The Role Of Farnesoid X Receptor On Hepatoprotection And Hepatocarcinogenesis

HCC development is multi-factorial, complex multi-gene andmulti-step process, HCC occurrence, the specific mechanisms ofdevelopment continues to be explored. Chronic hepatitis C virus infection(HBV/HCV) is the most important reason leading to cirrhosis, livercirrhosis is an important risk factor of HCC and nearly1/3patients withcirrhosis developed to HCC eventually. In recent years, more and moreevidence demonstrated that metabolism disorders also associated with theoccurrence of HCC, nuclear receptor FXR as an important regulator ofbile acids, carbohydrates, lipids play an important role in regulatinghomeostasis, and recent studies showed, that FXR also related with liverregeneration, inflammation and even hepatocellular carcinoma. In thisstudy, the important roles of FXR in hepatic inflammation, fibrosis, andhepatocellular carcinoma were further investigated in mice model andhuman HCC patients. Our study is not only to reveal the function of FXR,furthermore, it provides a new perspectives for the treatment and a newtarget of HCC.1.The role of FXR on DEN inducing acute liver injuryIn the first study, the different response to the acute effects of DENin FXR deficient mice and wild-type mice were compared, the resultsshowed livers of mice without the protection of FXR exhibited enhancedsensitivity to DEN, performance of higer transaminase levels, increasedapoptosis of hepatocytes and cytokines secreting. The expression mRNAslevels of inflammatory cytokines either TNF-α, IL-6, IL-1β or regeneration-associated factor TGF-α, HGF were elevated moreobviously in the livers absence of FXR in24h timepoints. NF-κBtranscriptional activity was further confirmed to be increased in FXR-/-mice by EMSA compared to wild-type mice. In order to clarify the role ofFXR in regμl ating NF-κB activity, in vitro study was further investigated.The resμl ts showed NF-κB-mediated inflammatory cytokines (TNF-α,IL-6, COX2) expression has significantly inhibit after activation of FXRin both primary hepatocellular carcinoma cell lines and mouse liver cells,but the same suppressive effect coμl d not seen in FXR deficient livercells. These results suggested that increased sensitivity to DEN inFXR-deficient mice may due to their lack of inhibition of NF-κB by FXR,resulting in NF-κB activation and the occurrence of excessiveinflammatory response.2.The role of FXR on formation of HCC mice model induced by DEN.In this study, DEN-induced HCC model in wild-type and FXRdeficient mice were successful established. With FXR gene deletion, themice showed increased susceptibility to DEN, which displayed a highertumor formation rate (100%VS.80%), more tumor nodules and greatertumor volume characteristics. In apoptosis and proliferation assays, wefound that apoptosis and proliferation were significantly increased inFXR-/-mice cancer cells compared with wild-type liver cells, in addition,liver inflammation was also significantly enhanced. For further tissuespecimens, we found that the cytokine TNF-α, IL-6, IL-1β expressionlevel of gene expression in cancer tissues were higher than thesurrounding normal tissue, and expressions of TNF-α and IL-6werehigher in both FXR-/-adjacent tissue and cancer tissue compared withwild-type mice, especially in cancer tissues. Increased transcriptionalactivity of NF-κB by EMSA was confirmed in either adjacent tissues or cancer tissue of FXR-/-mice compared with wild-type mice. To furtherinvestigate molecular phenotype differences in hepatocarcinogenesis, weexamined the expression of factors related to tumor occurrence, theresults showed that CyclinD1and CyclinE mRNA levels in tumor tissueswere higher than the corresponding peritumoral tissues in the two groupsof mice compared with wild-type mice, expression of cyclinE andcyclinD1in HCC tissue and peritumoral tissues were elevated in FXR-/-mice, especially in cancer tissues; P53protein levels were much lower inFXR-/-mice tumors and adjacent tissues than in wild-type mice,particularly its expression in tumor tissues was significantly inhibited.But the further results in vitro showed FXR can not affect P53expression after UV irridation. The mechanism needs to be futherinvestigated.3. Effect of FXR on liver fibrosisIn DEN chronic injury mouse model, TGF-β1and TIMP-1mRNAlevels in FXR-/-mice livers were significantly higher than wild-type mice,indicating that the degree of liver fibrosis is more obvious in the absenceof FXR conditions. To further investigate the relationship between FXRand liver fibrosis, we successfully separated the hepatic stellate cells fromthe livers of wild-type and FXR-/-mice, the mRNA expression of TGFβ1target gene PAI-1, α-SMA and Collagenα1were examined with TGF-β1stimμl ating after adding FXR activators--GW4064in HSCs, and theexpression of FXR target genes SHP has also been checked. The resultsdemonstrated that there FXR activation of hepatic stellate cells cansignificantly inhibit the expression of TGF-β1-induced expression offibrosis-related factors, and the significant inhibition can not be seen inthe absence of FXR group. Intrestingly, the expression of SHP can bedownregμl ated by TGF-β1inversely. Based on the results, FXR can protect TGF-β1-induced hepatic fibrosis, liver fibrosis and meanwhile thesecretion of TGF-β1after HSC activation can inhibit FXR activationwhich accelerate the process of the formation of hepatic fibrosis.4. FXR expression in Human HCC tissue specimensThe expression of FXR in human HCC tissues were furtherexamined. In the study,41specimens (including hepatocellular carcinomaand adjacent normal tissue) were selected according to differentpathological grade, both the mRNA expression of FXR and related genesand the protein levels of FXR in human HCC tissue were significantlydecreased compared to adjacent normal tissues by quantitative PCR andimmunohistochemical staining. We further checked some genes whichhad been confirmed to be significant differentially expressed genes inwild-type and FXR-/-mice liver using microarray assay previously, thesegenes related with cellular metabolism, proliferation and inflammation.Similar results were obtained in human HCC tissus compared withnormal liver tissue, eight genes in human HCC and FXR-/-mouse liversshowed similar upregulated or downregμl ated. CXCL1, DPYS, CD36,LIP1, and SOCS3which were decreased in FXR-/-mice liver also showeddownregμl ated in human HCC. FXR target gene SOCS3, CD36andLAGLS1has been confirmed closely associated with tumorigenesis andtumor development. Then we examined the level of methylation of FXRpromoter region, no typical CpG island is found in FXR promoter region.FXR gene expression is detected after methylation agent was added toHepG2cells for72h, the results showed that no significant change ofFXR mRMA expression with5-Aza-dC. The resμl ts demonstrated thatdownregulation of FXR expession was no significant correlated to thehypermethylation of FXR promoter.5. Inflammatory factors regulating the transcriptional activity of FXR To further investigate the mechanism of reduction of FXR in humorHCC tissues, primary hepatocytes were isolated and differentconcentrations of TNFα, IL-1β or IL-6were added to the cμl ture system.We observed that TNFα (10ng/ml), IL-1β (50ng/ml) and IL-6(1ng/ml) can significantly reduce the expression of FXR mRNA levels. Tofurther investigate whether the downregμl ated of FXR expression is dueto decreased activity of FXR promoter, the sequence of HNF-1α bindingsites (p-150/+129) in promoter of human FXR was inserted intorecombinant plasmid vector pGL-3which contain luciferase (luciferase)reporter gene. The reporter plasmids were transfected into HepG2cellsfor24h, respectively, in the culture system with TNFα, IL-1β, or IL-6(10ng/ml,50ng/ml) for24h, we used double fluorescence applicationsluciferase reporter system to detect luciferase activity, and the resultsshowed that the substrate of luciferase activity was significantly reducedwhich demonstrated the FXR promoter activity was inhibited bycytokines. The experiment was repeated in Hep3B cell line and the sameresults were obtained, the effect was more obvious at high concentrationsof IL-1β, TNF-α. Furthermore,, the substrate of luciferase activity was nosignificant change and FXR promoter activity was not inhibited if themutant sequence of FXR were inserted to promoter reporter plasmid(pGL3mut) then transfected HepG2cells under the same conditions. Theresults indicated that inflammatory cytokines can decrease FXRtranscriptional activity by HNF-1α. Then we examined the effect on geneexpression of HNF1α with inflammatory factors, the results show thatcytokines can not regulate the gene and protein expression of HNF1α.In order to further investigate the mechanism of inflammatorycytokines on FXR expression, chromatin immunoprecipitation (CHIP)experiments were applied to examined the effect of inflammatory cytokines TNFα on HNF1α DNA binding activity with FXR. The resultsshowed: FXR binding activity was not affected with TNFα stimulation inIgG antibody control group, but the binding activity of HNF-1promoterand FXR was significant decreased after TNFα stimulation in HNF-1antibody group compared with the untreated group. In order to determineits effect on liver cancer cells, we repeated the experiment above inanother tumro cell line Huh-7, and got the same result. These resultssuggest that inflammatory cytokines can impair the binding activity ofHNF1α promoter and FXR, which led to the transcriptional activity ofFXR decreased, resulting in decreased expression of FXR.