Accumulating evidence indicates that oxidative stress is involved in the physiopathology of liver fibrogenesis. However, amid the global context of hepatic oxidative stress, the specific role of hepatocyte mitochondrial dysfunction in the fibrogenic process is still unknown. The aim of this study was to determine whether a targeted protection of hepatocytes against mitochondrial dysfunction could modulate fibrosis progression. We induced liver fibrogenesis by chronic carbon tetrachloride treatment (3 or 6 weeks of biweekly injections) in transgenic mice expressing Bcl-2 in their hepatocytes or in normal control mice. Analyses of mitochondrial DNA , respiratory chain complexes, and lipid peroxidation showed that Bcl-2 transgenic animals were protected against mitochondrial dysfunction and oxidative stress resulting from carbon tetrachloride injury. Picrosirius red staining, ␣-smooth muscle actin immunohistochemistry, and real-time PCR for transforming growth factor-␤ and collagen ␣-I revealed that Bcl-2 transgenic mice presented reduced fibrosis at early stages of fibrogenesis. However, at later stages increased nonmitochondrial/ nonhepatocytic oxidative stress eventually overcame the capacity of Bcl-2 overexpression to prevent the fibrotic process. In conclusion, we demonstrate for the first time that specific protection against hepatocyte mitochondrial dysfunction plays a preventive role in early stages of fibrogenesis, delaying its onset. However, with the persistence of the aggression, this protection is no longer sufficient to impede fibrosis progression.

The role of hepatocyte apoptosis in the physiopathology of obstructive cholestasis is still controversial. Although some data have strongly suggested that hepatocellular cholestatic injury is due to Fas-mediated hepatocyte apoptosis, some others concluded that necrosis, rather than apoptosis, represents the main type of hepatocyte death in chronic cholestasis. Moreover, it has also been suggested that the reduced liver injury observed in the absence of Fas receptor after bile duct ligation was not due to lower hepatocyte apoptosis but to the indirect role of this receptor in non-hepatocytic cells such as cholangiocytes and inflammatory cells. The aim of this work was therefore to determine whether a protection against cell death limited to hepatocytes could be sufficient to reduce liver injury and delay cholestatic fibrosis. With this purpose, we performed bile duct ligation in transgenic mice overexpressing Bcl-2 in hepatocytes and in wild-type littermates. We found that, compared with necrosis, apoptosis was negligible in this model. Our results also showed that hepatocyte Bcl-2 expression protected hepatocytes against liver injury only in the early steps of the disease. This protection was correlated with reduced mitochondrial dysfunction and lipid peroxidation. However, in contrast to Fas receptor-deficient lpr mice, fibrosis progression was not hampered and liver inflammatory response was not reduced by Bcl-2 overexpression. These results therefore comfort the hypothesis that Fas-mediated apoptotic hepatocyte pathway is not a significant contributing factor to the clinical features observed in cholestasis. Moreover, in the absence of a blunted inflammatory response in transgenic mice, Bcl-2 protection against hepatocyte mitochondrial dysfunction and lipid peroxidation was not sufficient to block fibrosis progression.

Muscles are composed of multinucleated muscle fibers with different contractile and physiological properties, which result from specific slow or fast gene expression programs in the differentiated muscle cells. In the zebra fish embryo, the slow program is under the control of Hedgehog signaling from the notochord and floor plate. This pathway activates the expression of the conserved transcriptional repressor, Prdm1 (Blimp1), which in turn represses the fast program and promotes the slow program in adaxial cells of the somite and their descendants. In the mouse embryo, myogenesis is also initiated in the myotomal compartment of the somite, but the slow muscle program is not confined to a specific subset of cells. We now show that Prdm1 is expressed in the first differentiated myocytes of the early myotome from embryonic day (E)9.5-E11.5. During this period, muscle formation depends on the myogenic regulatory factors, Myf5 and Mrf4. In their absence, Prdm1 is not activated, in apparent contrast to zebra fish where Prdm1 is expressed in the absence of Myf5 and MyoD that drive myogenesis in adaxial cells. However, as in zebra fish, Prdm1 expression in the mouse myotome does not occur in the absence of Hedgehog signaling. Analysis of the muscle phenotype of Prdm1 mutant embryos shows that myogenesis appears to proceed normally. Notably, there is no requirement for Prdm1 activation of the slow muscle program in the mouse myotome. Furthermore, the gene for the transcriptional repressor, Sox6, which is repressed by Prdm1 to permit slow muscle differentiation in zebra fish, is not expressed in the mouse myotome. We propose that the lack of functional conservation for mouse Prdm1, that can nevertheless partially rescue the adaxial cells of zebra fish Prdm1 mutants, reflects differences in the evolution of the role of key regulators such as Prdm1 or Sox6, in initiating the onset of the slow muscle program, between teleosts and mammals.