Additional Supporting Information may be found in the online version of this article.

HEP_23412_sm_SuppFig1.tif2177KSupporting Figure 1. Analysis of the model of established liver cirrhosis. Cirrhosis was induced in male Sprague-Dawley rats of (180–200 g), with weekly intragastric administrations of carbon tetrachloride (CCl4, Riedel-de Haën) for 8 weeks, as described and 400mg/l of phenobarbital in the drinking water (Supporting Fig. 1A) (10). In brief, the initial CCl4 dose was 20 μl per rat. Subsequent doses were adjusted based on the change in body weight 48 hours after the last dose (10). All rats were observed at least twice daily until death. Following this protocol, ascites was apparent in some of the animals. Blood samples were collected from the retro-orbital plexus 8, 16, 25 and 33 weeks after the first administration of CCl4 (Supporting Fig. 1A). Serum transaminases (alanine aminotransferase and aspartate and alkaline phosphate), albumin and bilirubin, were measured (ABX diagnostics) in a Hitachi autoanalyzer (Roche). The results show that transaminases reached highest levels after 8 weeks of CCl4 administration (Supporting Fig. 1B). Then, transaminases decreased but levels were higher than healthy animals even 33 weeks after the first administration of CCl4 (Supporting Fig. 1B). The same result was obtained with bilirubin (data not shown). Also, albumin levels decreased after 8 weeks of CCl4 administration and remained lower than healthy controls 33 weeks after the initiation of the protocol (data not shown). Animals were sacrificed 16, 21 or 33 weeks after the first administrations of CCl4 (Supporting Fig. 1A). Liver samples were fixed in 4% paraformaldehyde, paraffin embedded and stained with hematoxylin-eosin to evaluate liver morphology (data not shown). Liver collagen content was assessed by Sirius red staining and scored by imaging analysis (AnalySIS 3.1, Soft Imaging System) (7). Liver fibrosis was apparent at 16 weeks and remained 21 and 33 weeks after the initiation of the protocol (Supporting Fig. 1C and D). Similar levels were observed in all cases as shown by quantification of Sirius red staining (Supporting Fig. 1D). We did not observed significant changes in the size of parenchymal nodules in the different samples.
HEP_23412_sm_SuppFig2.tif1811KSupporting Figure 2. Immunohistochemistry for vimentin and αSMA and expression of neurotrimin in IGF-I treated animals and controls. Consecutive liver sections, were used to evaluate active HSCs by αSMA staining and total HSCs by vimentin staining (Supporting Fig. 2A). ImageJ (1.40g) software has been used to calculate the ratio of vimentin/αSMA staining (Supporting Fig. 2B). Inactive HSCs were also evaluated by qRT-PCR of neurotrimin mRNA, which is a marker of non-activated HSCs (12) (Supporting Fig. 2C). Samples were taken from healthy animals, or CCl4-induced cirrhotic animals treated with saline (Ci), SVLuc (Ci+Luc) or SVIGF-I (Ci+IGF-I) for 8 weeks. αSMA staining has been previously described. For vimentin staining we used the same conditions with mouse anti-vimentin antibody (clone V9, SIGMA) diluted 1:100. qRT-PCR was done as previously described with the primers and conditions described in Sup. Table 1. The results show decreased αSMA staining in IGF-I treated animals compared to cirrhotic controls (see also Fig. 4). Interestingly, there are many αSMA negative vimentin positive cells in the remaining septum of IGF-I treated animals (Supporting Fig. 2A and B). Similarly, neurotrimin levels are higher in IGF-I treated animals than in cirrhotic controls (Supporting Fig. 2C). These data, suggest that HSCs become deactivated by IGF-I therapy. However, we cannot exclude an indirect mechanism by which dying activated HSCs are replaced by novel non-activated HSCs that localize to remaining septa. We have failed to detect convincing apoptosis or senescence (β-galactosidase staining) of HSCs (data not shown). We have examined tissue samples at 8 weeks after IGF-I vector administration and it is possible that we have missed HSC apoptosis occurring at earlier time points. In fact, HSC apoptosis has been detected immediately after 12 weeks of CCl4 administration but this process was almost absent 168 days later (26). Future experiments are required to address a putative HSCs desactivation and to evaluate the role of inactive HSCs in mediating IGF-I therapeutic effects.
HEP_23412_sm_SuppFig3.tif89KSupporting Figure 3. Analysis of c-met expression in IGF-I treated animals and controls. HGF expression could be crucial to IGF-I therapy (Fig. 6F and 8A). Improved HGF functionality would require increased expression of the HGF receptor c-met. Therefore, we have evaluated the levels of c-met mRNA by qRT-PCR in healthy animals or CCl4-induced cirrhotic animals treated with saline (Ci), SVLuc (Ci+Luc) or SVIGF-I (Ci+IGF-I) for 8 weeks. Primers and conditions are described in Sup. Table 1. The results shows that HGF and c-met share a similar pattern of expression. C-met mRNA is significantly increased in IGF-I treated animals compared to healthy individuals and cirrhotic controls.
HEP_23412_sm_SuppFig4.tif527KSupporting Figure 4. Analysis of SVIGF-I effects in healthy animals. To test whether IGF-I vector administration has unwanted effects on healthy animals, we performed a toxicological analysis. Healthy Sprague-Dawley rats were treated intraarterially with saline (Healthy) or with 1x1011 viral particles or SVLuc (He + Luc) or SVIGF-I (He + IGF-I). Blood extractions were performed at 4 days, and 1, 2, 3, 4, 6 and 8 weeks after vector administration, when the animals were sacrificed. Before sacrifice animals were weighted. Necropsy included organ weigh, histopathological analysis of several organs and specific molecular analysis of liver and serum samples. The results show no signs of toxicity after SVIGF-I expression. All animals showed similar serum levels of transaminases, bilirubin, albumin and total or free IGF-I before sacrifice (data not shown). Similar serum values were also observed for these parameters and for creatinine, total proteins (protein), cholesterol (chol), glucose, lactate dehydrogenase (LDH) and triglycerides 8 weeks after treatment (Supporting Fig. 4 A and B and data not shown). Total weight, and weight of liver, spleen, testis, kidney and heart were similar in all animals (data not shown). Histopathological analysis of kidney, lungs, brain, cerebellum, intestine, liver, muscle, testis, spleen, stomach and heart showed no significant differences in IGF-I treated animals compared to controls (data not shown). Preneoplastic lesions were not observed in any of the samples analyzed. Finally, we analyzed liver expression of IGF-I, IGF-IBP3 and HGF by qRT-PCR. As expected, animals treated with SVIGF-I showed higher levels of IGF-I mRNA than controls (Supporting Fig. 4C). IGF-I mRNA should be translated into a functional protein as IGF-I treated animals also show increased levels of IGF-IBP3 and HGF mRNAs compared to controls (Supporting Fig. 4 D and E).
HEP_23412_sm_SuppFig5.tif269KSupporting Figure 5. Analysis of MMPs and TIM-1 expression in rats with TAA-induced cirrhosis treated with SVIGF-I and controls. We have evaluated MMP1 (Supporting Fig. 5A) and MMP9 (Supporting Fig. 5B) levels by qRT-PCR and TIMP-1 by Western blot (Supporting Fig. 5C) in the liver of healthy rats and rats with TAA-induced liver cirrhosis treated with saline (Ci), SVLuc (Ci+Luc) or SVIGF-I (Ci+IGF-I). The results show increased expression of MMP1 and 9 mRNAs and reduced TIMP-1 levels in IGF-I treated animals compared to cirrhotic controls.
HEP_23412_sm_SuppTab1.tif664KSupporting Table 1. Table of primers and conditions used for quantitative RT-PCRs. Reverse transcription and real-time PCR was done using an iCycler (Bio-Rad, Hercules, CA) and the iQ SYBR Green Supermix (Bio-Rad). Conditions and primers used are summarized in Sup. Table 1. The amount of each transcript was expressed as the n-fold difference relative to the control gene gapdh (2δ Ct, where δ Ct represents the difference in threshold cycle between the control and target genes) as described previously (7).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.