Identification of NTCP as an HBV Receptor: The Beginning of the End or the End of the Beginning?
See “Hepatitis B and D viruses exploit sodium taurocholate co-transporting polypeptide for species-specific entry into hepatocytes,” by Ni Y, Lempp FA, Mehrle S, et al, on page 1070.
Considering that the hepatitis B virus (HBV) genome was cloned >3 decades ago, the journey toward the identification of its receptor has been extremely slow and tortuous. HBV expresses 3 co-terminal envelope proteins termed large (L), middle (M), and small (S), which contain preS1+preS2+S, preS2+S, and S domain alone, respectively. The preS1 domain is believed to mediate virus attachment to the high-affinity receptor, although anti-preS2 and anti-S antibodies can neutralize infectivity as well. In fact, the S domain mediates initial HBV recruitment to hepatocyte surface via heparan sulfate proteoglycans.1, 2, 3 In the last 2 decades, more than a dozen host-binding proteins to the preS1, preS2, or S domain have been identified.4 For a few such proteins, cDNA transfection conferred HBV infection based on the sensitive polymerase chain reaction, but without independent confirmation. Another approach is to determine whether silencing their expression impairs HBV infection in a susceptible cell line. Unfortunately, no such cell line was available until 2002, when Gripon et al5 reported that dimethyl sulfoxide (DMSO) treatment renders a human liver progenitor cell line, HepaRG, susceptible to low-level HBV infection. In this regard, DMSO could prolong the HBV susceptibility of cultured primary human hepatocytes by maintaining cellular differentiation.6 Adding polyethylene glycol during virus incubation also enhances infectivity of cell culture-derived HBV particles, probably by concentrating virions on the cell surface. Such improvements, together with the availability of primary Tupaia hepatocytes (PTH) as a more reliable source of in vitro infection, paved the way for the identification of sodium taurocholate cotransporting polypeptide (NTCP) as the candidate HBV receptor.7 Although this provocative finding from Li’s group attracted lots of attention, the report by Ni et al8 in the current issue of Gastroenterology represents the critical, independent confirmation.
In their original eLife report,7 Yan et al employed a novel technique to establish NTCP as a binding partner for the myristoylated peptide 2-48 of the preS1 domain. Lack of NTCP expression in 2 human hepatoma cell lines, HepG2 and Huh7 cells, is consistent with their resistance to HBV infection. Moreover, in vitro culture of PTH led to a rapid loss of NTCP expression, as expected for an HBV receptor. Importantly, silencing NTCP expression by shRNA reduced HBV infectivity in PTH, primary human hepatocyte, and HepaRG cells. It also diminished infectivity of hepatitis delta virus (HDV), which employs envelope proteins of HBV for its transmission. Conversely, transfection with human NTCP (hNTCP) cDNA enabled HBV infection of HepG2 cells, as well as HDV infection of both HepG2 and Huh7 cells. A limitation of that study was that they relied heavily on hepatitis B surface antigen (HBsAg, the secreted S protein) and hepatitis B e antigen (HBeAg, a secreted version of viral core protein) as indicators of successful HBV infection. Measurement of viral mRNA and replicative DNA was achieved mostly by polymerase chain reaction, thus raising issues about the robustness of infection.
Consistent with the Yan report, Ni et al from Stephan Urban’s laboratory observed a 3.6-fold induction of hNTCP mRNA in HepaRG cells differentiated by DMSO treatment.8 One of the 2 shRNAs nearly abolished hNTCP expression when delivered by lentivirus vector. Its ability to completely block HBV and HDV infection in HepaRG cells implicates NTCP as the dominant port of HBV/HDV entry in this particular cell line. Transfection of the hNTCP cDNA into HepG2, Huh7, and HepaRG cells conferred binding of myristoylated preS1 peptide. HDV infection was reconstituted more efficiently in hNTCP-transfected Huh7 and HepaRG cells, yet only HepG2 cells showed evidence of core protein expression as well as HBsAg and HBeAg secretion following HBV infection (summarized in Table 1). Therefore, molecules other than hNTCP are required for infectivity of both viruses, and such factors are virus specific and differentially expressed in the 3 cell lines. Treatment of NTCP-reconstituted cells with 2.5% DMSO markedly increased core protein expression, DNA replication, and HBeAg secretion in both HepG2 and Huh7 cell lines. It also enabled HBsAg secretion from Huh7 cells. In the presence of DMSO, NTCP transfected HepG2 cells showed a 100-fold higher level of HBeAg secretion than HepaRG cells. Consequently HBV RNAs, replicative DNA, and even the covalently closed circular DNA, the template for RNA transcription, could be detected by conventional hybridization technique. On the other hand, for unknown reasons, HBsAg secretion remained inefficient in hNTCP-transfected HepG2 cells despite DMSO treatment.
Interplay of Viral, Host, and Differentiation Factors in Sodium Taurocholate Cotransporting Polypeptide (NTCP)-Mediated Hepatitis B Virus (HBV) and Hepatitis Delta Virus (HDV) Infection of Hepatoma Cell Lines
hNTCP hNTCP + DMSO DMSO hNTCP hNTCP + DMSO mNTCPb mNTCP
HBV HBV HBV HDV HDV HBV HDV
HepG2 (human) + ++++++a - +/- ++++++ - -c
Huh7 (human) +/- ++ - ++ ++++ - ND
HepaRG (human) +/- NDb ++ ++ ND ND ND
Hep1-6 (mouse) - ND ND + ND ND ND
Hep56D (mouse) - ND ND + ND ND ND
TC5123 (rat) - ND ND ++ ND ND ND
View Table in HTML
DMSO, dimethyl sulfoxide; hNTCP, human NTCP; mNTCP, mouse NTCP; ND, not determined
aExcept for hepatitis B surface antigen secretion, which is inefficient.
bNot deficient in the binding of myristoylated preS peptide.
cBased on Ref12
Because the normal function of NTCP is to import bile salts, Ni et al asked whether HBV exploits this property for entry into hepatocytes. Indeed, the myristoylated preS1 peptide inhibited transport of taurocholate into hNTCP transfected HepG2 cells. Conversely, several bile salts inhibited binding of the preS1 peptide and consequently HBV infection of the hNTCP transfected Huh7 cells. Very similar results were obtained by the Li group and by Watashi et al.9, 10 Furthermore, Yan et al9 found residues in the hNTCP molecule critical for bile salt transport were also essential for hNTCP to serve as the HBV receptor. Further studies are needed to determine whether HBV infection interferes with bile salt metabolism.
Identification of NTCP as an HBV receptor raises the issue of how this molecule could account for the species specificity of HBV infection, and whether it is possible to establish a small animal model of HBV infection via NTCP reconstitution or humanization. hNTCP is a glycoprotein of 349 residues. The Tupaia homologue could serve as a receptor for HBV and related woolly monkey HBV despite a 30-amino acid insertion.11 In contrast, NTCP from crab-eating monkey lacked affinity for myristoylated preS1 peptide and failed to mediate HBV infection.7 Mouse NTCP failed to confer susceptibility to HBV infection despite its affinity for myristoylated preS1 peptide,8, 12 suggesting a role of NTCP beyond high-affinity binding to HBV virions. Unfortunately, hNTCP could not confer HBV infection in the 2 mouse hepatoma cell lines and rat hepatoma cell line tested (Table 1),8, 12 thus dashing the hopes for the establishment of a small animal model of HBV infection in the near future.
Besides the paper by Ni et al in the current issue of Gastroenterology others have also confirmed the ability of hNTCP to confer susceptibility to HBV infection in HepG2 cells10 (our unpublished observation). Moreover, 2 upcoming papers, one from the Urban laboratory, demonstrated that cyclosporine inhibits HBV infection by targeting NTCP.10, 13 Thus, NTCP serves as an HBV receptor in primary human hepatocytes, PTH, HepaRG, and HepG2 cells. However, one has to bear in mind that the current infection system is based on cell culture-derived HBV particles added at extremely high multiplicity of infection in the presence of polyethylene glycol. In vivo, a few virus particles injected through the bloodstream are sufficient to infect the entire liver of a chimpanzee.14 Thus, it remains to be seen whether the NTCP system of HBV infection can be improved by the addition of other host factors and whether NTCP can mediate the infection of serum-derived HBV particles. Another question is whether NTCP constitutes the only or major HBV receptor in vivo. In this regard, Li’s group found a naturally occurring polymorphism in the NTCP gene (S267F), which abrogated its ability to support HBV infection in cell culture.9 About 0.3% of East Asians harbor this mutation in both alleles. HBV infection in any of such individuals will cast doubt on the role of NTCP as the only HBV receptor. Nevertheless, some features of NTCP are compatible with its role as the HBV receptor. Both myristoylated and cholesterol-modified preS1 peptide 2-48, but not the unmodified peptide, could interact with NTCP.9 In this regard, cholesterol is important for HBV infectivity,15 and L protein is myristoylated at position 2 by a glycine residue, which is dispensable for HBV virion formation but essential for infectivity.16
Ni et al should be applauded for improving the in vitro HBV infection system, which will prove useful for many different applications, such as the screening of more potent entry inhibitors and testing of the infectivity of viral variants. Identification of NTCP as an HBV receptor should usher in an exciting era of research. An immediate question concerns the nature of co-receptors or co-factors for productive HBV infection, such as those responsible for differential infectivity between NTCP-reconstituted HepG2 and Huh7 cells, as well as those induced by DMSO in these cell lines. As post receptor binding steps for productive HBV infection include membrane fusion (which is most likely mediated by the S domain and may occur at plasma or endosomal membrane), capsid disassembly, nuclear targeting of the relaxed circular DNA genome, and its conversion to the covalently closed circular DNA, the HepG2-enriched and DMSO-induced factors could affect any of these steps. If the example of hepatitis C virus, an RNA virus targeting the liver, is any indication,17 many more host factors for HBV infection are to be discovered. It will be interesting in this regard to reexamine the previously identified binding partners of HBV envelope proteins for their contribution to efficient HBV infection.4 Finally, one may wonder whether the NTCP homologues in woodchucks and ducks mediate infection by the related woodchuck hepatitis virus and duck HBV (DHBV). The previously identified carboxypeptidase D and glycine decarboxylase for DHBV infection18, 19, 20 are unrelated to NTCP. They seem to be necessary for DHBV infection, but failed to reconstitute DHBV susceptibility, at least in a chicken hepatoma cell line.
Shuping Tong'Correspondence information about the author Shuping TongEmail the author Shuping Tong
Liver Research Center, Rhode Island Hospital, Brown University, Providence, Rhode Island
Medical Molecular Virology Laboratory Shanghai Medical College, Fudan University, Shanghai, China
Liver Research Center, Rhode Island Hospital, Brown University, Providence, Rhode Island
Identification of NTCP as an HBV Receptor: The Beginning of the End or the End of the Beginning? - Gastroenterology https://www.gastrojournal.org/article/S0016-5085(14)00235-2/fulltext