Expression of DC-SIGN and DC-SIGNR on human sinusoidal endothelium: a role for capturing hepatitis C virus particles

Abstract

Hepatic sinusoidal endothelial cells are unique among endothelial cells in their ability to internalize and process a diverse range of antigens. DC-SIGNR, a type 2 C-type lectin expressed on liver sinusoids, has been shown to bind with high affinity to hepatitis C virus (HCV) E2 glycoprotein. DC-SIGN is a closely related homologue reported to be expressed only on dendritic cells and a subset of macrophages and has similar binding affinity to HCV E2 glycoprotein. These receptors function as adhesion and antigen presentation molecules. We report distinct patterns of DC-SIGNR and DC-SIGN expression in human liver tissue and show for the first time that both C-type lectins are expressed on sinusoidal endothelial cells. We confirmed that these receptors are functional by demonstrating their ability to bind HCV E2 glycoproteins. Although these lectins on primary sinusoidal cells support HCV E2 binding, they are unable to support HCV entry. These data support a model where DC-SIGN and DC-SIGNR on sinusoidal endothelium provide a mechanism for high affinity binding of circulating HCV within the liver sinusoids allowing subsequent transfer of the virus to underlying hepatocytes, in a manner analogous to DC-SIGN presentation of human immunodeficiency virus on dendritic cells.

Figure 1

HSECs express both DC-SIGN and DC-SIGNR. Immunofluorescence was used to demonstrate the presence of DC-SIGN on HSEC. A: Primary cultures of HSEC-expressed DC-SIGN. B: Cells from the same preparation were also DC-SIGNR-positive.

Figure 2

Treatment with IL-4 increases the expression of DC-SIGNR and DC-SIGN on primary HSEC. HSECs were stimulated with IL-4 (100 ng ml⁻¹), TNF-α (10 ng ml⁻¹), or IL-17 (10 ng ml⁻¹) for 24 hours before assessing expression of DC-SIGNR (LS), DC-SIGN (DCS), CD31, and E-Selectin (CD62E) by enzyme-linked immunosorbent assay. An equivalent volume of complete media was added to the control wells. Data represent mean ± SE of four experiments using different HSEC isolates. IL-4 induced a significant increase in DC-SIGNR and DC-SIGN expression (P < 0.05 and P < 0.02; using Student’s t-test). * P < 0.05; ** P < 0.02.

Figure 3

DC-SIGNR and DC-SIGN on HSECs can bind HCV E2. Binding assays were performed using a strain of HCV HCJ4 with minimal interaction with CD81 when expressed as a truncated E2 species and flow cytometric detection with an antibody raised against E2, which fails to detect E2-CD81 complexes. A: Flow cytometric analysis of HCV E2 binding following treatment with IL-4 (100 ng ml⁻¹ for 24 hours) confirmed that IL-4 significantly increased the expression of both C-type lectins, which led to an increased HCV E2 interaction with the stimulated HSECs (red) compared to unstimulated HSECs (green) and control mock protein in black. B: HCV E2 bound to HSECs (red), and this interaction was inhibited by prior incubation of the cells with mAbs specific for DC-SIGN (gray) and DC-SIGNR (green line) alone or in combination (blue). Control binding with mock protein is in black.

Figure 4

Distribution of DC-SIGNR and DC-SIGN expression in liver tissue. The left panel is a representative immunofluorescent staining (original magnification, ×10) showing DC-SIGNR is not expressed on portal tracts (PT) as shown with alkaline phosphatase anti-alkaline phosphatase staining on the right panel (original magnification, ×20). DC-SIGN, however, does not have a patchy distribution and is expressed around portal tracts. CV, central vein.

Figure 5

DC-SIGN and DC-SIGNR colocalize with LYVE-1 staining on sinusoidal endothelium. Because LYVE-1 expression in the liver is restricted to sinusoidal endothelium, we used LYVE-1 staining to confirm the cellular localization of DC-SIGN staining. DC-SIGNR and DC-SIGN are stained red and LYVE-1 green. Colocalization is demonstrated in merged images by yellow staining confirming that DC-SIGN and DC-SIGNR are expressed on sinusoidal endothelial cells (original magnification, ×20).

Figure 6

DC-SIGN is detected on Kupffer cells in liver tissue. Immunofluorescent staining of mannose receptor and CD68 (green) was used to detect Kupffer cells, and DC-SIGN stained red. Colocalization of the two receptors is seen as yellow staining confirming that Kupffer cells express DC-SIGN in the liver (original magnification, ×40).

Figure 7

HCV E2 glycoprotein binding to sinusoids in normal liver tissue can be blocked by mannan. A: HCV E2 glycoprotein binding on normal liver tissue is seen on sinusoids and plump cells with Kupffer cell morphology. B: The HCV E2 binding can be inhibited by preincubating with mannan (20 μg/ml) for 1 hour before the addition of HCV E2 glycoprotein resulting in the attenuation of both the sinusoidal and Kupffer cell staining. Original magnification, ×20.

Figure 8

HCV E2 binds to Kupffer cells and sinusoidal endothelial cells in liver tissue. A: Kupffer cells labeled with macrophage mannose receptor (original magnification ×40, lower panel) and CD68 (original magnification ×20, upper panel) in red are seen clearly in the sinusoidal space, and HCV E2 (green) is shown binding to these cells. Colocalization of macrophage mannose receptor and CD68 with HCV E2 is demonstrated by the yellow staining in merged images. B: HCV E2 (green) also binds to DC-SIGNR-expressing sinusoidal endothelium (red) as confirmed by colocalization (yellow). Original magnification, ×20.

Figure 9

HCV E2 binds to normal sinusoidal endothelium and Kupffer cells via DC-SIGNR and DC-SIGN. A: Normal liver sections were preincubated with an isotype IgG2b control antibody an hour before labeling with HCV E2 glycoprotein (green). The staining pattern reveals sinusoidal and Kupffer cell distribution. B: When preincubated with DC-SIGN mAb (5 μg/ml) the binding of HCV E2 to sinusoidal and Kupffer cells was attenuated. C: Preincubation with DC-SIGNR (5 μg ml⁻¹) only attenuated binding to sinusoidal endothelium. Original magnification, ×20.

Figure 10

Sinusoidal endothelium does not support HCVpp viral infection. Because HSECs can support binding of HCV glycoproteins, we went on to determine whether they can also support viral infection. HSECs were prepared from three different livers, and all infections were preformed in triplicates. HCVpp bearing strain HCJ4 and H77 glycoproteins infected Hep3B hepatoma cells but failed to infect HSECs. Preincubating the HSECs with IL-4 (100 ng/ml) for 24 hours did not alter the results. HCV-MLV and HCV-No env served as positive and negative control experiments, respectively. Successful infection was measured by quantifying the luciferase reporter gene and expressed as relative light units (RLU).

Figure 11

Infection of Huh-7.5 cells, but not primary HSECs, with J6/JFH HCVcc in coculture. Huh-7.5 cells (A) and primary HSECs (B) have contrasting and distinctive morphologies in culture. To assess whether primary HSECs can be infected with J6/JFH, HSECs were cultured with persistently infected Huh-7.5 cells. The cells were mixed at an Huh-7.5:HSEC ratio of 1:4 and seeded at 200,000 cells per well in collagen coated 6-well plates. The following day the coculture had a distinct appearance in which HSECs and Huh-7.5 cells formed discrete foci (C). A control population of J6/JFH-infected Huh-7.5 cells was seeded at the same density in the absence of HSECs (D). Cocultures were maintained for 48 hours, and J6/JFH infection was monitored by staining for HCV NS5A antigen. Viral antigen was detected in the persistently infected Huh-7.5 cells (D) and within the coculture (E). Viral antigen was only detected in the Huh-7.5 cells and not within the HSECs, consistent with primary HSECs failing to support HCVpp infection (Figure 10).