HIV Envelope gp120 Alters T Cell Receptor Mobilization in the Immunological Synapse of Uninfected CD4 T Cells and Augments T Cell Activation

ABSTRACT

HIV is transmitted most efficiently from cell to cell, and productive infection occurs mainly in activated CD4 T cells. It is postulated that HIV exploits immunological synapses formed between CD4 T cells and antigen-presenting cells to facilitate the targeting and infection of activated CD4 T cells. This study sought to evaluate how the presence of the HIV envelope (Env) in the CD4 T cell immunological synapse affects synapse formation and intracellular signaling to impact the downstream T cell activation events. CD4 T cells were applied to supported lipid bilayers that were reconstituted with HIV Env gp120, anti-T cell receptor (anti-TCR) monoclonal antibody, and ICAM-1 to represent the surface of HIV Env-bearing antigen-presenting cells. The results showed that the HIV Env did not disrupt immunological synapse formation. Instead, the HIV Env accumulated with TCR at the center of the synapse, altered the kinetics of TCR recruitment to the synapse and affected synapse morphology over time. The HIV Env also prolonged Lck phosphorylation at the synapse and enhanced TCR-induced CD69 upregulation, interleukin-2 secretion, and proliferation to promote virus infection. These results suggest that HIV uses the immunological synapse as a conduit not only for selective virus transmission to activated CD4 T cells but also for boosting the T cell activation state, thereby increasing its likelihood of undergoing productive replication in targeted CD4 T cells.

IMPORTANCE There are about two million new HIV infections every year. A better understanding of how HIV is transmitted to susceptible cells is critical to devise effective strategies to prevent HIV infection. Activated CD4 T cells are preferentially infected by HIV, although how this is accomplished is not fully understood. This study examined whether HIV co-opts the normal T cell activation process through the so-called immunological synapse. We found that the HIV envelope is recruited to the center of the immunological synapse together with the T cell receptor and enhances the T cell receptor-induced activation of CD4 T cells. Heightened cellular activation promotes the capacity of CD4 T cells to support productive HIV replication. This study provides evidence of the exploitation of the normal immunological synapse and T cell activation process by HIV to boost the activation state of targeted CD4 T cells and promote the infection of these cells.

INTRODUCTION

Human immunodeficiency virus (HIV) infection leads to severe destruction of immune cells and functions. The helper CD4 T cell is one of the main cell types profoundly affected by HIV (1, 2). However, not all CD4 T cells are equally affected by HIV. Although HIV can infect resting naive CD4 T cells, these cells predominantly express the coreceptor CXCR4 and are less likely to express the coreceptor CCR5 required for the entry of the majority of transmitted and circulating HIV-1 isolates. In contrast, many memory CD4 T cells express the coreceptors CXCR4 and CCR5 (3, 4). The postentry steps in the HIV life cycle are also tightly linked to the activation status of CD4 T cells. Reverse transcription (5, 6), nuclear import (7), and integration (8) are inefficient unless CD4 T cells are activated and enter the cell cycle. Virus transcription is triggered via NF-κB (9), which is activated as a result of the specific signaling cascade triggered upon T cell receptor (TCR) engagement. Therefore, TCR-activated CD4 T cells are the optimal targets for HIV. Indeed, the recruitment of activated CD4 T cells to the genital or rectal mucosa associated with herpes simplex virus 2, gonorrhea, and other sexually transmitted diseases is considered to be one of the factors that increase the risk of HIV acquisition (10–12). Studies of simian immunodeficiency virus (SIV) and simian-human immunodeficiency virus infections in rhesus macaques also showed that the increased number of activated CD4 T cells at the site of virus entry constitutes one of the correlates of increased infection (13, 14). However, the mechanisms by which HIV preferentially targets the activated subsets of CD4 T cells are not fully understood.

CD4 T cell activation commences in an immunological synapse, a tight junction at the contact site between a CD4 T cell and an antigen-presenting cell (APC) formed when the CD4 T cell recognizes the cognate peptide-major histocompatibility complex class II (pMHC) complexes on the APC (reviewed in references 15 and 16). A CD4 T cell will stop migrating once it has formed an immunological synapse (17). At the periphery of the synaptic area, pMHC-TCR interactions form microclusters that quickly translocate to the center and converge to become the central supramolecular activation cluster (cSMAC) (18, 19). At the same time, ICAM-1–LFA-1 interactions begin clustering to form the peripheral SMAC (pSMAC). A mature stable synapse is thus created with a cSMAC and a pSMAC ring that are fully segregated and arrest cell migration for >1 h. Recent correlative optical/electron microscopy analyses of the immunological synapse formed on surrogate APCs based on lipid bilayers have provided higher-resolution pictures showing the cSMAC region as a cleft containing TCR-rich vesicles (20). These images are highly reminiscent of the electron tomographic data showing the accumulation of HIV virions in the contact area formed between a primary CD4 T cell target and an infected T cell (21).

The cell-cell contact area implicated in facilitating highly efficient HIV transmission from one cell to another is called the virological synapse (22–26). This synapse has a morphology similar to that of the immunological synapse and utilizes many of its components, although the two types of synapses also display distinctive features (27). The assembly of virological synapses is triggered by the binding of the HIV envelope (Env) on infected cells to CD4 on target cells and can occur independently of TCR-pMHC engagement (28). Similar to the immunological synapse, the formation of the HIV Env-induced synapse also results in the arrest of CD4 T cell migration (28). During this time, HIV Env-CD4 microclusters form and quickly converge to assemble a cSMAC-like cluster surrounded by the pSMAC-like ICAM-1/LFA-1 ring (29). The TCR alpha beta chains are also recruited to the cSMAC-like cluster, albeit in the absence of cognate pMHCs (29). However, when CD4 T cells encounter HIV-infected APCs, simultaneous HIV Env-CD4 and TCR-pMHC engagements ensue and the consequences of such interactions remain unclear.

As a result of TCR-pMHC engagement at the immunological synapse, T cells receive a series of activation signals (reviewed in reference 15). The TCR activation signals begin with the rapid recruitment and activation of the Src kinases Lck and Fyn (18, 19, 30). The signals propagate downstream, leading to the phosphorylation of phospholipase C-γ1 (PLC-γ1) and the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP₂). This results in the production of the second messengers diacylglycerol and inositol trisphosphate, the activation of transcription factor NF-κB via protein kinase C θ (PKCθ) and the mitogen-activated protein kinase/Erk pathways, the flux of intracellular Ca²⁺, and the transcription of interleukin-2 (IL-2) via Ca²⁺-associated calmodulin, calcineurin, and nuclear factor of activated T cells. Of note, the TCR-induced signaling cascade is transient; it is initiated at the newly formed TCR-pMHC microclusters and terminated at the cSMAC (18, 30). CD4 distribution is consistent with these data in that CD4 initially associates with the TCR-pMHC microclusters yet does not accumulate in the cSMAC (31). The lack of active signaling at the cSMAC is also explained by the finding that TCRs associated with microvesicles are detached and released from the cell into the synaptic cavity at the cSMAC (20).

A similar but atypical signaling cascade is triggered in the HIV Env-induced virological synapse (29). HIV Env-CD4 interactions at the synapse lead to the phosphorylation of Lck and the downstream TCR signalosome components CD3ζ, ZAP-70, LAT, SLP-76, Itk, and PLC-γ1 (29). These signaling molecules accumulate in the cSMAC-like central Env/CD4 cluster and retain their phosphorylation for more than 1 h, even after the synapse breaks and the cell resumes migration. Notably, the protracted signaling does not lead to full T cell activation, as indicated by lack of PKCθ activation and no elevation of Ca²⁺ or CD69. Nonetheless, the Env-induced signals trigger actin polymerization to create an F-actin ring with a depleted central zone suggested to serve as a channel for the efficient transport of HIV virions within the target CD4 T cell (29). These signals also activate LFA-1 on naive CD4 T cells (32).

In this study, we postulate that aberrant HIV Env-triggered signaling exerts a dominant effect on TCR-induced activation signals at the immunological synapse to impact the activation state of CD4 T cells targeted by the virus. We utilized a supported lipid bilayer (SLB) system that has been used to monitor the immunological synapses formed by CD4 T cells, CD8 T cells, and NK cells and the signaling molecules recruited to the various synapses (18, 19, 28, 29, 33, 34). To evaluate the specific contribution of the HIV Env in the CD4 T cell immunological synapse, we applied primary activated CD4 T cells to an SLB constituted with OKT3 (an anti-CD3 monoclonal antibody [MAb] that engages the TCR) and ICAM-1 (a key adhesion molecule integral to the immunological synapse) with or without HIV Env gp120. Although the HIV Env did not disrupt immunological synapse formation, its presence at the synapse affected the kinetics of TCR mobilization and synapse morphology. Further, the HIV Env induced temporal and spatial alterations to the most proximal TCR-associated signaling molecule Lck and augmented CD4 T cell activation and proliferation. These data support the notion that HIV co-opts the immunological synapse to disseminate to highly vulnerable CD4 T cells undergoing TCR-induced activation and promote its replication by heightening the T cell activation state.

MATERIALS AND METHODS

Cells.

Peripheral blood mononuclear cells (PBMCs) were isolated from leukopacks of anonymous healthy donors (purchased from the New York Blood Center) with Ficoll-Hypaque. Approval for the use of human specimens was obtained from the New York University School of Medicine Institutional Research Board. CD4 T cells or CD8 T cells were enriched by using CD4 or CD8 negative-selection magnetic bead kits (STEMCELL Technologies). The T cells were then activated on plates coated with anti-CD3 and anti-CD28 antibodies at 5 μg/ml each. After 48 h, activated cells were transferred onto new plates at a density of 1 × 10⁶/ml and supplemented with 20 U/ml IL-2. The cells were used in SLB experiments 5 to 12 days later.

SLBs.

SLBs were prepared as described previously (35), with some modifications. Liposomes containing 1,2-dioleoyl-sn-glycero-3-phosphocholine (Avanti Polar Lipids) supplemented with 12.5% 1,2-dioleoyl-sn-glycero-3-{[N-(5-amino-1-carboxypentyl)iminodiacetic acid]-succinyl} nickel salt (Ni-NTA-DOGS) and 0.005% 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(Cap biotinyl) were applied to clean coverslips placed onto sticky-Slide VI 0.4 chambers (Ibidi). After blocking with 5% casein containing 100 μM NiCl₂, His₁₂-tagged HIV gp120 SF162 (250/μm²) and ICAM-1 (200/μm²) were added to bind Ni-NTA-DOGS, while monobiotinylated OKT3 (30/μm²) was attached to Cap-biotinyl. All proteins were also fluorescently labeled prior to addition to the SLBs. The density of HIV gp120 tested on the SLBs was established in our previous studies (28). For comparison, the estimated Env density on the surface of HIV virions is 8 to 14 Env spikes per virion (36, 37), which correspond to 250 to 450/μm² for virions with a diameter of ∼100 nm. On the cell surface, the HIV Env glycoproteins are expressed in patches (38, 39), and the confocal microscopy measurements of the local densities of HIV gp120 at such patches on the HIV Env-transfected cell surfaces yielded an average density of 369 (median, 287; range, 0 to 2,490) molecules/μm² (reference 28 and data not shown). The ICAM-1 and OKT3 densities on the SLB were based on the respective expression levels of ICAM-1 and TCR ligand on the surfaces of APCs (28). The density of each protein was determined prior to use in SLBs by using the same SLB preparations to coat 5-μm-diameter silica beads and analyzing the beads by flow cytometry with calibration beads (Bangs Laboratories Inc., IN). The SLBs were then washed and incubated with 0.5 × 10⁶ cells in 200 μl of HEPES-buffered saline containing 1% human serum albumin, 2 mM MgCl₂, and 1 mM CaCl₂.

Immunofluorescence staining was used to detect total and phosphorylated Lck at the immunological synapse. At the time points designated, cells on SLBs were fixed with 2% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with antibodies specific for Lck phosphorylated on Y394 (R&D Systems) or Y505 (BD Transduction Laboratories) or total Lck (BD Phosflow), followed with an appropriate Alexa Fluor 546-labeled secondary antibody (Invitrogen, CA).

For antibody-blocking experiments, anti-gp120 MAbs obtained from Susan Zolla-Pazner (Icahn School of Medicine at Mount Sinai) and Miroslaw K. Gorny (NYU School of Medicine) were used (40–42). SLBs containing OKT3, ICAM-1, and gp120 were treated with 10 μg/ml MAb for 30 min prior to introduction of the cells and washed three times with the assay buffer to remove excess MAb. MAbs 654, 2158, and 2558 have potent neutralizing activities against HIV-1 SF162, from which the gp120 protein used in this study was derived (50% inhibitory concentrations of <1, 6, and <1 μg/ml, respectively [43]). Control MAb 1418 is specific for parvovirus and does not neutralize HIV.

Total internal reflection fluorescence microscopy.

T cells were loaded onto SLBs and allowed to interact for approximately 1 h or for the times indicated. Images were acquired either from live cells or after fixation and staining. Fluorescence microscopy and interference reflection microscopy (IRM) were performed with a Nikon Ti microscope with a total internal reflection fluorescence module and an Andor DU897 back-illuminated electron multiplier electron-multiplying charge-coupled device camera. Solid-state lasers provided illumination at 405, 488, 561, and 641 nm, and narrow-pass filters were used for detection.

Image analysis.

Acquisition settings were maintained constant throughout each imaging procedure and between samples. Image analysis was performed with ImageJ. To measure intensities, individual-cell contacts were traced with the region of interest (ROI) on the IRM channel. The background was subtracted from the images, and then cells in the subtracted images were marked with ROIs.

The densities of gp120 were calculated by measuring the fluorescence intensity values from the Alexa Fluor 488 (AF488)-gp120 channel within each cell and subtracting the background values from SLBs with no gp120. The gp120 densities on the OKT3+ICAM-1+gp120 SLBs in the areas with no cell contacts were normalized to 250 molecules/μm². The densities of OKT3 were measured on the basis of the fluorescence intensity values from the OKT3-AF647 channel minus the background values. The OKT3 background values were calculated from the zero time point and normalized to 30 molecules/μm².

Measurement of signaling molecules in cells on the SLBs was done on the entire cell-SLB contact area as detected in the IRM channel (29). Average fluorescence intensity was measured in each of the areas designated above. The background fluorescence was subtracted from the average intensity. The integrated fluorescence intensity was then calculated by multiplying the average intensity by the total pixel area measured in each cell. The pixel area was converted to square micrometers on the basis of the pixel/square micrometer ratio for the specific cameras and objectives that were used.

CD69 upregulation and IL-2 secretion.

CD4 T cells were activated with anti-CD3 and anti-28 MAbs as described above and used at least 7 days postactivation. CD4 T cells (0.5 × 10⁶ per condition) were incubated with silica beads 5 μm in diameter (Bangs Laboratories), the standard bead size for T cell stimulation and expansion (44). The beads were coated with SLBs containing OKT3 and ICAM-1 with or without CD80 or HIV gp120. After 6 h, cells were stained with phycoerythrin (PE)-conjugated anti-human CD69 antibodies (eBioscience) and analyzed with a FACSCalibur (BD). At 24 h, supernatants were collected and IL-2 concentrations were measured with a human IL-2 enzyme-linked immunosorbent assay (ELISA) kit (Thermo Scientific).

T cell proliferation and HIV replication.

To measure proliferation, CD4 T cells freshly isolated from peripheral blood were labeled with 5(6)-carboxyfluorescein N-hydroxysuccinimidyl ester (CFSE) and cultured on wells coated with OKT3 with or without HIV gp120. Anti-CD4 MAb Leu3A was also tested to block CD4 interaction with HIV gp120. After 5 days, cell proliferation was measured on the basis of CFSE dilution by flow cytometry and the data were analyzed with FlowJo. These stimulated CD4 T cells were also assessed for the ability to support virus replication. After culture with OKT3, OKT3+HIV gp120, or OKT3+anti-CD28 MAb for 5 days, the cells were infected with HIV-1 SF162 (multiplicity of infection of 0.1 as measured in MT4-TMZ-R5 reporter cells), washed, and incubated in the presence of the same stimuli for an additional 10 days without the addition of exogenous cytokines. Virus production was monitored in the supernatants at days 3 and 10 postinfection by p24 ELISA (XpressBio).

Statistics.

Graphs were drawn and statistical analyses were performed with Microsoft Excel and GraphPad Prism. Experiments were performed at least three times.

RESULTS

CD4 T cell immunological synapse formation and TCR recruitment to the synapse in the presence of HIV Env gp120.

In order to study the effect of the HIV Env on TCR-induced immunological synapse formation in primary human CD4 T cells, we used the SLB as an experimental model. This system enables the real-time visualization of microclusters and SMACs as they form in the TCR-induced immunological synapse and in the TCR-independent HIV Env gp120-induced virological synapse (18, 19, 28, 29). The SLBs were constituted with the anti-CD3 MAb OKT3, ICAM-1, and HIV Env gp120 to represent the surface of APCs expressing the HIV Env. Recombinant monomeric HIV-1 (SF162) gp120 with a His₁₂ tag at its C terminus was labeled with the fluorescent dye AF488 and attached to the SLB via Ni²⁺-chelating NTA lipids. ICAM-1 was also tagged with His₁₂ but was labeled with Alexa Fluor 405. Anti-CD3 MAb OKT3 was monobiotinylated, labeled with Alexa Fluor 647, and reacted sequentially with streptavidin and the biotin-CAP present in the SLBs. Activated CD4 T cells were then added to the SLBs, and live images were acquired by total internal reflection fluorescence microscopy (TIRFM) for 1 h (Fig. 1). The previously activated CD4 T cells were used to allow the evaluation of a relatively homogeneous population of primary target cells that are permissive to HIV infection and form well-defined immunological synapses on SLBs (18, 19, 28, 29).

FIG 1.

Recruitment of HIV Env gp120 and TCR to CD4 and CD8 T cell immunological synapses. (A, B) Activated CD4 T cells were introduced to SLBs containing fluorescence-tagged anti-CD3 antibody OKT3, ICAM-1, and gp120 or no gp120. Images of the cell contact area (as measured by IRM) and fluorescence signals were collected for 1 h. (A) Representative images from the 10-min time point are shown. Merge: combined fluorescence signals from OKT3 and gp120. White bar, 5 μm. (B) The distribution of gp120 (left) and TCR (right) in the CD4 T cell synapse at the 10-min time point. The densities of gp120 and OKT3 were measured radially from the center of each of 30 cells per experiment (white arrow around the circle in panel A merge, top). Data from representative cells are shown in the top right and left graphs. The averages and standard errors from experiments performed independently with cells from three different donors (>30 cells per experiment) are shown in the bottom right and left graphs. P values were determined by the Mann-Whitney test. (C, D) Activated CD8 T cells were added to SLBs presenting anti-CD3 MAb OKT3, ICAM-1, and gp120 or no gp120. Live images were acquired for 1 h. Data from the 10-min time point were analyzed. (C) Images from representative CD8 T cells on SLBs with or without gp120. Merge: combined fluorescence signals from gp120 and OKT3. (D) The densities of gp120 (left) and OKT3 (right) were measured along the cell radius as described above. Data from representative cells are shown in the top right and left graphs. The averages and standard errors from three different donors were similarly calculated and are shown in the bottom right and left graphs. ns, not significant by the Mann-Whitney test.

Within 10 min after the CD4 T cells had interacted with OKT3 and ICAM-1 on the SLBs, typical immunological synapses displaying the TCR-rich cSMAC and the ICAM-1 pSMAC rings were assembled in the absence of the HIV Env (Fig. 1A, top). In the presence of HIV gp120 on the SLBs, the cells also formed synapses with segregated cSMAC and pSMAC (Fig. 1A, bottom). These synapses were similar to those formed on SLBs without gp120. Notably, HIV gp120 accumulated in the center of the synapse and coalesced together with TCR at the cSMAC (Fig. 1A). The relative distribution of HIV gp120 and TCR in the synapse was quantified by determining the densities of gp120 and OKT3 along the cell radius. Data were acquired at the 10-min time point from a minimum of 30 cells per condition. These data confirmed the visual observation in Fig. 1A and demonstrated an increase in gp120 density from an initial level of 250 molecules/μm² on the SLBs to 400 molecules/μm² at the center of the cell (Fig. 1B left). Measurements also showed an increase in OKT3 density toward the center of the synapse and an accumulation of OKT3 at the cSMAC that was ∼3-fold higher (P < 0.0001) in the presence of HIV gp120 (Fig. 1B right), indicating that HIV gp120 may help augment TCR recruitment and clustering at the cSMAC.

For comparison, we also evaluated the effect of HIV gp120 on the immunological synapse formed by CD8 T cells. Activated CD8 T cells were applied to the SLBs containing OKT3 and ICAM-1 with or without gp120. The CD8 T cells formed immunological synapses with the typical bull's eye appearance comprising the TCR-rich cSMAC and the ICAM-1 pSMAC ring on the SLBs with or without gp120 (Fig. 1C, top and bottom). However, little gp120 was recruited to the centers of the immunological synapses formed by the CD8 T cells (Fig. 1C, bottom, and D, left). The OKT3 accumulation at the c-SMAC was also not significantly increased (Fig. 1D, right). These results indicate that HIV gp120 binds primarily to CD4 T cells and augments TCR accumulation in the CD4 T cell immunological synapse.

To further examine the alteration of TCR recruitment by HIV gp120, we compared the kinetics of OKT3 signal accumulation in the synapse over time in the presence versus the absence of gp120. CD4 T cells were introduced to SLBs bearing OKT3, ICAM-1, and gp120 or with SLBs with only OKT3 and ICAM-1, and OKT3 signals were measured for 50 min. In the absence of gp120, OKT3 signals increased to a plateau of approximately 60 to 70 molecules/μm² at 7 min (Fig. 2A and B; see Movie S1 in the supplemental material). This plateau was then maintained for up to 50 min. In contrast, when HIV gp120 was present, OKT3 signals increased more dramatically, peaking at 150 molecules/μm² at 7 min, but subsequently declined over time (Fig. 2A and B; see Movie S2 in the supplemental material).

FIG 2.

Kinetics of TCR accumulation in the CD4 T cell immunological synapse in the presence or absence of HIV Env gp120. Activated CD4 T cells were introduced to SLBs containing OKT3 and ICAM-1 with or without gp120. Live images were collected for 1 h. (A) Representative images from different time points during the 1-h observation period. Times (green numbers) are in minutes. IRM was used to determine cell contact area. Merge: combined fluorescence signals from gp120 and OKT3. White bar, 5 μm. (B) The kinetics of OKT3 recruitment into the synapse in the cells interacting with gp120 or no gp120 on SLBs. OKT3 densities were analyzed for ∼30 individual cells per condition for up to 50 min. The averages and standard deviations from all cells analyzed are shown in the left graphs, while data from randomly selected individual cells are shown in the right graphs. The data from two donors are presented. Statistical analysis was done with a two-tailed nonparametric Wilcoxon test comparing OKT3 recruitment in the presence versus the absence of HIV gp120. (C) Recruitment of gp120 to the synapse over time in cells interacting with gp120, OKT3, and ICAM-1 on SLBs. gp120 densities were analyzed as performed for OKT3. The left half shows the averages and standard deviations from ∼30 cells, and the right half shows the patterns displayed by representative individual cells. The data from two donors tested in independent experiments are shown.

On the SLBs with OKT3, ICAM-1, and gp120, the kinetics of gp120 accumulation was also evaluated. The HIV gp120 density increased to 400 molecules/μm² in approximately 10 min and continued to rise at a lower rate (Fig. 2C; see Movie S2 in the supplemental material). The gp120 signals fluctuated during the 50-min observation time, but they did not show a steady decline as seen with OKT3. In fact, from 20 to 50 min, the gp120 signals became more prominent than the OKT3 signals. Hence, the pattern of TCR accumulation did not follow that of gp120. Taken together with the data in Fig. 1, these data demonstrate that the presence of HIV gp120 in the CD4 T cell immunological synapse does not prevent synapse formation. However, HIV gp120 is recruited to the cSMAC along with the TCR, albeit with distinct patterns, and causes a dramatic change in the kinetics of TCR recruitment and retention in the synapse.

Effects of anti-gp120 antibodies on the recruitment of HIV gp120 and the TCR to the synapse.

To determine the interactions involved in HIV gp120 recruitment to the CD4 T cell immunological synapse, we pretreated SLBs containing gp120, OKT3, and ICAM-1 with human MAbs directed to different regions of gp120. The CD4 T cells were then added and allowed to interact with the SLBs. After 20 min, cells were fixed and HIV gp120 densities in the synapses were measured. The results showed that MAbs against the CD4-binding site (654), the V1V2 domain (2158), and the V3 loops (2558) significantly reduced the densities of HIV gp120 at the synapses, while the irrelevant control MAb (1418) did not have this effect (Fig. 3A). Representative images of the individual cells further showed that the anti-gp120 MAb treatment decreased the amounts of HIV gp120 at the synapses but did not block clustering (Fig. 3C). The location of the HIV gp120 clusters near the TCR-rich cSMAC also remained unchanged. Notably, low levels of gp120 interactions remained after treatment with the MAbs, including MAb 654, which blocks CD4-gp120 binding. These findings are similar to those of our past study in which the MAb against the CD4-binding site also did not completely abrogate the CD4 T cell interactions with HIV gp120 on SLBs presenting only gp120 and ICAM-1 (28). The data indicate that, in addition to CD4, other receptors are also engaged by HIV gp120 at these synapses and these interactions cannot be completely blocked by a single anti-gp120 MAb. Nonetheless, the non-CD4 receptors and their relative contributions have yet to be delineated.

FIG 3.

Reduction of HIV Env gp120 versus TCR accumulation by human anti-gp120 MAbs. Activated human CD4 T cells were introduced into SLBs with gp120, anti-CD3 MAb OKT3, and ICAM-1 pretreated with anti-gp120 MAbs that target the CD4-binding site (654), V2 (2158), and V3 (2558) or with an irrelevant MAb (1418). The MAbs were used at 10 μg/ml, a saturating concentration for gp120 binding on SLBs (28). Cells were fixed at 20 min, and images were collected. At least 30 cells were analyzed for each condition. Statistical analysis was performed by one-way ANOVA. Data from one of three independent experiments are shown. (A) Densities of gp120 in cell contact areas. ****, P < 0.0001; ***, P < 0.001; ns, not significant. Red bars represent mean values. (B) Densities of OKT3 in cell contact areas. ***, P < 0.001 (comparison with gp120+OKT3+ICAM plus or minus MAb); ns, not significant. Red bars, mean values. (C) Representative images show the distribution of gp120 and OKT3 with or without anti-gp120 MAb. IRM was used to determine the cell contact area. Merge: combined fluorescence signals from gp120 and OKT3.

Measurements of OKT3 density showed higher levels of OKT3 at the synapses in the presence of HIV gp120 (Fig. 3B), which is consistent with data in Fig. 1 and 2. Pretreatment with the anti-gp120 MAbs did not significantly affect the enhanced OKT3 accumulation. The OKT3 densities were slightly reduced by MAbs against the CD4-binding site (654) and the V3 loop (2558), but the decrease did not reach statistical significance. Hence, MAbs directed to the different sites of gp120 reduced HIV gp120 recruitment to the synapse without significantly curtailing TCR accumulation. These data suggest that the presence of very small amounts of HIV gp120 in the synapse is sufficient for increased OKT3 recruitment.

Effects of HIV gp120 on synapse morphology over time.

The ability of HIV gp120 to alter the kinetics of TCR accumulation and retention in the synapse suggests that HIV gp120 may affect synapse morphology over time. To evaluate this idea, we monitored the synapses 10 and 30 min after CD4 T cells made contact with SLBs presenting HIV gp120, OKT3, and ICAM-1 compared to the synapses on control SLBs containing only OKT3 and ICAM-1. On control SLBs, 70% of the CD4 T cells formed so-called mature immunological synapses within 10 min. These cells form stable contacts with SLBs, and the synapses displayed the hallmark central OKT3 cluster (cSMAC) in the center surrounded by a symmetrical peripheral ICAM-1 ring (pSMAC) (Fig. 4A, OKT3+ICAM-1, symmetrical). The ICAM-1 ring keeps the migratory T cells from crawling across the SLBs (17, 28). These synapses were maintained throughout the observation period; after 30 min, the percentage of cells with mature synapses remained at 75% (Fig. 4B). The percentages of cells with off-center TCR clusters and asymmetrical pSMAC structures also stayed at 25 to 30% throughout the observation period (Fig. 4A and B, OKT3+ICAM-1, asymmetrical).

FIG 4.

Alteration of CD4 T cell synapse morphology over time in the presence of HIV Env gp120. Activated CD4 T cells were introduced to SLBs containing OKT3, ICAM-1, and gp120 or no gp120 and monitored over time for changes in synapse morphology. (A). Representative images showing cells forming symmetrical versus asymmetrical synapses as determined by the OKT3 cluster position relative to the center of the cell. BF, bright field. Merge: merged images from gp120, OKT3, and BF. White bar, 5 μm. (B) The percentages of cells forming symmetrical or asymmetrical synapses at the 10- and 30-min time points on SLBs with or without gp120. Average values (shown by bar graphs) from independent experiments performed with cells from five different donors (depicted as scatter plots) are shown. *, P < 0.05 by one-way ANOVA with Dunn's multiple comparison.

In contrast, on SLBs with HIV gp120, OKT3, and ICAM-1, only 17% of the cells displayed mature synapses with central TCR clusters at 30 min, although at 10 min the proportion was comparable to that in the absence of gp120 (Fig. 4A and B, gp120+OKT3+ICAM-1, symmetrical). Hence, after 30 min, the number of cells with mature synapses on SLBs with gp120 declined. Most of the cells became elongated with off-center TCR clusters and no longer had well-defined ICAM-1 rings (Fig. 4A, gp120+OKT3+ICAM-1, asymmetrical), indicative of cells that had resumed migration. These data demonstrate that the presence of HIV gp120 not only alters TCR accumulation but also affects the propensity of cells to migrate after forming synapses.

Sustained Lck phosphorylation in the synapse in the presence of HIV gp120.

Given the ability of HIV gp120 to enhance TCR accumulation in the synapse, we next evaluated whether HIV gp120 affected the activation of TCR-proximal signaling. Since Lck activation is the first signal triggered upon TCR engagement and by gp120-CD4 binding (18, 29, 45, 46), we immunostained CD4 T cells interacting with OKT3 and ICAM-1 with or without gp120 on SLBs with anti-Lck antibodies. The cells were added to SLBs for 10 or 45 min, fixed, permeabilized, and stained with antibodies specific for total Lck, pLck (Y505), and pLck (Y394) (Fig. 5). TIRFM was then used to detect Lck phosphorylation and its recruitment to synapses. TIRFM allows the detection of fluorescence within a 100- to 200-nm slice and thus provides high lateral and axial resolution of Lck or pLck signals recruited specifically to the membrane-proximal areas at the T cell-bilayer interface (<200 nm of the T cell plasma membranes at cell-bilayer contact sites). The fluorescence signals beyond this thin area are excluded from the measurements. The data from individual cells from one of the donors are shown in Fig. 5A, and cumulative data from three different donors tested in separate experiments are presented in Fig. 5B.

FIG 5.

Altered Lck activation in CD4 T cell immunological synapse in the presence of HIV Env gp120. CD4 T cells were introduced to SLBs containing anti-CD3 MAb OKT3, ICAM-1, and gp120 or no gp120. After 10 or 45 min, cells were fixed and stained for total Lck, pLck (Y505), and pLck (Y394). The integrated fluorescence intensities detected by TIRFM were measured. (A) A total of 30 to 50 cells were analyzed for each condition, and the data from one representative donor are shown. (B) The average integrated intensities from independent experiments performed with cells from three different donors are shown. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P <0.05; ns, not significant. Red bars represent mean values. Iso Ctrl, staining with isotype control. Total and phosphorylated Lck staining at 10 and 45 min in the presence or absence of gp120 was significantly higher than in the isotype control (P > 0.05). Statistical analysis was performed with two-tailed t tests. (C) Images of representative cells on SLBs with or without gp120 to show the distribution of total Lck and phosphorylated Lck at two time points. Merge: merged gp120, OKT3, and Lck signals.

In the cells on control OKT3+ICAM-1 SLBs with no gp120, total Lck levels at the synapses were comparable at 10 and 45 min (Fig. 5A and B). In the presence of HIV gp120, a slightly higher level of total Lck staining was observed at 10 min, but overall, the total Lck staining remained relatively constant over time in the presence or absence of HIV gp120. In contrast, phosphorylated Lck staining (both pY505 and pY394) showed a declining trend from 10 to 45 min in cells on control SLBs. This is consistent with the fact that the physiologic Lck activation triggered via TCR is transient (18, 19, 46). In the presence of HIV gp120, however, pLck(Y505) staining showed an altered pattern: it increased from 10 to 45 min such that the pLck(Y505) staining at 45 min was higher in the presence versus the absence of HIV gp120. pLck (Y394) staining did not rise from 10 to 45 min but was sustained at higher levels at both time points in the presence of HIV gp120. Figure 5C (middle and right) further shows that, in the presence of HIV gp120, most pLck staining accumulated at the cSMAC after 45 min and localized near gp120 and OKT3. This pattern was not seen with the total Lck, which remained dispersed as small puncta throughout the cell-bilayer contact areas (Fig. 5C, left), indicating that the vast majority of Lck molecules were not phosphorylated and were not recruited to the membrane-proximal area of the gp120-containing cSMAC. In the absence of gp120, pLck and total Lck also did not converge to the cSMAC or colocalize with OKT3, consistent with past data (29).

These data indicate that, unlike the transient Lck activation induced by TCR in the peripheral areas of the immunological synapse, the presence of HIV Env gp120 in the synapse triggers a sustained level of phosphorylated Lck that accumulates in the synapse along with gp120 and OKT3 for up to 45 min. This pattern is reminiscent of the long-lasting Lck activation induced by HIV gp120 alone (29) and suggests the dominant effects of HIV gp120 over TCR on Lck activation in the CD4 T cell immunological synapse.

HIV gp120-mediated enhancement of CD4 T cell activation and virus replication.

The ability of HIV gp120 to augment TCR-induced Lck activation led us to investigate the downstream events associated with T cell activation. By itself, HIV gp120 only partially activates the early TCR signaling machinery on CD4 T cells, and the signals do not propagate to result in intracellular Ca²⁺ elevation or increased CD69 expression (29). However, the presence of HIV gp120 during TCR engagement may alter CD4 T cell activation. To test this idea, we compared the extents of CD69 upregulation, IL-2 secretion, and proliferation when CD4 T cells were triggered by OKT3 or by a combination of OKT3 and HIV gp120. To accommodate for the prolonged assay periods required to detect CD69 upregulation and IL-2 secretion beyond the 1-h period for the TIRFM experiments described above, SLBs were formed on standard 5-μm sterile silica beads; functionalized with OKT3, ICAM-1, and gp120 or no gp120; and then incubated with CD4 T cells for up to 24 h. SLB-beads coated with OKT3, ICAM-1, and CD80 (a costimulatory molecule binding to CD28) or with only ICAM-1 were also tested for comparison. The beads have been shown to provide stimulation to T cells equivalent to that of the planar surfaces (47). The cells were stained with anti-CD69 MAb at 6 and 24 h and analyzed by flow cytometry. The supernatants were collected at 24 h and tested for IL-2 by ELISA. Untreated cells served as a background control. T cell proliferation was also assessed on the basis of CFSE dilution 5 days after the cells were cultured in wells coated with OKT3 with or without gp120.

On CD4 T cells treated with OKT3+ICAM-1 beads, CD69 expression increased but declined rapidly (Fig. 6A). Stimulation with OKT3+ICAM-1 beads also induced no IL-2 production (Fig. 6B). With the addition of HIV gp120, CD69 upregulation was augmented and sustained over the 24-h observation period (Fig. 6A). These results were comparable to those achieved by OKT3-CD80 costimulation. HIV gp120 also significantly increased IL-2 production to the levels observed with OKT3-CD80 costimulation (Fig. 6B). Cells treated with ICAM-1-beads showed a response similar to that of untreated cells and did not upregulate CD69 expression or secrete IL-2. For a CD4 T cell proliferation assay (Fig. 6C), CFSE-labeled CD4 T cells were cultured for 5 days in wells coated with OKT3 with or without HIV gp120 (strains SF162 or CH040). Greater proliferation was observed with CD4 T cells stimulated with OKT3 plus HIV gp120 than in cells stimulated with OKT3 alone, and the enhanced response was inhibited by the Leu3A MAb that blocked HIV gp120-CD4 binding. These results indicate that HIV Env gp120 enhanced CD4 T cell activation to an extent similar to that of the bona fide costimulatory ligand.

FIG 6.

Enhanced TCR-induced activation in the presence of HIV Env gp120. CD4 T cells were incubated with silica beads coated with SLBs containing OKT3, ICAM-1, and gp120 or no gp120. Cells were also treated with beads coated with SLBs containing OKT3, ICAM-1, and CD80 or ICAM-1 alone for comparison. Untreated cells were used to establish background levels. CD4 T cells were initially activated on plates coated with anti-CD3 and anti-28 antibodies and tested >7 days after activation. (A) After 6 and 24 h of incubation with the SLB-coated beads, the cells were fixed, stained with anti-CD69 MAb, and analyzed by flow cytometry. (B) Culture supernatants were also collected after 24 h of incubation and tested by ELISA for IL-2 secretion. The averages and standard errors shown in the graphs are from three experiments performed independently with cells from different donors. (C) CFSE dilution assay of CD4 T cells cultured for 5 days in wells coated with OKT3 with or without HIV gp120 (strain SF162 or CH040). CD4 T cells freshly isolated from PBMCs were used. Anti-CD4 MAb Leu3a blocking CD4-gp120 binding was used to affirm gp120-mediated enhancement. Histograms from one representative experiment are shown. Division indexes were calculated from the histograms with FlowJo; the means and standard errors of two to five independent experiments with cells from different donors are shown. Statistical analysis was done with the Mann-Whitney test. (D) Virus production in cultures of CD4 T cells grown in wells coated with OKT3, OKT3-HIV gp120, or OKT3–anti-CD28 MAb for 5 days and infected with HIV-1 SF162. CD4 T cells freshly isolated from PBMCs were used. Virus p24 concentrations in supernatants were measured by ELISA on days 3 and 10 postinfection. Data from different donors are shown. Bkgrd: p24 concentrations in the supernatants of unstimulated cells. ****, P < 0.0001; **, P < 0.01; * P <0.05; ns, not significant.

To determine whether CD4 T cells stimulated by OKT3 in the presence of HIV gp120 can support virus infection better than those activated by OKT3 alone, we treated activated cells with HIV-1 SF162 and measured virus replication based on virus p24 in the culture supernatants. For comparison, p24 levels were also measured in the supernatants of unstimulated cells that supported minimal virus replication. These experiments were done without the addition of exogenous cytokines. The data show that the amounts of p24 produced by stimulated cells increased from day 3 to day 10 postinfection, indicative of active virus replication, although the concentrations varied among the different donors (Fig. 6D). Notably, higher levels of p24 were produced by infected CD4 T cells that had been stimulated with OKT3 in the presence of HIV gp120 than in those activated only with OKT3. The greater p24 production induced by OKT3 and HIV gp120 was detected as early as day 3 and became more evident at day 10. About 2- to 5-fold enhancement was attained in the absence of any exogenous cytokines and consistently observed with cells from the different donors. Importantly, virus replication after stimulation with OKT3 and HIV gp120 was comparable to or even higher than that seen in CD4 T cells stimulated with OKT3 and anti-CD28 MAb, indicating HIV gp120's potent costimulatory activity. All together, these results show that in the CD4 T cell immunological synapse, HIV Env may act as a costimulator that heightens the TCR activation of CD4 T cells and imparts to the cells the capacity to support higher levels of virus replication.

DISCUSSION

This study demonstrated the ability of HIV Env gp120 to alter the kinetics of TCR recruitment and the activation of TCR-proximal signaling at the CD4 T cell immunological synapse. While most earlier studies examined the changes in the immunological synapse formed by CD4 T cells already infected with the virus (48–50), this study evaluated the HIV Env interaction with target CD4 T cells in the immunological synapse prior to infection. Notably, the presence of the HIV Env did not interfere with immunological synapse assembly. TCR accumulation at the cSMAC of the synapse and its segregation from the adhesive pSMAC ring of ICAM-1–LFA-1 interactions proceeded similarly in the presence or absence of the HIV Env. The proportion of cells forming the synapse was also comparable. However, binding of the HIV Env to CD4 T cells resulted in accumulation of the HIV Env and phosphorylated Lck at the cSMAC, although the cSMAC is normally devoid of Lck in the absence of the HIV Env (18, 19, 31). Importantly, HIV Env accumulation at the synapse enhanced TCR-induced activation that predisposed the target CD4 T cells to become more receptive to HIV infection and productive replication. It should be noted that this activity was observed with the HIV Env gp120 monomers that cannot mediate virus infection. However, these monomeric gp120 proteins have receptor-binding activities. Moreover, when bound on membranes like the SLBs, the HIV Env gp120 monomers were able to cross-link the HIV Env receptors and cluster the HIV Env-receptor interactions to the cSMAC to result in Lck activation at the cSMAC. Past studies also have reported the activation of Lck and the downstream signals upon CD4 engagement by the HIV virions or the HIV Env expressed on cells (51–53). On the surfaces of virions and infected cells, heterogeneous species of the HIV Env glycoproteins are expressed (54–56), including nonfunctional Env proteins that cannot partake in the virus entry process but may have immunomodulatory activities demonstrated in this study.

The exact mechanism by which the HIV Env at the immunological synapse augments TCR-induced activation of CD4 T cells has yet to be delineated. However, the temporal pattern of TCR recruitment was significantly different in the presence of the HIV Env. Thereby, in the presence of the HIV Env, the TCR density in the synapse reached a higher peak (∼3-fold higher than that of the control) and declined over time. In contrast, in the absence of the HIV Env, the TCR level increased to a plateau that was maintained throughout the 1-h observation time. The alteration of TCR mobilization was induced by HIV Env binding to CD4 T cells. However, treatment with anti-Env MAbs directed to the CD4-binding site, the V1V2 domain, or the V3 loop reduced HIV Env accumulation in the CD4 T cell synapses without blocking enhanced TCR recruitment to the synapses. These data suggest the potential involvement of multiple receptors in affecting TCR recruitment and that a single anti-Env MAb is not sufficient to block this activity. We also observed that none of the individual anti-Env MAbs, including the 654 MAb efficiently blocking CD4-gp120 interaction, was able to totally prevent HIV Env recruitment to CD4 T cell synapses. Moreover, low levels of HIV Env were found to interact with CD8 T cells on SLBs, albeit without affecting TCR recruitment. Hence, in addition to CD4, other HIV Env receptors, such as the integrin α4β7 and the mannose receptors, are likely to be engaged at these synapses. Further, given that partial inhibition of Env recruitment was also seen with a MAb against the V3 loop known to be important for chemokine receptor binding (57), these coreceptors may also play a role in synapses. Nonetheless, the relative contributions of the different non-CD4 receptors remain to be determined.

The ability of the HIV Env to enhance TCR-induced T cell activation is rather unique and is in contrast to the inhibitory effects seen more commonly with other virus proteins. For example, the presence of HERV Env on dendritic cells reduces the ability of dendritic cells to form immunological synapses with T cells, resulting in blockage of T cell activation (58). Similarly, the Nef proteins of most SIV and HIV-2 strains have been shown to downregulate TCR and CD3 expression in virus-infected T cells, resulting in the suppression of immunological synapse formation with APCs (48). In contrast, HIV-1 Nef expression in infected CD4 T cells does not disrupt the formation of immunological synapses between infected CD4 T cells and APCs and instead affects synapse functions by inhibiting TCR-induced actin polymerization and TCR-proximal signaling (48, 59). The HCMV pUL125 protein also reduces the efficiency of immunological synapse formation by modifying the actin cytoskeleton required for the formation of immunological synapses with NK cells (60). Alternatively, the hRSV nucleoprotein interferes with the assembly of dendritic cell-CD4 T cell immunological synapses by decreasing TCR-pMHC binding and TCR-proximal signaling (61). The ability of these viruses to directly tamper with the immunological synapse and suppress immune cell activation serves as a mechanism of escape from immune surveillance. In contrast, HIV via its Env manipulates the immunological synapse to optimize the activation state of uninfected CD4 T cells to increase the capacity of these cells to support virus replication.

Although the significance of the HIV Env-induced alterations affecting the kinetics of TCR recruitment and retention in the immunological synapse is not fully understood, the data from this study show that TCR accumulation at the cSMAC was accompanied by HIV Env recruitment to the same central area of the synapse. In our earlier studies, HIV Env-CD4 interactions were also found to be mobilized to the synapse center to form a cSMAC-like structure independent of TCR engagement (28, 29). Importantly, the central Env-CD4 clusters were associated with long-lasting but partial activation of membrane-proximal signals that began with Lck phosphorylation (29). The present study also showed a similar pattern of durable Lck activation in the Env-containing TCR-induced immunological synapse. These observations indicate the dominant effect of the Env over transient TCR-induced signaling.

The presence of the HIV Env also enhanced TCR-induced CD4 T cell activation, as measured by CD69 upregulation and IL-2 secretion, and induced a greater level of CD4 T cell proliferation. These results are in concordance with a previous report by Zimmermann et al. (62) showing that simultaneous engagement of TCR by anti-CD3 antibody and CD4 by the HIV Env boosted CD4 T cell activation. In contrast, the interaction of CD4 T cells with soluble HIV gp120 retarded TCR-induced proliferation. HIV Env-CD4 interactions that occurred separately prior to TCR engagement were also found to undermine TCR-proximal signaling and downstream activation events (62). Similarly, pre-engagement of CD4 by HIV, followed by treatment with anti-CD3/CD28-coated beads, resulted in inhibition of Lck and F-actin recruitment to T cell immunological synapses (49). Taken together, these studies present evidence of the multifaceted role of the HIV Env in shaping the activation state and functionality of host CD4 T cells. It is also important to note that such HIV Env-mediated immunomodulation may play a part not only in the immune dysregulation seen during HIV infection but also in the inadequate elicitation and maturation of effector and memory immune responses to many of the HIV Env-based vaccine candidates evaluated thus far (63–65).

This study demonstrated the contribution of the HIV Env as would be expressed on virus-bearing APCs in augmenting the TCR-induced activation of cognate CD4 T cells that are targeted by the virus. We also observed that the HIV Env reduced the duration of the otherwise long-lasting immunological synapse, consistent with the transient nature of the T cell migration arrest induced by the HIV Env reported previously (28). A transient synapse may offer advantages for the virus. After a brief migration arrest to allow cell-cell contact and virus transfer, subsequent rapid detachment would prevent cell-cell fusion that terminates the virus life cycle while promoting the spread of newly infected cells to other sites in lymphoid tissues. Overall, the data presented here offer a model in which HIV, via its Env glycoproteins, alters the immunologic synapse and the activation state of CD4 T cells undergoing TCR activation in a way that is likely to facilitate viral replication and spread.

Funding Statement

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.