IGF-I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to...

Insulin-Like Growth Factor I Gene Transfer to CirrhoticLiver Induces Fibrolysis and Reduces Fibrogenesis

Leading to Cirrhosis Reversion in RatsLuciano Sobrevals,1 Carlos Rodriguez,1 Jose Lorenzo Romero-Trevejo,1 Gabor Gondi,1 Inaki Monreal,1 Astrid Paneda,1

Nerea Juanarena,1 Sara Arcelus,1 Nerea Razquin,1 Laura Guembe,1 Gloria Gonzalez-Aseguinolaza,1 Jesus Prieto,1,2 andPuri Fortes1

We investigated whether gene transfer of insulin-like growth factor I (IGF-I) to the hepatic tissue wasable to improve liver histology and function in established liver cirrhosis. Rats with liver cirrhosisinduced by carbon tetrachloride (CCl4) given orally for 8 weeks were injected through the hepaticartery with saline or with Simian virus 40 vectors encoding IGF-I (SVIGF-I), or luciferase (SVLuc).Animalsweresacrificed8weeksaftervector injection.Incirrhoticratsweobservedthat,whereasIGF-Iwas synthesized by hepatocytes, IGF-I receptor was predominantly expressed by nonparenchymalcells, mainly in fibrous septa surrounding hepatic nodules. Rats treated with SVIGF-I showed in-creased hepatic levels of IGF-I, improved liver function tests, and reduced fibrosis in association withdiminished�-smoothmuscleactinexpression,up-regulationofmatrixmetalloproteases (MMPs)anddecreased expression of the tissue inhibitors of MMPs TIM-1 and TIM-2. SVIGF-I therapy induceddown-regulation of the profibrogenic molecules transforming growth factor beta (TGF�), amphi-regulin,platelet-derivedgrowth factor (PDGF), connective tissuegrowth factor (CTGF), andvascularendotheliumgrowthfactor(VEGF)andinductionoftheantifibrogenicandcytoprotectivehepatocytegrowth factor (HGF). Furthermore, SVIGF-I-treated animals showed decreased expression of Wilmstumor-1 (WT-1; a nuclear factor involved in hepatocyte dedifferentiation) and up-regulation ofhepatocyte nuclear factor 4 alpha (HNF4�) (which stimulates hepatocellular differentiation). Thetherapeutic potential of SVIGF-I was also tested in rats with thioacetamide-induced liver cirrhosis.Also in this model, SVIGF-I improved liver function and reduced liver fibrosis in association withup-regulation of HGF and MMPs and down-regulation of tissue inhibitor of metalloproteinase 1(TIMP-1). Conclusion: IGF-I gene transfer to cirrhotic livers induces MMPs and hepatoprotectivefactors leading to reversion of fibrosis and improvement of liver function. IGF-I gene therapy may bea useful alternative therapy for patients with advanced cirrhosis without timely access to liver trans-plantation. (HEPATOLOGY 2010;51:912-921.)

Liver transplantation is the only curative optionfor patients with advanced liver cirrhosis. Thisprocedure can only be applied to a minority of

patients due to the presence of surgical contraindi-

cations and organ scarcity. In fact, the waiting list inthe USA includes �12,500 patients with a mediantime to transplantation of �300 days; more than 45%of the patients exceed 24 months on the waiting list,

Abbreviations: AR, amphiregulin; �SMA, �-smooth muscle actin; CCl4, carbon tetrachloride; CTGF, connective tissue growth factor; HGF, hepatocyte growth factor; HNF4�,hepatocyte nuclear factor 4 alpha; HSC, hepatic stellate cell; IGF-I insulin-like growth factor I; MAP, mitogen activated protein; MMPs, matrix metalloproteases; PDGF,platelet-derived growth factor; rIGF-I, recombinant IGF-I; TAA, thioacetamide; SVIGF-I, Simian virus 40 vectors encoding IGF-I; SVLuc, Simian virus luciferase; TGF�,transforming growth factor beta; TIMP-1, tissue inhibitor of metalloproteinase 1; VEGF, vascular endothelium growth factor; WT-1, Wilms tumor-1.

From the 1Department of Hepatology and Gene Therapy, Center for Applied Medical Research (CIMA), Pamplona, Spain; 2CIBERehd. Clınica Universitaria, Schoolof Medicine, University of Navarra, Pamplona, Spain.

Received January 21, 2009; accepted October 21, 2009.Supported by CICYT (SAF2003-01804, BIO2006-13225), Pedro Barrie de la Maza and Condesa de Fenosa Foundation, Instituto de Salud Carlos III and through the “UTE

project CIMA”.Address reprint requests to: Puri Fortes or Jesus Prieto, Pio XII 55, 31008 Pamplona, Spain. E-mail: [email protected] or [email protected]; fax: 34948194717.Copyright © 2009 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.23412Potential conflict of interest: Nothing to report.Additional Supporting Information may be found in the online version of this article.

912

where the mortality reaches 130 per 1,000 patients/year.1

Insulin-like growth factor I (IGF-I) is a potent cyto-protective and anabolic hormone, synthesized mainly inthe liver, which circulates bound to a set of binding pro-teins (IGFBPs) that regulate IGF-I biological activity.2

Degradation of IGFBPs releases free IGF-I which canbind and activate the IGF-I receptor (IGF-IR). Interac-tion with IGF-IR leads to activation of mitogen activatedprotein (MAP) kinase and PI3 kinase cascades that regu-late genes involved in cell survival, growth, and differen-tiation.3

In liver cirrhosis, as result of hepatocellular insuffi-ciency, there is a marked reduction in the levels of IGF-I.This hormonal deficiency may play a role in the systemicmetabolic derangement present in liver cirrhosis.4 In fact,treatment of cirrhotic rats with recombinant IGF-I(rIGF-I) promotes weight gain, nitrogen retention, andintestinal absorption of nutrients.5 In addition, rIGF-Ihas been shown to exert hepatoprotective activities in cir-rhotic rats.6 A recent pilot clinical trial showed that cir-rhotic patients treated with a daily dose of rIGF-I (100�g/kg bw) manifested a significant increase in serum al-bumin and an improvement of the Child-Pugh score.4

However, restoration of IGF-I levels in cirrhotic patientsusing recombinant protein entails consumption of highdoses of this molecule, making the treatment exceedinglycostly. It may be envisioned that the recombinant proteinmight be substituted by the use of viral vectors encodingIGF-I that allow sustained expression of the transgenewithin the cirrhotic liver. Previously, we showed that thetransfer of a recombinant Simian virus 40 vector encodingIGF-I (SVIGF-I) to a noncirrhotic liver reduced hepato-cellular damage induced by subsequent administration ofcarbon tetrachloride (CCl4).7 However, it remained to bedetermined whether the injection of the vector was able torevert established liver cirrhosis. In the present article weshow that administration of SVIGF-I to rats with estab-lished liver cirrhosis activates a robust tissue repair pro-gram characterized by stimulation of fibrolysis, down-regulation of profibrogenic factors, and induction ofcytoprotective molecules leading to improved hepatocel-lular function and reduced liver fibrosis. These findingssuggest that IGF-I gene transfer to the cirrhotic livermight be considered for the improvement of liver func-tion in patients without access to liver transplant or whodeteriorate while on the waiting list for transplantation

Materials and Methods

Isolation of Liver Cells. Hepatocytes, hepatic stellatecells (HSCs), and Kupffer cells were isolated from healthy

and cirrhotic male Sprague-Dawley rats as described.8,9

HSCs and Kupffer cells were separated in an invertedgradient. Cell viability was measured by trypan blue ex-clusion and exceeded 90%. The purity of HSC was higherthan 99%, as assessed by fluorescence of retinoid-contain-ing vacuoles under ultraviolet excitation.9

Model of Established Liver Cirrhosis and SV40Vectors Therapy. This study was performed followingthe regulations of the local Animal Care Ethical Commit-tee. Cirrhosis was induced by weekly intragastric admin-istration of CCl4 for 8 weeks (Supporting Fig. 1A)10 or byintraperitoneal administration of 200 mg/kg of thioacet-amide (TAA) 3 times per week for 7 weeks. SV40 vectorsencoding IGF-I (SVIGF-I) and luciferase (SVLuc) havebeen produced as described7 and a single dose of 1 � 1011

viral particles was administered through the hepatic artery1 week after the last dose of hepatotoxicant. For the CCl4model of liver cirrhosis four experimental groups of ani-mals were analyzed in two independent experiments:healthy rats (n � 11), cirrhotic rats injected with saline(Ci) (n � 14), and cirrhotic rats treated with either ofSVLuc (Ci�Luc) (n � 8) or SV-IGF-I (Ci�IGF-I) (n �16). For the TAA model animals were divided into thesame groups (6 healthy rats, 5 Ci, 5 Ci�Luc, 5 Ci�IGF-I). Animals were sacrificed 8 weeks after virus injection.Blood samples were collected at different timepoints andanalyzed as indicated (Supporting Fig. 1). Liver sampleswere processed for histology and purification of RNA andproteins for further analysis.

Liver Histology and Immunohistochemistry. Livercollagen content was assessed and quantified as described(Supporting Fig. 1).7 Immunohistochemical staining for�-smooth muscle actin (�SMA) was done with antibody1A4 (M0851, Dako) diluted 1:100, and for IGF-1R�with antibody sc-713 (Santa Cruz Biotechnology) diluted1:50.

Liver Proteins and RNA Analysis. Total liver IGF-I(OCTEIA Rat/mouse IGF-I, Vitro) was measured in se-rum and liver extracts by ELISA. Total MMP activity wasmeasured using a fluorogenic peptide substrate (R&DSystems). TIMP-1 was evaluated with antibody fromR&D Systems diluted 1:500 and the western blot wasquantified with Image Quant ECL (GE). Total RNA wasextracted as described.7 RNA was also extracted from laserdissected liver sections with Absolutely RNA nanoprep(Stratagene). Quantitative reverse-transcription polymer-ase chain reactions (qRT-PCRs) were done as described(Supporting Table 1).7

Statistical Analysis. Data are expressed as means �standard deviation. Statistical significance was estimatedwith Student’s t test. A P-value � 0.05 was considered

HEPATOLOGY, Vol. 51, No. 3, 2010 SOBREVALS ET AL. 913

significant (*). All statistical analyses were carried out withSPSS v. 11.0.

Results

IGF-I Gene Transfer to the Cirrhotic Liver. Toevaluate IGF-I effect in rat cirrhotic livers, cirrhosis wasinduced by intragastric administration of CCl4 for 8weeks (Supporting Fig. 1A). Transaminases increased atthe end of CCl4 treatment and remained higher thanhealthy controls more than half a year after completion ofcirrhosis induction (Supporting Fig. 1B). The cirrhoticlesion was rather stable because no significant changes inthe degree of liver fibrosis occurred in the same period oftime, indicating that there is no relevant spontaneous re-gression of hepatic fibrosis during the study (SupportingFig. 1C,D). After the induction of cirrhosis animals re-ceived saline, SVLuc, or SVIGF-I and were sacrificed 8weeks after virus injection.

As expected, both messenger RNA (mRNA) and pro-tein levels of IGF-I were significantly increased in theCi�IGF-I group and decreased in control cirrhotic livers(Ci and Ci�Luc) as compared to healthy rats (Fig. 1A,B).Liver IGF-I binding protein 3 (IGF-IBP3) mRNA, whoseexpression is activated by IGF-I was also increased in IGF-I-treated animals compared to controls (Fig. 1C).

In order to characterize the cell populations producingand responding to IGF-I, we determined IGF-I andIGF-IR mRNA levels in purified hepatocytes, HSCs, andKupffer/endothelial cells from healthy livers. We foundthat IGF-I is expressed mainly in hepatocytes and signif-icantly less in nonparenchymal cells, whereas IGF-IR isexpressed predominantly in HSCs and Kupffer/endothe-lial cells (Fig. 1D). In the cirrhotic liver, immunohisto-chemistry analysis showed that IGF-IR is mainly presentin septa surrounding cirrhotic nodules (Fig. 1E). Also,analysis of IGF-IR by qRT-PCR after laser dissection ofsepta and nodules indicated that levels of IGF-IR mRNAare significantly higher in septa than in nodules (Fig. 1F).Interestingly, IGF-IR expression was significantly in-duced in the septa of IGF-I-treated animals (Fig. 1F).Taken together, these data indicate that SVIGF-I vectorwas able to transduce the cirrhotic liver and to expressfunctional IGF-I protein. This hormone, in turn, stimu-lates the expression of IGFI-R in fibrotic tissue renderingthe cirrhotic liver more sensitive to IGF-I signals.

IGF-I Gene Transfer to the Cirrhotic Liver Im-proves Liver Function and Causes a Marked Reduc-tion of Liver Fibrosis. Cirrhotic rats treated withSVIGF-I showed ameliorated biochemical liver tests. Inthese animals serum aspartate aminotransferase (AST),

Fig. 1. Analysis of IGF-I, IGF-IBP3, and IGF-IR expression in IGF-I-treated livers and controls. IGF-I (A,D), IGFIBP3 (C), and IGF-IR (F) mRNAs and IGF-I(B) and IGF-IR (E) proteins were evaluated by qRT-PCR (A,C,D,F), ELISA (B), or immunohistochemistry (E). Samples are from healthy livers or cirrhotic liversfrom control animals (Ci) or from animals treated with SVLuc (Ci�Luc) or SVIGF-I (Ci�IGF-I) except in (D), where analysis was performed in extracts fromisolated hepatocytes, HSCs, or Kupffer and endothelial cells. Parenchyma and septum were obtained by laser dissection (F).

914 SOBREVALS ET AL. HEPATOLOGY, March 2010

alanine aminotransferase (ALT), alkaline phosphatase(ALP), and bilirubin were significantly lower and serumalbumin significantly higher than in control cirrhotic ratsand similar to healthy controls (Fig. 2A-C). These favor-able changes were accompanied by histological improve-ment with marked reduction of fibrosis and decreasedexpression of collagen I and IV and �SMA in SVIGF-I-treated rats (Figs. 3A-D, 4A). Immunohistochemicalanalysis of �SMA showed that this marker of HSC acti-vation was almost absent in IGF-I-treated animals,whereas it was conspicuous in the septa surrounding nod-ules in control cirrhotic animals (Ci and Ci�Luc) (Fig.4B). These findings indicate that treatment with SVIGF-Iefficiently reduces the presence of activated HSC in thedamaged liver. We were not able to detect apoptotic HSCat the timepoints where tissue sampling was performed(data not shown). Even when we cannot exclude apopto-sis, the antifibrogenic effect of SVIGF-I might also derivefrom deactivation of HSC without HSC loss. In fact, insepta from cirrhotic controls �SMA staining is concor-dant with vimentin expression (the latter labels active andinactive HSCs11), whereas in SVIGF-I-treated cirrhoticrats most of vimentin-positive cells found in the slenderremaining septa were �SMA-negative (Supporting Fig.2A). The increase of deactivated HSC in SVIGF-I-treatedlivers is also suggested by the enhanced expression of neu-rotrimin, a marker of nonactivated HSC12 (SupportingFig. 2B).

IGF-I Gene Expression Within the Cirrhotic LiverInduces Fibrolysis and Reduces the Expression of Pro-fibrogenic Factors. We investigated whether the decreasein fibrosis observed in SVIGF-I-treated rats was associatedwith activation of enzymes capable of removing collagen,such as MMPs. We found that as compared to normal rats,MMP1, 2, 9, and 14 mRNAs were down-regulated in con-trol cirrhotic livers and markedly up-regulated in the liversthat received SVIGF-I (Fig. 5A). In addition, liver expression

of the MMP inhibitors TIMP-1 and TIMP-2 showed a pat-tern opposite to that of MMPs. These TIMPs were inducedin the liver from Ci and Ci�Luc rats, whereas they weredown-regulated in Ci�IGF-I rats (Fig. 5B,C). In agreementwith these data we found decreased MMP activity in controlcirrhotic livers compared to healthy controls, whereas MMPactivity was significantly higher in IGF-I-treated rats than inhealthy controls. It seems possible therefore that increasedMMP activity may account for the efficient removal of fi-brous tissue from the cirrhotic liver of SVIGF-I-treated rats.

In agreement with the above data we observed reducedTGF� expression in the liver of Ci�IGF-I rats (Fig. 6A).Because TGF� is a powerful activator of HSCs and themost potent accelerator of liver fibrosis, its down-regula-tion by IGF-I might be a key factor underlying the anti-fibrogenic effect of the treatment.13 In addition to TGF�,other molecules that promote HSC growth and contrib-ute to liver fibrosis such as amphiregulin (AR), connectivetissue growth factor (CTGF), platelet-derived growth fac-tor (PDGF), and vascular endothelium growth factor(VEGF)14-16 were up-regulated in control cirrhotic liversbut markedly suppressed in those treated with SVIGF-I ascompared to control cirrhotic rats (Fig. 6B-E).

Together with the decrease of profibrogenic molecules,we found a significant increase in the expression of hepa-tocyte growth factor (HGF) and of the HGF receptorc-met in the liver of Ci�IGF-I rats as compared to con-trol animals (Fig. 6F and Supporting Fig. 3). BecauseHGF displays potent antifibrogenic activities, up-regula-tion of this molecule and of its receptor may contribute tothe regression of liver cirrhosis observed in IGF-I-treatedrats.17

Safety and Effect on Factors Influencing Hepato-cellular Differentiation of SVIGF-I Therapy. Wetested the safety of SVIGF-I therapy in cirrhotic rats formore than 8 months after vector injection and we foundno signs of toxicity for the entire observation period (data

Fig. 2. Analysis of serum parameters from IGF-I-treated rats and controls. Transaminases (AST, ALT, and ALP) (A), bilirubin (B), and albumin (C)were evaluated in serum from healthy rats or from cirrhotic control animals (Ci) or from cirrhotic animals treated with SVLuc (Ci�Luc) or SVIGF-I(Ci�IGF-I).


not shown). Necropsies revealed no apparent systemicabnormalities and liver histology confirmed the absenceof malignant or premalignant lesions. In fact, mRNA lev-els of Wilms tumor-1 (WT-1), a nuclear factor implicatedin tumorigenesis18 that promotes hepatocellular dediffer-entiation and which is overexpressed in liver cirrhosis,19

was found to be increased in control cirrhotic livers butdeeply down-regulated in rats treated with SVIGF-I (Fig.7A). On the contrary, we found that SVIGF-I-treated ratsexhibited significant up-regulation of HNF4�, a hepato-cyte nuclear factor that stimulates the expression of genescharacterizing the mature hepatocyte phenotype (Fig.7B).20 It seems possible that this effect might contributeto the improvement of liver function observed in IGF-I-treated cirrhotic rats.

We also assessed the safety of SVIGF-I in normal rats.To this aim, control healthy rats, rats injected withSVIGF-I or with SVLuc were sacrificed 8 weeks aftervector administration. IGF-I, IGFBP3, and HGF wereup-regulated in the liver of rats given SVIGF-I but his-topathological analysis of several organs and evaluation ofdifferent serum biochemical parameters showed no signif-icant differences between the groups (Supporting Fig. 4and data not shown).

IGF-I Gene Transfer Exerts Therapeutic Effects inRats Treated with TAA. To evaluate the robustness ofIGF-I therapy we tested SVIGF-I in a different model ofliver cirrhosis more difficult to revert. To this aim, weadministered saline or recombinant SV40 vectors encod-ing SVLuc or IGF-I (SVIGF-I) to rats in which cirrhosishad been previously induced by TAA administration for 7weeks. All animals and healthy controls were evaluated 8weeks after vector administration. Similar to what wasobserved in the CCl4 model, SVIGF-I vector was able toexpress functional IGF-I protein in the TAA cirrhoticliver. Thus, both IGF-I and IGF-IBP3 mRNAs were in-creased in IGF-I-treated animals compared to controls(Fig. 8A). This was accompanied by improved liver bio-chemistry, being the levels of serum ALP and serum bili-rubin significantly lower than in cirrhotic controls andsimilar to values found in healthy animals (Fig. 8B). Also,the liver of IGF-I-treated animals exhibited less nodular-ity macroscopically (data not shown) and on histologicalexamination showed a marked decrease of liver fibrosisand less ductular proliferation in portal tracts compared

Fig. 3. Assessment of liver fibrosis in IGF-I-treated rats and controls.Extracellular deposition was stained with Sirius Red (A) and quantified byimage analysis (B). Collagen I (C) and IV (D) mRNAs were quantified byqRT-PCR. All samples were obtained from healthy rats, control cirrhoticanimals (Ci), or from cirrhotic animals treated with SVLuc (Ci�Luc) orSVIGF-I (Ci�IGF-I).

Fig. 4. Analysis of activated HSCs in IGF-I-treated livers and controls.�SMA mRNA (A) and protein (B) were quantified by qRT-PCR (A) ordetected by immunohistochemistry (B) from healthy livers or cirrhoticlivers from control animals (Ci) or from animals treated with SVLuc(Ci�Luc) or SVIGF-I (Ci�IGF-I).


to cirrhotic controls (Fig. 8C,D). Reduced fibrosis corre-lated with a strong decrease of activated HSCs as detectedby immunohistochemistry and quantification of �SMAmRNA (Fig. 8C,E). In parallel to findings in the CCl4model, rats with TAA-induced cirrhosis treated withSVIGF-I showed in liver tissue up-regulation of HGFaccompanied by increased expression of MMPs and de-creased levels of TIMP-1 (Fig. 8A, Supporting Fig. 5).

DiscussionBecause liver transplantation can be offered to only a

limited number of cirrhotic patients, alternative therapiesfor advanced liver cirrhosis are urgently needed. In keep-ing with the fact that IGF-I deficiency is a key feature ofliver cirrhosis, a previous work by our group showed thatdaily administration of recombinant IGF-I to cirrhoticpatients induces a significant amelioration of liver func-tion.4 However, the amount of recombinant proteinneeded to accomplish hormone replacement therapy ishigh and a prolonged treatment would be exceedinglycostly. Thus, as an alternative therapeutic strategy wetested whether IGF-I gene transfer to hepatic tissue mightexert beneficial effects in a cirrhotic liver.

In a previous work we showed that transduction ofnormal rat liver with a SV40 vector encoding IGF-I con-ferred protection against CCl4 toxic injury.7 In that studytreated rats had a normal liver and resisted toxic injurywith less tissue damage than controls. However, it re-mained to be investigated whether IGF-I-based gene ther-apy was able to improve or to revert a previouslyestablished cirrhotic lesion. In this work we show that ratswith well-established liver cirrhosis treated with SVIGF-Iexperience an improvement of liver function and amarked reduction of liver fibrosis. These effects are ob-served not only in CCl4-induced cirrhosis but also in theTAA model, which represents a more difficult conditionto treat. The efficacy of the therapy in the two forms ofliver cirrhosis reinforces the concept that regression of thelesion is due to the therapeutic effect of SVIGF-I and notto spontaneous resolution of fibrosis.

IGF-I gene transfer to the cirrhotic liver was accom-plished using an SV40 vector. Although this vector has awide host range, liver specificity can be improved by hepaticartery administration as performed in our study. Advantagesof SV40 vector include low antigenicity, long duration oftransgene expression, ability to infect liver cells, and a small

Fig. 5. Analysis of the expressionof MMPs and MMP activity in IGF-I-treated livers and controls. MMP 1,2, 9, 14 (A) and TIMP-2 (C) mRNAswere quantified by qRT-PCR. TIMP-1was evaluated (B, upper photo-graph) and quantified (B, lowergraph) by western blot analysis.MMP activity was measured using afluorogenic substrate (D). Sampleswere from healthy livers or cirrhoticlivers from control animals (Ci) orfrom animals treated with SVLuc(Ci�Luc) or SVIGF-I (Ci�IGF-I).


virus particle size that would facilitate penetration throughthe collagenous extracellular matrix. In the present study,IGF-I expression was constant until half a year after SVIGF-Ivector administration in rats (data not shown). The level oftransgene expression using SV40 vectors is relatively low ascompared with other vectors.7,21 For our purposes this char-acteristic may be advantageous because low intrahepatic ex-pression of transgenic IGF-I would restrict the hormoneeffects to the liver without unduly increasing its serum val-ues. In fact, in our study we were able to increase intrahepaticIGF-I level (Fig. 1A,B) without raising serum concentration(data not shown).

We addressed the molecular mechanisms that couldmediate IGF-I therapeutic effects on liver cirrhosis. Liver

expression of the transgenic IGF-I should be sensed byIGF-IR, predominantly expressed by nonparenchymalliver cells within fibrous septa surrounding cirrhotic nod-ules. This receptor is expressed poorly by rat hepatocytes(Fig. 1D-F).22,23 Thus, it seems possible that IGF-I actson nonparenchymal cells to activate a tissue-repair pro-gram able to improve liver architecture and function. In-terestingly, we found that induction of IGF-I led to up-regulation of IGF-IR in the septa, suggesting the existenceof an amplification loop that would favor the efficacy ofthe therapy (Fig. 1F).

The reduction of liver fibrosis observed in SVIGF-I-treated rats may be related to the fact that IGF-I canincrease stability and expression of MMP 2, expression of

Fig. 6. Analysis of the expression of profibrogenic and antifibrogenic factors in IGF-I-treated livers and controls. TGF� (A), Amphiregulin (AR, B),PDGF (C), CTGF (D), VEGF (E), and HGF (F) mRNAs were quantified by qRT-PCR from healthy livers or cirrhotic livers from control animals (Ci) or fromanimals treated with SVLuc (Ci�Luc) or SVIGF-I (Ci�IGF-I).

Fig. 7. Analysis of maturation of liver cells treated with SVIGF-I. mRNAs from the dedifferentiation factor WT-1 (A) and the maturation markerHNF4� (B) were quantified by qRT-PCR. All samples were obtained from healthy rats, control cirrhotic animals (Ci), or from cirrhotic animals treatedwith SVLuc (Ci�Luc) or SVIGF-I (Ci�IGF-I).


MMP14, and production of HGF by HSCs, which seemto be the main producers of HGF in the intact liver invivo.24 HGF might be a key element in the process leadingto improvement of liver cirrhosis because this molecule isa powerful hepatoprotective factor and an inducer of hep-atocellular differentiation endowed with antifibrogenicactivities.17,25 HGF strongly inhibits the expression ofprofibrogenic factors such as TGF� and CTGF, whereasit activates MMP2, MMP9, and MMP14 and reducesTIMP-1 and TIM-2 expression in different cells (re-viewed25). These changes were found in SVIGF-I-treatedlivers making possible a participation of HGF/c-met in-duction in the antifibrogenic effect of IGF-I (Figs. 5, 6,Supporting Fig. 3). MMP1, MMP2, and MMP14 act asefficient collagen proteases, which could account for thefibrolysis observed with IGF-I therapy. Reduced activa-tion of HSCs could also result from decreased expressionof PDGF, VEGF, and TGF� (Fig. 6), which have beeninvolved in activation, proliferation, migration, and in-creased extracellular matrix deposition from HSCs.14,15

Therefore, IGF-I may favor resolution of the cirrhoticlesion by activation of fibrolysis and reduction of fibro-genesis. This is similar to what has been described forspontaneous reversion of cirrhosis in rats26 and oppositeto the mechanisms involved in cirrhosis progression inhumans,27 opening the possibility that IGF-I treatmentcould also decrease fibrosis progression in patients. Al-though the presence of cross-linking of collagen can makelong-standing liver cirrhosis in humans more difficult torevert than in experimental animals, still IGF-I-based

therapy may have a role in these cases as it may exerthepatoprotective activities, halt fibrogenesis, and partiallyrevert liver fibrosis.

Activated HSCs decrease in SVIGF-I-treated livers(Fig. 4). We could not determine with certainty whetheractivated HSCs undergo apoptosis and/or deactivation.Although we could not clearly detect HSC apoptosis atthe time of the sacrifice, this process might have occurredat an earlier timepoint taking into account the markedreduction in fibrous septa observed in SVIGF-I-treatedrats. However, as mentioned above, up-regulation of neu-rotrimin and the presence of vimentin-positive, �SMA-negative cells in the remaining septa of cirrhotic rats givenSVIGF-I argue in favor of HSC deactivation (SupportingFig. 2).

Importantly, IGF-I-based therapy also leads to im-provement of liver function probably due to the induc-tion of hepatoprotective factors, such as HGF, togetherwith up-regulation of HNF4�, which stimulates differen-tiated hepatocellular functions (Fig. 7).20 Up-regulationof HNF4� is paralleled by down-regulation of WT-1(Fig. 7), a nuclear factor that mediates the loss of differ-entiated hepatocellular functions.19 WT-1 is up-regulatedby TGF� and it seems possible that the reduction in theexpression of this cytokine may contribute to diminishWT-1 and to increase HNF4� abundance.19

A concern of IGF-I gene therapy of cirrhosis is thepossible induction of hepatocarcinogenesis. It should beconsidered, however, that fibrosis, inflammation, and in-creased cell turnover are key processes implicated in the

Fig. 8. Analysis of the therapeutic effects of SVIGF-I administration into TAA-induced cirrhotic animals. IGF-I, IGFIBP3, HGF (A), and �SMA (F)mRNAs were quantified by qRT-PCR. ALP and bilirubin (B) were evaluated in serum. Liver sections were evaluated by hematoxylin and eosin staining(H&E; C, left panel). Collagen deposits were stained with Sirius Red (C, middle panel) and quantified by image analysis (D). �SMA was also detectedby immunohistochemistry (C, right panel). All samples were taken from healthy livers or TAA-induced cirrhotic livers from control animals (Ci) or fromanimals treated with SVLuc (Ci�Luc) or SVIGF-I (Ci�IGF-I).


development of liver cancer.28 Therefore, by reducingliver fibrosis and exerting antiinflammatory effects thistherapy might in fact reduce the risk of liver carcinogen-esis. Interestingly, multifocal bile duct proliferation, aprocess that precedes carcinogenesis in the TAA model,was markedly decreased in SVIGF-I-treated rats.29 More-over, in rats with CCl4-induced cirrhosis, IGF-I therapyfavors hepatocellular differentiation (up-regulation ofHNF4� and down-regulation of WT-1), which may op-pose tumorigenesis (Fig. 7). In fact, up-regulation ofWT-1 has been shown to be associated with HCC devel-opment,18 suggesting that the effect of IGF-I in suppress-ing its expression may inhibit tumor formation.According to this idea it has been shown that a sharpdecrease of IGF-I production by the cirrhotic liver in-creases the risk of hepatocellular carcinoma (HCC).30 Al-though there is activation of IGF-IR in HCC, the levels ofIGF-I in tumor tissue are markedly diminished, whereasthere is up-regulation of IGF-II, suggesting that the latter,and not the former, is the ligand involved in IGF-IRsignaling in the tumor.31 As shown above, our data indi-cate that in the cirrhotic liver IGF-IR is mainly expressedby nonparenchymal cells present in fibrous septa. Thevery poor expression of this receptor in hepatocytes wouldminimize any direct effect of the molecule on these cells.However, IGF-I gene therapy should not be administeredto patients with cirrhosis who have already developedHCC because this therapy may enhance tumor growthdue to the abundance of IGF-IR in already transformedhepatocytes.

In summary, we show that gene transfer of IGF-I to thecirrhotic liver sets in motion a curative process involvingincreased expression of cytoprotective and antifibrogenicmolecules and reduced expression of profibrogenic fac-tors. IGF-I seems to be able to reprogram the liver from a“scar formation” response, which leads to cirrhosis, to a“tissue repair” circuit that favors cirrhosis regression. Ac-cording to these findings IGF-I gene therapy might be ofvalue for patients with advanced cirrhosis who do nothave access to timely liver transplantation.

Acknowledgment: We thank Pilar Perez and UxueLatasa for help with liver cells isolation, Maria Vera forSVIGF-I production, Ana Sandoval for the thioacetamidemodel, and Monica Enguita and Erkuden Casales.

References1. Freeman RB Jr, Steffick DE, Guidinger MK, Farmer DG, Berg CL, Me-

rion RM. Liver and intestine transplantation in the United States, 1997-2006. Am J Transplant 2008;8:958-976.

2. Mohan S, Baylink DJ. IGF-binding proteins are multifunctional and actvia IGF-dependent and -independent mechanisms. J Endocrinol 2002;175:19-31.

3. Riedemann J, Macaulay VM. IGF1R signalling and its inhibition. EndocrRelat Cancer 2006;13(Suppl 1):S33-S43.

4. Conchillo M, de Knegt RJ, Payeras M, Quiroga J, Sangro B, Herrero JI, etal. Insulin-like growth factor I (IGF-I) replacement therapy increases albu-min concentration in liver cirrhosis: results of a pilot randomized con-trolled clinical trial. J Hepatol 2005;43:630-636.

5. Castilla-Cortazar I, Prieto J, Urdaneta E, Pascual M, Nunez M, Zudaire E,et al. Impaired intestinal sugar transport in cirrhotic rats: correction by lowdoses of insulin-like growth factor I. Gastroenterology 1997;113:1180-1187.

6. 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 carbontetrachloride-induced cirrhosis. Gastroenterology 1997;113:1682-1691.

7. Vera M, Sobrevals L, Zaratiegui M, Martinez L, Palencia B, RodriguezCM, et al. Liver transduction with a simian virus 40 vector encodinginsulin-like growth factor I reduces hepatic damage and the developmentof liver cirrhosis. Gene Ther 2007;14:203-210.

8. Avila MA, Carretero MV, Rodriguez EN, Mato JM. Regulation by hyp-oxia of methionine adenosyltransferase activity and gene expression in rathepatocytes. Gastroenterology 1998;114:364-371.

9. Friedman SL, Roll FJ. Isolation and culture of hepatic lipocytes, Kupffercells, and sinusoidal endothelial cells by density gradient centrifugationwith Stractan. Anal Biochem 1987;161:207-218.

10. Runyon BA, Sugano S, Kanel G, Mellencamp MA. A rodent modelof cirrhosis, ascites, and bacterial peritonitis. Gastroenterology 1991;100:489-493.

11. de Leeuw AM, McCarthy SP, Geerts A, Knook DL. Purified rat liverfat-storing cells in culture divide and contain collagen. HEPATOLOGY 1984;4:392-403.

12. De Minicis S, Seki E, Uchinami H, Kluwe J, Zhang Y, Brenner DA, et al.Gene expression profiles during hepatic stellate cell activation in cultureand in vivo. Gastroenterology 2007;132:1937-1946.

13. Friedman SL. Mechanisms of disease: mechanisms of hepatic fibrosis andtherapeutic implications. Nat Clin Pract Gastroenterol Hepatol 2004;1:98-105.

14. Friedman SL. Molecular regulation of hepatic fibrosis, an integrated cel-lular response to tissue injury. J Biol Chem 2000;275:2247-2250.

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

16. Perugorria MJ, Latasa MU, Nicou A, Cartagena-Lirola H, Castillo J, GoniS, et al. The epidermal growth factor receptor ligand amphiregulin partic-ipates in the development of mouse liver fibrosis. HEPATOLOGY 2008;48:1251-1261.

17. Ueki T, Kaneda Y, Tsutsui H, Nakanishi K, Sawa Y, Morishita R, et al.Hepatocyte growth factor gene therapy of liver cirrhosis in rats. Nat Med1999;5:226-230.

18. Perugorria MJ, Castillo J, Latasa MU, Goni S, Segura V, Sangro B, et al.Wilms’ tumor 1 gene expression in hepatocellular carcinoma promotes celldedifferentiation and resistance to chemotherapy. Cancer Res 2009;69:1358-1367.

19. Berasain C, Herrero JI, Garcia-Trevijano ER, Avila MA, Esteban JI, MatoJM, et al. Expression of Wilms’ tumor suppressor in the liver with cirrhosis:relation to hepatocyte nuclear factor 4 and hepatocellular function. HEPA-TOLOGY 2003;38:148-157.

20. Ryffel GU. Mutations in the human genes encoding the transcriptionfactors of the hepatocyte nuclear factor (HNF)1 and HNF4 families:functional and pathological consequences. J Mol Endocrinol 2001;27:11-29.

21. Vera M, Fortes P. Simian virus-40 as a gene therapy vector. DNA Cell Biol2004;23:271-282.

22. Caro JF, Poulos J, Ittoop O, Pories WJ, Flickinger EG, Sinha MK. Insulin-like growth factor I binding in hepatocytes from human liver, humanhepatoma, and normal, regenerating, and fetal rat liver. J Clin Invest 1988;81:976-981.

23. Scharf JG, Knittel T, Dombrowski F, Muller L, Saile B, Braulke T, et al.Characterization of the IGF axis components in isolated rat hepatic stellatecells. HEPATOLOGY 1998;27:1275-1284.


24. Long L, Navab R, Brodt P. Regulation of the Mr 72,000 type IV collage-nase by the type I insulin-like growth factor receptor. Cancer Res 1998;58:3243-3247.

25. Liu X, Hu H, Yin JQ. Therapeutic strategies against TGF-beta signalingpathway in hepatic fibrosis. Liver Int 2006;26:8-22.

26. Issa R, Zhou X, Constandinou CM, Fallowfield J, Millward-Sadler H,Gaca MD, et al. Spontaneous recovery from micronodular cirrhosis: evi-dence for incomplete resolution associated with matrix cross-linking. Gas-troenterology 2004;126:1795-1808.

27. Asselah T, Bieche I, Laurendeau I, Paradis V, Vidaud D, Degott C, et al.Liver gene expression signature of mild fibrosis in patients with chronichepatitis C. Gastroenterology 2005;129:2064-2075.

28. Elsharkawy AM, Mann DA. Nuclear factor-kappaB and the hepatic in-flammation-fibrosis-cancer axis. HEPATOLOGY 2007;46:590-597.

29. Yeh CN, Maitra A, Lee KF, Jan YY, Chen MF. Thioacetamide-inducedintestinal-type cholangiocarcinoma in rat: an animal model recapitulatingthe multi-stage progression of human cholangiocarcinoma. Carcinogenesis2004;25:631-636.

30. Mazziotti G, Sorvillo F, Morisco F, Carbone A, Rotondi M, Stornaiuolo G, etal. Serum insulin-like growth factor I evaluation as a useful tool for predictingthe risk of developing hepatocellular carcinoma in patients with hepatitis Cvirus-related cirrhosis: a prospective study. Cancer 2002;15:2539-2545.

31. Scharf JG, Braulke T. The role of the IGF axis in hepatocarcinogenesis.Horm Metab Res 2003;35:685-693.


IGF-I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to...

Documents

Transcript of IGF-I gene transfer to cirrhotic liver induces fibrolysis and reduces fibrogenesis leading to...