Steric shielding of surface epitopes and impaired immune recognition induced by the ebola virus glycoprotein

Abstract

Many viruses alter expression of proteins on the surface of infected cells including molecules important for immune recognition, such as the major histocompatibility complex (MHC) class I and II molecules. Virus-induced downregulation of surface proteins has been observed to occur by a variety of mechanisms including impaired transcription, blocks to synthesis, and increased turnover. Viral infection or transient expression of the Ebola virus (EBOV) glycoprotein (GP) was previously shown to result in loss of staining of various host cell surface proteins including MHC1 and β1 integrin; however, the mechanism responsible for this effect has not been delineated. In the present study we demonstrate that EBOV GP does not decrease surface levels of β1 integrin or MHC1, but rather impedes recognition by steric occlusion of these proteins on the cell surface. Furthermore, steric occlusion also occurs for epitopes on the EBOV glycoprotein itself. The occluded epitopes in host proteins and EBOV GP can be revealed by removal of the surface subunit of GP or by removal of surface N- and O- linked glycans, resulting in increased surface staining by flow cytometry. Importantly, expression of EBOV GP impairs CD8 T-cell recognition of MHC1 on antigen presenting cells. Glycan-mediated steric shielding of host cell surface proteins by EBOV GP represents a novel mechanism for a virus to affect host cell function, thereby escaping immune detection.

Figure 1. Transient expression of the EBOV glycoprotein results in loss of surface staining of β1 integrin and MHC1.

(A) 293T cells were transfected with empty pCAGGS (vector) or vector encoding GP. Floating and adherent cells were harvested 24 h after transfection, pooled, and stained for GP using the KZ52 antibody, followed by FITC-labeled secondary antibodies, and co-stained for β1 integrin or MHC1 with PE-Cy5 conjugated monoclonal antibodies and assayed by flow cytometry. (B) Following transfection with vector encoding GP, floating 293T cells were removed from adherent cells, stained for β1 integrin and assayed by flow cytometry (left panel). Similarly treated cells were mounted on coverslips, fixed, permeabilized and stained for GP with mouse monoclonal antibodies, followed by Alexa 594 conjugated antibodies and assayed by immunofluorescence microscopy. A representative cell is shown (right panel).

Figure 2. Steady-state levels of β1 integrin and MHC1 are unchanged in GP-expressing cells.

293T cells were transfected with empty vector (A) or vector encoding wt GP (B). Floating and adherent cells were harvested 24 h after transfection, pooled, and stained for GP, β1 integrin, and MHC1 as described earlier and assayed by flow cytometry. Prior to staining, a portion of cells were fixed and permeabilized to expose occluded surface and internal epitopes.

Figure 3. The mucin and globular domains of the EBOV glycoprotein masks the KZ52 epitope on the cell surface.

(A) 293T cells were transfected with empty vector or vector encoding wt GP, CmucAU1 GP, or NmucAU1 GP. Lysates were harvested in RIPA buffer after 24 h and subjected to SDS-4 to 15% PAGE, transferred to PVDF, and immunoblotted with anti-GP polyclonal rabbit antibodies (top blot) or anti-AU1 antibodies (bottom blot) and anti-GAPDH antibodies. (B) HeLa cells were transfected with vector encoding wt GP, CmucAU1 GP, or NmucAU1 GP. 24 h after transfection, cells were fixed, permeabilized, and stained for GP with mouse monoclonal antibodies, and the AU1 epitope with anti-AU1 antibodies, followed by Alexa 594 and Alexa 488 conjugated antibodies, respectively and assayed by immunofluorescence microscopy. Scale bars are 10.6 µm. (C, D) 293T cells were transfected with vector encoding wt GP, CmucAU1 GP, or NmucAU1 GP. Floating and adherent cells were harvested 24 h after transfection, pooled, and stained for GP using the KZ52 antibody or the AU1 tag, β1 integrin, and MHC1, as described earlier and assayed by flow cytometry. (C) β1 integrin vs. GP or AU1 surface staining. (D) MHC1 vs. GP or AU1 surface staining. Cartoon depictions of the KZ52 epitope (red star), CmucAU1 epitope (green star) or NmucAU1 epitope (yellow star) are shown below their respective flow cytometry plots. The globular region of GP is shown shaded blue; the mucin domain is shown shaded green.

Figure 4. Removal of the GP₁ subunit from the cell surface results in exposure of previously occluded surface epitopes.

(A-C) 293T cells were transfected with empty vector or vector encoding wt GP. Floating and adherent cells were harvested 24 h after transfection, pooled, and either left untreated, or incubated at 37°C for 20 minutes in 150 mM DTT. (A) Western blot analysis of the GP₁ subunit shed into the supernatant of untreated or DTT-treated cells. (B) Cells were transfected with empty vector and assayed by flow cytometry to show baseline differences in surface staining for β1 integrin and MHC1 between untreated cells (grey shading) and DTT-treated cells (black trace). (C) Cells were transfected with vector encoding wt GP and assayed by flow cytometry for the internal proteins GM130 and calnexin to show the effect of DTT on cell permeabilization. Untreated cells are shown in the grey shading; DTT-treated cells are shown in the black trace, and, as a positive control, fixed/permeabilized cells are shown in the dashed trace. (D) Cells were transfected with vector encoding wt GP, were mock- or DTT-treated, stained for GP and β1 integrin or MHC1 as described above, and assayed by flow cytometry. (E, F) 293T cells were transfected with empty vector or vector encoding wt GP or GP cl(-). Cells were harvested and treated as above. (E) Western blot analysis using rabbit polyclonal antibodies of GP₁ or GP₀ shed into the supernatant of untreated or DTT-treated cells. (F) Transfected and treated cells were surface stained for MHC1 and assayed by flow cytometry. Empty vector-transfected, untreated cells are shown in the grey shading; GP-expressing cells are shown after mock treatment (black trace) or DTT treatment (dashed trace).

Figure 5. Surface N- and O- linked glycans contribute to GP-mediated shielding.

(A, left plot) Cells transfected with empty vector were mock incubated (grey shading) or incubated with neuraminidase and PNGaseF (black trace), stained for β1 integrin and assayed by flow cytometry. (A, right plot) Cells were treated with DMSO, transfected with empty vector, and mock incubated (grey shading) or treated with benzyl-α-GalNAc (bz-GalNAc), transfected with empty vector, and incubated with PNGaseF (black trace), then stained for β1 integrin and assayed by flow cytometry. (B) Cells were left untreated and untransfected (grey shading), or treated with bzGalNAc, transfected with vector encoding GP, and incubated with PNGaseF (black trace), then assayed by flow cytometry for the internal proteins GM130 and calnexin to show the effect of these treatments on cell permeabilization. As a positive control, bz-GalNAc-treated, fixed/permeabilized cells are shown (dashed trace). (C, D) Floating cells in cultures transfected with vector encoding GP were mock incubated or incubated with neuraminidase and/or PNGaseF, then analyzed by Western blot with rabbit polyclonal antibodies to GP (C, left blot) or stained for β1 integrin and assayed by flow cytometry (D). (C, E) Cells were treated with DMSO or bz-GalNAc, then transfected with empty vector, or vector encoding GP. Cells were then mock-incubated, or incubated with PNGaseF, then analyzed by Western blot (C, right blot) or stained for β1 integrin and assayed by flow cytometry (E). For D and E, mock-incubated cells or DMSO-treated cells transfected with empty vector =  grey shading; GP-transfected and DMSO-treated or mock-incubated cells =  black traces; GP-transfected and bz-GalNAc- and glycosidase-treated cells =  dashed traces. In Western blot panels (C), selected samples were lysed and denatured before incubation with PNGaseF for comparison, as indicated.

Figure 6. EBOV GP-induced disruption of MHC1 prevents activation of CD8⁺ T cells.

OV79 SL9 target cells were mock transduced (no Ad) or transduced with Adenoviral vectors expressing GFP (Ad GFP) or GFP and EBOV GP (Ad GP) at an MOI of 300. 48 h after transduction, cells were assayed for GFP expression (A); No Ad =  blue trace; Ad GFP =  green trace; Ad GP =  orange trace. Cells were also stained for MHC1 (B); isotype antibody =  shaded peak; No Ad =  blue trace; Ad GFP =  green trace; Ad GP =  orange trace. In parallel, CD8 T cells expressing a transgenic TCR (868TCRwt) that recognizes the SL9 HLA-A2 complex were incubated alone or with mock- (no Ad) or Ad- transduced target cells in a 2∶1 ratio. After co-culture, T cells were surface stained for CD8, then fixed and permeabilized, and stained for 868TCRwt and MIP-1β with APC-H7, FITC, and PE- conjugated antibodies, respectively, and assayed by flow cytometry. (C) CD8⁺ and 868TCRwt⁺ events were analyzed for MIP-1β staining. (D) Bar graph depicts percent cells positive for MIP-1β, normalized to the No Ad target cell sample.