Hepatitis C virus targets DC-SIGN and L-SIGN to escape lysosomal degradation

Abstract

Hepatitis C virus (HCV) is a major health problem. However, the mechanism of hepatocyte infection is largely unknown. We demonstrate that the dendritic cell (DC)-specific C-type lectin DC-SIGN and its liver-expressed homologue L-SIGN/DC-SIGNR are important receptors for HCV envelope glycoproteins E1 and E2. Mutagenesis analyses demonstrated that both HCV E1 and E2 bind the same binding site on DC-SIGN as the pathogens human immunodeficiency virus type 1 (HIV-1) and mycobacteria, which is distinct from the cellular ligand ICAM-3. HCV virus-like particles are efficiently captured and internalized by DCs through binding of DC-SIGN. Antibodies against DC-SIGN specifically block HCV capture by both immature and mature DCs, demonstrating that DC-SIGN is the major receptor on DCs. Interestingly, internalized HCV virus-like particles were targeted to nonlysosomal compartments within immature DCs, where they are protected from lysosomal degradation in a manner similar to that demonstrated for HIV-1. Lewis X antigen, another ligand of DC-SIGN, was internalized to lysosomes, demonstrating that the internalization pathway of DC-SIGN-captured ligands may depend on the structure of the ligand. Our results suggest that HCV may target DC-SIGN to "hide" within DCs and facilitate viral dissemination. L-SIGN, expressed by THP-1 cells, internalized HCV particles into similar nonlysosomal compartments, suggesting that L-SIGN on liver sinusoidal endothelial cells may capture HCV from blood and transmit it to hepatocytes, the primary target for HCV. We therefore conclude that both DCs and liver sinusoidal endothelial cells may act as reservoirs for HCV and that the C-type lectins DC-SIGN and L-SIGN, as important HCV receptors, may represent a molecular target for clinical intervention in HCV infection.

FIG. 1.

DC-SIGN specifically binds HCV envelope glycoproteins E1 and E2. (A) DC-SIGN binds both H. polymorpha and mammalian cell-produced HCV envelope glycoproteins. DC-SIGN interaction with HCV envelope glycoproteins E1 and E2 (0.25 μg/well) produced by either the yeast H. polymorpha or mammalian cells with a recombinant vaccinia virus, was determined in an Fc-based ELISA. The specificity of the binding was confirmed with the DC-SIGN-specific blocking antibody AZN-D1 (50 μg/ml), mannan (100 μg/ml), and the calcium chelator EGTA (5 mM). (B) DC-SIGN binds more strongly to HCV envelope glycoprotein E2 than to E1. HCV glycoproteins E1 and E2 were titrated (0 to 16 and 0 to 14 μM, respectively), and DC-SIGN-Fc binding was determined as described above. DC-SIGN-specific antibody AZN-D1 (50 μg/ml) was used to specifically block the interaction. (C) DC-SIGN has a higher affinity for envelope glycoprotein E2 than for envelope glycoproteins E1 and gp120. Mammalian cell-produced glycoproteins E1 and E2 (10 nM) and gp120 (2 nM) were coated, and DC-SIGN-Fc binding was determined as described above. Mannan was titrated (0 to 1,000 μg/ml) to block the interaction. Binding is represented as a percentage of maximal binding. (D) DC-SIGN binding to HCV envelope glycoproteins E1 and E2 and HIV-1 envelope gp120 is equally dependent on calcium. Viral envelope glycoproteins E1 and E2 (10 nM) and gp120 (2 nM) were coated, and DC-SIGN-Fc binding was measured as described above in the presence of calcium (0 to 5 mM). Binding is represented as a percentage of maximal binding at 5 mM calcium. Standard deviations were <2%.

FIG. 2.

HCV envelope glycoproteins are bound by cellular DC-SIGN and L-SIGN. (A) K562 transfectants express similar levels of DC-SIGN and L-SIGN. K562 cells were transfected with DC-SIGN or L-SIGN as described in Materials and Methods. Expression was measured by fluorescence-activated cell sorting staining with the DC-SIGN- and L-SIGN-specific antibody AZN-D2. Dotted lines represent the isotype control, and black lines represent AZN-D2 staining. Mean fluorescence indices were >700. (B) Cellular DC-SIGN and L-SIGN bind to both envelope glycoproteins E1 and E2. HCV envelope glycoprotein binding by K562 cells expressing DC-SIGN or L-SIGN was measured with a fluorescently coated bead adhesion assay. The L-SIGN- and DC-SIGN-specific blocking antibody AZN-D2 (50 μg/ml) was used to determine the specificity of the interaction. (C) The DC-SIGN Val³⁵¹ mutant binds both HCV envelope glycoproteins E1 and E2. Binding of K562 cells transfected with DC-SIGN V351G to glycoprotein E1- and E2-coated beads was investigated in the presence and absence of the DC-SIGN-specific antibody AZN-D2 (50 μg/ml) and EGTA (10 μM). Standard deviations were <5%.

FIG. 3.

Internalization pathway of DC-SIGN and L-SIGN is dependent on cellular background. (A) DC-SIGN and L-SIGN expressed by K562 and THP-1 transfectants bind HCV VLPs similarly. Binding of HCV E1/E2 VLP to K562 and THP-1 transfectants with both DC-SIGN and L-SIGN was measured. Interaction was blocked by the DC-SIGN- and L-SIGN-specific antibody AZN-D2 (50 μg/ml). Standard deviations were <5%. (B to E) DC-SIGN- or L-SIGN-bound HCV is targeted to the early endosomes (transferrin positive) in THP-1 transfectants, in contrast to the lysosomal (LAMP-1 positive) targeting in K562 transfectants. K562 (B and C) and THP-1 (D and E) cells expressing DC-SIGN or L-SIGN were incubated overnight with HCV VLPs. HCV was detected with a human anti-HCV antibody and a FITC-labeled secondary antibody. Intracellular targeting was determined by staining the endosomal compartments with a mouse antibody against the lysosomal and late endosomal LAMP-1-specific and Alexa Fluor 594-labeled secondary antibody (B and D) or by coincubating the cells for 15 min with Alexa Fluor 594-labeled transferrin, which is specifically transported to the early endosomes (C and E). Cells were analyzed by fluorescence microscopy.

FIG. 4.

DCs strongly bind to HCV glycoprotein E1 and E2 through DC-SIGN. (A) Immature DCs express high levels of DC-SIGN. LPS-matured DCs express lower levels of DC-SIGN. Monocyte-derived DCs were isolated as described in Materials and Methods. Expression of DC-SIGN was measured by fluorescence-activated cell sorting staining with the DC-SIGN-specific antibody AZN-D2. Open histograms, isotype control; solid histograms, AZN-D2 staining. (B and C) Immature DCs and mature DCs bind strongly to HCV envelope glycoproteins E1 and E2 and mixed HCV E1/E2 VLPs via DC-SIGN. Immature (B) and LPS-matured (C) DC binding to HCV envelope glycoproteins was determined by a fluorescent bead adhesion assay. Specificity was determined by anti-DC-SIGN antibody AZN-D2 (50 μg/ml), mannan (100 μg/ml), EGTA (10 μM), and anti-mannose re- ceptor antibody (clone 19) (50 μg/ml). Standard deviations were <5%. (D) HCV VLPs do not affect DC activation or maturation. Immature DCs were incubated with HCV (12.5 μg/ml) alone, LPS (10 μg/ml) alone, or HCV and LPS together for 48 h, and activation was determined by measuring the expression of CD83 and CD86. Dotted lines, isotype controls; solid histograms with thin lines, incubations without HCV; open histograms with thick lines, incubations without or with LPS in combination with HCV VLPs. Upper panel, incubations without LPS; lower panel, incubations with LPS. One representative experiment out of three is shown.

FIG. 5.

DC-SIGN on immature DCs targets HCV VLPs to early endosomes but Lewis X antigen to lysosomes. (A) Immature DCs were incubated with HCV VLPs (30 μg/ml) for 4 h or overnight. HCV was detected with a human anti-HCV antibody and a FITC-labeled secondary antibody. Intracellular targeting was determined by staining the endosomal compartments with a mouse antibody against the lysosome- and late endosome-specific marker LAMP-1 or the early endosome-specific marker EEA-1 and an Alexa Fluor 594-labeled secondary antibody or by coincubating the cells for 15 min with Alexa Fluor 594-labeled transferrin, which is specifically transported to early endosomes. (B) Immature DCs were incubated with HCV VLPs (30 μg/ml) for 4 h or overnight at 37 or 4°C. HCV was detected as described for panel A. Localization of DC-SIGN was determined with the DC-SIGN-specific antibody AZN-D2 and an Alexa Fluor 594-labeled sec- secondary antibody. Cells were analyzed on a 3i Marianas digital imaging microscopy workstation with SlideBook software. (C) DC-SIGN on immature DCs targets its ligand Lewis X antigen to late endosomes and lysosomes. Immature DCs were incubated with Lewis X (10 μg/ml) for 4 h. Intracellular targeting was determined by staining the endosomal compartments with a mouse antibody against the lysosome- and late endosome-specific marker LAMP-1 or by coincubating the cells for 15 min with Alexa Fluor 594-labeled transferrin, which is specifically transported to early endosomes. Cells were analyzed by fluorescence microscopy.

FIG. 6.

HCV interacts with L-SIGN-expressing LSECs in situ. (A) L-SIGN is expressed by human LSECs, as was determined by staining of liver tissue with an L-SIGN-specific antibody. (B) Binding of HCV VLPs by liver tissue was determined by incubating liver sections with HCV VLPs (10 μg/ml) for 2 h at 37°C. HCV binding was detected with a human anti-HCV antibody and FITC-labeled secondary antibody. (C) HCV VLP binding to LSECs is specifically blocked by the L-SIGN-specific antibody AZN-D2. Sections were incubated with AZN-D2 (50 μg/ml) for 30 min at room temperature before HCV VLPs were added as described for panel A. Sections were analyzed by fluorescence microscopy with a 20× objective. Bars, 50 μm.