Hepatitis B virus: The “metabolovirus” highjacks cholesterol and bile acid metabolism

The detection of the Na/taurocholate cotransporter, NTCP (SLC10A1) as an entry receptor for the hepatitis B virus (HBV) (and its satellite Delta virus) clearly represents a milestone in our understanding of this infection. NTCP represents the main transporter for sodium-dependent bile acid uptake from portal blood into hepatocytes.1 Located at the basolateral plasma membrane of hepatocytes, NTCP has been identified as the binding site for the HBV pre-S domain of the HBV envelope protein,2 a finding that was recently confirmed by an alternative bioinformatics-assisted array approach.3 The molecular determinants critical for viral entry overlap with those for bile acid uptake and interfere with normal function of NTCP.4 Natural NTCP substrates at high concentration functionally interfere with HBV infection, indicating competition for the same sodium-driven transport process. In turn, inactivation of NTCP-mediated bile acid transport by Myrcludex B, a peptide ligand derived from the HBV L-protein, raised the question whether HBV infection affects bile acid homeostasis in HBV-infected patients.3 Oehler et al.5 now demonstrate that indeed HBV infection alters bile acid and cholesterol metabolism as a consequence of impaired bile acid uptake (Fig. 1). Bile acid homeostasis is mediated upon binding to the nuclear receptor FXR (farnesoid X receptor), which in turn activates the nuclear repressor small heterodimer partner (SHP) to modulate the gene expression of NTCP and cholesterol 7α-hydroxylase (CYP7A1), the rate-limiting enzyme in bile acid synthesis, according to the intracellular availability of the ligand.6 Using human liver-chimeric uPA/SCID mice the authors showed that (ligand-induced) nuclear FXR localization and SHP expression are decreased in mice with chronic HBV infection, whereas CYP7A1 expression is increased. These findings could be validated in liver biopsy samples from HBV-infected patients. Expression of hNTCP is modestly increased (1.4-fold) in chronically infected humanized mice. This finding is in line with hNTCP gene activation by the glucocorticoid receptor and peroxisome proliferator-activated receptor (PPAR)-gamma coactivator-1 alpha (PGC-1α) and FXR/SHP-dependent suppression by bile acids.7 Together with SREBP2 as the key regulator of cholesterol metabolism, genes involved in cholesterol uptake (LDL receptor) and biosynthesis (HMG-CoA reductase) are induced upon HBV infection. Treatment of uninfected humanized mice with pre-S1 derived Myrcludex resulted in the same changes in bile acid and cholesterol metabolism as described for HBV-infected mice. These findings provide further functional evidence that binding of the pre-S1 domain to its cellular receptor NTCP is sufficient to interfere with intracellular nuclear receptor activation. It remains open whether the level of HBV viremia is critical for this interference with bile acid transport. Preliminary data from a small group of patients in the current study suggest that even low levels of HBV DNA may be sufficient to suppress NTCP and FXR function. Given the substantial changes in cholesterol uptake and biosynthesis, subsequent activation of oxysterol-dependent pathways by the nuclear receptor LXR would be expected. Of note, the viral HBx protein itself contributes to LXR activation and subsequent alterations in cholesterol metabolism.8 However, in the study by Oehler et al. no increase in LXR as well as the lipogenic transcription factor SREBP1c could be observed at the mRNA expression level and the up-regulation of its target gene fatty acid synthase (FASN) also did not reach statistical significance. Increased expression of apolipoprotein A1 and PPARγ even indicated a favorable gene expression profile. Obviously, the observed changes in cholesterol uptake and synthesis do not cause the same changes in lipid metabolism as described in mice with genetic hepatocellular Fxr-deficiency under a high cholesterol diet which develop hepatic steatosis.9 Further studies of serum bile acid, cholesterol, and triglyceride levels as well as metabolic consequences in target organs such as the liver are clearly warranted. Whether lipogenic gene activation and hepatic steatosis actually appear in human livers upon HBV infection remains unclear. Conflicting data exist with regard to the association of HBV infection with hepatic steatosis.10-12 Interestingly, an increased likelihood of HBs antigen loss and decreased HBV DNA levels have been described with extreme obesity in the REVEAL-HBV study database.13, 14 This association may explain the observation that chronic HBV carriers have been characterized by low serum cholesterol and triglyceride levels and partly associated with the absence of hepatic steatosis.15-17 From the teleological standpoint, the question remains whether the observed changes are in favor of the virus. HBV assembly is dependent on cholesterol and restriction to cholesterol uptake by ezetimib has been shown to suppress viral infection.18 Furthermore, LXR has been identified as one of the nuclear factors activating HBV transcription.19 Decreased nuclear localization of FXR and absent SHP activation in HBV-infected hepatocytes as observed by Oehler et al. contributes to a maintained expression of NTCP as the entry receptor on the basolateral membrane. In addition, bile acids and their nuclear receptor FXR increase HBV transcription and gene expression, at least in cell culture experiments.20, 21 In favor of the host, it can be noted that a shift in energy metabolism towards a lipogenic profile should go along with an inactivation of PGC-1 alpha. This coreceptor plays a central role in HBV transcription as an integrator of several nuclear receptors including HNF4 and FOXO1.22, 23 It could be speculated that the discussed metabolic changes generate a permissive milieu early after HBV infection but long-term consequences such as hepatic steatosis may ultimately contribute to viral eradication, which typically occurs late in the course of the infection. Metabolic changes induced by HBV infection of hepatocytes. Alterations in related signaling pathways and their implications on HBV replication are highlighted at the bottom. The influence of HBV infection on triglyceride metabolism and its impact on viral replication remains uncertain. ∼ indicates no change. Appropriate measure to address metabolic changes induced by the infection may not be food restriction. From several epidemiological studies it became clear that individual diet and nutrition state play a relevant role for the observed diversity in disease severity of HBV infection, which is particularly unfavorable in Asia. It has been proposed that short-term starvation and subsequent activation of PGC-1 alpha may be disadvantageous due to an increased transcription and gene expression of the “metabolovirus.”24 Furthermore, fasting could also increase the susceptibility of hepatocytes since it led to a PGC-1a-mediated overexpression of Ntcp, at least in rats.25 Finally, the implications of inherited and acquired conditions of bile acid retention as well as functional NTCP variants are unclear. NTCP expression is down-regulated as a protection against bile acid overload in states of cholestasis and a decreased expression has been described in patients with primary biliary cirrhosis.6, 26 It could be assumed that viral entry is at least less effective in cholestatic hepatocytes and that individuals carrying the cholestatic V444A polymorphism in the bile salt export pump ABCB11 may be less likely to acquire HBV infection. This would be in contrast to HCV, where bile acid retention may even decrease the likelihood of viral clearance during acute infection, as hypothesized from an overrepresentation of the same polymorphism in chronically infected individuals.27, 28 To date, no such data are available for HBV. Several NTCP variants have been described which interfere with taurocholate uptake function, HBV preS binding, and viral uptake.4 As a proof-of-concept, a study from Asia reports in HBV carriers at least a clear trend towards decreased CT genotype and T allele frequency of the NTCP S267F (c.800C>T) variant, which has been associated with decreased bile acid uptake function.29 Importantly, this S267F variant could neither bind to the preS1 region nor support HBV infection in cell culture studies.4 Without doubt, the present study represents an important step towards the understanding of the complex interplay between HBV infection and liver metabolism and offers attractive new concepts for future host-targeted antiviral strategies. Andreas Geier, M.D.