Hepatitis C Virus Induces CD81 and Claudin-1 Endocytosis

Cell lines, antibodies, and reagents.

Huh-7.5 cells (Charles Rice, The Rockefeller University, New York, NY) (9), Huh-7 Lunet cells (Thomas Pietschmann; Twincore, Hanover, Germany) (7), and 293T cells (American Type Culture Collection) were propagated in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS)–1% nonessential amino acids (Invitrogen, California) and grown in a humidified atmosphere at 37°C in 5% CO₂. Huh-7.5 cells were transduced to express AcGFP.CD81, as previously reported (26). Huh-7 Lunet cells were transduced to express wild-type CD81 and a CD81 mutant lacking the C-terminal cytoplasmic domain (CD81_ΔC) (6) (a gift from Martin Hemler, Harvard, Cambridge, MA). Claudin-1 expression in Huh-7.5 cells was siRNA silenced (Dharmacon) using Lipofectamine RNAiMAX according to the manufacturer's instructions.

Anti-CD81 MAbs 2s66 and 2s131 were generated after immunization of mice with recombinant full-length CD81 (32). Purified mouse immunoglobulin was a gift from M. Goodall (University of Birmingham). Anti-CD81 JS81 (BD Biosciences) and 1.3.3.22 (Santa Cruz) MAbs were purchased from their respective suppliers. Anti-CD81 affinity for recombinant MBP-CD81 large extracellular loop (LEL) (19) was assessed using the Biacore 3000 system (Biacore AB). Recombinant MBP-CD81 was immobilized to a CM5 chip using a standard amine coupling method (Biacore, GE Healthcare) and increasing concentrations of anti-CD81 MAbs flowed over the chip surface (3.9 to 500 μM) at 5 μl/min at 25°C for 50 min and dissociation recorded for 60 min. To control for nonspecific protein interactions, all MAbs were flowed over an “empty” channel. The binding affinity (K_d) for each MAb was quantified by fitting a one-site binding (hyperbola) curve to the data. Anti-claudin-1 clone 421203 was purchased from R&D Systems. Anti-EEA1 clone1G11 was obtained from Abcam. Anti-tubulin clone T9026 was purchased from Sigma. Alexa-594 conjugated transferrin (Tfn-594) and secondary Alexa 488/594-labeled goat anti-mouse, goat anti-rabbit, and goat anti-rat immunoglobulins were obtained from Invitrogen. Species-specific secondary antibodies were used to detect the colocalization of anti-receptor antibodies, and their specificities were confirmed in each staining experiment.

Virus genesis, infection, and anti-CD81 neutralization.

Cell culture-derived HCV (HCVcc) J6/JFH was generated as previously described (40). Briefly, RNA was transcribed in vitro from full-length genomes (RiboMax t7 kit; Promega) and electroporated into Huh-7.5 cells. Viral stocks were generated by two sequential passages through Huh-7.5 cells with a specific infectivity in the range of 1/500 to 1/780 (21). Supernatants were collected, pooled, and stored at −80°C. Infected cells were detected by methanol fixation and staining for HCV-encoded NS5A with MAb 9E10 (a gift from Charles Rice and Tim Tellinghuisen Rockefeller University); bound antibody was detected with anti-mouse IgG-Alexa 488 and quantified by enumerating NS5A-expressing cells.

Huh-7.5 cells were seeded at 1.5 × 10⁴ cells/cm², and the following day the medium was replaced with ice-cold 3% FBS-DMEM containing J6/JFH, followed by incubation on ice for 30 min. The cells were washed, the medium was replaced with prewarmed 3% FBS-DMEM and, at 0, 20, 60, and 120 min, anti-CD81 MAbs, at a concentration known to neutralize >95% virus infectivity (5 μg/ml) or proteinase K (PK; 50 μg/ml) were added in serum-free DMEM. PK-treated cells were subsequently incubated with 3% FBS-DMEM to inactivate its activity, as previously reported (59). At 48 h postinfection, the cells were fixed, and the infectivity was quantified (59). Virus escape from anti-CD81 or PK was calculated relative to the maximum neutralization at 120 min, and the time taken for 50% of the virus (t_1/2) to escape treatments was determined.

Pseudoviruses expressing a luciferase reporter were generated as previously described (30). Briefly, 293T cells were transfected with a 1:1 ratio of plasmids encoding HIV provirus expressing luciferase and HCV strain H77 E1E2 envelope glycoproteins (HCVpp), murine leukemia virus envelope glycoprotein (MLVpp), vesicular stomatitis virus G glycoprotein (VSV-Gpp), or empty vector (Env⁻pp). Supernatants were harvested 48 h posttransfection and filtered, and virion-associated p24 was determined using a commercial assay (Aalto Bioreagents). Infection was quantified by measuring the luciferase activity (relative lights units [RLU]), and the specific infectivity was determined by subtracting the mean Env⁻pp signal from the HCVpp, MLVpp, or VSV-Gpp values.

Cell surface biotinylation endocytosis assay.

Huh-7.5 cells were seeded at 1.5 × 10⁴ cells/cm² and the following day serum starved for 60 min and, where indicated, incubated with Dynasore (80 μM) for 30 min, prior to adding Sulfo-NHS-SS-biotin (Pierce) at 0.5 mg/ml for 30 min at 4°C. The cells were washed and maintained at 4°C or transferred to 37°C for 60 min in the presence or absence of antibodies or viral pseudoparticles, as indicated. After the cells were returned to 4°C, they were washed with ice-cold buffer (50 mM Tris [pH 7.5], 100 mM NaCl [pH 8.6]) and biotin cleaved from surface-labeled proteins by incubation with the cell-impermeable reducing agent 2-mercaptoethanesulfonic acid at 15 mg/ml (MesNA; Sigma) for 30 min or incubation with ice-cold buffer in the absence of MesNA to determine the total protein expressed at the cell surface. The cells were lysed in lysis buffer (1% Brij97, 10 mM Tris [pH 7.5], 150 mM NaCl, 1 mM CaCl₂, 1 mM MgCl₂, 0.01% sodium azide). Biotinylated proteins were immunoprecipitated from defined quantities of cell lysate with streptavidin-coupled Dynabeads (Invitrogen). Precipitates and whole-cell lysates were separated by SDS-PAGE, transferred onto polyvinylidene difluoride membrane, and probed for CD81, claudin-1, or tubulin expression by enhanced chemiluminescence (GeneFlow). Densitometric analysis was performed using ImageJ software and the percentage of internalized CD81 or claudin-1 presented relative to the total CD81 or claudin-1 expressed at the cell surface.

Quantitative imaging of CD81 and claudin-1 endocytosis.

Parental Huh-7.5 cells or those transduced to express AcGFP.CD81 were seeded on glass coverslips at 1.5 × 10⁴ cells/cm² and the following day incubated with anti-CD81 or anti-claudin-1 MAbs in 3% FBS-DMEM–0.5% bovine serum albumin (BSA) at 37°C. The cells were fixed in 3.6% electron microscope-grade formaldehyde (EM-F; TAAB, United Kingdom), and protein-antibody complexes were detected using Alexa Fluor-conjugated antibody diluted in phosphate-buffered saline–0.05% saponin–0.5% BSA for 60 min at room temperature. The cells were counterstained with DAPI (4′,6′-diamidino-2-phenylindole; Invitrogen) for 5 min, and coverslips were mounted onto glass slides (ProLong Gold Antifade; Invitrogen). The cells were imaged by laser scanning confocal microscopy (LSCM; Zeiss LSM510) using a ×100 oil immersion objective lens. To avoid any bias in the selection of cells for quantitation, the cells were initially identified by their DAPI nuclear stain and imaged for CD81 or claudin-1 expression. Microscope settings were optimized for each fluorescent protein to obtain the highest signal to noise ratio while controlling for cross talk. z-sections were collected, and the central section found to give optimal resolution. Consequently, a single 1-μm z-section across the cell center were selected for quantitation. CD81 and claudin-1 localization at the plasma membrane and intracellular locations were quantified by using the overlay and intensity frequency tools included with the Zeiss LSCM software. Cellular protein expression (arbitrary fluorescence units [AFU]) was calculated by combining the number of fluorescent pixels and multiplying this value by the average intensity. The fluorescence intensity at the plasma membrane and the intracellular locations were determined by extracting the regions of interest from the whole-cell image. The results are expressed as a percentage of the total fluorescence and represent the mean intracellular fluorescence intensity ± the standard error of the mean (SEM) of between 10 and 20 cells. Statistical analyses were performed using nonparametric one-way analysis of variance (ANOVA; Kruskal-Wallis test) or the Student t test in Prism 4.0, where necessary corrections for multiple comparisons were made. Intracellular colocalization was determined using ImageJ colocalization finder plug-in software, and the overlap coefficient is presented as a percentage of the colocalized pixels averaged from analyzing 10 cells.

Mechanism of CD81 endocytosis.

Huh-7.5 cells were transfected with green fluorescent protein (GFP) control plasmid or GFP-tagged transdominant mutants of Eps15 (GFP-EH29; Alexandre Benmerah, Institute Cochin, Paris, France) and Dynamin2 (K44A-GFP; Mark Niven, The Mayo Clinic, Rochester, MN) using Lipofectamine 2000 according to the manufacturer's instructions (Invitrogen). GFP-, GFP-EH29-, and K44A-GFP-expressing cells were seeded onto glass coverslips at 1.5 × 10⁴ cells/cm² and incubated in serum-free medium for 60 min at 37°C; where indicated, naive Huh-7.5 cells were preincubated with Dynasore (80 μM) or dimethyl sulfoxide (DMSO) control for 60 min at 37°C, followed by anti-CD81 or Tfn-594 in the continued presence or absence of Dynasore for the indicated time intervals. To assess the role of Rho GTPases and epidermal growth factor (EGF) in CD81 endocytosis, Huh-7.5 cells were serum starved overnight and then incubated with C3 transferase (Cytoskeleton) at 5 μg/ml for 16 h, Rac-1 inhibitor (Calbiochem) at 100 μM for 60 min, EGF at 1 μg/ml, or Erlotinib at 10 μM in serum-free medium for 60 min, respectively, prior to adding anti-CD81 2s66 at 2 μg/ml for 60 min at 37°C. The cells were immediately fixed with EM-F and CD81, visualized, and quantified as detailed above.

Live cell imaging of CD81 and claudin-1 trafficking.

Parental Huh-7.5 and those transfected to express GFP-Rab5 (Alexandre Benmerah) were seeded in glass-bottom culture dishes at 1.5 × 10⁴ cells/cm². The following day, the cells were incubated for 30 min at 37°C with fluorescence-labeled anti-CD81 2s66 and anti-claudin-1 labeled with the ATTO fluorophores 647N and 565, respectively (ATTO-TEC, Seigen, Germany). The cells were washed and placed in phenol red-free 3% FBS-DMEM supplemented with 20 mM HEPES. Real-time LSCM was performed (inverted Zeiss 780) of a 1-μm z-section through the center of the cell using a ×100 oil immersion objective lens at 37°C over 5 min, and the images were processed using Zeiss software.

HCV-induced receptor endocytosis.

Huh-7.5 cells expressing AcGFP.CD81 were seeded at 1.5 × 10⁴/cm² on glass coverslips and incubated the following day with lentiviral pseudoparticles or HCVcc J6/JFH for 1 h at 37°C at an approximate multiplicity of infection of 0.3. J6/JFH was incubated with 100 μg of anti-HCV IgG or control IgG/ml for 1 h at 37°C or at 65°C for 1 h, prior to inoculating the cells. The cells were fixed in EM-F using MAbs, intracellular CD81/claudin-1 imaged, and quantified as described above. In parallel plates, unbound J6/JFH was removed by washing, and the medium was replenished with 3% FBS-DMEM for 48 h, after which the virus infection was quantified by enumerating NS5A-expressing cells.

Anti-CD81 MAbs neutralize HCV infection after virus internalization.

Several reports have studied the time for HCV to escape the neutralizing activity of anti-receptor antibodies to investigate the temporal involvement of cellular molecules in the virus entry process (13, 21, 23, 37, 75). The majority of these reports have used the commercially available anti-CD81 MAbs JS81 and 1.3.3.22 that bind conformation-dependent epitopes in the large extracellular loop (LEL) (29). We recently generated two MAbs 2s66 and 2s131 that recognize a linear epitope in CD81 (amino acids 173 to 192) and demonstrate a significantly higher affinity for CD81 LEL than JS81 or 1.3.3.22 (2s66, K_d = 3.33 μM; 2s131, K_d = 2.09 μM; JS81, K_d = 31.69 μM; and 1.3.3.22, K_d = 76.2 μM) (Fig. 1A). To ascertain whether the affinity or epitope specificity of CD81 MAbs plays a role in their mode of neutralization, we compared the time for HCV to escape the neutralizing activity of these MAbs. HCVcc J6/JFH was bound to Huh-7.5 cells on ice for 30 min, unbound virus removed by washing and entry initiated by transferring the cells to 37°C. Anti-CD81 MAbs were added at a concentration that inhibited >95% of virus infectivity at the indicated times, and the half-time (t_1/2) for HCV to escape their neutralizing activity was assessed (Fig. 1B). HCV J6/JFH escaped the neutralizing activity of 2s66 and 2s131 at significantly later times (52 and 60 min, respectively) than MAbs JS81 and 1.3.3.22 at 36 and 37 min, respectively (Fig. 1C). To independently assess HCV internalization, we measured the time for HCV to escape the proteolytic activity of the serine protease, PK. J6/JFH acquired resistance to PK after only 18 min (t_1/2) (Fig. 1C), a finding consistent with our earlier report (59). The observation that anti-CD81 MAbs could neutralize virus infection up to 60 min after internalization is initiated and long after PK has ceased to have an effect suggests intracellular sites for anti-CD81 MAb neutralization and a role for endosomal CD81 in virus infection.

CD81 endocytosis.

The majority of CD81 in Huh-7.5 hepatoma cells localizes to the plasma membrane with a limited pool of intracellular protein detected (Fig. 2A, upper panel). To study CD81 endocytosis, Huh-7.5 cells were incubated with a saturating concentration of anti-CD81 MAbs JS81, 1.3.3.22, 2s66, or 2s131 at 37°C for 60 min. Cell surface-bound MAb was removed by a low pH treatment, the cells were fixed and permeabilized, and intracellular CD81-MAb complexes were visualized by confocal microscopy. Significantly greater amounts of intracellular 2s66 and 2s131 were detected compared to JS81 and 1.3.3.22 (Fig. 2A, lower panel), which is consistent with our earlier neutralization data showing later escape times for MAbs 2s66 and 2s131 (Fig. 1C). To confirm we were studying intracellular CD81, we demonstrated that saponin permeabilization was required to visualize CD81 puncta and low pH treatment designed to remove cell surface-bound MAb had minimal effect on secondary antibody detection of intracellular CD81 (Fig. 2B).

To confirm that antibody ligation promoted CD81 endocytosis, we surface labeled Huh-7.5 cells with Sulfo-NHS-SS-biotin on ice according to published protocols and maintained the cells on ice to prevent endocytosis or incubated at 37°C for 1 h in the presence of irrelevant mouse immunoglobulin or anti-CD81 2s66. Subsequent incubation of the cells with MesNA cleaves biotin from surface-labeled proteins, while internalized proteins are protected. Total cell surface-labeled proteins or biotinylated proteins that have entered the cell are detected by streptavidin capture, and CD81 was visualized by Western blotting. Incubation with anti-CD81 2s66 significantly stimulated CD81 endocytosis (19.5%), whereas control IgG treated cells showed a low level of constitutive endocytosis (6%) (Fig. 3A). Under identical conditions we failed to precipitate the intracellular protein tubulin, confirming specific labeling of cell surface proteins (Fig. 3B).

We noted that antibody/biotin removal from the cell surface was on occasion incomplete. Therefore, to further study the mechanism of CD81 internalization, we developed an imaging-based approach to quantify plasma membrane and intracellular CD81, allowing the use of fluorescently tagged transdominant-negative proteins/antibodies. Initial experiments demonstrated that a 1-h treatment with MAb 2s66 increased intracellular localization of AcGFP.CD81 by 16.8% (Fig. 3C). Comparable results were obtained from analyzing 10 or 20 cells, and the data showed a Gaussian distribution, demonstrating an adequate sampling of the cell population (Fig. 3C).

The increased detection of intracellular CD81 observed with MAbs 2s66 and 2s131 compared to JS81 and 1.3.3.22 (Fig. 2A) may reflect differences in binding affinity or epitope-specific effects of MAb ligation on CD81 trafficking. The observation that 2s66 FAb bound CD81 LEL with a reduced K_d of 44 μM compared to 3.33 μM for the complete IgG molecule (Fig. 1A) allowed us to study the role of ligand affinity on CD81 endocytosis. Both IgG and FAb antibody preparations at equimolar concentrations induced comparable levels of AcGFP.CD81 internalization (16.8 and 15.8%, respectively) (Fig. 3C), demonstrating that bivalent antibody induced cross-linking is not required to promote CD81 internalization, and ligand-induced CD81 internalization may not solely be driven MAb binding affinity. Brazzoli et al. reported that antibody ligation of CD81 promoted protein movement to cell-cell contacts in Huh-7 cells, suggesting a mechanism for CD81 tethered virus to interact with tight-junction-associated claudin-1 and occludin proteins (11). In contrast, we found CD81 to preferentially localize at cell-cell contacts in naive untreated cells (Fig. 2A) and failed to observe any significant effect of MAb engagement on CD81 localization at cell-cell contacts (Fig. 3D).

Mechanism of CD81 internalization.

HCV has been reported to enter cells in a clathrin-dependent manner (8, 48). To assess the clathrin dependency of CD81 endocytosis, we transfected Huh-7.5 cells to express GFP or GFP-tagged transdominant mutants of Eps15 (EH29), a regulatory protein essential for the formation of clathrin-coated pits (4), or Dynamin2 (K44A), a protein necessary for the scission of vesicles from the plasma membrane (46) (reviewed in reference 57). Both mutants reduced Alexa 594-labeled transferrin uptake (Fig. 4A), confirming clathrin-dependent endocytosis. CD81 internalization was reduced by 51.5 and 59.5% in GFP-EH29- and K44A-GFP-expressing cells, respectively, compared to control GFP-transfected cells (Fig. 4A), demonstrating a clathrin- and dynamin-dependent internalization process. In agreement with these data, treating Huh-7.5 cells with the dynamin inhibitor Dynasore (34, 44, 50, 65) reduced both transferrin and CD81 uptake (Fig. 4B). Decreased CD81 endocytosis in the presence of Dynasore was confirmed by surface biotinylation (Fig. 4C). Furthermore, Dynasore reduced HCVcc, HCVpp, and VSV-Gpp infection (33) but had minimal effect(s) on MLVpp infection (38) (Fig. 4D). We conclude that CD81 internalizes via a clathrin- and dynamin-dependent process.

CD81 endocytosis is independent of the intracellular C-terminal domain.

Cell surface receptor endocytosis is often regulated by motifs within the cytoplasmic tail that bind adaptor proteins involved in the formation of clathrin coated pits or caveolae (39). The tetraspanins CD63, CD151, and CD82 have been reported to internalize by diverse cellular routes (41, 56, 71) via an endocytic motif YXXØ in their C-terminal domains (10, 27). However, CD81 does not contain this endocytic motif. To determine whether the C-terminal cytoplasmic domain is required for CD81 endocytosis, a mutant protein lacking this region (CD81_ΔC) was expressed in Huh-7 Lunet cells that express low levels of endogenous CD81 detectable by flow cytometry (Fig. 5A). We confirmed cell surface expression of exogenous wild-type and CD81_ΔC proteins and observed comparable levels of internalized CD81 after 2s66 MAb treatment (Fig. 5A and B), suggesting a role for partner proteins to regulate CD81 endocytosis. The GTPases Rho and Rac-1 were reported to play a role in HCV entry (11); however, their mode of action was not defined. To study their involvement in CD81 endocytosis, Huh-7.5 cells were incubated with inhibitors of Rho (C3 transferase) or Rac-1. Inhibition of Rho significantly reduced antibody stimulated CD81 endocytosis compared to control cells (Fig. 5C), whereas Rac1 inhibition had no observable effect, supporting a role for Rho GTPase in CD81 endocytosis. Both inhibitors reduced HCVpp infection as previously published (11).

EGF receptor (EGF-R) was recently reported to play a role in HCV entry by promoting particle internalization (43). Reports that EGF-Rs associate with the tetraspanins CD9, CD81, and CD82 (51) and that EGF promotes CD82 endocytosis (71) prompted us to investigate the effects of EGF-R activation on CD81 endocytosis. We confirmed the stimulatory and inhibitory effect of EGF and Erlotinib, respectively, on HCVpp entry in Huh-7.5 cells (data not shown). However, treating cells with EGF or receptor kinase inhibitor Erlotinib had no effect on CD81 endocytosis (Fig. 5D), demonstrating that EGF-R does not promote HCV entry by regulating CD81 endocytosis.

CD81 and claudin-1 coendocytosis.

We (26) and others (73) previously reported that CD81 associates with claudin-1 and the receptor complex plays a role in HCV entry (25). Tight-junction proteins such as claudin-1 are known to readily endocytose via stimulus-specific endocytic pathways (67, 74). To ascertain whether CD81 and claudin-1 coendocytose, Huh-7.5 cells were incubated with anti-CD81 and anti-claudin-1 MAbs, surface-bound antibody was removed by a low pH treatment, and the cells were fixed and permeabilized prior to the addition of a secondary antibody. We observed that 46% of the intracellular CD81 colocalized with claudin-1 internalized from the plasma membrane (Fig. 6A). Similarly, incubating Huh-7.5 cells with anti-CD81 MAb alone, followed by low pH and permeabilization steps and anti-claudin-1 staining, demonstrated intracellular vesicular costaining for CD81 and claudin-1 (Fig. 6A), suggesting the coendocytosis of both receptor molecules. To confirm CD81/claudin-1 coendocytosis, Huh-7.5 cells were incubated with directly labeled fluorescent anti-CD81 2s66*647N and anti-claudin-1*565 MAbs at 37°C for 30 min, unbound MAbs were removed by washing, and the cells were imaged over a 5-min time period. We observed CD81 and claudin-1 trafficking at the plasma membrane prior to their coendocytosis (Fig. 6B) (see Movie S1 in the supplemental material). HCV-cell fusion is reported to occur within early endosomes (48, 53). To study the localization of internalized CD81 and claudin-1, Huh-7.5 cells were transfected with a GFP-tagged early endosomal marker Rab5 (GFP-Rab5), and recently internalized CD81 and claudin-1 proteins were visualized with fluorescence-labeled antibodies. Real-time imaging showed CD81- and claudin-1-containing vesicles trafficking to and colocalizing with Rab5 expressing early endosomes (Fig. 6C) (see Movie S2 in the supplemental material), demonstrating the recruitment of both receptors to early endosomes. Costaining GFP-Rab5 expressing vesicles with anti-EEA1 confirmed correct targeting of GFP-Rab5 to early endosomes (Fig. 6D). To ascertain whether claudin-1 expression is essential for CD81 endocytosis, we siRNA silenced claudin-1 expression and monitored the effects of MAb 2s66 on CD81 endocytosis. There was no difference in anti-CD81 stimulated endocytosis in claudin-1 silenced cells (Fig. 6E), suggesting CD81 endocytosis is not regulated by claudin-1 expression per se.

HCV promotes CD81 and claudin-1 endocytosis.

To investigate whether HCV can promote CD81 and claudin-1 endocytosis, Huh-7.5 cells were incubated with cell culture-grown HCV particles (HCVcc), and AcGFP.CD81 and claudin-1 internalization was compared to that in mock-infected (control) cells (Fig. 7A). HCVcc promoted a significant increase in intracellular CD81 (21.4%) and claudin-1 (15.2%), resulting in pixel saturation that may underestimate CD81 endocytosis. Heat inactivation or anti-HCV immunoglobulin inhibited HCVcc infectivity (Fig. 7B) and ablated the effect(s) of virus on CD81 and claudin-1 internalization, suggesting that infectious virus is required to initiate receptor endocytosis. We confirmed HCVcc stimulation of endogenous CD81 endocytosis by antibody detection (Fig. 7C). After HCVcc treatment we observed 89% of intracellular CD81 colocalizing with claudin-1 internalized from the plasma membrane (Fig. 7D).

To confirm that this effect was mediated via the virus-encoded glycoproteins, we studied the effect of recombinant soluble HCV E2 glycoprotein (sE2) or lentiviral pseudoparticles expressing HCV E1E2 glycoproteins (HCVpp) or envelope-deficient particles (Env⁻pp) on AcGFP.CD81 internalization in Huh-7.5 cells. Both sE2 and HCVpp induced a significant increase, 15.6 and 27.8%, respectively, in intracellular AcGFP.CD81 compared to Env⁻pp-treated cells (Fig. 8A and B). Furthermore, we confirmed HCVpp stimulation of CD81 (30.9%) and claudin-1 (27.4%) endocytosis by surface biotinylation (Fig. 8C). In conclusion, these data illustrate HCV envelope glycoprotein stimulated CD81 and claudin-1 endocytosis.