

23. Guanabens N, Pares A, Alvarez L, Martinez de Osaba MJ, Monegal A, Peris P, et al. Collagen-related markers

of bone turnover reflect the severity of liver fibrosis in patients with primary biliary cirrhosis. J Bone Miner Res

1998 Apr;13(4):731-738.

24. Knauper V, Lopez-Otin C, Smith B, Knight G, Murphy G. Biochemical characterization of human collagenase-

3. J Biol Chem 1996 Jan 19;271(3):1544-1550.

25. Takahara T, Furui K, Yata Y, Jin B, Zhang LP, Nambu S, et al. Dual expression of matrix metalloproteinase-2

and membrane-type 1-matrix metalloproteinase in fibrotic human livers. Hepatology 1997 Dec;26(6):1521-1529.

26. Rodriguez-Juan C, de la Torre P, Garcia-Ruiz I, Diaz-Sanjuan T, Munoz-Yague T, Gomez-Izquierdo E, et al.

Fibronectin increases survival of rat hepatic stellate cells--a novel profibrogenic mechanism of fibronectin. Cell

Physiol Biochem 2009;24(3-4):271-282.

27. Sokolovic M, Sokolovic A, Wehkamp D, Ver Loren van TE, de Waart DR, Gilhuijs-Pederson LA, et al. The

transcriptomic signature of fasting murine liver. BMC Genomics 2008;9:528.

28. Lee SR, Guo SZ, Scannevin RH, Magliaro BC, Rhodes KJ, Wang X, et al. Induction of matrix

metalloproteinase, cytokines and chemokines in rat cortical astrocytes exposed to plasminogen activators.

Neurosci Lett 2007 Apr 24;417(1):1-5.

29. Taras D, Blanc JF, Rullier A, Dugot-Senant N, Laurendeau I, Vidaud M, et al. Pravastatin reduces lung

metastasis of rat hepatocellular carcinoma via a coordinated decrease of MMP expression and activity. J Hepatol

2007 Jan;46(1):69-76.

30. Murphy FR, Issa R, Zhou X, Ratnarajah S, Nagase H, Arthur MJ, et al. Inhibition of apoptosis of activated

hepatic stellate cells by tissue inhibitor of metalloproteinase-1 is mediated via effects on matrix

metalloproteinase inhibition: implications for reversibility of liver fibrosis. J Biol Chem 2002 Mar

29;277(13):11069-11076.

31. van Nieuwkerk CM, Elferink RP, Groen AK, Ottenhoff R, Tytgat GN, Dingemans KP, et al. Effects of

Ursodeoxycholate and cholate feeding on liver disease in FVB mice with a disrupted mdr2 P-glycoprotein gene.

Gastroenterology 1996 Jul;111(1):165-171.

32. Sokolovic M, Sokolovic A, van Roomen CP, Gruber A, Ottenhoff R, Scheij S, et al. Unexpected effects of

fasting on murine lipid homeostasis - Transcriptomic and lipid profiling. J Hepatol 2010 Mar 4.

chapter 5

94

33. Smit JJ, Schinkel AH, Oude Elferink RP, Groen AK, Wagenaar E, van DL, et al. Homozygous disruption of

the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver

disease. Cell 1993 Nov 5;75(3):451-462.

34. van Nieuwerk CM, Groen AK, Ottenhoff R, van WM, van den Bergh Weerman MA, Tytgat GN, et al. The role

of bile salt composition in liver pathology of mdr2 (-/-) mice: differences between males and females. J Hepatol

1997 Jan;26(1):138-145.

35. Krausgruber T, Blazek K, Smallie T, Alzabin S, Lockstone H, Sahgal N, et al. IRF5 promotes inflammatory

macrophage polarization and TH1-TH17 responses. Nat Immunol 2011 Mar;12(3):231-238.

36. Spanheimer RG, Peterkofsky B. A specific decrease in collagen synthesis in acutely fasted, vitamin C-

supplemented, guinea pigs. J Biol Chem 1985 Apr 10;260(7):3955-3962.

37. Spanheimer R, Zlatev T, Umpierrez G, DiGirolamo M. Collagen production in fasted and food-restricted rats:

response to duration and severity of food deprivation. J Nutr 1991 Apr;121(4):518-524.

38. Berg RA, Kerr JS. Nutritional aspects of collagen metabolism. Annu Rev Nutr 1992;12:369-390.

39. Friedman SL. Hepatic fibrosis -- overview. Toxicology 2008 Dec 30;254(3):120-129.

40. Zaman AK, French CJ, Schneider DJ, Sobel BE. A profibrotic effect of plasminogen activator inhibitor type-1

(PAI-1) in the heart. Exp Biol Med (Maywood ) 2009 Mar;234(3):246-254.

41. Bueno M, Salgado S, Beas-Zarate C, Armendariz-Borunda J. Urokinase-type plasminogen activator gene therapy

in liver cirrhosis is mediated by collagens gene expression down-regulation and up-regulation of MMPs, HGF

and VEGF. J Gene Med 2006 Nov;8(11):1291-1299.

42. Sun C, Li DG, Chen YW, Chen YW, Wang BC, Sun QL, et al. Transplantation of urokinase-type plasminogen

activator gene-modified bone marrow-derived liver stem cells reduces liver fibrosis in rats. J Gene Med 2008

Aug;10(8):855-866.

43. Shimizu M, Hara A, Okuno M, Matsuno H, Okada K, Ueshima S, et al. Mechanism of retarded liver

regeneration in plasminogen activator-deficient mice: impaired activation of hepatocyte growth factor after Fas-

mediated massive hepatic apoptosis. Hepatology 2001 Mar;33(3):569-576.

44. Hertig A, Berrou J, Allory Y, Breton L, Commo F, Costa De Beauregard MA, et al. Type 1 plasminogen

activator inhibitor deficiency aggravates the course of experimental glomerulonephritis through overactivation

of transforming growth factor beta. FASEB J 2003 Oct;17(13):1904-1906.

45. Kallis YN, Robson AJ, Fallowfield JA, Thomas HC, Alison MR, Wright NA, et al. Remodelling of extracellular

matrix is a requirement for the hepatic progenitor cell response. Gut 2011 Apr;60(4):525-533.

46. Uchinami H, Seki E, Brenner DA, D'Armiento J. Loss of MMP 13 attenuates murine hepatic injury and fibrosis

during cholestasis. Hepatology 2006 Aug;44(2):420-429.

47. Kelly MD, Styles JA, Pritchard NR. Analysis of cytological changes in hepatocytes from rats dosed with 3'-

methyl-4-dimethylaminoazobenzene: initial response appears to involve cytokinesis of binucleated cells. Cancer

Lett 1990 Aug;53(1):1-4.

48. Lu MH, Warbritton A, Tang N, Bucci TJ. Dietary restriction alters cell proliferation in rats: an

immunohistochemical study by labeling proliferating cell nuclear antigen. Mech Ageing Dev 2002

Feb;123(4):391-400.

49. James SJ, Muskhelishvili L. Rates of apoptosis and proliferation vary with caloric intake and may influence

incidence of spontaneous hepatoma in C57BL/6 x C3H F1 mice. Cancer Res 1994 Nov 1;54(21):5508-5510.

50. Hakvoort TB, Moerland PD, Frijters R, Sokolovic A, Labruyere WT, Vermeulen JL, et al. Interorgan

coordination of the murine adaptive response to fasting. J Biol Chem 2011 May 6;286(18):16332-16343.

51. Lin S, Saxena NK, Ding X, Stein LL, Anania FA. Leptin increases tissue inhibitor of metalloproteinase I

(TIMP-1) gene expression by a specificity protein 1/signal transducer and activator of transcription 3

mechanism. Mol Endocrinol 2006 Dec;20(12):3376-3388.

52. Frederich RC, Lollmann B, Hamann A, Napolitano-Rosen A, Kahn BB, Lowell BB, et al. Expression of ob

mRNA and its encoded protein in rodents. Impact of nutrition and obesity. J Clin Invest 1995 Sep;96(3):1658-

1663.

53. Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos-Flier E, et al. Role of leptin in the

neuroendocrine response to fasting. Nature 1996 Jul 18;382(6588):250-252.

54. Faggioni R, Moser A, Feingold KR, Grunfeld C. Reduced leptin levels in starvation increase susceptibility to

endotoxic shock. Am J Pathol 2000 May;156(5):1781-1787.

fibrosis in fasting

95

55. Pruett SB, Fan R, Myers LP, Wu WJ, Collier S. Quantitative analysis of the neuroendocrine-immune axis: linear

modeling of the effects of exogenous corticosterone and restraint stress on lymphocyte subpopulations in the

spleen and thymus in female B6C3F1 mice. Brain Behav Immun 2000 Dec;14(4):270-287.

56. Pruett SB, Fan R. Quantitative modeling of suppression of IgG1, IgG2a, IL-2, and IL-4 responses to antigen in

mice treated with exogenous corticosterone or restraint stress. J Toxicol Environ Health A 2001 Feb

9;62(3):175-189.

57. Munck A, Guyre PM, Holbrook NJ. Physiological functions of glucocorticoids in stress and their relation to

pharmacological actions. Endocr Rev 1984;5(1):25-44.

58. Khansari DN, Murgo AJ, Faith RE. Effects of stress on the immune system. Immunol Today 1990

May;11(5):170-175.

59. Klebanov S, Diais S, Stavinoha WB, Suh Y, Nelson JF. Hyperadrenocorticism, attenuated inflammation, and

the life-prolonging action of food restriction in mice. J Gerontol A Biol Sci Med Sci 1995 Mar;50(2):B78-B82.

60. Nakamura H, Kouda K, Tokunaga R, Takeuchi H. Suppressive effects on delayed type hypersensitivity by

fasting and dietary restriction in ICR mice. Toxicol Lett 2004 Feb 2;146(3):259-267.

61. Krenning G, Zeisberg EM, Kalluri R. The origin of fibroblasts and mechanism of cardiac fibrosis. J Cell Physiol

2010 Nov;225(3):631-637.

62. Zeisberg EM, Kalluri R. Origins of cardiac fibroblasts. Circ Res 2010 Nov 26;107(11):1304-1312.

63. Oviedo-Orta E, Bermudez-Fajardo A, Karanam S, Benbow U, Newby AC. Comparison of MMP-2 and MMP-9

secretion from T helper 0, 1 and 2 lymphocytes alone and in coculture with macrophages. Immunology 2008

May;124(1):42-50.

64. Duffield JS, Forbes SJ, Constandinou CM, Clay S, Partolina M, Vuthoori S, et al. Selective depletion of

macrophages reveals distinct, opposing roles during liver injury and repair. J Clin Invest 2005 Jan;115(1):56-65.

c h a p t e r 6

closure:

summarizing discussion and conclusions

summaries

abbreviations

chapter 6

98

SUMMARIZING DISCUSSION

SUBJECT AND AIM

Fibrosis is a wound healing response to a variety of acute and/or chronic stimuli (ethanol, viral

infection, drugs and toxins, cholestasis, and metabolic disease) (1, 2). It develops due to an increase

in fibrillar collagen synthesis and deposition, along with insufficient remodeling (3, 4). This process is

associated with a number of pathological and biochemical changes that lead to structural and

metabolic abnormalities, and increased hepatic scarring (5, 6). The progression of liver fibrosis leads

to cirrhosis, characterized by distortion of the normal architecture, formation of septae and nodule,

altered blood flow, portal hypertension, HCC, and ultimately liver failure (7). Understanding the

mechanisms and the pathways involved in the pathogenesis of the fibrogenic response and

identification of potentially new therapeutic agents could provide novel therapeutic approaches for

liver diseases in which fibrosis is a detrimental component.

The overarching aim of the studies described in this thesis, therefore, was to improve the

understanding of the mechanisms underlying liver fibrosis, and to identify ways to slowdown the

development and/or reduce liver fibrosis.

APPROACHES

To live up to the challenge, we used a battery of experimental models and approaches. In our in vitro

studies, we used LX2 cells, a model system for activated HSCs, and primary MFs. In vivo, we studied

Abcb4-/- mice, a model for chronic cholangiopathies that spontaneously develop liver fibrosis, as well

as Abcb4-/- SMP8-Igf1+/-, a transgenic model created in one of the studies. These various models

were challenged either by directly affecting the expression of genes of interest (IGFBP5, IGF1R, and

IGF1) to scrutinize the potential roles of these gene products in the development of liver fibrosis.

Protein expression levels were perturbed in vitro using adenoviral vector for overexpression, or

siRNA for depletion. In vivo, we either used an AAV vector to deliver IGFBP5 to the hepatocytes of

Abcb4-/- mice, or crossbred these mice with a transgenic strain to ensure IGF1-overexpression in

activated HSCs. To test the hypothesis that nutrient deprivation benefits the fibrotic pathology, we

studied the effects of fasting on existing liver fibrosis in Abcb4-/- mice. A broad range of molecular

biology and biochemical techniques were employed, coupled with transcriptomics and system

genetics approaches.

FINDINGS

Igf axis in vitro and in vivo

Our first effort to shed some light on the role of IGFBP5 in liver fibrosis has revealed its

involvement in the survival of HSCs. Earlier studies have indicated cell- and tissue-type specific

effects of IGFBP5. Its increased expression in fibrotic lung, skin and liver (8-11), in myogenesis, in

vitro breast cancer cell growth, and in gingival epithelial cells (12-15), seems to be coupled with

promotion of cell survival. Quite oppositely, in e.g. mammary gland involution IGFBP5 promotes

closure

99

apoptosis of epithelial cells in vivo and in vitro (16-19). In liver fibrosis, apoptosis is reduced in

activated HSCs that play a pivotal role in the initiation and perpetuation of pathology (20). Since

IGFBP5 is spontaneously induced in HSCs cells upon activation (8), we hypothesized that it played a

role in proliferation of HSCs in fibrotic liver. Chapter 2 presents a study of the effects of IGFBP5 in

vitro in primary human MFs and LX2 cells, a model for partially activated human HSCs (21-23). In

both cell lines an increased presence of IGFBP5, either by lentiviral overexpression or by addition of

recombinant protein, promoted survival by lowering apoptosis. In support, silencing of endogenous

IGFBP5 decreased the viability of LX2 cells, and of primary hepatic MFs, by an increase in

apoptosis, pointing at IGFBP5 as a pro-survival factor in these two pro-fibrotic cell types.

The same study has addressed another important question raised whenever an IGF-binding

factor is studied, namely what is the net effect of the interplay between IGF1 and IGFBP5? IGFBP5

binds IGF1 with high affinity and protects it from rapid degradation (24). Simultaneously, however,

it prevents activation of the IGF1R and thereby inhibits prosurvival signaling by IGF1. Our in vitro

studies showed that, in contrast to IGFBP5 that acted via lowering apoptosis, IGF1 increased

proliferation of LX2 cells. Moreover, we found no effect of IGF1 on IGFBP5 action, nor of

IGFBP5 on IGF1 signaling, suggesting that in LX2 cells these two factors exert their effect via

different, independent, routes. Several mechanisms may be involved in the IGF1 independent effect

of IGFBP5. It was first shown in vitro that IGFBP5 could increase activation of TGFβ1, a major pro-

fibrotic cytokine, by plasmin via its interaction with tissue plasminogen activator (25). Consistently,

mice overexpressing IGFBP5 had elevated tissue concentrations of plasmin (25). IGFBP5 directly

activates tPA, which in turn activates plasmin. This further leads to activation of several MMPs

involved in cellular migration/invasion, and in activation of TGFβ1 (25, 26). IGFBP-5 can also

interact with TNFR1 (in vivo and in vitro) and block cell proliferation via inactivation of the NFκB

signalling pathway (27). Furthermore, IGFBP5 interacts directly with several proteins from the

matricellular family of proteins (osteopontin, thrombospondin-1 and tenascin-C) (28, 29), which are

implicated in cellular adhesion, migration, wound healing or metastasis. In breast cancer cells

IGFBP5 induces integrin-mediated Cdc42-dependent cell survival pathway, which leads to an

adhesive, anti-migratory, phenotype (involves stimulation of integrin-linked kinase (ILK), Akt and

p53) and restricts epithelial–mesenchymal transgressions. As such, IGFBP5 might play role in

limiting metastasis (30). IGFBP5 provokes formation of a filamin A (FLNa)-based nuclear shuttle

that binds IGFBP5, recruits transcription factors, translocates to the nucleus and induces laminin

(major component of the epithelial basement membrane involved in cell migration) gene

transcription (31). In addition, IGFBP5 contains a nuclear localization sequence that mediates

nuclear translocation of IGFBP5 to influence transcription of genes involved in osteoblast cell

proliferation/differentiation (32). Evidence suggests that IGFBP5 binds and phosphorylates a

putative IGFBP5 receptor on the osteoblast cell surface and stimulates downstream signaling

pathways (33).

Much to our surprise, the following in vivo study (presented in Chapter 3) revealed an

antifibrotic effect of IGFBP5 in the livers of Abcb4-/- mice. Our initial finding that IGFBP5

expression strongly increased during development of liver fibrosis in Abcb4-/- mice (8), that IGFBP5

acts as a prosurvival signal in vitro in LX2 cells, and two recent studies that reported a high expression

of IGFBP5 in HCC and intrahepatic cholangiocarcinoma, suggested that it could have a profibrotic

role in vivo, and promote the survival of cancer cells and tumor formation (34, 35). The current study

chapter 6

100

has, however, shown a clearly reduced liver fibrosis in IGFBP5-treated animals, with considerably

reduced portal bridging, and demonstrated the role of IGFBP5 in reduction of the existing scar

tissue. To clarify the different role of IGFBP5 in liver and other organs, we analyzed its steady-state

expression in the BxD genetic reference population (Chapter 3). The gene-phenotype correlations

indicated an organ-specific regulation of IGFBP5, with a role that differed between organs but

possibly protective in the liver. This could explain why IGFBP5 was found a profibrotic factor in

lung and skin (10, 36), as opposed to its protective effect in liver fibrosis. A protective effect of

IGFBP5 on hepatocytes in vivo (as shown by our research) is consistent with its prosurvival effect

seen in activated HSC in vitro (37). Hence, even within the liver the prosurvival signal in stellate cells

vs. hepatocytes may have counteractive effects. Apparently, in the in vivo setting of the Abcb4-/-

mouse the protective effect on hepatocytes prevails over the increased survival of stellate cells.

The most likely mechanism of the anti-fibrotic effect of IGFBP5 expression in the Abcb4-/-

mice is reduction of portal inflammation, which is initially caused by leakage of toxic bile, one of the

most prominent pathological features in this model for chronic cholangiopathy (38). Overexpression

of IGFBP5 lowers the influx of inflammatory cells into the area, lowering the production of

inflammatory cytokines, known to stimulate collagen synthesis and to enhance hepatocyte damage.

Fewer damaged hepatocytes and fewer activated inflammatory cells means less oxidative stress (39,

40) that enhances the expression of profibrotic genes in human HSCs and MFs (41, 42), which may

additionally alleviate liver fibrosis.

IGFBP5, however, also may have a more direct effect. The increased amount of p21 in

IGFBP5-overexpressing livers may be instrumental for the reduced proliferation. p21 serves as a

negative cell cycle regulator under stress conditions caused by various factors (growth factor

deficiency, DNA damage, and exposure to heavy metals or antiproliferative cytokines, particularly

TGFβ). Upon DNA damage, p21 blocks the transition from G1 into S phase or from G2 into

mitosis, enabling the repair of damaged DNA (43). Furthermore it regulates PCNA that binds to

damaged DNA (44) reducing its nuclear presence as seen in our study. In the cytoplasm, p21 protein

has an anti-apoptotic effect through binding to and inhibiting caspase 3, as well as the apoptotic

kinases ASK1 and JNK (45). p21 can act as a tumor suppressor, in particular by participating in the

launch of a senescence program (replicative senescence as well as stress-induced premature

senescence) (46) More and more evidence points to IGFBP5 as an important player in cellular

senescence, process that allows an escape from uncontrolled cell proliferation, thereby halting

tumorigenesis (47, 48). Regardless the cause of senescence, two tumor suppressor pathways, the p53-

p21 and p16-pRB, are typically responsible for the following growth arrest and senescence (48). Also

several cytokines, including IFNα, IFNγ, TGFβ, IGFBP3, IGFBP5 have been reported to regulate

senescence (49-52). Increased IGFBP5 expression is associated with senescence in vitro in pulmonary

fibrosis, human dermal fibroblasts and endothelial cells (53-57). Knockdown of IGFBP5 partially

reverses senescence in aged human umbilical-vein endothelial cells (HUVEC), whereas IGFBP5

induces and accelerates premature senescence in young HUVEC cells. IGFBP5-induced senescence

is associated with induction of the tumour suppressor p53 (51).

Following the path in the IGF-axis upstream of IGFBP5, in Chapter 4 we tested the

influence IGF1 on liver fibrosis in (a model for) chronic cholangiopaties. It has been shown by two

independent studies, that increased expression of IGF1significantly reduced liver fibrosis in CCl4

closure

101

treated rats (58, 59). In a follow-up phase I-II clinical trial, IGF1 administration improved liver

function in patients suffering from liver cirrhosis (60). The potential use of IGF1 in chronic

cholangiopathies, however, has not been studied. Since the biliary epithelium expresses IGF1R (61),

and IGF1 protects cholangiocytes against cholestatic injury in vitro (62), we postulated that it could be

beneficial in chronic cholangiopathy. To study it, we crossbred the Abcb4-/- mice with a transgenic

mouse that overexpresses IGF1 in activated HSCs. This cross showed increased expression of pro-

fibrotic factors (Tgfβ and Pdgf), an increased hydroxyproline level (2-fold higher than previously

reported (63)), as well as an enhanced deposition of ECM. IGF1 overexpression also increased the

expression of Timp1, which could reduce MMP activity and ECM degradation in our model.

These observations all point to a more prominent fibrosis in Abcb4-/- in the presence of

IGF1. The increased serum ALP and bilirubin levels after 12 weeks of cholate diet in the current

study indicate that IGF1 could increase proliferation and survival of cholangiocytes, and

independently cause cholestasis in hepatocytes. Though these responses are aimed at repair of the

bile duct epithelium, increase of IGF1 above the normal levels induces extensive proliferation

resulting in bile duct abnormalities. Although unexpected, the different effect of IGF1

overexpression between the previously studied CCl4 model (58, 59) and this model for chronic

cholangiopathy is most likely due to the difference in severity of the models. To clarify this, we have

performed bile duct ligation (BDL) in the SMP8-IGF1. The data revealed an absence of protective

effect of overexpression of IGF1 also in BDL model. In this model, however, no difference in bile

duct proliferation was seen between the WT and the SMP8-IGF1 mice.

A difference in pathogenesis between the models may explain the discrepancy between

Abcb4-/- mice and BDL-induced mouse models of liver fibrosis. Ligation of the common bile duct

results in an acute interstitial (biliary) fibrosis, associated with massive proliferation of bile ducts,

(only rarely observed in man). On the other hand, in Abcb4-/- mice progression of fibrosis is

spontaneous due to cholangiocyte proliferation and massive up-regulation of profibrogenic genes,

and bears more resemblance to human biliary fibrosis (63). The damage in the Abcb4-/- model is

chronic and less severe and localized to the portal areas (64), which results in a persistent increase in

IGF1 signaling enhancing cholangiocyte proliferation (61). In PBC patients suffering from chronic

bile duct damage, activation of the IGF1 system plays a role in the survival of cholangiocytes (65). In

addition, IGF1 enhances the proliferation of liver cystic epithelium and reduction of its expression

reduces the size of liver cysts in a mouse model (66). This may explain the formation of extremely

large, but ill-developed bile-duct structures observed in that model, leading to the formation of

strictures. Presence of strictures (though not specifically demonstrated) could explain the cholestasis

that, according to serum bilirubin and ALP level, has eventually developed in our transgenic mice.

Fasting in abcb4-/- mice

Driven by the data from human and rodent studies (67, 68), in which fibrotic profiles in obese

individuals were improved by weight loss, we postulated that the adaptation to nutrient deprivation

would also be beneficial for fibrosis caused by chronic cholangiopathy. The fasting study in Abcb4-/-

mice (Chapter 5) has indeed exposed a surprisingly quick improvement in pathology of fibrotic liver

- already after 12h of food deprivation. Starvation provokes a massive response of the body to

chapter 6

102

prevent an irreversible loss of resources (69). In mice, within hours after the last meal, the organs

respond with changes in gene expression mainly in general metabolism (70). The role of the liver is

to provide energy for glucose-dependent tissues, by glycogenolysis, gluconeogenesis, ketogenesis,

and fatty-acid β-oxidation (71). The basic architecture of the lobules and the zonation are not

affected, but the cell size declines in prolonged fasting, when murine liver restores partly its glycogen

deposits, and much of gene expression returns to control values (72). In Abcb4-/- mice, collagens,

fibronectin and vimentin, responsible for the structural integrity of the ECM, were strongly affected

by fasting. The downregulation of collagen mass in fasted fibrotic livers occurred in early stages of

starvation, before the other energy sources were depleted, pointing to an active ECM remodeling,

rather than to an ad hoc degradation for mere provision of amino acids. As a part of tissue

remodeling, fibrotic livers of Abcb4-/- mice responded to caloric restriction by a decrease in their

pathologically high proliferation. Liver fibrosis was further attenuated by reduced number of

activated HSCs, most likely driven by downregulation of TGFβ1 (73). The impressive decrease in

inflammation in fasted Abcb4-/- mice, and migration of macrophages from the periportal tract

through the acinar zone 2 towards the central vein, seem a likely driving force for improved

pathology. This change in number and localization of resident and infiltrated macrophages may be

causal in resolution of the parenchymal fibrosis (by altered production of MMPs (74, 75)).

A caveat of the studies presented in this thesis could be seen in the fact that liver fibrosis was

addressed mainly in a single in vivo model (of chronic cholangiopathy). Clearly, the effects of IGFBP5

should be further tested in in additional animal models including a viral administration of IGFBP5 to

the newly established IGF1 overexpressing model of Abcb4-/-. In addition, apart from hepatocytes,

other cell types involved in fibrosis should be specifically targeted in Abcb4-/- mice, using different

specific AAV vectors (HSCs by AAV5. Similarly, the effect of fasting should be studied upon

subjecting animals to short(er) and/or alternate day fasting in more than one fibrotic model. The

mechanism behind the antifibrotic effects seen in this thesis should be further elucidated.

CONCLUSIONS

In a nutshell, the studies presented in this thesis have shown that progression of liver fibrosis in

Abcb4-/- mouse model of chronic cholangiopathies can be halted and even reversed by targeted gene

transfer of IGFBP5, or by starvation.

We have demonstrated different, cell-type specific effects of IGFBP5. Firstly by

overexpressing it in vitro, in activated HSCs and MFs, where it improved survival by reducing

apoptosis (in an IGF1-independent manner), suggesting its possible profibrotic role. However,

prolonged expression of IGFBP5 in vivo, in hepatocytes of Abcb4-/- mice, decreased their

proliferation, reduced inflammation and oxidative stress, and lowered number of activated HSCs and

ECM deposition, thereby evidently reducing liver fibrosis.

The star of the IGF-axis, IGF1, did not live up to the expectations in the Abcb4-/- model. In

a new transgenic mouse model created in our study, increased expression of IGF1 in Abcb4-/- mice

aggravated the pathology by increasing cholangiocyte proliferation, opposite to the anti-fibrotic

action of IGF1 in acute parenchymal liver damage caused by BDL or CCl4. In contrast to drug-

closure

103

induced liver fibrosis, IGF1 administration therefore does not seem an option for treating fibrosis

caused by chronic cholangiopathies.

Furthermore, fasting appeared a strong metabolic prosurvival stimulus that should not be

underestimated. We showed that as short as 12 hours of food deprivation positively influenced

pathology in fibrotic Abcb4-/- mouse liver, by decreasing inflammation, and by actively increasing

matrix remodeling.

These studies were the first to tackle the problem of liver fibrosis by specifically targeting

IGFBP5, and by introducing fasting challenge. Though we have not eliminated the cause of fibrosis

by any of the approaches (i.e. bile composition remained unchanged), we did alleviate its

consequences. This may have opened a door for new therapies of liver fibrosis.

References

1. Anthony PP, Ishak KG, Nayak NC, Poulsen HE, Scheuer PJ, Sobin LH. The morphology of

cirrhosis. Recommendations on definition, nomenclature, and classification by a working group

sponsored by the World Health Organization. J Clin Pathol 1978 May;31(5):395-414.

2. Schuppan D. Structure of the extracellular matrix in normal and fibrotic liver: collagens and

glycoproteins. Semin Liver Dis 1990 Feb;10(1):1-10.

3. Friedman SL. Liver fibrosis -- from bench to bedside. J Hepatol 2003;38 Suppl 1:S38-S53.

4. George J, Rao KR, Stern R, Chandrakasan G. Dimethylnitrosamine-induced liver injury in rats: the

early deposition of collagen. Toxicology 2001 Jan 2;156(2-3):129-138.

5. Bissell DM. Cell-matrix interaction and hepatic fibrosis. Prog Liver Dis 1990;9:143-155.

6. George J, Chandrakasan G. Biochemical abnormalities during the progression of hepatic fibrosis

induced by dimethylnitrosamine. Clin Biochem 2000 Oct;33(7):563-570.

7. Schuppan D, Afdhal NH. Liver cirrhosis. Lancet 2008 Mar 8;371(9615):838-851.

8. Boers W, Aarrass S, Linthorst C, Pinzani M, Elferink RO, Bosma P. Transcriptional profiling reveals

novel markers of liver fibrogenesis: gremlin and insulin-like growth factor-binding proteins. J Biol

Chem 2006 Jun 16;281(24):16289-16295.

9. van Nieuwkerk CM, Elferink RP, Groen AK, Ottenhoff R, Tytgat GN, Dingemans KP, et al. Effects

of Ursodeoxycholate and cholate feeding on liver disease in FVB mice with a disrupted mdr2 P-

glycoprotein gene. Gastroenterology 1996 Jul;111(1):165-171.

10. Yasuoka H, Jukic DM, Zhou Z, Choi AM, Feghali-Bostwick CA. Insulin-like growth factor binding

protein 5 induces skin fibrosis: A novel murine model for dermal fibrosis. Arthritis Rheum 2006

Sep;54(9):3001-3010.

11. Yasuoka H, Zhou Z, Pilewski JM, Oury TD, Choi AM, Feghali-Bostwick CA. Insulin-like growth

factor-binding protein-5 induces pulmonary fibrosis and triggers mononuclear cellular infiltration.

Am J Pathol 2006 Nov;169(5):1633-1642.

12. Cobb LJ, Salih DA, Gonzalez I, Tripathi G, Carter EJ, Lovett F, et al. Partitioning of IGFBP-5

actions in myogenesis: IGF-independent anti-apoptotic function. J Cell Sci 2004 Apr 1;117(Pt

9):1737-1746.

13. Hung PS, Kao SY, Liu CJ, Tu HF, Wu CH, Lin SC. Insulin-like growth factor binding protein-5

enhances the migration and differentiation of gingival epithelial cells. J Periodontal Res 2008 Jun 25.

14. McCaig C, Perks CM, Holly JM. Intrinsic actions of IGFBP-3 and IGFBP-5 on Hs578T breast cancer

epithelial cells: inhibition or accentuation of attachment and survival is dependent upon the presence

of fibronectin. J Cell Sci 2002 Nov 15;115(Pt 22):4293-4303.

chapter 6

104

15. Perks CM, Bowen S, Gill ZP, Newcomb PV, Holly JM. Differential IGF-independent effects of

insulin-like growth factor binding proteins (1-6) on apoptosis of breast epithelial cells. J Cell Biochem

1999 Dec 15;75(4):652-664.

16. Flint DJ, Boutinaud M, Tonner E, Wilde CJ, Hurley W, Accorsi PA, et al. Insulin-like growth factor

binding proteins initiate cell death and extracellular matrix remodeling in the mammary gland.

Domest Anim Endocrinol 2005 Aug;29(2):274-282.

17. Lochrie JD, Phillips K, Tonner E, Flint DJ, Allan GJ, Price NC, et al. Insulin-like growth factor

binding protein (IGFBP)-5 is upregulated during both differentiation and apoptosis in primary

cultures of mouse mammary epithelial cells. J Cell Physiol 2006 May;207(2):471-479.

18. Ning Y, Hoang B, Schuller AG, Cominski TP, Hsu MS, Wood TL, et al. Delayed mammary gland

involution in mice with mutation of the insulin-like growth factor binding protein 5 gene.

Endocrinology 2007 May;148(5):2138-2147.

19. Tonner E, Allan G, Shkreta L, Webster J, Whitelaw CB, Flint DJ. Insulin-like growth factor binding

protein-5 (IGFBP-5) potentially regulates programmed cell death and plasminogen activation in the

mammary gland. Adv Exp Med Biol 2000;480:45-53.

20. Iredale JP. Hepatic stellate cell behavior during resolution of liver injury. Semin Liver Dis 2001

Aug;21(3):427-436.

21. Anan A, Baskin-Bey ES, Bronk SF, Werneburg NW, Shah VH, Gores GJ. Proteasome inhibition

induces hepatic stellate cell apoptosis. Hepatology 2006 Feb;43(2):335-344.

22. Shi YF, Fong CC, Zhang Q, Cheung PY, Tzang CH, Wu RS, et al. Hypoxia induces the activation of

human hepatic stellate cells LX-2 through TGF-beta signaling pathway. FEBS Lett 2007 Jan

23;581(2):203-210.

23. Wang XM, Yu DM, McCaughan GW, Gorrell MD. Fibroblast activation protein increases apoptosis,

cell adhesion, and migration by the LX-2 human stellate cell line. Hepatology 2005 Oct;42(4):935-945.

24. Beattie J, Allan GJ, Lochrie JD, Flint DJ. Insulin-like growth factor-binding protein-5 (IGFBP-5): a

critical member of the IGF axis. Biochem J 2006 Apr 1;395(1):1-19.

25. Sorrell AM, Shand JH, Tonner E, Gamberoni M, Accorsi PA, Beattie J, et al. Insulin-like growth

factor-binding protein-5 activates plasminogen by interaction with tissue plasminogen activator,

independently of its ability to bind to plasminogen activator inhibitor-1, insulin-like growth factor-I,

or heparin. J Biol Chem 2006 Apr 21;281(16):10883-10889.

26. Hayashi H, Sakai T. Biological Significance of Local TGF-beta Activation in Liver Diseases. Front

Physiol 2012;3:12.

27. Hwang JR, Huh JH, Lee Y, Lee SI, Rho SB, Lee JH. Insulin-like growth factor-binding protein-5

(IGFBP-5) inhibits TNF-alpha-induced NF-kappaB activity by binding to TNFR1. Biochem Biophys

Res Commun 2011 Feb 25;405(4):545-551.

28. Clemmons DR, Busby WH, Arai T, Nam TJ, Clarke JB, Jones JI, et al. Role of insulin-like growth

factor binding proteins in the control of IGF actions. Prog Growth Factor Res 1995;6(2-4):357-366.

29. Nam TJ, Busby WH, Jr., Rees C, Clemmons DR. Thrombospondin and osteopontin bind to insulin-

like growth factor (IGF)-binding protein-5 leading to an alteration in IGF-I-stimulated cell growth.

Endocrinology 2000 Mar;141(3):1100-1106.

30. Sureshbabu A, Okajima H, Yamanaka D, Tonner E, Shastri S, Maycock J, et al. IGFBP5 induces cell

adhesion, increases cell survival and inhibits cell migration in MCF-7 human breast cancer cells. J Cell

Sci 2012 Apr 1;125(Pt 7):1693-1705.

31. Abrass CK, Hansen KM. Insulin-like growth factor-binding protein-5-induced laminin gamma1

transcription requires filamin A. J Biol Chem 2010 Apr 23;285(17):12925-12934.

closure

105

32. Amaar YG, Tapia B, Chen ST, Baylink DJ, Mohan S. Identification and characterization of novel

IGFBP5 interacting protein: evidence IGFBP5-IP is a potential regulator of osteoblast cell

proliferation. Am J Physiol Cell Physiol 2006 Mar;290(3):C900-C906.

33. Andress DL. Insulin-like growth factor-binding protein-5 (IGFBP-5) stimulates phosphorylation of

the IGFBP-5 receptor. Am J Physiol 1998 Apr;274(4 Pt 1):E744-E750.

34. Nishino R, Honda M, Yamashita T, Takatori H, Minato H, Zen Y, et al. Identification of novel

candidate tumour marker genes for intrahepatic cholangiocarcinoma. J Hepatol 2008 Aug;49(2):207-

216.

35. Umemura A, Itoh Y, Itoh K, Yamaguchi K, Nakajima T, Higashitsuji H, et al. Association of

gankyrin protein expression with early clinical stages and insulin-like growth factor-binding protein 5

expression in human hepatocellular carcinoma. Hepatology 2008 Feb;47(2):493-502.

36. Pilewski JM, Liu L, Henry AC, Knauer AV, Feghali-Bostwick CA. Insulin-like growth factor binding

proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular

matrix deposition. Am J Pathol 2005 Feb;166(2):399-407.

37. Sokolovic A, Sokolovic M, Boers W, Elferink RP, Bosma PJ. Insulin-like growth factor binding

protein 5 enhances survival of LX2 human hepatic stellate cells. Fibrogenesis Tissue Repair 2010;3:3.

38. Mauad TH, van Nieuwkerk CM, Dingemans KP, Smit JJ, Schinkel AH, Notenboom RG, et al. Mice

with homozygous disruption of the mdr2 P-glycoprotein gene. A novel animal model for studies of

nonsuppurative inflammatory cholangitis and hepatocarcinogenesis. Am J Pathol 1994

Nov;145(5):1237-1245.

39. Greenwel P, Dominguez-Rosales JA, Mavi G, Rivas-Estilla AM, Rojkind M. Hydrogen peroxide: a

link between acetaldehyde-elicited alpha1(I) collagen gene up-regulation and oxidative stress in mouse

hepatic stellate cells. Hepatology 2000 Jan;31(1):109-116.

40. Nieto N. Oxidative-stress and IL-6 mediate the fibrogenic effects of [corrected] Kupffer cells on

stellate cells. Hepatology 2006 Dec;44(6):1487-1501.

41. Bataller R, Sancho-Bru P, Gines P, Brenner DA. Liver fibrogenesis: a new role for the renin-

angiotensin system. Antioxid Redox Signal 2005 Sep;7(9-10):1346-1355.

42. Novo E, Marra F, Zamara E, Valfre di BL, Caligiuri A, Cannito S, et al. Dose dependent and

divergent effects of superoxide anion on cell death, proliferation, and migration of activated human

hepatic stellate cells. Gut 2006 Jan;55(1):90-97.

43. Cazzalini O, Scovassi AI, Savio M, Stivala LA, Prosperi E. Multiple roles of the cell cycle inhibitor

p21(CDKN1A) in the DNA damage response. Mutat Res 2010 Apr;704(1-3):12-20.

44. Cazzalini O, Perucca P, Savio M, Necchi D, Bianchi L, Stivala LA, et al. Interaction of

p21(CDKN1A) with PCNA regulates the histone acetyltransferase activity of p300 in nucleotide

excision repair. Nucleic Acids Res 2008 Mar;36(5):1713-1722.

45. Cmielova J, Rezacova M. p21Cip1/Waf1 protein and its function based on a subcellular localization

[corrected]. J Cell Biochem 2011 Dec;112(12):3502-3506.

46. Zhao B, Benson EK, Qiao R, Wang X, Kim S, Manfredi JJ, et al. Cellular senescence and organismal

ageing in the absence of p21(CIP1/WAF1) in ku80(-/-) mice. EMBO Rep 2009 Jan;10(1):71-78.

47. Kuilman T, Michaloglou C, Mooi WJ, Peeper DS. The essence of senescence. Genes Dev 2010 Nov

15;24(22):2463-2479.

48. Rodier F, Campisi J. Four faces of cellular senescence. J Cell Biol 2011 Feb 21;192(4):547-556.

49. Debacq-Chainiaux F, Pascal T, Boilan E, Bastin C, Bauwens E, Toussaint O. Screening of

senescence-associated genes with specific DNA array reveals the role of IGFBP-3 in premature

senescence of human diploid fibroblasts. Free Radic Biol Med 2008 May 15;44(10):1817-1832.

chapter 6

106

50. Frippiat C, Chen QM, Zdanov S, Magalhaes JP, Remacle J, Toussaint O. Subcytotoxic H2O2 stress

triggers a release of transforming growth factor-beta 1, which induces biomarkers of cellular

senescence of human diploid fibroblasts. J Biol Chem 2001 Jan 26;276(4):2531-2537.

51. Kim KS, Seu YB, Baek SH, Kim MJ, Kim KJ, Kim JH, et al. Induction of cellular senescence by

insulin-like growth factor binding protein-5 through a p53-dependent mechanism. Mol Biol Cell 2007

Nov;18(11):4543-4552.

52. Kim KS, Kang KW, Seu YB, Baek SH, Kim JR. Interferon-gamma induces cellular senescence

through p53-dependent DNA damage signaling in human endothelial cells. Mech Ageing Dev 2009

Mar;130(3):179-188.

53. Sureshbabu A, Okajima H, Yamanaka D, Shastri S, Tonner E, Rae C, et al. IGFBP-5 induces

epithelial and fibroblast responses consistent with the fibrotic response. Biochem Soc Trans 2009

Aug;37(Pt 4):882-885.

54. Garcia CK, Wright WE, Shay JW. Human diseases of telomerase dysfunction: insights into tissue

aging. Nucleic Acids Res 2007;35(22):7406-7416.

55. Kojima H, Kunimoto H, Inoue T, Nakajima K. The STAT3-IGFBP5 axis is critical for IL-6/gp130-

induced premature senescence in human fibroblasts. Cell Cycle 2012 Feb 15;11(4):730-739.

56. Yoon IK, Kim HK, Kim YK, Song IH, Kim W, Kim S, et al. Exploration of replicative senescence-

associated genes in human dermal fibroblasts by cDNA microarray technology. Exp Gerontol 2004

Sep;39(9):1369-1378.

57. Hampel B, Fortschegger K, Ressler S, Chang MW, Unterluggauer H, Breitwieser A, et al. Increased

expression of extracellular proteins as a hallmark of human endothelial cell in vitro senescence. Exp

Gerontol 2006 May;41(5):474-481.

58. Castilla-Cortazar I, Garcia M, Muguerza B, Quiroga J, Perez R, Santidrian S, et al. Hepatoprotective

effects of insulin-like growth factor I in rats with carbon tetrachloride-induced cirrhosis.

Gastroenterology 1997 Nov;113(5):1682-1691.

59. Sobrevals L, Rodriguez C, Romero-Trevejo JL, Gondi G, Monreal I, Paneda A, et al. Insulin-like

growth factor I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to

cirrhosis reversion in rats. Hepatology 2010 Mar;51(3):912-921.

60. Conchillo M, de Knegt RJ, Payeras M, Quiroga J, Sangro B, Herrero JI, et al. Insulin-like growth

factor I (IGF-I) replacement therapy increases albumin concentration in liver cirrhosis: results of a

pilot randomized controlled clinical trial. J Hepatol 2005 Oct;43(4):630-636.

61. Alvaro D, Metalli VD, Alpini G, Onori P, Franchitto A, Barbaro B, et al. The intrahepatic biliary

epithelium is a target of the growth hormone/insulin-like growth factor 1 axis. J Hepatol 2005

Nov;43(5):875-883.

62. Gatto M, Drudi-Metalli V, Torrice A, Alpini G, Cantafora A, Blotta I, et al. Insulin-like growth

factor-1 isoforms in rat hepatocytes and cholangiocytes and their involvement in protection against

cholestatic injury. Lab Invest 2008 Sep;88(9):986-994.

63. Popov Y, Patsenker E, Fickert P, Trauner M, Schuppan D. Mdr2 (Abcb4)-/- mice spontaneously

develop severe biliary fibrosis via massive dysregulation of pro- and antifibrogenic genes. J Hepatol

2005 Dec;43(6):1045-1054.

64. Fickert P, Fuchsbichler A, Wagner M, Zollner G, Kaser A, Tilg H, et al. Regurgitation of bile acids

from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology

2004 Jul;127(1):261-274.

65. Onori P, Alvaro D, Floreani AR, Mancino MG, Franchitto A, Guido M, et al. Activation of the IGF1

system characterizes cholangiocyte survival during progression of primary biliary cirrhosis. J

Histochem Cytochem 2007 Apr;55(4):327-334.

closure

107

66. Spirli C, Okolicsanyi S, Fiorotto R, Fabris L, Cadamuro M, Lecchi S, et al. Mammalian target of

rapamycin regulates vascular endothelial growth factor-dependent liver cyst growth in polycystin-2-

defective mice. Hepatology 2010 May;51(5):1778-1788.

67. Peng L, Wang J, Li F. Weight reduction for non-alcoholic fatty liver disease. Cochrane Database Syst

Rev 2011;(6):CD003619.

68. Promrat K, Kleiner DE, Niemeier HM, Jackvony E, Kearns M, Wands JR, et al. Randomized

controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis. Hepatology 2010

Jan;51(1):121-129.

69. Diamond J. The double puzzle of diabetes. Nature 2003 Jun 5;423(6940):599-602.

70. Hakvoort TB, Moerland PD, Frijters R, Sokolovic A, Labruyere WT, Vermeulen JL, et al. Interorgan

coordination of the murine adaptive response to fasting. J Biol Chem 2011 May 6;286(18):16332-

16343.

71. Cahill GF, Jr. Fuel metabolism in starvation. Annu Rev Nutr 2006;26:1-22.

72. Sokolovic M, Sokolovic A, Wehkamp D, Ver Loren van TE, de Waart DR, Gilhuijs-Pederson LA, et

al. The transcriptomic signature of fasting murine liver. BMC Genomics 2008;9:528.

73. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005 Feb;115(2):209-218.

74. Oviedo-Orta E, Bermudez-Fajardo A, Karanam S, Benbow U, Newby AC. Comparison of MMP-2

and MMP-9 secretion from T helper 0, 1 and 2 lymphocytes alone and in coculture with

macrophages. Immunology 2008 May;124(1):42-50.

75. Uchinami H, Seki E, Brenner DA, D'Armiento J. Loss of MMP 13 attenuates murine hepatic injury

and fibrosis during cholestasis. Hepatology 2006 Aug;44(2):420-429.

chapter 6

108

SUMMARY

As a normal wound-healing response to liver injury, fibrosis is reversible - normal architecture is

restored by fibrolysis, and ECM producing cells are removed by apoptosis. However, as reviewed in

Chapter 1, chronic liver injury increases production of fibrogenic signals by resident and infiltrating

liver cells, causing an imbalance between fibrogenesis and fibrolysis, scar formation, architectural

distortion, cirrhosis, and eventually liver failure. With liver transplantation as the only effective

treatment, the lack of donors and surgical contraindications impel for treatments to halt the

progression of disease.

HSCs, mesenchymal cells vital to hepatic function and its response to injury, play a pivotal

role in the development of liver fibrosis. IGFBP5 was shown to be highly expressed in fibrotic livers

of Abcb4-/- mice. Therefore, in Chapter 2, we examined the influence of IGFBP5 on HSCs and MFs

in vitro. Using gain- and loss-of-function approaches (overexpression by lentiviral transduction, or

silencing by siRNA), we showed that IGFBP5 influenced the survival of human LX2 cells, a model

for (partially) activated HSCs, and of hepatic MFs. Their endurance was improved without enhancing

proliferation, by lowering the level of apoptosis, via an IGF1-independent mechanism. Moreover,

IGFBP5 increased the expression of genes involved in ECM deposition.

The finding that IGFBP5 promotes survival of HSCs and MFs in vitro has led us to

investigate its role in vivo, in Abcb4−/−mice. These mice spontaneously progress to severe biliary

fibrosis, due to absence of biliary phospholipids that leads to primary sclerosing cholangitis. Abcb4-/-

mice are also a model for human MDR3 deficiency, ranging from progressive familial intrahepaic

cholestasis type 3 to adult liver cirrhosis, which makes them an attractive model for testing potential

antifibrotics. In Chapter 3 we demonstrate that prolonged liver-specific overexpression of IGFBP5

alleviated the hepatocyte damage, as demonstrated by improved biomarkers of liver injury, and

decreased their proliferation, possibly by arresting cell cycle, accompanied by senescence.

Furthermore, overexpression of IGFBP5 reduced inflammation, indicated by decreased presence of

markers for infiltrating and resident macrophages, neutrophils and monocytes. Consequential

lowered release of proinflammatory cytokines may explain the decreased oxidative stress in these

livers. The resulting reduced presence of activated HSC/MFs and reduced expression of collagens

led to a decreased amount of ECM, ameliorating pathology in the model for chronic cholangiopathy.

A recent study has indicated that biliary epithelium expresses IGF1R, and that IGF1 protects

cholangiocytes against cholestatic injury in vitro. To establish the effect of IGF1 on the existing

cholestatic injury in vivo, we subjected the Abcb4-/- mice to a prolonged increase in hepatic IGF1

expression, by creating a transgenic animal. Chapter 4 shows that sustained overexpression of IGF1

in fibrotic livers increased cholangiocyte proliferation, enhanced inflammation and reduced matrix

remodeling, bringing about an increase in deposition of scar tissue and progression of liver fibrosis.

IGF1 administration therefore does not seem an option for treating fibrosis caused by chronic

cholangiopathies.

The research presented in Chapters 2-4 has challenged the Abcb4-/- mice directly, by

affecting the expression of genes of interest (IGFBP5 and IGF1), to scrutinize their potential roles in

the development of liver fibrosis. Along the same line, we studied how nutrient deprivation could

closure

109

affect liver fibrosis in the same model. In Chapter 5 we show that food deprivation causes a rapid

adaptive response in Abcb4-/- mice, already after 12h. A striking decrease in inflammation in fasted

Abcb4-/- mice seems a likely driving force for a cascade of events, including decreased hepatocyte

proliferation, lowered number of activated HSCs/MFs, decreased production of ECM components,

and increased expression of genes involved in tissue remodeling.

The studies described in this thesis embarked upon the problem of biliary fibrosis by

delineating the roles of specific players in IGF-axis, and by introducing a fasting challenge. Though

we have not eliminated the cause of fibrosis by any of the approaches (i.e. bile composition remained

unchanged), we did alleviate the consequences, which leaves the door to new therapies for liver

fibrosis slightly ajar.

chapter 6

110

NEDERLANDSE SAMENVATTING

Door schade aan de lever kan littekenvorming in de lever ontstaan, dit proces noemen we fibrose.

De strengen littekenweefsel in de lever bestaan voor een groot deel uit bindweefsel zoals collageen.

Lever fibrose kan beschouwd worden als een ontstekings of wondhelingsproces en hoeft daarom

niet noodzakelijkerwijs tot permanente leverschade te leiden. Collageen bevattende fibrotische

littekens kunnen ook weer opgeruimd worden.

Wanneer er sprake is van constante leverschade gaat er iets mis en wordt de fibrotische

littekenvorming versterkt. In Hoofdstuk 1 wordt beschreven dat door chronische leverschade de

afbraak van collageen niet goed meer verloopt. Hierdoor gaat het proces van littekenvorming verder

waardoor de leverstructuur verwoest wordt, met als gevolg cirrose en uiteindelijk een complete

teloorgang van de lever.

Het is al lang bekend dat stellaatcellen cruciaal zijn voor het proces van lever fibrose. Omdat

wij vonden dat in fibrotische muizenlevers het IGFBP5 eiwit sterker werd aangemaakt hebben we in

Hoofdstuk 2 onderzocht wat het effect van IGFBP5 op stellaatcellen is. We konden aantonen dat

stellaatcellen gestabiliseerd worden door IGFBP5 zonder dat ze sneller gingen delen. Belangrijk is

ook de vinding dat stellaat cellijnen door IGFBP5 meer collageen gaan produceren.

Deze experimenten zijn allemaal gedaan met stellaatcellen in celkweek, de volgende stap was

het onderzoeken van het effect van IGFBP5 in een diermodel van leverfibrose. In Hoofdstuk 3

laten we zien dat de schade aan de lever van muizen met fibrose verminderd kan worden door de

productie van IGFBP5 in de levers te verhogen. In tegenstelling tot onze vindingen in stellaatcellen

in celkweek in Hoofdstuk 2, blijkt in muizen het IGFBP5 eiwit juist minder ontsteking en

collageenvorming in fibrotische levers te veroorzaken.

In de wetenschappelijk literatuur wordt gesuggereerd dat het IGF1 eiwit een andere

belangrijke factor is die de lever tegen schade door fibrose zou kunnen beschermen. In Hoofdstuk 4

beschrijven we daarom het gevolg van het verhogen van de hoeveelheid IGF1 eiwit in de lever van

muizen met fibrose. Tot onze verassing bleek door IGF1 de fibrose juist erger te worden, er was

meer litteken- en collageenvorming in de lever. Het toedienen van IGF1 is dus geen oplossing voor

het probleem van leverfibrose.

In Hoofdstukken 2-4 hebben we leverfibrose in modelsystemen, cellen en muizen,

bestudeerd en proberen te behandelen door de eiwitten IGFBP5 en IGF1. In Hoofdstuk 5 hebben

we een andere weg behandeld door muizen met leverfibrose te laten vasten. We vonden hierbij dat

het verminderen van de voedselinname in muizen met leverfibrose een snelle vermindering van de

leverontsteking tot gevolg had. Al 12 uur na het begin van het dieet konden we aantonen dat er

minder collageen in de lever werd gevormd.

In dit proefschrift hebben we laten zien dat IGFBP5 leverfibrose vermindert en dat IGF1

leverfibrose juist verergert. We hebben ook gevonden dat vasten leverfibrose kan verminderen.

Alhoewel we met deze vindingen de oorzaak van leverfibrose niet aan hebben kunnen pakken, leven

onze resultaten toch goede aanknopingspunten op waarmee wellicht nieuwe geneesmiddelen voor

fibrose gevonden kunnen worden.

closure

111

REZIME

Fibroza kao normalan odgovor na povredu tkiva, je u osnovi reverzibilan proces. Normalna

arhitektura jetre se obnavlja razgradnjom fibroznog matriksa (ožiljačnog tkiva) u procesu zvanom

fibroliza, dok ćelije koje ga produkuju bivaju odstranjene putem apoptoze. Hronično oštećenje jetre,

medjutim, kako je elaborirano u Poglavlju 1, dovodi iznova i iznova do aktiviranja fibrogenih

signala, kako od strane oštećenih parenhimskih ćelija jetre (hepatocita), tako i od strane lokalnih i

infiltrirajućih ćelija imunog odgovora, imajući za posledicu narušavanje fine ravnoteže koja postoji

između fibrogeneze i fibrolize. Taj disbalans izmedju dva procesa dovodi do nagomilavanja

ožiljačnog tkiva, narušavanja normalne arhitekture, ciroza i, na kraju, prestanka rada jetre.

Transplatacija jetre, koja je još uvek jedini efikasni tretman, je suočena sa teško premostivim

izazovima, pre svega manjkom donora, a potom i kontraindikacijama koje prate sam hiruški zahvat,

što nameće potrebu za hitnim razvojem tretmana koji bi zaustavili napredovanje same bolesti.

Poznato je da stelatne ćelije, vitalne za normalno funkcionisanje jetre i njen odgovor na

oštećenje/povredu, imaju vodeću ulogu u razvoju fibroze. Naša prethodna istraživanja su pokazala

da je proizvodnja IGFBP5 proteina značajno povišena u fibrozi kod miševa sa holestazom jetre.

Stoga je istraživanje prikazano u Poglavlju 2 imalo za cilj da prouči specifičan uticaj IGFBP5 na

stelatne ćelije i miofibroblaste. IGFBP5, ispostavilo se, poboljšava njihovo preživljavanje smanjenjem

nivoa apoptoze, ne utičući na ćelijsku deobu. Pokazalo se i da IGFBP5 povećava prisustvo gena

uključenih u proizvodnju proteina ožiljačnog tkiva, ukazujući na njegovu moguću profibrotičku

ulogu.

Ovi rezultati iz eksperimenata u ćelijkoj kulturi su bili uvod u sledeći korak – da se prouči

uloga IGFBP5 u životinjskom modelu fibroze jetre. Korišćeni miševi se karakterišu potpunim

nedostatkom fosfolipida u žuči, što pojačava njena detergentska svojstva, čini je toksičnom i dovodi

do spontanog razvoja fibroze. U Poglavlju 3 smo pokazali da povišeno prisustvo IGFBP5 u jetri,

dovodi do smanjenja oštećenja hepatocita, koje za posledicu ima smirivanje upalnih procesa i,

shodno tome, smanjen oksidativni stres. Kao posledica, došlo je do smanjenja broja aktiviranih

stelatnih ćelija i miofibroblasta, smanjenja količine ekstracelularnog matriksa, što je na kraju dovelo

do neočekivanog ublažavanja patologije, sugerišući da IGFBP5 u jetri, za razliku od izolovanih

stelatnih ćelija, ima antifibrotičku ulogu.

Odnedavno je poznato da epitelijalne ćelije u žučnim kanalima (holangiocite) imaju i receptor

za IGF1 protein i da ih, u ćelijskoj kulturi, IGF1 štiti od oštećenja izazvanog toksičnom žuči. Da bi

smo ustanovili uticaj IGF1 na već postojeće ostećenje u jetri, proizveli smo transgenog miša, sa

povišenom količinom ovog proteina u fibrotičkoj jetri. U Poglavlju 4 smo pokazali da stalna

povišena proizvodnja IGF1 u ovim miševima stimuliše deobu holangiocita, pojačava inflamaciju i

smanjuje remodelovanje matriksa, izazivajući povećano deponovanje ožiljačnog tkiva i napredovanje

fibroze. Tako je, nasuprot očekivanjima baziranim na akutnim modelima fibroze, ovo istraživanje

pokazalo da IGF1 nije najbolji kandidat za tretiranje bilijarne fibroze izazvane hroničnom

holangijopatijom.

U studijama opisanim u Poglavljima 2 - 4 direktno smo uticali na ekspresiju gena od interesa

(IGFBP5 i IGF1), da bi proučili njihovu moguću ulogu u razvoju fibroze. Paralelno, proučavali smo i

chapter 6

112

kako smanjeno unošenje hrane utiče na fibrozu u istom modelu. Rezultati prikazani u Poglavlju 5

ukazuju na vrlo brz adaptivni odgovor na gladovanje. Upadljivo smirivanje zapaljenjskih procesa u

jetrama izgladnjvanih životinja već posle 12 sati je, čini se, zamajac za kaskadu dogadjaja koja

uključuje smanjenje brze deobe hepatocita, inače karakteristične za ove miševe, smanjenje broja

aktiviranih stelatnih ćelija i miofibroblasta, kao i smanjeno skladištenje komponenata ekstracelularnog

matriksa, i povišenu ekspresiju gena uključenih u remodelovanje tkiva.

Istraživanja prikazana u ovoj tezi su, dakle, bila pokušaj da se rasvetli problem bilijarne

fibroze – bilo odredjujući uloge pojedinih gena, bilo uvodeći gladovanje kao izazov za fibrotičnu

jetru. Iako direktan uzrok fibroze nije eliminisan ni jednim od pristupa (tj. sastav žuči ostao je

nepromenjen), posledice fibroze su ublažene, što ostavlja odškrinuta vrata za razvoj novih terapija.

closure

113

ABBREVIATIONS

AAV adeno-associated virus

ABCB4 ATP-binding cassette, sub-family B member 4

ALT alanine aminotransferase

BCL2 B-cell CLL/lymphoma 2

BDL bile duct ligation

BrdU bromo-2’-deoxy-uridine

CB1 endocannabinoid receptor

CC cholangiocarcinoma

CCL4 carbon tetrachloride

CD95 cell surface death receptors

CLD chronic liver diseases

ECM extracellular matrix

EGF epidermal growth factor

F4/80 EGF-like module-containing mucin-like, hormone receptor-like 1

FCS fetal calf serum

FGF2 fibroblast growth factors 2

GFP green fluorescent protein

GSH glutathione

GSSG glutathione disulfide

HCC hepatocellular carcinoma

HGF hepatocyte growth factor

HSCs hepatic stellate cells

IFNα interferon alpha

IFNγ interferon gamma

IGF1 insulin-like growth factor 1

IGF1R insulin-like growth factor 1 receptor

IGFBP5 insulin-like growth factor binding protein 5

ILK integrin-linked kinase

JNKs c-jun N-terminal kinases

chapter 6

114

Mac1 integrin alpha M

MFs myofibroblasts

MMPs matrix metalloproteinases

MPO myeloperoxidase

NAFLD non-alcoholic fatty liver disease

NASH non-alcoholic steatohepatitis

NFκB nuclear factor κB

NGF nerve growth factor

p21 cyclin-dependent kinase inhibitor 1A

p53 tumor protein 53

PCNA proliferating cell nuclear antigen

PDGF platelet-derived growth factor

PDGFβ-R platelet derived growth factor receptor β

PI3 phosphatidylinositol 3-kinases

rIGF1 recombinant insulin-like growth factor 1

rIGFBP5 human recombinant insulin-like growth factor binding protein 5

ROS reactive oxygen species

SAβ-gal senescence-associated β-galactosidase

siRNA small interfering RNA

TGFα transforming growth factor α

TGFβ1 transforming growth factor, beta 1

TIMP1 tissue inhibitor of metalloproteinase 1

TNFα tumor necrosis factor alpha

uPA urokinase-type plasminogen activator

VEGF vascular endothelial cell growth factor

α-SMA actin, alpha 2, smooth muscle

closure

115

personalia

acknowledgements

curriculum vitae

list of publications

personalia

118

ACKNOWLEGEMENTS

The mid-nineties found me in Belgrade, leaving the University at the very last year and setting off to

work, amidst the raging wars, with hyperinflation that made every effort worthless and turned life

into mere survival. The romantic idea from the late eighties of becoming a biologist seemed surreal

and pointless, kind of alien. Ten years later and two thousand kilometers away, hard work and luck

have finally met in the AMC in Amsterdam, and my dream of being a scientist started coming true.

This unexpected adventure would be impossible without special people that I would like to thank –

those that have directly contributed to this thesis, and those who have made the exciting and

eventful road to it worth remembering.

Ronald, thanks for offering me an opportunity to peak into the world of science, for decisive and

proficient guidance, for critical suggestions, for the door always opened for my questions, and for

help in wrapping up this thesis.

Piter, I appreciate your idea to kick off by IGFBP5. Many thanks for daily supervision, for prompt

reviewing of the manuscripts, and for leaving me freedom to grow.

Wout, thanks for showing faith in me by giving me a chance to volunteer, which put a foot in the

door towards the doctorate. Thanks for the constructive suggestions, for contagious enthusiasm, and

for demonstration how to be a Professor (capital letter is not a lapsus clavis).

Distinguished committee members, professors Lamers, Beuers, Prieto, Moshage, Poelstra and

Rothuizen, thanks for your readiness to invest time into reading yet another thesis, and to spend yet

another morning in the Agnietenkapel, as opponents at my defense. Jesus, many thanks for

accepting the invitation to come over from Spain, but also for the successful collaboration and

prompt, useful comments on the manuscript.

Carlos, thanks for the immense effort to organize and perform the experiments in Pamplona and for

all your help during preparation of the manuscript. Thanks for your South European charm and

kindness. I owe my gratitude to all your collaborators from CIMA who took part in IGF1-Abcb4

project.

Milka, my „undercover“ mentor, thanks for the unexpected years of science, for all the techniques

that I have learned, from pipetting to organizing presentations, and, finally, for leading the research

in the last study of this thesis.

Dad, thanks for the countless hours of listening to my (not always clear) ideas, and for transforming

them into drawings that illustrate the thesis. The outcome is amazing! Working together on such an

important project was a special, priceless experience!

Willem, without you there would be a whole different beginning. Thank you for the unforgettable

day crowned by Italian dinner that has marked your retirement and the beginning of my work. I am

glad that I could contribute with a bit of understanding of liver fibrosis in the legendary Mdr2s.

Vassilis, thanks for your endless energy, for the ability to turn ordinary into unusual, for your gift to

make people feel special (if you only want so). Thanks for Nana, moussaka and tzatziki (salad not

personalia

119

sause!), but most of all thanks for Eirikur! I wonder if my witnessing the phone call from Nana (that

you haven’t answered?!) was to blame for accepting to be my paranymph.

Jurgen, thanks for your delicate talent to be at hand when something goes wrong – in science and in

life. Thanks for the Dutch resume, and thanks for eventually accepting to be exposed to Ulrich’s

artillery fire during the defense. I am glad that Wout’s wild geese have opened your gastronomic

world and have led to pleasant afternoons with our families.

I owe huge gratitude to Susanne, for perfect organization and execution of the animal experiments,

for the long hours of patience finished with her typical “no thanks!” Big thanks to: Cindy – for

delicate surgeries, Lysbeth – for willingness to help, Rudi, the fastest man in the world – for

HPLCs, Kam – for chemicals always appearing where they should, Mona – for all the complicated

bureaucratic actions around the defense, Coen – for the opportunity to work with someone whom I

first saw on TV as a rock-star. Honest and warm Dineke, thanks for becoming a friend, even though

you snatched the position in front of my nose (and became the best young researcher of the ALC). I

do hope you’ll stay in science. To Paula, promoter and defender of Portugal against the rest of the

world (that lacks understanding of “our” South European ways), thanks for the immunostainings.

Johan, apart from thanking you for the evident technical help, I am glad that you, quiet and simple,

with a smile and a good word for everybody, have become my friend. Andy (and Anita) thanks for

the wonderful evening and dinners with flawless attention to detail. And, of course, for

demonstration of proverbial German efficiency!

Though I was officially a part of the ALC’s “downstairs”, I lived and worked (with pleasure)

“upstairs”. Dear special Wil and Jacqueline, thanks for your patience and help with the techniques

that refused to cooperate (they outnumber the compliant ones), for help with the microscope and

histology, for an unusually spacious working place, for… Wil and Ron, thanks for making us feel

(even more) welcome. Jan, thanks for the primers, antibodies and infectively good mood. Theo,

Ingrid, thanks for your suggestions and ideas on Wednesday meetings, and always pleasant

encounters. Farah is acknowledged for a reason to trust her SOPs and solutions that I used routinely

(and every so often secretly). Farah and Marijn, thank you for our warm friendship and a chance to

enjoy the exciting mix of Surinamese and Dutch. Youji, our friendship has started oddly, with a

pounding heart due to a tiny innocent tea-leave, and has grown into something special. Thanks for

every moment we spent with you, Ning and the kids!

Many thanks to the colleagues from Biochemistry, Cindy, Roelof, Saskia and Bert (the latter from

the ALC, Biochemistry and Groningen), for their generous help with the “hungry study”, and Jan

Aten from Pathology for the patience with all the painstaking immunostainings.

It is, obviously, impossible to name everyone who has contributed to me writing these

acknowledgements today – all the family members (inherited and acquired) warm enough to make

the distances shorter; friends who have remained so ever since making my school and faculty days

(more) pleasant; or those that could be classified as „others“, who have, e.g. scanned the illustrations

for this thesis (tens of times!). I have to mention but a few who have left a mark on creation of the

thesis in a very special way. To all the others, simply many thanks!

personalia

120

Among people who can help, rare and special are the ones that are also willing to. Mile Ivanović is

one of them. Mile, thanks to your help our life is so different than it would be without my

graduation!

Jan, thanks for the cozy months of our Dutch life together, and even cozier years of friendship.

Golub, thanks for your quiet, unpretentious support during these thirtyfive years of friendship,

though I was the one who has “spoiled” you! Siki, our quarter of a century of friendship was neither

quiet not unpretentious, but it was certainly special. Thanks for choosing to spend time with us when

you could have been at more exciting places. Dare, you are a rare soul with enough readiness and

energy to go along with all my actions! Thanks for always having answers, for your friendship and

support for life.

Anne and Arvid, thanks for the warm Norwegian hospitality during my first months abroad, for the

long years of friendship that have followed, and for the countless (unearned) birthday presents. I

wish my gratitude could reach grandma Elsa, for the long talks at the time when my English

vocabulary contained no more than fifty words.

Nočka and Øistein, thanks for pushing that vocabulary to finally start growing in my fourth (?!)

decade. Even more, thanks for being there for us even when we did not think we needed help.

Jojka and Vujo, I wish I could express my pleasure and gratitude for you taking such a special place

in my life. Thanks for becoming such a family!

Mom, thanks for your watchfulness, for all the care and infinite kindness for me. Dad, thanks for

understanding my (extreme) ideas and desires. Thank you both for the uplifting feeling of being

unconditionally loved! Also thanks for high expectations, for always raising the bar, and for the

support for getting past the obstacles. Duca, my dear “little” brother with infinite charm and subtle

sensitivity for the world around you, thanks for all the love and patience that you must have had to

live (and survive) next to me!

Milka, thanks for all the incredibly rich and exciting years since I first saw you in Belgrade’s

Botanical Garden! For all the compassion, understanding, patience and determination that you have

woven into our life, so we could be where we are now. Especially for your courage and persistence to

turn one dream into Mia. Mia, thanks for unveiling an entire new world to me, and for letting me

adore you!

personalia

121

ZAHVALNICA

Kada sam sredinom devedesetih u Beogradu, na poslednjoj godini fakulteta, u vrtlogu ratova koji su

besneli oko nas, sa hiperinflacijom koja je obesmišljavala rad i svodila život na preživljavanje, odlučio

da ostavim studije i počnem da radim, romantična ideja o bavljenju biologijom s kraja osamdesetih

postala je nerealana i besmislena, nekako tuđa. Desetak godina kasnije i dve hiljade kilometara dalje, u

amsterdamskom AMC-u, naporan rad i sreća su se konačno susreli i moj mladalački san o nauci

počeo da biva realnost. Ova potpuno neočekiva avantura ne bi bila moguća bez posebnih ljudi

kojima bih voleo da se zahvalim - onima koji su direktno doprineli nastanku ove teze, i onima koji su

učinili da ovaj uzbudljiv i pun obrta put postane i ostane vredan pamćenja.

Ronalde, hvala na prilici da zavirim u svet nauke, na odlučnom i stručnom usmeravanju, na

kritičnim sugestijama, na vratima uvek otvorenim za moja pitanja i na pomoći da se teza privede

kraju.

Piter, uz zahvalnost za početnu ideju o IGFBP5, hvala za dnevno rukovođenje, za brzu recenziju

napisanog i za slobodu da rastem.

Vaute, hvala na pruženom poverenju i prilici da volontiram u tvojoj grupi, koja je otškrinula vrata u

doktorat. Hvala za konstruktivne sugestije, za zarazni entuzijazam i pokaznu vežbu kako biti

Profesor (veliko slovo nije omaška).

Uvaženim članovima komisije, profesorima Lamers, Bojers, Prieto, Moshage, Pulstra i

Rothauzen, hvala na spremnosti da ulože vreme u još jedno iščitavanje još jedne teze i da provedu

(još jedno) prepodne u Agnietenkapel, kao oponenti tokom moje odbrane. Hesusu posebno hvala

što je prihvatio poziv da za tu priliku doputuje iz Španije, kao i na uspešnoj saradnji i brzim i

konstruktivnim komentarima tokom pisanja rada.

Karlos, hvala za lavovski trud oko organizacije i izvođenja experimenata u Pamploni i za svu pomoć

prilikom pisanja rada. Hvala za južnoevropski šarm i ljubaznost. Hvala i svima iz CIMA-e koji su

učestvovali u radu na IGF1-Abcb4 projektu.

Milki, mom „tajnom“ mentoru, hvala za neočekivane godine bavljenja naukom, za sve tehnike koje

sam naučio - od pipetiranja do organizovanja prezentacija, i, na kraju, na rukovođenju istraživanima

iz poslednje studije ove teze.

Tata, hvala za bezbrojne sate strpljivog slušanja mojih (ne uvek jasnih) ideja i za njihovo pretakanje u

crteže koji ilustruju tezu. Rezultat je očaravajući! Naš zajednički rad na za mene tako važnom

projektu je bio posebno dragoceno iskustvo.

Vileme, da nije bilo tebe kao prethodnice, ja bih morao da počnem od nekog drugog početka. Hvala

za nezaboravan dan krunisan italijanskom večerom, koji je obeležio tvoj odlazak u penziju i početak

mog rada. Drago mi je što sam bar malo doprineo razumevanju fibroze u famoznim Mdr2-ovima.

Vasili, hvala za neizmernu energiju, za mogućnost da od obicnog napraviš neobično, za sposobnost

da (kad hoćeš) učiniš da se ljudi osećaju posebnima. Hvala za Nanu, musaku i caciki (salatu, ne sos!) i,

najvažnije od svega, za Eirikura! Pitam se da li je moje prisustvo tokom Naninog telefonskog poziva

(na koji nisi odgovorio!?) zaslužno što si prihvatio da mi svedočiš tokom odbrane.

personalia

122

Jurhen, hvala za delikatan talenat da se nađeš pri ruci kad zapne – u nauci i u životu. Hvala za rezime

na holandskom i, na kraju, za hrabar pristanak da se izložiš Ulrihovoj paljbi tokom odbrane. Drago

mi je što su me Woutove divlje guske uvele u tvoj svet gastronomije i dovele do naših prijatnih

porodičnih poslepodneva.

Suzan dugujem beskrajnu zahvalnost za savršenu organizaciju i izvođenje eksperimenata na

životinjama, za sve duge sate strpljenja završavane famoznim “bez zahvaljivanja!”. Veliko hvala:

Sindi – za delikatne operacije, Lizbet – za spremnost da pomogne, Rudiju, najbržem čoveku na

svetu – za HPLCove, Kam – za to što su se hemikalije uvek stvarale gde treba, Moni – za sve

komplikovane birokratske radnje oko odbrane, Kunu – na prilici da radim sa nekim koga sam prvo

video na TVu kao rok-zvezdu. Iskrena i topla Dineke, hvala sto si postala prijatelj, iako si mi ispred

nosa ugrabila poziciju (i postala najbolji mladi ALC-ovski istrazivač). Nadam se da ćeš u nauci i

ostati. Pauli, braniocu Portugalije od ostatka sveta (kome nedostaje razumevanje za “naš”

juznoevropski način), hvala za immuno bojenja. Johane, osim za evidentnu tehničku pomoć, hvala

što si mi, ćutljiv i jednostavan, sa osmehom i lepom rečju za svakoga, postao prijatelj. Endi (i Anita),

hvala za divne večeri i večere, sa pažnjom posvećenom i najsitnijem detalju. Naravno, i za

demonstraciju poslovične nemačke efikasnosti!

Iako sam zvanično pripadao ALC-ovskom “dole”, živeo sam i sa zadovoljstvom radio “gore”. Drage

i posebne Vil i Žaklin, hvala za strpljenje i pomoć kad god bi tehnika odbila da sarađuje (a takvih je

više od kooperativnih), za pomoć oko mikroskopa i histologije, za neubičajeno prostrano radno

mesto, za... Vil i Ron, hvala što smo se uz vas osećali još dobrodošlije. Janu hvala na prajmerima,

antitelima i zarazno dobrom raspoloženju, Teu i Ingrid, na savetima i idejeama na sredskim

sastancima i na uvek prijatnim susretima. Fari hvala na razlogu za bezgranično poverenje u njene

SOP-ove i rastvore, koje sam (ponekad tajno) nemilo koristio, a njoj i Marajnu na lepom druženju i

prilici da uživam u uzbudljivoj mešavini surinamskog i holandskog. Juđi, naše druženje je započelo

neobično – uzlupavanjem srca od naizgled bezazlenog spiralnog listića čaja, a preraslo u nešto vrlo

posebno. Hvala na svakom trenutku koji smo doživeli sa tobom, Ningom i decom!

Hvala kolegama sa Biohemije, Sindi, Rulofu, Saskiji i Bertu (poslednji iz ALCa, sa Biohemije i iz

Groningena), na nesebičnoj pomoći oko „gladne studije“, i Janu Atenu sa Patologije za strpljenje

tokom mukotrpnih imunobojenja.

Nemoguće je, naravno, imenovati sve koji su uticali na to da danas pišem ovu zahvalnicu – članove

porodice (nasleđene i stečene) koji imaju dovoljno topline da daljine ne budu predaleke, prijatelje koji

su to ostali od kad su mi školu i fakultet činili prijatni(ji)m, ili one koji se vode pod „ostali“, a koji su,

npr. skenirali ilustracije za ovu tezu (na desetine puta!). Pomenuću samo nekoliko njih koji su

nastanak ove teze obeležili na vrlo poseban način. Ostalima, jednostavno, veliko hvala!

Medju ljudima koji mogu da pomognu, retki su i posebni oni koji to i hoće. Mile Ivanović je jedan

od njih. Mile, hvala što je po mom diplomiranju naš život mnogo drugačiji nego što je mogao biti!

Jan, hvala na prijatnim mesecima zajedničkog holandskog života i još lepšim godinama druženja.

Golube, hvala na tihoj, odmerenoj podršci tokom svih 35 godina druženja, iako sam ja bio taj koji te

je „pokvario“! Siki, naših četvrt veka druženja koje nije bilo ni tiho ni odmereno je svakako posebno.

Hvala što si birala da budeš s nama i kad si mogla da budeš na zanimljivijim mestima. Dare, ti si

personalia

123

jedan od retkih sa dovoljno volje i energije da isprate svaku moju akciju! Hvala na uvek spremnim

odgovorima, na prijateljstvu i osloncu za ceo život.

Ane i Arvide, hvala za toplo norveško gostoprimstvo tokom mojih prvih meseci u inostranstvu, za

duge godine prijateljstva koje su usledile i za nebrojene (ničim zaslužene) rodjendanske poklone.

Voleo bih da moja zahvalnost može da stigne i do bake Else, za duge razgovore u vreme kad moj

rečnik engleskog nije sadžavao više od pedesetak reči.

Nočka i Ejstajne, hvala što je taj rečnik u mojoj četvrtoj deceniji (!?) ipak počeo da raste. Posebno

hvala što ste bili uz nas i kad nismo mislili da nam je pomoć potrebna!

Jojka i Vujo, voleo bih da imam reči da opišem zadovoljstvo i zahvalnost za tako posebno mesto

koje zauzimate u mom životu. Hvala što ste mi postali takva porodica!

Mama, hvala za bdenje nada mnom, za strepnju i beskrajnu nežnost. Tata, hvala na razumevanju za

moje (ekstremne) ideje i želje. Hvala oboma na uspokojavajućem osećaju da sam bezuslovno voljen!

Hvala i za visoka očekivanja, stalno podizanje lestvice i podršku da se hvatam u koštac sa

problemima. Duco, moj dragi „mali” brate sa beskrajnim šarmom i istančanim osećajem za svet oko

sebe, hvala za svu ljubav i strpljenje koje si morao imati da bi živeo (i preživeo) sa mnom!

Milka, hvala za sve ove nestvarno bogate i uzbudljive godine od kada sam te prvi put ugledao u

beogradskoj Botaničkoj bašti! Za svu nežnost, razumevanje, strpljenje i odlučnost koje si utkala u naš

život, da bi smo danas bili ovde gde jesmo. Posebno hvala za hrabrost i upornost da jedan san

postane Mia. Mia, hvala za jedan novi svet koji si mi otkrila i što mi dozvoljvaš da te obožavam!

personalia

124

CURRICULUM VITAE

Aleksandar Sokolović attended secondary school majoring Biochemistry and Molecular Biology in

his hometown Lebane in Serbia. He graduated as General Biologist at Faculty of Biology, University

of Belgrade, Serbia, focusing on applied genetics. His diploma work on genetic variability of yield

and yield components of soy bean was performed at Maize Research Institute Zemun Polje,

Belgrade, under supervision of Prof. Mile Ivanović. He has worked in a world far distant from

biology for a number of years. Upon coming to the Netherlands, he started volunteering at the AMC

Liver Center in Amsterdam, in the group of Prof. Wout Lamers, on transcriptomics signature of

fasting. A year later, he started his PhD studies under supervision of Prof. Ronald Oude Elferink, at

Tytgat Institute for Liver and Intestinal Research in Amsterdam. His studies on liver fibrosis are

described in this thesis.

Aleksandar Sokolović je srednju školu na smeru Biohemija i molekularna biologija završio u rodnom

Lebanu u Srbiji. Zvanje diplomiranog biologa je stekao na Biološkom fakultetu Univerziteta u

Beogradu, na smeru Primenjena genetika. Diplomski rad na genetičkoj varijabilnosti prinosa i

komponenti prinosa soje je uradio na Institutu za kukuruz u Zemun Polju, pod vođstvom profesora

Mileta Ivanovića. Dugi niz godina je radio u svetu vrlo dalekom od biologije. Po dolasku u

Holandiju, počinje da volontira u grupi prof. Vauta Lamersa na transkriptomskom potpisu

gladovanja, u Centru za jetru Akademskog medicinskog centra u Amsterdamu. Godinu dana kasnije,

započinje doktorske studije pod vođstvom prof. Ronalda Aude Elferinka u Tajthat institutu za

proučavanje jetre i creva u Amsterdamu. Njegova istraživanja na temu fibroze jetre opisana su u ovoj

tezi.

personalia

125

LIST OF PUBLICATIONS

1) Fasting reduces liver fibrosis in mouse model for chronic cholangyopathies

Sokolović A, van Roomen CP, Ottenhoff R, Scheij S, Hiralall JK, Claessen N, Aten J, Oude Elferink RP, Groen AK, Sokolović M

under revision

2) Insulin-like growth factor 1 enhances bile-duct proliferation and fibrosis in Abcb4-/- mice

Sokolović A, Rodrigues C, ten Bloemendaal L, Prieto J, Oude Elferink RPJ, Bosma PJ

accepted for publication in

Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease; (2013)

3) Unexpected reduction of murine liver fibrosis by insulin like growth factor binding protein 5

Sokolović A, Miranda PSM, de Waart DR, Cappai RMN, Duijst S, Hiralall JK, ten Bloemendaal L, Sokolović M, Bosma PJ

Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease; (2012)

4) Interorgan coordination of the murine adaptive response to fasting

Hakvoort TBM, Moerland PD, Frijters R, Sokolović A, Labruyère WT, Vermeulen JLM, Ver Loren van Themaat E, Breit TM, Wittink FR, van Kampen AH , Verhoeven AJ, Lamers WH, Sokolović M

Journal of Biological Chemistry; (2011)

5) Unexpected effects of fasting on murine lipid homeostasis - transcriptomic and lipid profiling

Sokolović M, Sokolović A, van Roomen CP, Gruber A, Ottenhoff R, Scheij S, Hakvoort TB, Lamers WH, Groen AK.

Journal of Hepatology; (2010)

6) Insulin-like growth factor binding protein 5 enhances survival of LX2 human hepatic stellate cells

Sokolović A, Sokolović M, Boers W, Elferink RP, Bosma PJ,

Fibrogenesis and Tissue Repair; (2010)

7) The transcriptomic signature of fasting murine liver

Sokolović M, Sokolović A, Wehkamp D, Ver Loren van Themaat E, de Waart DR, Gilhuijs-Pederson LA, Nikolsky Y, van Kampen AHC, Hakvoort TBM, Lamers WH,

BMC Genomics; (2008)

8) Fasting induces a biphasic adaptive metabolic response in murine small intestine

Sokolović M, Wehkamp D, Sokolović A, Vermeulen J, Gilhuijs-Pederson LA, van Haaften RIM, Nikolsky Y, Evelo CTA, van Kampen AHC, Hakvoort TBM, Lamers WH,

BMC Genomics, (2007)