NTCP Oligomerization Occurs Downstream of the NTCP-EGFR Interaction during Hepatitis B Virus Internalization

Abstract

Sodium taurocholate cotransporting polypeptide (NTCP) is a receptor that is essential for hepatitis B virus (HBV) entry into the host cell. A number of HBV entry inhibitors targeting NTCP have been reported to date; these inhibitors have facilitated a mechanistic analysis of the viral entry process. However, the mechanism of HBV internalization into host cells after interaction of virus with NTCP remains largely unknown. Recently, we reported that troglitazone, a thiazolidinedione derivative, specifically inhibits both HBV internalization and NTCP oligomerization, resulting in inhibition of HBV infection. Here, using troglitazone as a chemical probe to investigate entry process, the contribution of NTCP oligomerization to HBV internalization was evaluated. Using surface plasmon resonance and transporter kinetics, we found that troglitazone directly interacts with NTCP and noncompetitively interferes with NTCP-mediated bile acid uptake, suggesting that troglitazone allosterically binds to NTCP, rather than to the bile acid-binding pocket. Additionally, alanine scanning mutagenesis showed that a mutation at phenylalanine 274 of NTCP (F274A) caused a loss of HBV susceptibility and disrupted both the oligomerization of NTCP and HBV internalization without affecting viral attachment to the cell surface. An inhibitor of the interaction between NTCP and epidermal growth factor receptor (EGFR), another host cofactor essential for HBV internalization, impeded NTCP oligomerization. Meanwhile, coimmunoprecipitation analysis revealed that neither troglitazone nor the F274A mutation in NTCP affects the NTCP-EGFR interaction. These findings suggest that NTCP oligomerization is initiated downstream of the NTCP-EGFR interaction and then triggers HBV internalization. This study provides significant insight into the HBV entry mechanisms. IMPORTANCE Hepatitis B virus (HBV) infection is mediated by a specific interaction with sodium taurocholate cotransporting polypeptide (NTCP), a viral entry receptor. Although the virus-receptor interactions are believed to trigger viral internalization into host cells, the exact molecular mechanisms of HBV internalization are not understood. In this study, we revealed the mode of action whereby troglitazone, a specific inhibitor of HBV internalization, impedes NTCP oligomerization and identified NTCP phenylalanine 274 as a residue essential for this oligomerization. We further analyzed the association between NTCP oligomerization and HBV internalization, a process that is mediated by epidermal growth factor receptor (EGFR), another essential host cofactor for HBV internalization. Our study provides critical information on the mechanism of HBV entry and suggests that oligomerization of the viral receptor serves as an attractive target for drug discovery.

Keywords: EGFR; HBV; NTCP; entry; internalization; multimerization; oligomerization; preS1; receptor; troglitazone.

FIG 1

Troglitazone directly interacts with NTCP to inhibit NTCP oligomerization. (A) Surface plasmon resonance (SPR) analysis of the interaction between compounds (troglitazone or pioglitazone) and recombinant His-tagged NTCP (His-NTCP) or BSA. Various concentrations (50, 25, 12.5, 6.25, and 3.125 μM) of compounds were injected from 0 to 120 s onto a sensor chip on which the indicated protein had been immobilized, followed by application of compound-free buffer. The SPR responses are indicated in resonance units (RU). (B) NTCP-dependent bile acid uptake was measured by quantifying [³H]taurocholic acid (TCA) as a substrate in HepG2-hNTCP-C4 cells in the presence or absence of preS1 peptide (100 nM), troglitazone (25 μM), or pioglitazone (25 μM) in either sodium-free (gray) or sodium-containing (black) buffer. (C) (a and b) Michaelis-Menten plot in which reduced maximum activity indicates a noncompetitive mode of inhibition (green), while little or no reduction in the maximum activity indicates competitive inhibition of the target protein (red). Kinetics of NTCP transporter activity were analyzed following treatment with or without cyclosporin A (c) or troglitazone (d) at either 0 (black), 5 (light gray), or 10 (dark gray) μM and various concentrations of TCA. (D) Interactions between recombinant full-length His-NTCP and the fragment peptides derived from NTCP were examined by pulldown assay. Biotinylated (bio) peptides of preS1 (aa 2 to 48) or NTCP fragments (20-aa lengths corresponding to the indicated regions) immobilized on streptavidin (SA)-agarose beads were incubated with His-NTCP in the presence or absence of troglitazone. Coimmunoprecipitated His-NTCP was detected with anti-His antibody by immunoblotting. Quantitative densitometry is shown in the upper graph. Data are means and standard deviations (SD). Statistical significance was determined using a two-tailed nonpaired Student's t test (*, P < 0.05; **, P < 0.01).

FIG 2

Identification of regions responsible for NTCP oligomerization. (A) HBV inhibition activity of alanine-variant NTCP peptides (20-aa lengths, with the mutated regions indicated in red) was evaluated by HBV infection assay. HepG2-hNTCP-C4 cells treated with or without NTCP peptides were inoculated with HBV for 16 h. HBV infection was determined by HBs antigen level in the culture supernatant at 12 days postinoculation. (a) Inhibition activity of NTCP peptide (aa 221 to 240) (red dotted line) was compared with those of seven mutant peptides (221–223A, 224–226A, 227–229A, 230–232A, 233–235A, 236–238A, and 239–240A). (b) Inhibition activity of NTCP peptide (aa 271 to 290) (red dotted line) was compared with those of seven mutant peptides (271–273A, 274–276A, 277–279A, 280–282A, 283–285A, 286–288A, and 289–290A). (B) Interactions between recombinant full-length His-NTCP and the NTCP peptides were examined by pulldown assay, as described in the Fig. 1D legend. Quantitative densitometry is shown in the upper graph. Data are means and SD. Statistical significance was determined using a two-tailed non-paired Student’s t-test.

FIG 3

NTCP residue F274 is essential for both NTCP oligomerization and HBV infection. (A) Schematic representation of the amino acid sequences of NTCPs possessing single mutations (in red) within each element (aa 274 to 276, aa 283 to 285, and aa 289 to 290) identified as regions important for NTCP oligomerization in Fig. 2. (B) HBV preS1 attachment to host cells was examined in HepG2 cells transfected with expression plasmids encoding the wild-type or mutant NTCP. The cells were exposed for 30 min at 37°C to TAMRA-labeled preS1 peptide (preS1 probe) (red) and then were stained for the nucleus (blue) by DAPI (a). PreS1 fluorescence intensities were quantified using Dynamic Cell Count (Keyence) (b). (C) HepG2 cells overexpressing the wild-type or mutant NTCP were infected with HBV, and the supernatant HBs antigen was quantified at 12 days postinoculation. (D) HepG2 cells overexpressing the wild-type or mutant NTCP (F274A) were exposed for 8 h at 37°C to preS1 probe (red) and then were stained for the nucleus (blue) by DAPI. Merged patterns are shown in lower panels. (E) Proximity ligation assay (PLA) to evaluate NTCP oligomerization. HepG2 cells overexpressing both myc-NTCP (WT or F274A) and HA-NTCP (WT or F274A) were treated for 30 min at 4°C with or without preS1 probe (red), and then the PLA signal (green) produced by the proximity of anti-myc and anti-HA antibodies was detected with Duolink PLA. The nucleus was also stained with DAPI (blue). (a) The lower panels show the PLA signal (green) only, and the upper panels show the merged images of green (PLA signal), red (preS1 probe), and blue (nucleus) signals. (b) PLA signals in the experiment reported in panel a were quantified using Dynamic Cell Count. (F) NTCP transporter activity was measured by using [³H]taurocholic acid as a substrate in HepG2 cells overexpressing the wild-type or mutant NTCP (F274A) in either sodium-free or sodium-containing buffer, as described in Materials and Methods. Intracellular ³H counts taken up into the cells are shown. Data are means and SD. Statistical significance was determined using a two-tailed nonpaired Student's t test (**, P < 0.01).

FIG 4

Functional relevance of the association between NTCP oligomerization and EGFR-mediated HBV internalization. (A) Effect of NTCP peptide (aa 131 to 150) and gefitinib on NTCP oligomerization was evaluated by PLA, as described in the Fig. 3E legend. The PLA signal was detected in HepG2 cells overexpressing both myc-NTCP and HA-NTCP following treatment with or without the indicated compounds for 30 min at 4°C in the presence of preS1 probe. (a) The lower panels show the PLA signal (green) only, and the upper panels are the merged images of green (PLA signal), red (preS1 probe), and blue (nucleus). (b) PLA signals in the experiment reported in panel a were quantified using Dynamic Cell Count (Keyence). (B) Effect of mutations in NTCP (G144A/G148A) on NTCP oligomerization was evaluated by PLA, as described in the Fig. 3E legend. The PLA signal was detected in HepG2 cells overexpressing both myc-NTCP (WT or G144A/G148A) and HA-NTCP (WT) following treatment with preS1 probe. (a) The lower panels show the PLA signal (green) only, and the upper panels are the merged images of green (PLA signal), red (preS1 probe), and blue (nucleus). (b) PLA signals in the experiment reported in panel a were quantified using Dynamic Cell Count. (C) 293T cells overexpressing HA-EGFR and myc-NTCP (WT [a and b] or F274A [b]) were treated with or without NTCP peptide (aa 131 to 150) or troglitazone and then harvested for immunoprecipitation with anti-myc antibody (or normal mouse IgG as a negative control). The precipitates or total cell lysates were assessed for NTCP, EGFR, and actin by immunoblotting. Data are means and SD. Statistical significance was determined using a two-tailed nonpaired Student's t test (*, P < 0.05; **, P < 0.01).

FIG 5

Proposed model of HBV internalization involving NTCP oligomerization and EGFR. Upon HBV attachment, NTCP first interacts with EGFR, followed by NTCP oligomerization. Troglitazone impedes the NTCP oligomerization, and the phenylalanine at position 274 of NTCP is essential for the oligomerization. Formation of the complex consisting of HBV virion, EGFR, and NTCP oligomer, followed by EGFR autophosphorylation, allows its internalization via endocytosis. Results from Fig. 1 to 4 are summarized at the bottom.