Anatomy of a viral entry platform differentially functionalized by integrins α3 and α6

Abstract

During cell invasion, human papillomaviruses use large CD151 patches on the cell surface. Here, we studied whether these patches are defined architectures with features for virus binding and/or internalization. Super-resolution microscopy reveals that the patches are assemblies of closely associated nanoclusters of CD151, integrin α3 and integrin α6. Integrin α6 is required for virus attachment and integrin α3 for endocytosis. We propose that CD151 organizes viral entry platforms with different types of integrin clusters for different functionalities. Since numerous viruses use tetraspanin patches, we speculate that this building principle is a blueprint for cell-surface architectures utilized by viral particles.

Figure 1

Integrin α3 or integrin α6 knockdown inhibits viral capsid processing and infection. Integrin α3 or integrin α6 were knocked down in HaCaT cells by siRNA transfection (for knockdown efficiency see Fig. S1). (A) Two days after transfection, cells were incubated for 24 h with HPV16 PsVs, washed, lysed and analyzed by Western blot for the viral protein L1 and its ~25 kDa cleavage product. For clarity, lanes were cropped from original blots shown in full in Fig. S14 (L1) and S15 (actin). Values were related to the control which was set to 100% and are given as means ± SD (n = 3 independent experiments). (B,C) Two days after transfection, cells were incubated for 5 h with HPV16 PsVs, washed fixed, stained with an antibody that recognizes L1 after capsid disassembly, and imaged by confocal microscopy taking an optical section from the cell body. (B) Actin (cyan) and L1–7 (inverted greyscale) each are displayed at the same arbitrary scaling (linear lookup tables). From the optical section (B), an image analysis algorithm counted the number of detected vesicles per cell (C) and quantified the vesicle staining intensity (Fig. S2). Values are given as means ± SD (n = 60 analysed cells collected from three biological replicates). (D) HaCaT cells were transfected and incubated with PsVs as in (A) with the difference that on the encapsidated plasmid luciferase expression is under the control of the HPV16 promoter instead of the CMV promoter. One day after adding PsVs, cells were lysed and the infection rate was assessed by analysing the luciferase activity. For normalization to cell number, luciferase activity was related to the dehydrogenase activity. Values are expressed as percent of control (average of control was set to 100%). Values are given as means ± SD (n = 20–21 technical replicates collected from five biological replicates). Unpaired Student’s t-test, comparing control to knockdown conditions (***p < 0.001; **p < 0.01).

Figure 2

Knockdown of integrin α6 but not of integrin α3 strongly inhibits PsV binding. Two days after siRNA transfection, cells were incubated for 1 h at 0 °C with HPV16 PsVs, scraped from the substrate, washed, lysed and analyzed by Western blot. Full Western blots are shown in Fig. S16 (L1) and S17 (actin). Binding of viral particles was assayed by quantification of the L1 ~55 kDa band. The control value was set to 100% and used for normalization. Values are given as means ± SD (n = 4 independent experiments). Unpaired Student’s t-test, comparing control to knockdown conditions (***p < 0.001; *p < 0.05).

Figure 3

Characteristics of CD151 and integrin maxima. CD151-GFP was transfected into HaCaT cells. CD151-GFP and endogenous CD151 localize to the same domains (Fig. S7E,F) and CD151-GFP is as functional as non-tagged CD151 (Fig. S8). One day after transfection, cells were treated for 5 h without or with PsVs (Fig. S9), washed, membrane sheets were generated, stained and analyzed by three channel STED microscopy. Green lookup table, CD151-GFP visualized by nanobodies; red and cyan lookup tables, integrin α6 and integrin α3 stained by antibody labeling, respectively. Images are displayed at arbitrary intensity scalings (linear lookup tables). Staining with the integrin antibodies is highly specific (Fig. S1). (A) Large panel, membrane sheet (channel overlay). The white box marks an area from which magnified views of the individual channels are shown. From regions of interest (ROIs) we measured (B) maxima size and signal overlap (C). For (C), for all three channels within a ROI the pixels with an intensity higher than the average ROI intensity were selected. Then, the number of pixels positive in all three channels were related to the number of all positive pixels in one specific channel as indicated. (D) Shortest inter-maxima distances of CD151-GFP to CD151-GFP, integrin α3 or integrin α6. Values are given as means ± SD (n = 60 membrane sheets collected from three biological replicates).

Figure 4

Viral particle distance to CD151 and integrin maxima. (A–C) Membrane sheets were generated from HaCaT cells treated for 5 h with EdU-PsVs. They were stained for EdU-PsVs (cyan lookup table; click-labelled with fluorescein), CD151 (green lookup table; immunostaining) and integrin α6 (red lookup table; immunostaining), followed by STED microscopy imaging. (A) Images from the three channels and an overlay are shown and displayed at arbitrary intensity scalings (linear lookup tables). Circles mark identical pixel locations. (B) Pearson correlation coefficient between PsVs and CD151 (left) and integrin α6 (right). A control value for randomized distribution (‘flipped’) was calculated after flipping of one channel. (C) For individual PsVs, we plotted their nearest distance to a CD151 and an integrin α6 maximum (the analysis includes 4452 PsVs from 57 membrane sheets collected from three biological replicates). (D–F) As in (A–C) but cells were analysed and we stained for integrin α3 (the plot in F includes 8519 analyzed PsVs from 60 cells collected from three biological replicates). The pattern of PsV binding to the cell membrane looks the same on cells and membrane sheets (Fig. S10). Values are given as means ± SD (n = 57 membrane sheets or 60 cells, each collected from three biological replicates). Statistical analysis was performed employing the unpaired Student’s t-test comparing the original to the flipped images (***p < 0.001). Red boxes frame viral particles potentially used for further analysis of the platform area (Fig. 5). Only PsVs were considered that were closer than 250 nm to bright CD151/integrin maxima.

Figure 5

Local crowding of clusters at virus attachment sites. To characterize the anatomy of a possibly forming viral platform area, we selected those PsVs from Fig. 4C,F with a distance smaller than 250 nm to both a bright CD151 and integrin maximum (for examples see centrally located PsVs in the left panels in A and B). In a PsV-centred 1 µm × 1 µm ROI the number of bright maxima (see methods) in the CD151 and integrin channels were counted. The same was performed after flipping the CD151/integrin channels, causing a randomized relationship between PsV-attachment sites and maxima (A and B, right panels). In the original images more closely associated maxima are present at sites of PsV-attachment. Images are displayed in cyan (PsV) green (CD151) and red (integrin) at arbitrary intensity scalings (linear lookup tables). (C,D), quantification of the bright maxima in the 1 µm × 1 µm ROI. Values are given as means ± SD (n = 678 (C) and 283 (D) 1 µm × 1 µm ROIs, respectively). Statistical analysis was performed employing the unpaired Student’s t-test comparing the original to the randomized condition (***p < 0.001). Please note, that the number of clusters per attachment site was highly variable. For a histogram illustrating the variability see Fig. S12.

Figure 6

Integrin α3 and integrin α6 knockdown differentially affect the CD151 level of the cell membrane and PsV-binding. HaCaT cells were transfected with siRNA and two days later incubated without or with HPV16 PsVs for 5 h. Cells were fixed, permeabilized, immunostained for CD151 (green) and L1 (red), and imaged by STED microscopy. Images are displayed employing linear lookup tables. For A and D different scalings were applied, although all panels in A or D have the same arbitrary scaling. (A) Integrin α6 knockdown. For each condition, overlays from an overview and a magnified view are shown. (B) Average CD151 immunostaining intensity and (C) the PsVs particle density (100% correspond to 1.4 particles/μm²). (D–F) As (A–C), showing the integrin α3 knockdown (100% PsV density correspond to 1.2 particles/μm²). Values are shown as means ± SD (n = 60 cells collected from three biological replicates). The unpaired Student’s t-tests compare the conditions as indicated by the bars (B,E) or PsV to PsV and knockdown (C,F) (***p < 0.001; **p < 0.01; *p < 0.05).

Figure 7

Large CD151 patches coincide with intracellular actin accumulations. CD151-GFP transfected HaCaT cells were treated for 3 h without or with PsVs. Membrane sheets were generated, stained, and imaged by confocal microscopy. Green (CD151-GFP; GFP signal was enhanced by nanobodies), red (PsVs visualized by L1 antibody labeling) and cyan (filamentous actin; fluorescently labelled phalloidin). Images are displayed using a linear lookup table. For each channel, the same arbitrary scaling was applied. (A) For each condition a membrane sheet is shown. Magnified views from the white boxes are shown, illustrating the individual channels. (B) For the CD151 and the actin channels, within a freehand ROI excluding membrane edges, the pixels with an intensity higher than the average ROI intensity were selected. Then, the number of pixels positive in both channels were related to the number of all positive pixels in the CD151 channel. Values are given as means ± SD (n = 45 membrane sheets collected from three biological replicates). Unpaired Student’s t-test (***p < 0.001).