TRIF as a Novel Modulator of Liver Inflammation and Fibrosis.

The central role of toll-like receptor 4 (TLR4) activation in nonalcoholic steatohepatitis (NASH) has been well-recognized, specifically its role in the activation of innate immune responses, hepatocyte apoptosis, and fibrosis. TLR4 could be activated by various signals; in the context of NASH dysregulation of gut microbial homeostasis, gut leakiness and consequent increase in bacteria-derived lipopolysaccharide (LPS) seem to play a major role. Activation of TLR4 involves adapter proteins including MyD88-adaptor-like and TRIF-related adaptor molecule (TRAM) recruiting Myd88 and TRIF, respectively, and ultimately leading to the activation of nuclear factor kappa B and/or interferon-regulatory factor 3 (IRF3). In addition, the sterile α- and armadillo-motif-containing protein (SARM) was shown to interact with TRIF, modulating its function. In models of liver injury and fibrosis TRIF has been studied as the main inducer of IRF3. Although earlier reports have demonstrated that stimulator of interferon genes (STING)–mediated IRF3 induction is a key to apoptosis in early alcoholic injury, others concluded that IRF3 is actually protective against steatosis and insulin resistance in nonalcoholic fatty liver disease. To further analyze the reason for these seemingly controversial findings and to delineate the role of the TLR4/TRIF axis in NASH, Yang et al analyzed the cell and pathway-specific roles of TRIF that are responsible for inflammation, steatosis, and fibrosis in NASH. The authors have used TLR4–/–, TLR4–/–/BM chimeric, and TRIF–/– mice on choline-deficient amino acid defined NASH diet for these studies. Although these experiments revealed that intact TLR4/TRIF in both hepatocytes and BM-derived monocytic cells were required for steatosis, surprisingly, TRIF–/– mice exhibited worse inflammation with an increase in alanine aminotransferase, tumor necrosis factor α levels, and hepatocyte ballooning. Consistently, fibrosis was also worse in these mice. Interestingly however, both TLR4–/– and TRIF–/– hepatocytes were more resistant to palmitate and LPS-induced apoptosis. To identify the possible mechanism for the increased inflammation in TRIF–/– mice, the authors focused on chemokine production and noticed an exacerbated release of CXCL1 and CCL3 in TRIF–/– hepatic stellate cells (HSCs) that were exposed to LPS. On the basis of this, it is plausible that TRIF by a yet unidentified pathway represses the LPS/TLR4/Myd88 axis in HSCs and controls chemokine production and thereby the recruitment of BM-derived monocytic cells in NASH.

The potential direct role of TRIF in hepatocytes also deserves some attention. Notably, TRIF was recently examined in a mouse model of acute sterile inflammation and hepatocyte cell death. In this model TRIF–/– mice had a lower inflammatory response with a reduced recruitment of myeloid cells accompanied by a decrease in CCL2 and CXCL1 in the tissue and the serum. Injured hepatocytes through TRIF were thought to be responsible for the intense inflammatory response through the induction of myeloid recruiting factors. However, in a subset of TRIF–/– mice no protection was seen, and in these mice a potential tip toward Myd88-related signaling was thought to be responsible for the inflammation. The discrepancy between the above studies could be due to several factors; whereas in the acute sterile injury model the effects of TRIF were independent of TLR4 and type I interferon, in a chronic model such as NASH LPS/TLR4 play a significant role. It is possible that other factors such as TRAM recruitment or SARM action could be responsible for the different outcomes in the acute and chronic models. One might also argue that the role of TRIF could be cell type–specific, and the effect on inflammation could be differentially regulated depending on the chronicity and extent of the pathologic insult. Activated HSCs in the NASH model potentially can have a key role modulating CXCL1 and/or CCL3-mediated recruitment of inflammatory cells. Supporting this notion are the findings of Iracheta-Vellve et al showing that in the CCl4-mediated fibrosis model IRF3, the downstream target of TRIF, by its association to STING, is central to hepatocyte apoptosis and subsequent fibrosis. TRAM or TRIF deficiency in this model could not prevent fibrosis, consistent with the study in NASH. Finally, it is also worthwhile to note that throughout the NASH study total TRIF knockouts were used; thus the potential role of adipose tissue inflammation or effects on intestinal barrier function could not be excluded. In fact, TRIF in intestinal epithelial cells serves as an important regulator of antimicrobial defense, and TRIF–/– mice had lower antimicrobial peptide expression. In summary, TRIF evolves to be an important regulator of host immune responses in acute or chronic liver injury. To better understand its specific roles and targets, further studies will be necessary in cell or tissue specific knockout or knock-in models.