BDI causes biliary acid flow obstruction, increases bile deposition in hepatic tissue, and causes hepatocyte degeneration, cell death, and release of liver enzymes such as aminotransferases and AP. Additionally, inflammatory cells migrate to the injured tissue, which contributes to secondary biliary cirrhosis[7, 8].
TGF-β is a pleiotropic molecule involved in ECM proteins expression in fibrogenesis[24, 25]. There is some controversy about whether the serum TGF-β concentration can be correlated with hepatic injury in different liver diseases[26, 27]. Despite this controversy, we found that BDI patients have increased serum concentrations of this protein, perhaps reflecting the overexpression of this cytokine in the liver. mRNA TGF-β expression also increases in rat liver cholestasis[20, 28].
Dooley et al demonstrated that Smad7 (overexpressed through an adenovirus containing Smad7 cDNA) inhibits TGF-β signal transduction and influences the extent of fibrosis when used as preventive therapy in a bile duct ligation (BDL) model of fibrosis. These results suggest that Smad7 is a potential therapeutic tool for liver fibrosis and cirrhosis. The aim of our study was to determine whether Smad7 expression correlates with the gene expression of TGF-β, Col I, Col III, Col IV, and PAI-1 in patients with liver fibrosis caused by BDI.
Tahashi et al showed that Smad7 protein has a critical role in acute and chronic liver injury and that the amount of this protein increases in acute and decreases in chronic liver injury, along with an associated decrease in collagen expression. Our results showing increased Smad7 mRNA expression along with the collagens I, III and IV seem to contradict those of Tahashi et al. However, they used a CCl4 model that produces damage similar to that caused by alcohol in humans and not a BDL model.
We found that TGF-β mRNA expression was 11-fold higher in BDI patients than in liver donors. This contrasts with a 6-fold higher Smad7 expression level in BDI patients. An increase in the expression of Smad7 must correlate with a reduction of the expression of fibrogenic molecules because of the direct negative feedback in TGF-β signaling. Although we found higher Smad7 mRNA expression in patients, expression of the mRNA for collagens and PAI-1 also increased, demonstrating that Smad7 expression is not enough to inhibit TGF-β signaling. These results are similar to those reported recently by Seyhan et al who found continuous Smad7 augmentation along with liver fibrosis cholestasis induced in an experimental rat model was insufficient to blunt fibrogenesis. This result may reflect the observation that the transcriptional regulation of ECM proteins can be directed through another TGF-β dependent pathway, such as connective tissue growth factor (CTGF), a growth factor that is recognized as a fibrogenic molecule in the liver[31, 32]. Smad7 expression is regulated by two mechanisms: transcription and protein degradation. The Smad7 transcriptional regulation is mediated by the Ski protein, which inhibits the Smad7 promoter, and the posttransductional regulation is mediated through the effects of Smurf and Jab1 proteins, which promote Smad7 degradation[34, 35].
We found a strong positive correlation of Smad7 mRNA expression with TGF-β, Col I and Col III mRNA expression. The mRNA expression levels were 6.69, 11.59, 26.86, and 18.84 REU, respectively. We also observed a significant amount of fibrotic tissue deposition in the liver biopsies of BDI patients. In the literature, there is one report on increased Smad7 levels in tissue biopsies. In this study, we were limited to analyzing only mRNA expression, and the expression of Smad7 and of ECM proteins need to be measured in BDI patients.
We emphasize that most of the information about the fibrotic process and the molecules involved had been obtained from experimental models (animals and cell cultures). This is the first work in humans to focus on the mRNA expression of ECM proteins and their relationship to Smad7.
The clinical parameters of biliary obstruction evaluated in the patients (AST, ALT, DB, and ALP concentrations) did not correlate with the gene expression levels, serum TGF-β concentration, or fibrosis percentage. We found no correlation between the lesion type and fibrosis percentage. Type 3 lesion was the most frequent lesion in the patients (11 from 14), but this type was not related to the mRNA expression or liver fibrosis.
Additionally, gene polymorphisms of TGF-β (non coding -509 C/T and coding codon 10, codon 25, codon 263) have been associated with modified serum levels of this cytokine in different diseases [37–39]. We showed increased serum levels of TGF-β in patients with BDI. In this study, we did not evaluate any of the TGF-β polymorphisms. It will be important to plan further studies to determine if serum TGF-β levels are more affected by the polymorphisms rather than the presence of bile duct injury.
A recent report demonstrated the predominance of a single nucleotide polymorphism at codon 25 in cirrhotic, but not healthy subjects of Mexican origin. It will also be interesting to determine if there exists a strong relationship between Smad7 and TGF-β dependent genes such as collagens, PAI-1, TIMPs, and others that lead to extracellular matrix deposition in patients with different polymorphisms for TGF-β.