FXR, intestinal FiXeR of hepatocellular carcinoma?

Potential conflict of interest: Nothing to report. See Article on Page 161 Despite increasing insight into the molecular pathogenesis, treatment options for hepatocellular carcinoma (HCC) are limited.1 Late‐stage diagnosis, frequently rendering patients ineligible for curative treatments, and high recurrence rates add to the poor prognosis. Strategies to detect and initiate treatment of HCC at earlier stages are likely to result in improved patient outcomes. In this issue, Degirolamo et al. demonstrate that development of HCC in aged mice lacking farnesoid x receptor (FXR) can be prevented by re‐expressing a constitutive active form of FXR in the intestine.2 FXR is a ligand‐activated transcription factor with its target genes controlling metabolic homeostasis, inflammation, and tissue repair.3 Bile salts are the endogenous ligands of FXR, and FXR serves as a master regulator that prevents cytotoxic effects of excessive intracellular bile salts. FXR is expressed in tissues in contact with bile salts as part of normal physiology (e.g., intestine and liver) and in tissues that encounter bile salts primarily under conditions of pathophysiological overload such as cholestasis (e.g., kidneys). Genetic deficiency of Fxr in mice results in dysregulated bile salt homeostasis with elevated bile salt synthesis, expansion of the circulating and hepatic bile salt pool, and progressive hepatocellular injury.4Fxr−/− mice (12‐15 months old) of either gender develop spontaneous tumors in their liver (hepatocellular adenoma and carcinoma), but not elsewhere. Hepatocarcinogenesis in these mice is preceded by inflammatory and fibrogenic stages and is accompanied by activation of the Wnt/β‐catenin pathway6 and large‐scale alterations in expression of genes engaged in metabolism, inflammation, extracellular matrix remodeling, cell proliferation, and cell death.4 Thus, a chronic setting of bile‐salt–induced liver damage, ensuing inflammation and compensatory proliferation, and an exaggerated tissue repair response appears to underlie hepatocarcinogenesis in the absence of Fxr. FXR controls hepatic synthesis, uptake, biotransformation, and biliary secretion of bile salts. Gut‐liver crosstalk involving ileal FXR and the ileal enterokine, fibroblast growth factor (FGF) 15/19 (in rodents and humans, respectively), is crucial for control of hepatic bile salt synthesis.7 FGF15/19 signaling through hepatocytic FGF receptor 4 (FGFR4), results in down‐regulation of cytochrome P450 (CYP)7A1, the initiating and rate‐limiting enzyme in the dominant pathway of bile salt synthesis. Degirolamo et al. tested whether restoration of intestinal FXR action in a background of whole‐body Fxr deficiency prevents hepatocarcinogenesis.2 To this end, Fxr−/− mice were crossbred with transgenic mice expressing human FXR with a swapped transactivation domain (VP16) under control of the villin promoter. This strategy obviates the requirement of a ligand to activate FXR and restricts transgene expression to differentiated enterocytes. Groups of young (4 months) and aged (16 months) Fxr−/− mice expressing the VP16 transactivation domain only (Fxr−/−iVP16, controls), or as fusion with the DNA‐binding domain of human FXR (Fxr‐/‐iVP16FXR), were studied. Intestinal FXR reactivation resulted in up‐regulation of ileal Fgf15 and activation of a downstream effector of hepatic Fgf15/Fgfr4 signaling (i.e., c‐Jun N‐terminal kinase7) in young mice. This was accompanied by reduced Cyp7a1 messenger RNA/protein expression and lowering of serum, hepatic, and fecal bile salts. Thus, constitutive activation of intestinal FXR improves or corrects bile salt dyshomeostasis in the Fxr−/− background, and this appeared sufficient to prevent the development of HCC. None of the Fxr−/−iVP16FXR mice displayed hepatic tumors, whereas multiple tumors (4‐12 per mice) were observed in all control mice after 16 months. Histological features that were noted in young (disrupted hepatic parenchyma and collagen deposition) and aged (lobular inflammation, collagen deposition, and marked necrosis) control mice were largely absent or less striking in Fxr−/− mice with intestinal FXR reactivation. At the molecular level, improved hepatic inflammatory state in aged Fxr−/−iVP16FXR mice was apparent from reduced hepatic expression of interferon‐ɣ and the activated form of its downstream signaling mediator, the protumorigenic transcription factor, signal transducer and activator of transcription 3.8 Effects on both liver (reduction of bile‐salt–induced inflammation) and intestine (improved antibacterial defense and barrier function) may contribute to amelioration of hepatic inflammation. Furthermore, transcriptome analysis revealed that intestinal FXR reactivation affected hepatic expression of cyclinD1 (repression) and small heterodimer partner (Shp; up‐regulation) in both young and aged mice. CyclinD1 expression is negatively regulated by Shp, an Fxr target gene. CyclinD1 overexpression as well as Shp deficiency result in spontaneous HCC in mice, and Shp overexpression can reduce tumor malignancy in Fxr−/− mice.9 The combined findings indicate that multiple molecular pathways contribute to the hepatoprotective effect of intestinal FXR reactivation in Fxr−/− mice. Although the etiology of human HCC is diverse, HCC usually develops in a setting of chronic liver injury and ensuing inflammation, fibrosis, and cirrhosis. Hepatocarcinogenesis in Fxr−/− mice resembles HCC in pediatric patients with genetic defects of the bile salt export pump, where an etiological role of bile‐salt–induced hepatotoxicity is also foreseen.10 However, this is a rare condition and one could ask whether FXR is dysregulated in more common forms of HCC developing against a background of chronic inflammation as in hepatitis B virus, hepatitis C virus, or nonalcoholic steatohepatitis. Chronic exposure to cytotoxic bile salts, combined with abrogated defense mechanisms in the gut and liver, may be an important and underestimated theme in hepatocarcinogenesis. Decreased expression of hepatic FXR and FXR target genes is observed in human HCC.4 However, of greater relevance is the question of whether FXR activity is deficient in the cirrhotic stage that often precedes HCC and, if so, in what forms of cirrhosis. Proinflammatory cytokines interleukin‐6 and tumor necrosis factor alpha, both induced during experimental carcinogenesis and elevated in human cirrhosis and HCC, repress hepatic FXR gene expression or alter the ratio of relevant FXR isoforms.4 It is unexplored whether post‐translational modifications that modulate FXR activity contribute to (presumed) hepatic FXR dysfunction in cirrhosis and HCC.14 Fecal bile salts are lowered,15 whereas urinary and serum bile salts are elevated, in liver cirrhosis and HCC,16 indicative of disturbed bile salt homeostasis. This is in line with impaired function of FXR. Thus, (gradual) loss of FXR‐mediated hepatoprotection appears relevant for human hepatocarcinogenesis and may also play a role in HCC recurrence.17 It remains to be established whether the hepatoprotective effect of constitutive activation of intestinal FXR is maintained in other animal models of HCC and whether progression of established tumors can be halted or reversed by this approach. Nonetheless, Degirolamo et al. raise the intriguing possibility of (intestinal‐selective) FXR modulation for prevention of HCC. A potential safety concern for such therapeutic approach in humans is the possible interaction between FGF19 (induced by intestinal FXR activation) and human HCC. Ectopic overexpression of FGF19 induces HCC in mice in an Fgfr4‐dependent manner.18 In human HCC, FGF19 was coamplified with CCND1 (encoding cyclinD1) in 14% of cases with representative etiology and was identified as an equally important driver gene.19 Moreover, frequent overexpression (30%‐47% of cases18) of FGFR4 is observed in HCC and is implicated in its malignant nature.20 Thus, deregulation of the FGF19‐FGFR4 pathway may play a role in hepatocarcinogenesis in at least a subset of patients. In mice, genetic intestinal FXR reactivation resulted in only modest induction of ileal Fgf15 expression while completely preventing HCC development in otherwise Fxr‐deficient mice.2 It remains imperative to determine whether long‐term pharmaceutical stimulation of (intestinal) FXR gives rise to undesirable hepatoproliferative effects, in particular in patients with liver cirrhosis. Author names in bold designate shared co‐first authorship.