HIV-1-infected blood mononuclear cells form an integrin- and agrin-dependent viral synapse to induce efficient HIV-1 transcytosis across epithelial cell monolayer

Abstract

The heparan sulfate proteoglycan agrin and adhesion molecules are key players in the formation of neuronal and immune synapses that evolved for efficient communication at the sites of cell-cell contact. Transcytosis of infectious virus across epithelial cells upon contact between HIV-1-infected cells and the mucosal pole of the epithelial cells is one mechanism for HIV-1 entry at mucosal sites. In contrast, transcytosis of cell-free HIV-1 is not efficient. A synapse between HIV-1-infected cells and the mucosal epithelial surface that resembles neuronal and immune synapses is visualized by electron microscopy. We have termed this the "viral synapse." Similarities of the viral synapse also extend to the functional level. HIV-1-infected cell-induced transcytosis depends on RGD-dependent integrins and efficient cell-free virus transcytosis is inducible upon RGD-dependent integrin cross-linking. Agrin appears differentially expressed at the apical epithelial surface and acts as an HIV-1 attachment receptor. Envelope glycoprotein subunit gp41 binds specifically to agrin, reinforcing the interaction of gp41 to its epithelial receptor galactosyl ceramide.

Figure 1.

(A) Cell-free HIV-1 is not efficient for transcytosis and uninfected T-lymphocytes cannot substitute for the synapse formation. HIV-1-infected cells (1 million; Standard), cell-free HIV-1 (500 pg/ml p24), cell-free virus produced upon contact with the epithelial cell apical pole (500 pg/ml p24; virus on epithelial cells), in the absence or presence (1 million) of uninfected PBMC (uninfected cells) were inoculated at the apical pole of epithelial cells. Transcytosis was permitted for 2 h at 37°C. Transcytosis, estimated by the p24 content in the basolateral medium, was expressed as % of transcytosis observed in standard conditions. (B) Virological synapse between HIV-1 (R5 JR-CSF molecular clone)-infected PBMCs and the apical pole of epithelial cells. HIV-1-infected cells were inoculated at the apical pole of epithelial cells and allowed to initiate HIV-1 transcytosis for 45 min at 37°C. Cells were then fixed and processed for electron microscopy examination after ultrathin sectioning. (a, b, c) Low magnification micrograph (×6200) showing an HIV-1-infected cell (HIV-1+ cell) interacting with the epithelial surface (EC). (a′-c′ and a″-c″) the black boxed areas in (a-c) have been enlarged by ×5 and ×25, respectively. In a′: black arrows indicate viral particles that are shown enlarged in the inset. In a″-c″, the microvilli are evidenced around the cell, outside the synaptic cleft. Bar, a′ 100 nm; a′ inset 50 nm

Figure 2.

(A) Monomeric RGD peptides inhibit transcytosis induced upon HIV-1+ cell contact with epithelial cells. Monomeric RGD peptide, at the indicated concentration, was preincubated at the apical pole of epithelial cell monolayers before inoculation of HIV-1(X4 NDK clones)-infected cells. RGL was used as control of specificity. Transcytosis was measured as described above. (B) Transcytosis induced upon HIV-1+ cell contact with epithelial cells depends on beta-1 integrin. Blocking (DE9) or nonblocking (LM534) anti-human beta-1 integrin antibodies were preincubated at the apical pole of epithelial cell monolayers, for 20 min at 37°C, before inoculation of (X4 NDK clones) HIV-1-infected PBMC (Anti-beta-1, blocking or nonblocking). When indicated the apical epithelial surface was washed three times before virus inoculation to eliminate unbound blocking beta-1 antibodies (Anti-beta-1, blocking on Epithelial Cells). Alternatively, HIV-1-infected PBMC were preincubated with blocking anti-human beta-1 integrin antibodies followed by three washes to eliminate unbound antibodies, before inoculation onto the epithelial apical surface of the endothelial monolayers (Anti-beta-1, blocking on HIV-1+ PBMC). Transcytosis was allowed to occur for 2 h at 37°C and was measured as described above and compared with standard conditions. Error bars represent the mean of at least three independent experiments performed using X4 (NDK clone) or R5 (YU2 clone), HIV-1-infected CD4⁺ cells or the cell-free X4 HIV-1 (LAI clone) or R5 HIV-1 (YU2 clone) with similar results. (C) Efficient cell-free virus transcytosis is induced by oligomeric RGD peptides. Cell-free HIV-1 (the X4 LAI or the R5 YU2 clone; 500 pg/ml) were inoculated at the apical pole of epithelial cells alone (virus) or after apical treatment for 20 min at 37°C with monomeric biotinylated-RGD (100 μM; RGD-b), streptavidin (25 μM; SA), RGD-b tetramers complexed with streptavidin (RGD-b+SA), or synthesized RGD octamers (RGDx8). Transcytosis was estimated after 1 h 30 min at 37°C by quantifying HIV-1 p24 in the basolateral medium. Error bars represent the means of at least three independent experiments performed using the cell-free X4 HIV-1 (LAI clone) or R5 HIV-1 (YU2 clone) with similar results.

Figure 3.

(A) Cell-free HIV-1 bind to the apical epithelial surface in an HSPG-dependent but GalCer-independent mechanism. Cell-free HIV-1 (the R5 JR-CSF or the R5 YU2 clone; 500 pg/ml) were directly inoculated at the apical pole of epithelial cells (Standard). Alternatively, virus inoculation was preceded by apical treatment for 20 min at 37°C with anti-GalCer mAb or anti-HSPG IgM. HIV-1 was allowed to bind for 1 h at 4°C with gentle rocking. In one case a washing step with acidic buffer (glycine 100 mM, pH 2.2) was performed after the HIV-1 binding step. Binding was estimated by quantification of HIV-1 p24 associated with the cell monolayer after extensive washing of unbound virus and also expressed as % of virus bound in absence of treatment (Standard). Error bars represent the mean of at least four independent experiments performed using R5 HIV-1 (JR-CSF or YU2 clone). (B) Heparan sulfate and heparin enhance cell-free HIV-1 binding to the apical epithelial surface. In a similar binding assay as described in panel a, epithelial cells were preincubated with HS of different size from disaccharide subunit (dp) 8-18 or with heparin 9 kDa. Error bars represent the mean of at least three independent experiments performed using the cell-free R5 HIV-1 (YU2 clone).

Figure 4.

(A) Heparan sulfate and heparin increase the interaction between P1 and GalCer. P1 (125 μM) was preincubated with heparan sulfate (HS) from 14 to 18 disaccharide units (dp; 20 nM) or Heparin, either 9 kDa or Choay (20 nM), before addition of BpGalCer^*. Fluorescence Resonance Energy Transfer (FRET), from the tryptophan present in P1 to the Bodipy of BpGalCer^* was measured. P2 and a scrambled peptide containing the same number of tryptophan as P1 were used as control. Results are expressed as % of FRET from P1 to BpGalCer^* (Standard). (B) HIV-1 transcytosis is stimulated by heparin and dependent on an HSPG. Transcytosis of HIV-1 from HIV-1-infected PBMC or cell-free virus transcytosis initiated by RGDx8 was measured as described above after preincubation of the apical surface of the epithelial cell monolayer with heparin (33 μM), heparinase, or anti-HSPG antibodies for 20 min at 37°C. Transcytosis was measured as described above and compared with standard conditions. Error bars represent the mean of at least three independent experiments performed using X4 HIV-1 (NDK clone) or R5 (YU2 clone)-infected CD4⁺ cells or the cell-free R5 HIV-1 (YU2 or JR-CSF clone).

Figure 5.

(A) Epithelial cells express agrin. Filter grown HT29 (left lane) and HEC-1 (right lane) cells were solubilized, proteins separated by SDS-PAGE 5% and after electrotransfer, agrin was immunodetected by Western blot using a rabbit polyclonal anti-chick agrin antibody (antibody 3240). The migration pattern of molecular-weight standards run simultaneously is shown on the right. Note that no trace of agrin is visible at the loading position. (B) Agrin is expressed at the apical pole of epithelial cells. Filter grown HT29 (top row) and HEC-1 (bottom row) cells were fixed and labeled for agrin by indirect immunofluorescence with a rabbit polyclonal anti-chick agrin antibody revealed by FITC-labeled secondary IgG. To visualize the basolateral region of the monolayer, epithelial cells were double-labeled with Texas Red-labeled phalloidin. Agrin and actin labeling were detected by dual-color confocal microscopy. Agrin “Apical” corresponds to the labeling detected in the upper 2 μm (5 consecutive sections, 0.5 μm apart) of the cell above the tight junction and “Basolateral” to the labeling detected below the tight junction down to the filter (15 consecutive sections, 0.5 μm apart). To visualize the basolateral region the actin labeling is shown (Actin Basolateral). Bars, 10 μm. (C) Agrin is expressed at the apical pole of human colon epithelial cells. Human colon biopsies were frozen, sectioned, fixed, and labeled for galactosyl ceramide or agrin by indirect immunofluorescence with mouse monoclonal anti-GalCer antibody and rabbit polyclonal anti-chick agrin antibody, respectively. Labeling was detected by confocal microscopy. GalCer (green labeling) and agrin (red labeling) colocalized (merged) at the apical pole of epithelial cells (EC), facing the mucosal lumen (mucosal) in a single optical section (0.5 μm). Magnification, ×100.

Figure 6.

(A) HIV-1 binding to and transcytosis across epithelial cells involves agrin. HIV-1-infected cell derived transcytosis, cell-free virus transcytosis initiated by RGDx8, or cell-free HIV-1 binding was measured as described above, after preincubation of a rabbit polyclonal anti-chick agrin antibody (antibody 3240), at the epithelial cell apical pole, for 20 min at 37°C. Transcytosis was measured as described above and compared with standard conditions. Error bars represent the mean of at least three independent experiments performed using X4 (NDK clone) or R5 (YU2 clone) HIV-1-infected CD4⁺ cells or the cell-free R5 HIV-1 (YU2 or JR-CSF clone). (B) (a) and (b) Agrin is expressed apically on epithelial cells. HT29 (1) and HEC-1 (2) cells grown on filters were fixed, labeled for agrin, and processed with a rabbit polyclonal anti-mouse agrin antibody (antibody 204) whose binding was detected by 5-nm gold labeled anti-rabbit antibody and then preembedded and processed for ultrastructural analysis. Gold labeling is detected apically (a), but not basolaterally (b). Magnification: (a) ×70,000; (b) ×40,000. (c) Controls: control human and rabbit IgG were used together as described above as primary antibodies, followed by 5-nm gold labeled anti-rabbit and 10-nm gold labeled anti-human antibodies. No labeling is detected on HT29 (1) and Hec-1 (2) cells. Magnification, ×40,.000. (d) HIV-1 colocalized with agrin on epithelial cells. HIV-1 (NDK clone)-infected CD4⁺ cells were inoculated apically on filter-grown HT29 (1) or Hec-1 (2) cells for 50 min at 37°C. Cells were fixed and processed for agrin (5-nm gold) and HIV-1 gp41 (10-nm gold) labeling as described above. Apical detection of both colocalized markers, is shown. Magnification, ×60,000. (C) Gp41 on HIV+ PBMCs colocalized with epithelial agrin at the cell-cell contact zone. HIV-1 (JR-CSF)-infected PBMCs were inoculated apically on filter-grown HT29 cells for 45 min at 37°C. Cells were fixed and processed for labeling of epithelial agrin (revealed by FITC-labeled anti-rabbit IgG) and HIV-1 gp41 at the surface of HIV+ cells or on viral particles (revealed by Texas Red anti-human IgG), as described above. Cells were observed by two-color confocal microscopy. Overlapped images of only the cell-cell contact zone are shown (section 1-10: 0.3 μm steps) onto which colocalization, analyzed with the Bio-Rad 1024 software, appears as yellow dots. Position of the sections shown is boxed in the schematized coculture system (I). Two examples are shown (II and III). For the first of the series, the vertical projection is shown (P). Bars, 10 μm.

Figure 7.

(A) P1 binds to glycosylated agrin. Secretory component, glycosylated and nonglycosylated agrin were spotted onto nitrocellulose membrane. Spotted proteins were by Ponceau red transient staining (Ponceau) before membrane saturation and incubated with monomeric biotinylated-P1 (15 μM) or oligomeric biotinylated-P1 (50 μM). Biotinylated-P1 was revealed with HRP-streptavidin. (B) Glycosylated agrin specifically potentiates P1 interaction with BpGalCer^*. P1 (125 μM) was preincubated with 10 nM of glycosylated agrin, nonglycosylated agrin, or syndecan-1, or the unrelated heavily glycosylated secretory component, before addition of BpGalCer^*. FRET from the W present in P1 to the Bodipy of BpGalCer^*, was measured. Results are expressed as % of FRET (Standard) from P1 to BpGalCer^*. Only glycosylated agrin (B/z+ or B/z-) increases FRET.