Department of Medicine, Center for Internal Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
German Center for Infection Research (DZIF), Hamburg and Heidelberg sites, Germany
Address reprint requests to: Maura Dandri, Ph.D., Department of Internal Medicine, Center for Internal Medicine, University Medical Center Hamburg-Eppendorf, Martinistraße 52, D-20246 Hamburg, Germany. E-mail: email@example.com; fax: + 49-40-7410 57232.
Potential conflict of interest: Nothing to report.
The study was supported by the German Research Foundation (DFG) by a grant to M.D., J.H., and B.F. (SFB 841: A5; B6; SP2) and to S.U. (UR72-5-1/2). M.D. and S.U. also received funding from the German Center for Infection Research (DZIF-BMBF; TTU05.804).
See Editorial on Page 1458
Chronic hepatitis B virus (HBV) infection has been associated with alterations in lipid metabolism. Moreover, the Na+-taurocholate cotransporting polypeptide (NTCP), responsible for bile acid (BA) uptake into hepatocytes, was identified as the functional cellular receptor mediating HBV entry. The aim of the study was to determine whether HBV alters the liver metabolic profile by employing HBV-infected and uninfected human liver chimeric mice. Humanized urokinase plasminogen activator/severe combined immunodeficiency mice were used to establish chronic HBV infection. Gene expression profiles were determined by real-time polymerase chain reaction using primers specifically recognizing transcripts of either human or murine origin. Liver biopsy samples obtained from HBV-chronic individuals were used to validate changes determined in mice. Besides modest changes in lipid metabolism, HBV-infected mice displayed a significant enhancement of human cholesterol 7α-hydroxylase (human [h]CYP7A1; median 12-fold induction; P < 0.0001), the rate-limiting enzyme promoting the conversion of cholesterol to BAs, and of genes involved in transcriptional regulation, biosynthesis, and uptake of cholesterol (human sterol-regulatory element-binding protein 2, human 3-hydroxy-3-methylglutaryl-coenzyme A reductase, and human low-density lipoprotein receptor), compared to uninfected controls. Significant hCYP7A1 induction and reduction of human small heterodimer partner, the corepressor of hCYP7A1 transcription, was also confirmed in liver biopsies from HBV-infected patients. Notably, administration of Myrcludex-B, an entry inhibitor derived from the pre-S1 domain of the HBV envelope, provoked a comparable murine CYP7A1 induction in uninfected mice, thus designating the pre-S1 domain as the viral component triggering such metabolic alterations. Conclusion: Binding of HBV to NTCP limits its function, thus promoting compensatory BA synthesis and cholesterol provision. The intimate link determined between HBV and liver metabolism underlines the importance to exploit further metabolic pathways, as well as possible NTCP-related viral-drug interactions. (Hepatology 2014;60:1483–1493)
Infection with the hepatitis B virus (HBV) still represents a major health burden, with approximately 350 million individuals chronically infected worldwide who are at risk of developing liver cirrhosis and hepatocellular carcinoma. Although HBV is not directly cytopathic, the establishment of a complex network of virus-host interactions permits the virus to meet its replication requirements and persist within infected hepatocytes. Previous studies indicated that transcription factors involved in activation of hepatic metabolic processes, such as hepatocyte nuclear factors (HNFs), farnesoid X receptor (FXR), cyclic adenosine monophosphate response element-binding protein, and peroxisome proliferator-activated receptors (PPARs), can be recruited to the HBV genome, whereas studies in transgenic (Tg) mice have shown that the viral regulatory protein, HBV X protein (HBx), can induce activation of lipogenic genes and fatty acid (FA) accumulation, which, in turn, could contribute to disease progression by promoting steatosis, generation of oxidative stress, and liver inflammation.[3, 4] Although patient studies indicated an association between chronic hepatitis B (CHB) and hepatic steatosis,[5, 6] knowledge about the effect of HBV infection on metabolic profiles, and on the molecular processes that may be involved in the alteration of lipid and cholesterol pathways in the course of infection, remains limited.
The infectious HBV particle consists of a small DNA-containing enveloped particle, which is characterized by a very high tissue and species specificity. The viral membrane contains three envelope proteins that are named, according to their size, pre-S1 (or large), pre-S2 (or middle), and S (or small). All three proteins share the same C-terminal S domain, which contains the hepatitis B surface antigen (HBsAg), whereas the pre-S2 and pre-S1 proteins display progressive N-terminal extensions. Characteristic of HBV infection is the presence of noninfectious subviral particles, exclusively composed of viral envelope proteins, which are typically secreted in large excess into the blood of infected individuals. The myristoylation and integrity of the first 77 amino acids of the pre-S1 domain of the large envelope protein were shown to be essential for infectivity. Notably, the entry of both HBV and hepatitis delta virus (HDV), which also needs the HBV envelope for infection and propagation, was shown to be blocked by a myristoylated lipopeptide (Myrcludex-B) derived from the pre-S1 domain of the large envelope protein.[8-10] Previous studies indicated that cell polarization, in addition to the differentiation status of hepatocytes, plays a fundamental role in the infection process. Recently, Li et al. could identify the Na+-taurocholate cotransporting polypeptide (NTCP) as the functional cellular receptor permitting HBV and HDV to enter the primary human liver cells, a finding that was confirmed by other groups. These studies revealed that viral binding to NTCP is mediated by the pre-S1 domain of the HBV envelope protein. NTCP is a transmembrane transporter localized to the basolateral membrane of highly differentiated primary hepatocytes, which mediates most of the hepatocellular Na+-dependent uptake of bile salts. Notably, recent in vitro studies showed that Myrcludex-B binding to NTCP inhibits uptake of bile salts.[14-16] These discoveries attribute new importance to possible HBV-mediated changes in bile acid (BA), cholesterol, and lipid metabolism.
We employed immune-deficient (severe combined immunodeficiency [SCID]/beige) urokinase plasminogen activator (uPA) Tg mice harboring livers partially reconstituted with human hepatocytes to investigate whether HBV infection can affect the metabolic profile of human hepatocytes. We found that humanized mice stably infected with HBV displayed remarkable alterations in key genes of cholesterol and BA metabolism and provided evidence that these changes are mostly triggered by the binding of HBV to its cellular receptor.
Materials and Methods
Generation of Human Chimeric Mice, HBV Infection, and Treatments
Generation of humanized mice was conducted as previously described. Briefly, 3-week-old homozygous uPA/SCID/beige mice (USB for short) were anesthetized with isofluoran and injected intrasplenically with 1 million viable thawed human hepatocytes that were obtained from four different donors. Animals were housed and maintained under specific pathogen-free conditions, according to institutional guidelines under authorized protocols. All procedures were approved by the ethical committee of the city and state of Hamburg and accorded with the principles of the Declaration of Helsinki. Human hepatocyte repopulation levels were determined by measuring human serum albumin (HSA) in mouse serum with the human Albumin ELISA Quantitation Kit (Bethyl Laboratories, Biomol GmbH, Hamburg, Germany). Human chimeric mice displaying HSA concentrations between 1 and 4 mg/mL, corresponding to levels of human chimerism ranging from 30% to 70% were employed for the study. To establish HBV infection, animals received a single intraperitoneal (IP) injection of HBV-infectious serum (5 × 107 HBV DNA copies per mouse; genotype D). Four different human hepatocyte donors were used for transplantation. Thirteen stably HBV-infected animals (>10 weeks post-HBV injection) and 13 uninfected control animals were employed to analyze gene expression profiles, so that each group contained at least 3 mice reconstituted with the same human donor. As indicated in the Results, mice were subcutaneously injected with 2 μg/g body weight (b.w.) of Myrcludex-B and sacrificed either 8 or 24 hours after receiving one single drug injection, or after 6 weeks of daily treatment. Uninfected mice received an IP injection of 12.5 ng/g b.w. of simvastatin (SMV; Cayman Chemical Company, Ann Arbor, MI) dissolved in dimethyl sulfoxide (DMSO) and were sacrificed after 6 hours. Control mice received injections with the same DMSO-containing buffer.
Virological Measurements and Analysis of Gene Expression
Extraction of HBV DNA from serum samples was conducted with the QiAmp MinElute Virus Spin kit (Qiagen, Hilden, Germany). For quantification, real-time polymerase chain reaction (RT-PCR) was performed using a Lightcycler (Roche Applied Science, Mannheim, Germany) and HBV-specific probes, whereas cloned HBV-DNA references were amplified in parallel to establish a standard curve for quantification. Viral RNA was extracted from mouse liver samples using the RNeasy RNA purification kit (Qiagen) and was reverse transcribed from 1 μg of total RNA using oligo-dT primers and the Transkriptor kit (Roche Applied Science). To determine gene expression levels, human- and mouse-specific primers from the TaqMan Gene Expression Assay System were used and samples were analyzed in the ViiA™ 7 Real-Time PCR System (both Life Technologies, Carlsbad, CA). The mean of the human housekeeping genes, glyceraldehyde 3-phosphate dehydrogenase and ribosomal protein L30, was used for normalization. The same protocols were also used to determine gene expression in NTCP-transduced HepG2 cells (see also the Supporting Information). Mouse gene expression levels were normalized using primers specifically recognizing the murine β-actin transcripts. To validate the species specificity of the primers, reverse-transcribed complementary DNAs (cDNAs) from murine, human, and human chimeric samples were employed.
Analysis of Human Liver Biopsies
Human liver tissues were obtained from needle liver biopsy specimens taken from 6 HBV chronically infected patients and 6 uninfected individuals, which were observed in the outpatient clinic of the University Medical Center Hamburg-Eppendorf and underwent biopsy to determine grading and staging of liver disease. All HBV-infected patients were HBsAg positive, negative for HCV, human immunodeficiency virus, and HDV serological markers of infection, and were not receiving antiviral treatments. A small piece of tissue not needed for histological examination was immediately frozen in liquid nitrogen, stored at −80°C, and utilized for nucleic acid extraction. The protocol for the study was performed according to the principles of the Declaration of Helsinki and approved by the ethical committee of the city and state of Hamburg (OB-042/06 and PV4081). RNA extraction, reverse transcription, and gene expression analysis were conducted as described above.
Protein Analysis and Immunofluorescence Staining in Humanized Mice
Total proteins (280 μg) extracted from liver tissue of HBV-infected and uninfected humanized mice were used for immunoprecipitation following the manufacturer's instructions (Dynabeads Protein G Kit; Invitrogen GmbH, Karlsruhe, Germany). Precipitates of polyacrylamide gel electrophoresis (CYP7A1) and alpha-antitrypsin (AAT) were further analyzed by western blotting on 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and blotted onto a nitrocellulose membrane (Hybond ECL Nitrocellulose Membrane; GE Healthcare, Buckinghamshire, UK). Both the rabbit anti-CYP7A1 (C-terminal; precipitation, 1:250; detection, 1:1000; Abcam, Cambridge, UK) and the mouse AAT antibody (Ab; precipitation, 1:10; detection, 1:1,000; Biotrend Chemikalien GmbH, Cologne, Germany) used for immunoprecipitation and detection do not exhibit cross-reactivity with mouse proteins. Immunofluorescence (IF) staining was performed on acetone-fixed cryostat liver sections from humanized mouse livers using a human-specific FXR mouse Ab (1:100 dilution; NR1H4, clone 1B10; Abnova, Taipei City, Taiwan). Nuclear staining was achieved by Hoechst 33258 (Invitrogen, Eugene, OR). Stained sections were analyzed by fluorescence microscopy (Biorevo BZ-9000; Keyence, Osaka, Japan) using the same settings for all groups.
Statistical analysis was performed with GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, CA). Mann-Whitney's U test was performed for nonparametric pair-wise comparisons. Kruskal-Wallis' nonparametric test was applied using Dunn's test to compare group-wise as a posttest. P values <0.05 were considered statistically significant.
Effect of HBV Infection on the Expression Profile of Genes of the Lipid, Cholesterol, and BA Metabolism in Human Hepatocytes
To determine the expression profile of human hepatocytes within mouse liver, primers specifically recognizing human transcripts and not cross-reacting with murine genes were validated by RT-PCR using artificial mixtures of human and mouse cDNAs, as previously reported on. We assessed whether persistent HBV infection altered the steady-state expression of genes involved in lipid, cholesterol, and BA metabolism in human liver-chimeric mice by comparing intrahepatic RNA expression levels of the human apolipoproteins, sterol-regulatory element-binding protein (hSREBP), FA synthase (hFAS), and PPARs (hPPARs) in 13 HBV stably infected versus 13 uninfected animals (Table 1). Although HBV-infected mice displayed only modest changes in genes regulating FA and lipoprotein metabolism at the transcriptional level, our analysis revealed a significant enhancement of human apolipoprotein (hAPO) A1 (P = 0.005) and hPPAR-γ (p = 0.01) genes, whereas expression levels of other human apolipoproteins (hAPO B, hAPO C, and hAPO E), hPPAR-α, and hSREBP1c remained unchanged or differences did not reach statistical significance (human liver X receptor [hLXR]-α, hLXR-β, and hFAS). Notably, we found that various human genes involved in transcriptional regulation, biosynthesis, and uptake of cholesterol, such as hSREBP2 (P = 0.0001), hydroxymethylglutaryl-coenzyme A (CoA) reductase (hHMGCR; P = 0.0009), and low-density lipoprotein receptor (hLDLr; P = 0.001) were significantly enhanced in infected mice, compared to uninfected controls (Fig. 1A-C).
Table 1. Expression Profile of Genes Involved in Lipid, Cholesterol, and Bile Acid Metabolism
Fold Difference (Median/Median)
Determined by quantitative RT-PCR in uninfected (n = 13) versus HBV stably infected (n = 13) humanized mice. P values reaching significance are shown as bolded terms; *P < 0.05; **P < 0.01; ***P < 0.001.
Abbreviation: NS, not significant.
Genes of lipid metabolism
Genes of cholesterol metabolism
Genes of bile acid metabolism
The most striking change regarded the strong induction of hCYP7A1 (median, 12-fold; P = 0.0001), the rate-limiting enzyme converting cholesterol into BAs in hepatocytes (Fig. 2A). The clear enhancement of hCYP7A1 in HBV-infected mice could be appreciated also when the four different groups of mice harboring distinct human donor hepatocytes were analyzed separately (Supporting Fig. 1). The hCYP7A1 increase determined at the transcription level could be confirmed also at the protein level (Fig. 2B). Further gene expression alterations regarded a reduction (P = 0.0355) of the expression levels of the nuclear receptor, small heterodimer partner (hSHP), the corepressor of hCYP7A1 transcription (Table 1; Fig. 2C-E), as well as a modest enhancement of human hFXR (P = 0.0018) and of liver receptor homolog 1 (hLRH-1; P = 0.0240) among the HBV-infected group.
Taking into account the crucial role played by NTCP in mediating both viral entry and uptake of BAs in hepatocytes, we determined whether HBV altered steady-state transcript levels of human NTCP. Our measurements indicated that HBV infection did not affect hNTCP at the RNA level (Supporting Fig. 2). Also, expression levels of hBSEP (Table 1), which regulates the canalicular efflux of bile salts from the infected human hepatocytes, did not differ between HBV-infected and uninfected humanized mice, indicating that HBV did not alter, at least at the transcriptional level, the total amounts of the two important bile salt transporters required for bile salt homeostasis.
Importantly, hCYP7A1 was also found significantly induced (P = 0.0022) in liver biopsy samples obtained from patients chronically infected with HBV. Intrahepatic messenger RNA (mRNA) expression of hCYP7A1 was comparatively analyzed by RT-PCR in 6 uninfected and 6 high-HBV chronic carriers (median, 5x10E9 HBV-DNA/mL; range, 8x10E7-4x10E9; Fig. 3). A significant enhancement of CYP7A1 expression was also determined in a small group of patients displaying low viremia (data not shown), suggesting that even lower amounts of circulating virions and/or HBsAg may affect NTCP function. The greater than 1 log induction of hCYP7A1 and significant reduction of the repressor, hSHP (P = 0.009), determined in infected patients strongly support the findings obtained in human-liver chimeric mice.
Effect of HBV on Expression of Mouse Genes in Humanized Mice
To assess whether HBV infection, or rather the presence of circulating virions and subviral particles in mouse blood, could also alter the expression pattern of murine genes involved in BAs and cholesterol metabolism, we employed mouse-specific primers not recognizing human transcripts and the same two groups of animals (13 HBV-infected vs. 13 uninfected) were analyzed. These analyses revealed significant up-regulation of murine (m)CYP7A1 (P = 0.0002), mLDLr (P = 0.0002), mHMGCR (P = 0.0003), and mSREBP2 (P = 0.0056; Supporting Fig. 3). The observation that the induction of genes involved in the synthesis of BAs and cholesterol provision occurred also in murine hepatocytes, which are not susceptible to HBV infection, but were shown to permit the binding of the pre-S1-derived lipopeptide to the mNTCP receptor, suggested that the binding of HBV to mNTCP, rather than intracellular viral replication, may trigger the alterations observed in cholesterol and BA metabolism.
The HBV Pre-S1-Derived Lipopeptide, Myrcludex-B, Induces CYP7A1 Expression in Humanized Mice
To narrow down the viral components responsible for the described expression profile shift in metabolic genes, we assessed whether treatment of uninfected humanized mice (n = 6) with the HBV pre-S1-derived peptide, Myrcludex-B, which can bind both to human and with lower affinity also to murine NTCP,[19, 20] could induce similar alterations. A strong enhancement of human CYP7A1 expression (median, 35-fold) could be induced after one single administration of the HBV entry inhibitor, Myrcludex-B, both 8 and 24 hours posttreatment (Fig. 4A; P = 0.0238), whereas mCYP7A1 appeared to be induced to a lesser extent (Fig. 4B; P = not significant). Moreover, in agreement with previous studies showing remarkable stability in serum and accumulation of Myrcludex-B in mouse liver, hCYP7A1 mRNA levels appeared even higher after 24 hours (n = 3). Notably, repeated administration of Myrcludex-B in HBV-infected mice (3 weeks; daily injection: 2 mg/kg) did not induce further alterations in hCYP7A1 mRNA expression, because hCYP7A1 levels determined in Myrcludex-B-treated, HBV-infected mice were comparable to levels found in untreated HBV-infected littermates (Fig. 4C). In line with the results presented here, expression levels of hSHP, hSREBP2, and hLDLr were also affected upon Myrcludex-B administration (Supporting Fig. 4). But, again, these changes were comparable to those detected in HBV-infected animals.
To assess whether a limited bile salt uptake, mediated by the binding of the pre-S1 domain to NTCP, may lead to up-regulation of hCYP7A1, we also analyzed CYP7A1 expression levels in vitro after incubating NTCP-transduced HepG2 cells with Myrcludex-B or sodium taurocholate (TC). Despite the generally low expression levels of this hepatocyte-specific enzyme in hepatoma cells, CYP7A1 expression was enhanced upon Myrcludex-B administration and further reduced after incubating cells with TC (Supporting Fig. 5). Altogether, these results provide evidence that binding of the pre-S1 domain of HBV envelope to its cellular receptor, NTCP, can trigger gene expression changes in bile acid metabolism.
Several U.S. Food and Drug Administration-approved drugs, including simvastatin (SMV), were shown to limit NTCP-mediated uptake of bile salts. To explore whether binding of SMV to NTCP in humanized mice may also induce metabolic gene alterations resembling those determined in HBV stably infected mice, we administered SMV to 2 uninfected humanized mice and sacrificed them 6 hours after injection. Gene expression analysis showed that SMV administration also increased hCYP7A1 expression (Fig. 4D), suggesting that binding of the pre-S1 component of HBV to NTCP induces effects resembling, at least in part, those observed after administration of compounds that were shown to act as NTCP inhibitors.
Human FXR Cellular Distribution Differs in HBV-Infected Humanized Mouse Liver
BAs also serve as intracellular sensor molecules that can activate FXR transcription factor. Upon nuclear translocation, activated FXR was shown to induce SHP expression, the suppressor of CYP7A1 transcription. To assess whether HBV-mediated reduction of NTCP function may also lead to a decreased FXR activation and hence nuclear localization, we analyzed, by IF, the intracellular distribution of hFXR in liver tissues of HBV-infected and uninfected humanized mice. We observed a clear decrease of FXR amounts in human hepatocyte nuclei of HBV-infected mice (Fig. 5C), in comparison to uninfected controls (Fig. 5A). However, no clear differences could be appreciated in uninfected mice upon one single Myrcludex-B administration (Fig. 5B), indicating that repeated exposure to the pre-S1 peptide may be needed to detect a clear localization shift at protein level, by IF techniques. Moreover, the FXR cellular distribution pattern in HBV-infected and Myrcludex-B-treated mice was similar to the pattern observed in untreated HBV-infected animals (data not shown).
The discovery of the BA transporter, NTCP, as the functional cellular receptor mediating viral entry placed new emphasis on analysis of the involvement of HBV in BA metabolism. However, the function and integrity of metabolic pathways and hepatocyte-specific transporters are known to be altered or lost in cell lines and primary human hepatocyte cultures and the narrow tissue and host tropism of HBV limited studies about interactions established by HBV in infected human hepatocytes. Thus, it is not clear to which extent HBV infection can affect cellular pathways related to host metabolism. In an attempt to overcome these limitations, we have analyzed the expression profile of human genes known to play key roles in the regulation, synthesis, and provision of FAs, cholesterol, and BAs, using four groups of mice harboring distinct human donor hepatocytes, which were either stably infected with HBV or left untreated for comparison. The maintenance of a high differentiation status with production of hepatocyte-specific factors and transport proteins in chimeric mice has already proven valuable, both for infection studies with human hepatotropic viruses[8, 9, 23] and for investigations of human liver-mediated metabolism.
Studies performed in vitro and in Tg mice reported that the expression of HBx protein can cause lipid accumulation by up-regulation of various lipogenic genes.[3, 4] However, our analysis shows only modest alterations of the activity of these genes in HBV-infected human hepatocytes. These differences were mainly limited to a light, but significant, enhancement of hAPO A1 (P = 0.005), a component of high-density lipoprotein particles that plays an important role in reverse cholesterol transport, and of hPPAR-γ (P = 0.01), a nuclear receptor protein known to regulate FA storage and adipocyte differentiation. The less pronounced induction of lipogenic genes determined in HBV-infected humanized mice could be a result of the lower concentrations of HBx protein expected in the infection setting, compared to the levels reached in HBx Tg mice. Notably, and in line with a previous study determining gene expression in HBV Tg mice, we found that various human genes involved in transcriptional regulation (hSREBP2), biosynthesis (hHMGCR), and uptake of cholesterol (hLDLr) were up-regulated in HBV-infected mice. The observed enhancement of cholesterogenic genes suggests that HBV may alter cholesterol metabolism to meet its replication requirements, because depletion of cholesterol was shown to inhibit HBV secretion in hepatoma cells.[28, 29] Thus, promotion of cholesterol synthesis and uptake underlines the importance of these pathways in HBV infection and morphogenesis.
The most dramatic change determined in this study regarded induction of hCYP7A1 (median, 12-fold up-regulation) in HBV-infected mice, which was demonstrated both at the transcriptional and protein level. CYP7A1 is the rate-limiting enzyme of BA synthesis and is responsible for conversion of cholesterol into 7-α-hydroxycholesterol and hence into BAs within hepatocytes. The strong enhancement of hCYP7A1, which was also observed in vitro upon treatment of NTCP-transduced HepG2 with Myrcludex-B, prompted us to examine the expression profile of factors that are involved in regulation of CYP7A1 transcription. Interestingly, significant alterations regarded the lower expression levels of hSHP, which is known to act as a corepressor of hCYP7A1 transcription. Notably, the strong induction of hCYP7A1 transcription and suppression of hSHP was also confirmed in liver biopsy samples obtained from CHB patients.
Enhancement of cholesterogenic genes could be, at least in part, a consequence of CYP7A1 induction, because previous studies indicated that overexpression of CYP7A1 increased transcription levels of SREBP2, LDLr, and HMGCR. Within hepatocytes, BAs operate as signaling molecules that activate FXR, a BA sensor that can initiate various signaling cascades involved in regulation of CYP7A1. Nuclear translocation of the FXR/retinoid X receptor complex was shown to induce expression of SHP, which, in turn, can suppress CYP7A1 transcription also by interacting with LRH-1 and HNF-4α. The reduced presence of hFXR proteins observed by IF in the nuclei of HBV-infected human hepatocytes, along with the transcriptional changes determined upon HBV infection, suggest that the lower levels of hSHP and induction of hCYP7A1 may reflect the attempt of hepatocytes to maintain BA homeostasis, whereas the light up-regulation of hFXR could be a result of feedback mechanisms.
Up-regulation of mCYP7A1 and genes involved in cholesterol and BA metabolism was determined also in murine hepatocytes present within livers of the same HBV-infected humanized mice. Because mouse hepatocytes do not support HBV infection, but permit binding of HBV to mNTCP,[19, 20, 32] we assumed that interactions occurring between circulating viral components and the hepatocyte surface, rather than intracellular viral replication, may provoke the alterations determined in hepatocytes of both species. Because NTCP is responsible for the majority of hepatocellular uptake of conjugated bile salts, we hypothesized (Fig. 6) that binding of HBV to its receptor limits BA uptake and, as a means of compensation, activates pathways promoting BA synthesis.
Myrcludex-B is a synthetic lipopeptide derived from the pre-S1 domain of HBV, which has entered clinical trials after demonstration of its capacity to block HBV and HDV infection in humanized uPA/SCID mice[8-10] upon binding to NTCP. We showed here that also administration of Myrcludex-B induced a strong enhancement of human and, to a lesser extent, murine CYP7A1 expression in uninfected humanized mice, thus providing evidence that binding of the pre-S1 domain of the HBV envelope to its cellular receptor is sufficient to limit BA uptake, thereby triggering alterations of BA metabolism. It is worth noting that repeated administration of Myrcludex-B in HBV stably infected mice did not cause further changes in gene expression, because hCYP7A1 expression levels determined in Myrcludex-B-treated, HBV-infected mice were comparable to those achieved in untreated HBV-infected controls.
Our findings raise important questions about the fate and transport function of hNTCP after viral binding, as well as the molecular mechanisms by which pre-S1-mediated binding to NTCP triggers the observed expression changes in vivo. Besides transporting bile salts, NTCP is involved in the uptake of various drugs that have been shown to affect its capacity to uptake BAs. To this regard, SMV was shown to limit, but not completely inactivate, NTCP-mediated uptake of bile salts, and we showed here that administration of SMV to uninfected humanized mice also enhanced hCYP7A1 expression, thus strongly suggesting that binding of HBV to NTCP induces effects resembling those observed upon administration of substances that are known to act as NTCP inhibitors.
Altogether, these results support previous in vitro observations reporting a limited uptake of bile salts upon Myrcludex-B administration[14, 15] and suggest that, even in the setting of CHB infection, hepatocellular uptake of BAs may be affected. This may lead to different levels of compensatory metabolic alterations. As a result of serum sampling limitations and differences in levels of human liver chimerism, which are known to affect the ratio and type of circulating BA, we could not detect clear bile salt concentration differences among the different groups of humanized mice (M. Haag, personal communication). Future studies will be needed to address such questions and evaluate the consequences that the described metabolic alterations may have on liver disease progression, drug-drug interactions, and other metabolic pathways. Moreover, it will be important to assess whether viral replication may be also affected by factors interfering with BA and cholesterol metabolism. It is tempting to speculate that the intimate link shown between HBV and liver metabolism may be further exploited for host-targeted therapeutic strategies.
The authors are grateful to A. Alexandrov for providing Myrcludex-B, to M. Haag and M. Schwab (IKP, Stuttgart, Germany) for bile salt measurements, to A. Groth and R. Reusch for their excellent assistance with the mouse colony, and to G. Apitzsch, S. Ehret, and C. Dettmer for their great technical help.