The Orientation of HIV-1 gp120 Binding to the CD4 Receptor Differentially Modulates CD4⁺ T Cell Activation

Healthy immune systems allow humans to cope lifelong with multiple persistent virus infections (1, 2). However, HIV-1 infection causes progressive deterioration of the immune system that ultimately leads to death due to opportunistic infections. Systemic immune activation as well as the continuous decline of a functional CD4⁺ T cell response are main contributors to disease progression (3–5). The capacity of HIV-1 to progressively debilitate the immune system is multifaceted and incompletely understood. For example, the rapid evolution of HIV-1 owing to its high mutation rate (6–8), as well as its potential to establish viral reservoirs in resting memory CD4⁺ T cells (9, 10), helps HIV-1 to evade functional immune responses. Furthermore, the HIV-1 surface envelope gp120 renders CD4⁺ T cells prime targets for infection (11, 12) via the high-affinity gp120–CD4 receptor interaction (13) and thereby enables HIV-1 uniquely among persistent viruses to directly interfere with CD4⁺ T cells, which are known to take a central position in adaptive immunity. Interestingly, the extent of qualitative (14) and quantitative (5, 15) defects within the CD4⁺ T cell population in HIV-1–infected individuals is disproportionally high in comparison with the levels of infectious virus and productively infected cells (16, 17), arguing against the fact that CD4⁺ T cell–directed viral cytopathicity is the major contributor to disease progression. However, HIV-1 replication, which is fundamental to disease progression, is not only associated with virion-associated envelope glycoprotein but also with soluble gp120 and gp160 due to shedding from the virus or from infected cells (18, 19). Consequently, soluble gp120 is present in the plasma (20–22) or lymphoid tissues of HIV⁺ patients (23, 24), and together with the appearance of dysfunctional CD4⁺ T cells, it is conceivable that binding of soluble gp120 to CD4⁺ T cells may be a central parameter in HIV-1 pathogenesis that could at least partially explain the high proportion of dysfunctional CD4⁺ T cells.

CD4 receptors play a crucial role in enhancing sensitivity of TCR-triggered T cell activation by interacting with MHC class II (MHC II) molecules on APCs (25, 26) and by their noncovalent interaction with the src family tyrosine kinase p56^lck whose activation initiates TCR signaling progression (26–29). In line with this, a plethora of in vitro studies provide experimental proof that gp120 binding to the CD4 receptor interferes with TCR-induced CD4⁺ T cell activation (30–47). However, the effect of noninfectious gp120–CD4 receptor interaction on CD4⁺ T cell activation is still a contradictory issue in the literature, and the gp120 levels measured in plasma of HIV⁺ patients may be below those that have functional effects on cells in vitro (48). On the one hand, gp120 interaction with CD4 receptors was suggested to enhance the activation of CD4⁺ T cells in terms of increased calcium signaling and IL-2R expression (44), transiently enhanced p56^lck activity (46, 47), activation of the transcription factors AP-1 (45), and elevated proliferation (49). On the other hand, gp120 binding to CD4⁺ T cells was shown to reduce their activation mediated through TCR stimulation (30–43). Mechanisms to explain gp120-induced negative effects on CD4⁺ T cell activation are manifold: gp120-mediated cross-linking of CD4 receptors was suggested to induce CD4 receptor endocytosis (33, 35, 36, 41, 46), to prevent peptide–MHC II–CD4 receptor interaction (37, 50) through interference of gp120 with the MHC II binding site on the CD4 receptor (51), to abolish proximal TCR signaling (32, 34, 38, 40, 52), or to redistribute p56^lck away from TCRs or from the immunological synapse (IS) (39, 53).

Based on these controversial reports and the possibility that gp120 binding to the CD4 receptor interferes with CD4⁺ T cell function, we investigated in detail whether and how non–virion-associated gp120 modulates TCR-induced CD4⁺ T cell activation. With the aim to mimic the in vivo situation as close as possible, we performed our in vitro experiments with primary CD4⁺ T cells and gp120–anti-gp120 immune complexes (ICs), as these are the predominant in vivo form of non–virion-associated gp120 (54, 55). To assess the impact of gp120–anti-gp120 ICs on CD4⁺ T cell activation, we stimulated cells either with agonistic anti-human CD3 mAb or with HLA class II–presented peptide Ags.

Our mechanistic study revealed that gp120–anti-gp120 ICs were rapidly and quantitatively transferred from APCs to CD4⁺ T cells. Importantly, following this process, IC-engaged CD4 receptors proved to substantially prevent subsequent TCR-mediated activation of CD4⁺ T cells. In contrast, CD4 and TCR cross-linking induced in vitro by immobilized gp120 and anti-human CD3 mAb on a planar substrate in close proximity promoted full CD4⁺ T cell activation. Thus, the microanatomical environment of how gp120 cross-links CD4 receptors appears to be crucial for the ensuing effect on CD4⁺ T cell activation. Our data thereby offer an explanation for the contradictory results of previous in vitro studies, reporting either enhanced or reduced CD4⁺ T cell activation upon gp120–CD4 receptor interaction. Furthermore, we demonstrate that CD4⁺ T cells from HIV-1 patients are covered with ICs that at least partly contain gp120, as also reported in Refs. 54–59, and therefore suggest that gp120-containing ICs impair CD4⁺ T cell activation and hence represent a crucial driving force of HIV-1 pathogenesis.

Patients were enrolled in the Swiss HIV Cohort Study (60) or the Zurich Primary HIV Infection Study at the Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich (http://clinicaltrials.gov, ID no. 5 NCT00537966) (61). Approval and written informed consent from patients were obtained according to the guidelines of the Ethics Committee of the University Hospital Zurich.

Citrate-phosphate-dextrose anticoagulated buffy coats of healthy human adults were purchased from the Swiss Red Cross (Blutspende Zürich, Schlieren, Switzerland), and EDTA anti-coagulated blood was obtained from HIV⁺ and HIV⁻ individuals from the University Hospital Zurich (Zurich, Switzerland). Viral load in HIV-1⁺ blood was determined at the Institute for Medical Virology at the University of Zurich (see Table I for detailed information). PBMCs were isolated by density centrifugation on lymphocyte separation media. Cells were washed in PBS and resuspended in RPMI 1640 supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine (all reagents from PAA Laboratories, Pasching, Austria), which is referred to as RPMI 10 throughout. When PBMCs were not directly used for experiments, they were cryopreserved in RPMI 1640 supplemented with 20% FBS and 10% DMSO (Sigma-Aldrich Chemie, Buchs, Switzerland). Cell viability was determined by 0.4% trypan blue exclusion (Invitrogen, Basel, Switzerland) and assessed to be >90%. CD4⁺ T cells and CD14⁺ monocytes were isolated from PBMCs using anti-CD4 and anti-CD14 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions.

Plasma samples from HIV-1 patients (see Table II for detailed information) were heat inactivated at 56°C for 60 min and stored at −80°C until processing. Plasma viral loads (copies/ml), IgG anti-gp120 (JR-FL) titer, as well as IgG anti-gp120–CD4 binding site (CD4BS) titer were determined as described previously (62).

For CFSE labeling, CD4⁺ T cells were incubated at a concentration of 5 × 10⁶ cells/ml in PBS supplemented with 10% FBS and 5 μM CFSE (Molecular Probes/Life Technologies Europe, Zug, Switzerland) for 11 min at 37°C. Afterward, cells were washed twice with cold RPMI 10 to quench residual CFSE and finally resuspended with RPMI 10 at an appropriate cell concentration.

ICs were prepared by incubating mammalian cell–derived recombinant HIV-1 gp120 JR-FL (endotoxin level < 1 endotoxin units/mg; Progenics Pharmaceuticals, Tarrytown, NY) or gp120 LAI together with non-CD4BS anti-gp120 mAbs [clone 2G12 (63) and clone 1-79 (64)] or with CD4BS anti-gp120 mAbs [clone b6 (65) and b12 (66)] at a molar gp120/anti-gp120 mAb ratio of ∼4:5 (final concentration of gp120 [f.c._gp120] 5 μg/ml) to prepare gp120–anti-gp120 IC and gp120–anti-gp120–CD4BS control IC, respectively. Gp120 LAI and anti-gp120 Abs 2G12, 1-79, b6, and b12 were provided by A. Trkola. For the generation of gp120 containing ICs with plasma from HIV-1⁺ donors, plasma from HIV-1⁻ and HIV-1⁺ donors and recombinant HIV-1 gp120 JR-FL were both diluted 1:100 and incubated for at least 2 h.

CD4⁺ T cells (2 × 10⁶/ml) were activated with plate-immobilized stimuli in a final volume of 100 μl RPMI 10. Plate immobilization of stimuli was done by incubating anti-human CD3 mAb (clone OKT3), anti-human CD28 mAb (clone CD28.2 or clone CD28.6), anti-human CD4 mAb (clone SK3; all from eBioscience, Vienna, Austria) recombinant gp120 JR-FL/LAI/CAAN (provided by A. Trkola) or ICs/control ICs (prepared as described previously) in a total volume of 100 μl PBS on F-bottom 96-well plates (Nunc Maxisorp; Sigma-Aldrich Chemie) for at least 18 h. Anti-human CD3 mAbs (50 ng/well) were used alone to induce subthreshold T cell activation or in combination with 0.8 μg/well gp120 or 1 μg/well anti-human CD28 mAb/CD4 mAb. To induce full T cell activation by the TCR alone, 1 μg anti-human CD3 mAb was coated on each well. For the stimulation of mouse CD4⁺ T cells, anti-mouse CD3ε mAb and anti-mouse CD28 mAb (both from BioLegend, Lucerna Chem AG, Lucerne, Switzerland) were used at similar concentrations as human-specific Abs. For internal assay controls, CD4⁺ T cells were left unstimulated or mitogenically stimulated with PMA (50 ng/ml) in combination with ionomycin (500 ng/ml, both from Sigma-Aldrich Chemie). When indicated, the gp120 binding site on the CD4 receptor was blocked by preincubation of the CD4⁺ T cells with anti-human CD4 mAbs (10 μg/ml, clone SK3) for 1 h at 37°C or actin polymerization was blocked by preincubation of the CD4⁺ T cells with cytochalasin D (50 μM, Sigma-Aldrich Chemie) for 30 min at 37°C prior to stimulation. For intracellular analysis of IL-2 expression, brefeldin A (10 μg/ml, Sigma-Aldrich Chemie) was added during the final 4 h of stimulation. Expression of CD40L, CD69, IL-2, IFN-γ, and TNF-α was analyzed after 6 h, and expression of CD25 and CD38 as well as proliferation by CFSE dilution after 5 d of stimulation at 37°C was as described below under “Flow cytometric analysis.”

CD4⁺ T cells (2 × 10⁵) were stimulated with autologous CD14⁺ monocytes (1 × 10⁵) in round-bottom 96-well cell culture plates in a final volume of 100 μl RPMI 10. Monocytes were loaded by sequential incubation of anti-human CD3 mAb (0.5–10 ng/ml, clone OKT3) for 30 min at 4°C or with CMV lysate (1:50, Virion, Rüschlikon, Switzerland) overnight at 37°C followed by incubation with ICs/control ICs (f.c._gp120 of 5 μg/ml) or gp120 containing ICs generated with plasma from HIV-1⁻ and HIV-1⁺ donors (final dilution factor 100–400) (as described for IC preparation) for 30 min at 4°C. Expression of CD69 was analyzed after 6 h, and expression of CD25 and CD38 as well as proliferation by CFSE dilution after 5 d of stimulation at 37°C was as described below under “Flow cytometric analysis.”

For the analysis of levels of phosphorylated and unphosphorylated linker for activation of T cells (LAT), cells were cultured in RPMI 1640 supplemented with 10% human serum (type AB), 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine (all reagents from PAA Laboratories). CD4⁺ T cells were preactivated for 2 d with plate-immobilized anti-human CD3 mAb (50 ng/well) in combination with anti-human CD28 mAb (1 μg/well), rested for 3 d in fresh medium, and ultimately stimulated with plate-immobilized stimuli as described above under “In vitro CD4⁺ T cell stimulation assay with plate-immobilized stimuli.” Cells (6 × 10⁵) were lysed in 30 μl cell lysis buffer (Cell Signaling Technologies/BioConcept, Allschwil, Switzerland) in the presence of phosphatase inhibitor tablets (Thermo Fisher Scientific, Reinach, Switzerland). Proteins were separated on a 10% SDS polyacrylamide gel and transferred to 0.2-μm pore size nitrocellulose transfer membranes (Schleicher and Schuel, Dassel, Germany). The amount of phosphorylated LAT was determined by polyclonal anti–phospho-LAT Ab (Tyr¹⁷¹, 1:1000) followed by HRP-conjugated goat anti-rabbit Ab (1:5000, Jackson ImmunoResearch Laboratories, Suffolk, U.K.) using Amersham ECL technology (GE Healthcare, Glattbrugg, Switzerland). To measure the level of nonphosphorylated LAT protein, the membrane was reprobed after stripping with a polyclonal anti-LAT Ab (Cell Signaling Technology). Signal intensities were quantified by ImageJ (National Institutes of Health, Rockville, MD). After background subtraction, the signal intensity of phosphorylated LAT was normalized to the amount of nonphosphorylated LAT, and fold increase of normalized levels versus the unstimulated controls was calculated.

For analysis of microtubule organizing center (MTOC) polarization in the context of plate stimulation, CD4⁺ T cells were incubated for 2 h on 12-mm round coverslips (Karl Hecht Assistent, Altnau, Switzerland) that have previously been coated for at least 18 h with anti-human CD3 mAb (50 ng/ml) in combination with ICs/control ICs (f.c._gp120 of 5 μg/ml) diluted in PBS containing 0.0025% poly-l-lysine (Sigma-Aldrich Chemie). MTOC polarization analysis in the context of cellular stimulation was performed as described above under “In vitro CD4⁺ T cell stimulation assay with autologous monocytes” except that stimulation was performed for 2 h only and cells were adhered onto poly-l-lysine (0.0025%) coated coverslips. After stimulation, cells were fixed with PBS containing 4% paraformaldehyde for 15 min at room temperature, followed by permeabilization (20% lysing solution [BD Biosciences, Allschwil, Switzerland] and 0.05% Tween 20 [National Diagnostics, Chemie Brunschwig, Basel, Switzerland]) for 10 min at room temperature. MTOC structures were visualized by staining with an anti–tubulin-β mAb (clone 9F3, Cell Signaling Technology/Bioconcept, Allschwil, Switzerland) overnight at 4°C in a humid chamber and Cy3-conjugated anti-rabbit ab (Jackson ImmunoResearch Laboratories) for 30 min at room temperature.

For the analysis of IC localization by immunofluorescence confocal microscopy, CD14⁺ monocytes were loaded with ICs/control ICs (as described above under “In vitro CD4⁺ T cell stimulation assay with autologous monocytes”) and cocultured for 30, 60, or 120 min with CD4⁺ T cells on poly-l-lysine (0.0025%; Sigma Aldrich Chemie)–coated coverslips. When indicated, CD4⁺ T cells were preincubated with anti-human CD4 mAbs (10 μg/ml, clone SK3; eBioscience) for 1 h at 37°C. ICs were visualized by staining with a Cy3-conjugated anti-human IgG Ab (Jackson ImmunoResearch Laboratories) for 30 min at room temperature. For the analysis of IC localization by flow cytometry, CD4⁺ T cells were cocultured with IC/control IC–loaded CD14⁺ monocytes or directly with gp120 containing ICs generated with plasma from HIV-1⁻ and HIV-1⁺ donors (final dilution of 1:100). After 6 h, IC localization was investigated by staining with an FITC-conjugated anti-human IgG (H+L) (Jackson ImmunoResearch Laboratories) as described below under “Flow cytometric analysis.”

After staining, cells were fixed with PBS containing 4% paraformaldehyde for 15 min at room temperature and briefly rinsed with distilled H₂O before mounting in VectaShield (Vector Laboratories, Burlingame, CA) containing 0.1% DAPI (Sigma-Aldrich Chemie) for visualization of nuclear DNA. Confocal immunofluorescence microscopy was performed with an inverted confocal microscope (Axiovert 200, Carl Zeiss) equipped with an oil-phase contrast objective (×63 oil objective, Plan Neofluar, 1.25 numerical aperture, Carl Zeiss) and a CSU-X1 spinning-disk confocal unit (Yokogawa) and a solid-state laser unit with four laser lines (405, 488, 561, an 647 nm; Toptica). Data analysis was done with Volocity 5.0.3 (Improvision, Coventry, U.K.). Quantification of IC staining was done by evaluation of the integrated signal density by ImageJ (National Institutes of Health), with values being normalized to cell size. Analysis of MTOC polarization was done by visual scoring as described in Fig. 6B, 6E.

Cells were surface stained with fluorochrome-conjugated anti-CD3 (clone UHCT-1), anti-CD4 (clone SK3), anti-CD8 (clone SK1), anti-CD14 (clone 61D3), anti-CD25 (clone BC96), anti-CD38 (clone HIT-2a), anti-CD40L (clone TRAP1), anti-CD69 (clone FN50), or anti-human IgG (H+L) (Jackson ImmunoResearch Laboratories) for 30 min at 4°C. Gp120 binding on CD4⁺ T cells was assessed by FITC (Sigma-Aldrich Chemie)-conjugated gp120 staining for 30 min at 4°C. For the analysis of intracellular IL-2 expression, surface staining was followed by a permeabilization step of 10 min at room temperature (20% lysing solution and 0.05% Tween 20) and staining with anti–IL-2 (clone MQ1-17H12), anti–IFN-γ (clone B27), and anti–TNF-α (clone 6401.1111) for 30 min at room temperature. Abs were from BioLegend, Lucerna Chem, or BD Biosciences unless specified otherwise. When staining was performed in whole blood (Fig. 7A–C), RBCs were lysed with lysing solution directly after staining for 10 min at room temperature. For compensation, a combination of anti-mouse Igκ/negative control compensation particles (BD Biosciences) or PBMCs were used. Cells were resuspended in PBS containing 1% paraformaldehyde before acquisition. Acquisition was done on an LSR II flow cytometer (BD Biosciences) using FACSDiva software. Samples were acquired on the day of the analysis and usually 1–3 × 10⁵ cells of interest were acquired. Doublets were excluded using side and forward scatter area and width parameters, and negative biological or fluorescence minus one controls were used to set gates. Data analysis was done with FlowJo software (Tree Star, San Carlos, CA). Cells were sequentially gated on lymphocytes, singlets, and CD3⁺CD4⁺ cells unless mentioned otherwise in the figure legend.

PBMCs were surface stained using anti-CD3 (clone UHCT-1), anti-CD4 (clone RPA-T4), anti-CD8 (clone SK1), and anti-CD14 (clone 61D3) for 30 min at 4°C. Abs were from BioLegend, Lucerna Chem, or BD Biosciences. Cells were isolated on a FACSAria (BD Biosciences) using FACSDiva software.

Data were analyzed and plotted with GraphPad Prism (GraphPad Software, La Jolla, CA) and results are illustrated as means ± SEM. Statistical analysis was performed as specified in the figure legend. A p value <0.05 was considered statistically significant. In the figure legends, n values indicate individual donors unless specified otherwise.

CD4 receptors play an essential role in TCR-mediated activation of CD4⁺ T cells through their intracellular association with the T cell–specific kinase p56^lck (26–29). Because gp120 binds with high affinity to the CD4 receptor (13), a potential modulating capacity of TCR-driven CD4⁺ T cell activation is attributed to gp120. Based on the importance of CD4⁺ T cells as central players of the adaptive immune system and the high availability of soluble gp120 in the plasma or lymphoid tissue of HIV-1 patients (20–24), we set out to investigate how non–virion-associated gp120 affects TCR-induced CD4⁺ T cell activation. In a first in vitro assay, CD4⁺ T cells from healthy donors were stimulated with plate-immobilized recombinant gp120 in combination with low amounts (50 ng/well) of agonistic anti-human CD3 mAb, which we previously determined to induce only marginal CD4⁺ T cell activation by itself. In combination, however, simultaneous CD4 receptor/TCR cross-linking by plate-immobilized gp120/anti-human CD3 mAb significantly enhanced CD4⁺ T cell activation compared with gp120 and anti-CD3 stimulation alone to a similar level as CD28-induced costimulation. Activation was determined by the expression of early T cell activation markers CD69 and CD40L within 6 h (Fig. 1A–D) and levels of proliferation were measured by CFSE dilution after 5 d (Fig. 1E, 1F). Importantly, CD4 receptor cross-linking by gp120 in the absence of concomitant TCR stimulation did not have any activating capacity on CD4⁺ T cells (Fig. 1A, 1B). Of note, a low level of CD4⁺ T cell activation is not only enhanced by R5-tropic HIV-1 envelope glycoprotein gp120 strain JR-FL but also by gp120 originating from the R5-tropic strain CAAN, the X4-tropic strain LAI, as well as by anti-CD4 mAbs (Fig. 1C), supporting the fact that TCR-induced CD4⁺ T cell activation is enhanced by concomitant CD4 receptor cross-linking rather than a gp120 intrinsic property. Ligation of CD4 receptors supports TCR signaling in a quantitative manner that allows the expression of activation markers and proliferation; however, cytokine expression in T cells critically depends on a qualitative, distinct costimulatory signal such as induced by the ligation by CD28 (67, 68). In line with a selective role of CD4 receptors in reinforcing proximal TCR signaling, IL-2, INF-γ, and TNF-α production was not robustly induced by gp120-mediated CD4 receptor cross-linking in contrast to CD28 costimulatory conditions (Fig. 1G, 1H). These data suggest that CD4 receptor cross-linking quantitatively supports TCR-induced signaling, yielding a high level of CD4⁺ T cell activation when signals are provided in close temporal and spatial proximity.

The capacity of gp120 to selectively enhance TCR-induced activation of human (Fig. 1) and not mouse CD4⁺ T cells (Fig. 2A), whose CD4 receptor does not bind gp120 (69), indicates that gp120–CD4 receptor interaction is crucial for gp120-mediated enhancement of CD4⁺ T cell activation and also excludes the possibility that impurities of the gp120 preparation influenced the observed CD4⁺ T cell activation. To substantiate the prerequisite of gp120–CD4 receptor interaction in promoting CD4⁺ T cell activation, we used in a next step Abs that specifically blocked the gp120 binding site on the CD4 receptor (70). As expected, gp120 binding to CD4⁺ T cells was reduced to background levels in the presence of the CD4 receptor blocking Ab (clone SK3) (Fig. 2B, 2C), and such blocking of the gp120 binding site on the CD4 receptor selectively abrogated anti-CD3/gp120– but not anti-CD3/anti-CD28–induced activation of CD4⁺ T cells (Fig. 2D). Furthermore, CD8⁺CD4⁺ T cells, which are present in the blood at low frequencies (71), were efficiently activated by gp120 in the presence of low amounts of anti-human CD3 mAb, whereas this was not the case for CD8⁺CD4⁻ T cells (Fig. 2E).

The CD4 receptor–associated p56^lck is crucial for proximal TCR signaling (26, 27, 72) by initiating a sequential cascade of tyrosine phosphorylation on signaling molecules, as for example on LAT, which is the nucleating site for multiprotein signaling complexes (73). We found that simultaneous CD4 receptor/TCR cross-linking by plate-immobilized gp120 and low amounts of plate-immobilized anti-human CD3 mAb induced slightly elevated levels of phosphorylated LAT, similar as with CD28 costimulatory conditions, in relationship to CD4⁺ T cells that received weak TCR stimulation alone (Fig. 2F, 2G).

These experiments support the notion that gp120 binding to the CD4 receptor not only represents the first step of viral entry, but it also has the potential to significantly increase CD4⁺ T cell activation by quantitatively supporting proximal TCR signaling.

Because gp120 (20–24) as well as anti-gp120 Abs (74–76) are present in HIV-1 patients, gp120–anti-gp120 ICs are likely to be generated and abundant in vivo (54, 55). We therefore generated such gp120–anti-gp120 ICs by coincubating recombinant gp120 with two different human IgG anti-gp120 mAb at a molar gp120/anti-gp120 mAb ratio of ∼4:5. As schematically shown in Fig. 3A, either two different anti-gp120 mAbs that block the CD4BS [clones b6 (65) and b12 (66, 77)] or anti-gp120 mAbs that spare the CD4 binding site (non-CD4BS) [clones 2G12 (63) and 1-79 (64)] were used for the formation of gp120–anti-gp120–CD4BS control IC (left) or gp120–anti-gp120 IC (right), which are henceforth referred to as control IC or IC, respectively. Consequently, the control IC, which contains anti-gp120 Abs that target the CD4BS, does not allow any interaction with the CD4 receptor and is not expected to have any functional effect on CD4⁺ T cells. Our results clearly showed that similarly to non–virion-associated gp120 alone, plate-immobilized IC significantly enhanced CD4⁺ T cell activation induced by weak TCR stimulation as measured by CD69 expression (Fig. 3B, 3C) after 6 h. In contrast, the control IC failed to do so (Fig. 3B, 3C). Additionally, proliferation of CD4⁺ T cells determined by CFSE dilution as well as the expression of late T cell activation markers CD25 and CD38 were significantly increased in the presence of IC compared with the control IC in combination with low level of TCR stimulation (Fig. 3D, 3E) after 5 d. Thus, gp120 within the context of ICs enhances a low level of TCR-driven CD4⁺ T cell activation in a CD4 receptor–dependent manner similar to gp120 alone.

In a next step, we set out to investigate whether the activating potential of gp120 can be translated to a more physiological situation. To mimic a condition in which ICs made of gp120 and anti-gp120 Abs interact with CD4 receptors during TCR-induced CD4⁺ T cell stimulation by APCs, CD4⁺ T cells from healthy donors were stimulated with autologous monocytes that had been previously incubated with gp120 containing ICs as well as anti-human CD3 mAb. Initially, flow cytometric determination of IC localization was performed by the analysis of human IgG binding after 6 h of coculture on CD4⁺ T cells or CD14⁺ monocytes. Control ICs were predominantly found on the CD14⁺ monocytes whereas the IC was almost exclusively found on CD4⁺ T cells (Fig. 4A). A similar localization pattern of control IC and IC was confirmed by the analysis of human IgG localization after 2 h by confocal microscopy (Fig. 4B). These experiments showed that only the IC and not the control IC was efficiently transferred to the CD4⁺ T cells from monocytes on which they were initially localized. The cell–cell transfer of ICs proved to be very fast and efficient, as already after 60 min most ICs were localized on CD4⁺ T cells (Fig. 4C). Furthermore, the transfer of ICs to CD4⁺ T cells was abrogated in the presence of an anti-human CD4 mAb, which blocks the gp120 binding site on the CD4 receptor (Fig. 4D, 4E).

We next investigated whether ICs, which were transferred from monocytes to CD4 receptors on CD4⁺ T cells, would also reinforce responsiveness toward TCR stimulation induced by anti-human CD3 mAb–coated monocytes, as this was the case when the TCR and CD4 receptor were cross-linked in close proximity at a defined site, as for example upon immobilization of the stimuli on a plate (Figs. 1–3).

Toward this end, CD4⁺ T cells were stimulated with monocytes that had been previously loaded with ICs that contain gp120 in combination with anti-human CD3 mAb. Gp120 within ICs either allows an interaction with the CD4 receptor (IC) or does not allow any interaction with the CD4 receptor (control IC). Interestingly, ICs that were transferred to CD4⁺ T cells, and consequently not only cross-linked CD4 receptors at the site of TCR engagement but on the whole CD4⁺ T cell surface, impaired CD4⁺ T cell activation induced by anti-human CD3 mAb–loaded monocytes in comparison with control conditions with control IC–loaded monocytes. The reduction was manifested by the expression of CD69 after 6 h (Fig. 5B, 5C) or CD25 and CD38 expression as well as proliferation measured by CFSE dilution after 5 d (Fig. 5D, 5E). Importantly, the inhibitory effect of IC on TCR-induced CD4⁺ T cell activation was not limited to polyclonal activation induced by anti-human CD3 mAb but also held true for CMV-specific CD4⁺ T cell activation induced by HLA class II–bound CMV peptides on monocytes, as the levels of CD25 and CD38 expression as well as proliferation of CMV-specific CD4⁺ T cells were significantly reduced in the presence of IC compared with control IC (Fig. 5F). The fact that IC-coated CD4⁺ T cells were still susceptible to mitogenic stimulation with PMA and ionomycin (Fig. 5G) implies that IC interferes with a proximal TCR signaling event rather than rendering CD4⁺ T cells completely incompetent to respond to activating stimuli in general.

Thus far, our data indicate that the way of CD4 receptor cross-linking by gp120 dictates in which way TCR-driven CD4⁺ T cell is affected. Whereas simultaneous TCR/CD4 receptor cross-linking confined to a specific site leads to enhanced CD4⁺ T cell activation, CD4 receptor cross-linking by IC not only at the site of TCR engagement manifested impaired responsiveness toward TCR-induced CD4⁺ T cell activation. We next aimed at elucidating how the manner of CD4 receptor cross-linking with respect to TCR engagement determines whether TCR signaling leads to enhanced or reduced CD4⁺ T cell activation.

TCR triggering induces a clustering of signaling complexes beyond the engaged TCRs that upon sustained signaling gives rise to the formation of IS (78–81). Such accumulation of surface and signaling molecules depends on dynamic reorganization of the actin- and tubulin-based T cell cytoskeleton (82–86), which is also reflected by the repositioning of the MTOC toward the site of stimulation (87, 88). CD4 receptors are essential in proximal TCR signaling (26–29) by recruiting p56^lck to the establishing IS (89, 90). Because many TCR signaling molecules (e.g., the CD4 receptor) are indirectly or directly linked to the cytoskeleton (91–93), this opens the possibility that gp120 binding to the CD4 receptor interferes with IS formation. Accordingly, when cytoskeleton dynamics were abrogated by disrupting actin polymerization by cytochalasin D (94), TCR signaling induced by concomitant CD4/TCR cross-linking at a defined site did not promote full CD4⁺ T cell activation, whereas a functional, dynamic cytoskeleton supported the accumulation of signaling complexes underneath the engaged TCRs such that local signaling strength resulted in full CD4⁺ T cell activation (Fig. 6A). Importantly, TCR- and CD4-independent stimulation by PMA and ionomycin was not dependent on a functional, dynamic actin cytoskeleton (Fig. 6A).

We therefore speculated that spatially and temporally linked engagement of TCR and CD4 (as in the case of plate-immobilized anti-human CD3 mAb and gp120) would favor IS formation. However, when CD4 receptors are cross-linked over the entire surface of the CD4⁺ T cell upon capturing ICs from monocytes, we suggest that the cytoskeletal dynamics that are required for IS formation might be impeded in such IC-decorated CD4⁺ T cells. To test this hypothesis, we investigated MTOC positioning toward the growing synapse as a marker for IS formation (87, 88) by staining for β-tubulin. Fig. 6B shows representative confocal z-stack images of CD4⁺ T cells stimulated by plate-immobilized anti-human CD3 mAb and gp120. The left lane shows a CD4⁺ T cell in which the MTOC is oriented toward the stimuli (referred to as polarized), whereas the right lane depicts a CD4⁺ T cell in which the MTOC is not polarized. Fig. 6E shows representative immunofluorescence images of CD4⁺ T cells that were stimulated by anti-human CD3 mAb–loaded monocytes. On the top, a situation is shown in which the MTOCs of both APC-engaging CD4⁺ T cells point toward the site of cell–cell contact (filled arrow, polarized), whereas the MTOC of the APC-engaging CD4⁺ T cell shown in the bottom image points away from the cell–cell contact (open arrow, not polarized). In line with the increased CD4⁺ T cell activation induced by plate-immobilized anti-human CD3 mAb and gp120 or IC (Figs. 1–3), MTOC polarization toward the site of TCR stimulation site was increased in the presence of IC (Fig. 6C, bottom, 6D) in contrast to the control IC (Fig. 6C, top, 6D). Note that MTOC polarization was defined as MTOC being localized in the cell hemisphere close to the planar surface. Because MTOC position is random in resting, unstimulated CD4⁺ T cells, it will by chance point toward the site of stimulation (i.e., the planar surface) in ∼50% of the cells (data not shown).

In striking contrast to the results with the plate-immobilized stimuli, reduced TCR-induced CD4⁺ T cell activation in the presence of IC on monocytes (Fig. 5) was associated with impaired MTOC polarization toward the CD4⁺ T cell–APC interface, whereas the presence of the control IC allowed efficient polarization of MTOCs toward the site of TCR engagement (Fig. 6F, 6G).

These data indicate that the manner of gp120/CD4 receptor cross-linking relative to the site of TCR engagement is crucial in the decision whether MTOCs can polarize to the site of engaged TCRs in a way that IS formation and hence CD4⁺ T cell activation are supported.

Based on the fact that gp120 is present in plasma from HIV-1 patients (20–22) and that HIV-1 infection is associated with robust Ab titers specific for HIV-1 surface structures (74–76, 95), gp120–anti-gp120 ICs are suggested to be the predominant form of gp120 in vivo (54, 55). Such complexes might either circulate or more likely be bound by FcR or complement receptor–expressing host cells. Based on the observation that monocyte-bound ICs were efficiently transferred to CD4⁺ T cells in our in vitro experiments (Fig. 4), we hypothesized that circulating CD4⁺ T cells from HIV-1 patients might be decorated with gp120-containing ICs. In line with this, we found that CD4⁺ T cells isolated from blood of HIV-1 patients (Table I) exhibited elevated anti-human IgG staining compared with CD4⁺ T cells isolated from blood of healthy controls (Fig. 7A–C), indicative of the presence of ICs on their surface. Because this difference was absent on CD8⁺ T cells (Fig. 7B), we speculated that the increased IgG signal on CD4⁺ T cells, at least partially, originates from gp120-containing ICs. These ICs on the surface of CD4⁺ T cells from HIV⁺ donors might indeed contain gp120 (as also suggested in Ref. 59), because staining with two human anti-gp120 mAbs in combination with an anti-human IgG Ab yielded increased IgG staining on CD4⁺ T cells from HIV⁺ donors compared with IgG levels without the addition of the two human anti-gp120 mAbs (Fig. 7C).

Finally, we addressed the question whether the repertoire of polyclonal anti-gp120 Abs present in the plasma of HIV-1 patients would support the formation of gp120–anti-gp120 ICs that would interfere with TCR-mediated CD4⁺ T cell activation. Toward this end, plasma from HIV-1 patients (Table II) was incubated with soluble gp120 to allow generation of gp120–anti-gp120 ICs, followed by adsorption to CD4⁺ T cells from healthy donors. Staining for anti-human IgG revealed that even in plasma from HIV⁺ donors with high titers of anti-gp120 Abs targeting the CD4BS, such ICs could be detected on CD4⁺ T cells, whereas plasma from healthy controls did not lead to the formation of ICs, which specifically bound to CD4⁺ T cells (Fig. 7D). When such gp120 ICs composed of polyclonal Abs from HIV⁺ patients, in combination with anti-human CD3 mAb, were loaded on CD14⁺ monocytes, the activation of autologous CD4⁺ T cells was significantly impaired, as shown by the analysis of CD69 expression after 6 h (Fig. 7E).

An exceptional peculiarity of HIV-1 is that its envelope surface glycoprotein has a very high affinity for the CD4 receptor (13), enabling HIV-1 and/or gp120 to directly interact and potentially interfere with CD4⁺ T cell function. Such direct interference with the function of central regulators of the immune system is unique among human persistent viruses and represents a crucial factor in undermining HIV-specific and heterologous immune control. CD4 receptor binding by the HIV-1 envelope mediates not only viral infection (11, 12), but it can also amplify TCR-induced signaling by recruiting p56^lck to the engaged TCRs (26–29, 89, 90), attributing the noninfectious interaction between gp120 and the CD4 receptor a potential role in modulating CD4⁺ T cell responses (30–47). Several studies demonstrated the presence of soluble gp120 in the plasma or lymphoid tissue of HIV-1 patients (20–24), but its effect on TCR-induced CD4⁺ T cell activation is still a contentious issue in the literature.

Our data clearly show that the spatial orientation of CD4 receptor cross-linking by gp120 relative to the site of TCR engagement substantially affects TCR-mediated activation of CD4⁺ T cells. Namely, simultaneous provision of plate-immobilized agonistic anti-human CD3 mAb together with gp120 promotes CD4⁺ T cell activation, indicating that concomitant cross-linking of CD4 receptors and TCRs at a confined site augments the signaling of the engaged TCRs, whose cross-linking alone would not allow optimal signal propagation required for inducing full T cell activation. Gp120 provided in a simultaneous manner with TCR stimulation does not increase CD4⁺ T cell activation in a gp120-intrinsic manner, because CD4 cross-linking by Abs similarly enhances stimulation of CD4⁺ T cells in the presence of low levels of agonistic anti-CD3 mAbs, indicating that any CD4 cross-linking modality would promote CD4⁺ T cell activation when provided in a temporally and spatially linked manner. Conversely, ICs adsorbed to CD4 receptors all around CD4⁺ T cells led to substantial impairment of CD4⁺ T cell activation upon interaction with APCs providing TCR engagement either by agonistic anti-human CD3 mAb or antigenic HLA class II–peptide complexes.

We suggest that the basis for gp120-mediated interference with TCR-induced CD4⁺ T cell activation relies on the fact that CD4 receptors not only interact with the TCR–CD3 complex and p56^lck but also with elements of the cytoskeleton (91), and consequently substantially interfere with the cytoskeletal-dependent rearrangement of signaling molecules toward the site of TCR stimulation, which ultimately regulates CD4⁺ T cell activation (82–86). This reasoning is supported by the observation that cross-linking of TCR–CD4 receptors in close proximity at a confined site only supports CD4⁺ T cell activation when the dynamics of the cytoskeleton are highly dynamic but not when actin polymerization is blocked by cytochalasin D. MTOC orientation toward the site of TCR engagement is one hallmark of a functional IS (87, 96), and our results clearly demonstrate that MTOC positioning is preferentially oriented toward plate-coimmobilized anti-human CD3 mAb and gp120. This indicates that cross-linking of CD4–TCR in close proximity allows cytoskeleton-directed accumulation of signaling molecules toward the site of engagement, thereby supporting full CD4⁺ T cell activation.

In contrast, IC binding to CD4 receptors of CD4⁺ T cells likely restricts cytoskeletal rearrangement by impairing lateral mobility of CD4 receptors and their cytoskeleton-associated framework. In line with this, we demonstrate that cross-linking of CD4 receptors by ICs not just at the site of TCR engagement hinders the accumulation of signaling units beyond the engaged TCRs, reflected by impaired MTOC orientation toward the site of TCR stimulation and hence compromised T cell activation. These results attribute gp120 a role in interfering with CD4⁺ T cell activation at the level of IS formation, which is in line with a previous report that describes defective recruitment of p56^lck to the IS upon CD4 receptor engagement prior to TCR engagement (39).

Although many studies agree on a role of gp120 in negatively impacting CD4⁺ T cell activation (30–41, 46, 50, 52, 53), the underlying mechanism was so far not clearly elucidated. Gp120 binding to CD4 receptors was described to induce their downregulation and consequently impairing CD4–TCR complex assembly required for CD4⁺ T cell activation (33, 35, 36, 41, 46). Our results did not confirm CD4 receptor downregulation upon exposure to ICs within the time of analysis (data not shown). Furthermore, our data cannot confirm that gp120 prevents CD4⁺ T cell activation by interfering with MHC II–CD4 receptor interaction required for optimal CD4⁺ T cell activation (37, 50), because we observed the inhibitory effect of gp120 on CD4⁺ T cells activated upon MHC II–independent TCR engagement by anti-human CD3 mAb. Gp120 binding was suggested to interfere with proximal TCR signaling (32, 34, 38, 40, 52, 97), being possibly a direct consequence of impaired IS formation and consequent TCR signaling propagation, as demonstrated by our results. Moreover, we exclude that ICs manifest reduced CD4⁺ T cell activation due to TCRs being sterically hindered from contacting their activating ligands on the APC, because both the control IC and IC were localized comparably at the site of cell–cell contact, and only the latter ones affected CD4⁺ T cell activation.

HIV-1 replication is one major driving force of progressive immunodeficiency (3). One constituent of HIV-1, namely gp120, has the capacity to directly modulate CD4⁺ T cell function, as shown previously (30–47) and in the present study. This effect has mostly been shown in vitro using relatively high concentrations of gp120, which leads to the question whether these effects might similarly operate in an in vivo scenario. Gp120 can be shed from the virion or from infected cells (18, 19) and is detectable in the plasma of HIV-1 patients, albeit there is some uncertainty about the actual gp120 concentrations (48). Levels in the plasma, however, might be much lower compared with those in lymphoid tissues where gp120 is thought to accumulate (23, 24). Thus, it is conceivable that gp120 availability in vivo might be sufficient to affect CD4⁺ T cell function. The robust Ab response against the envelope surface structure of HIV-1 (74–76, 95, 98) results in the formation of gp120–anti-gp120 ICs, which are the most abundant form of gp120 in vivo (54, 56). Consistent with this notion, we and others (54–58) have demonstrated that CD4⁺ T cells isolated from the blood of HIV-1 patients are covered with ICs, based on increased IgG deposition on CD4⁺ T cells compared with healthy controls. Because no increase of IgG binding was observed on CD8⁺ T cells, we speculate that the increased IgG signal on CD4⁺ T cells, at least partially, originates from gp120-containing ICs. Experimental proof is provided by us and others (59) by disclosing surface staining of gp120 on a substantial proportion of CD4⁺ T cells.

ICs that had been loaded on monocytes were rapidly and quantitatively transferred to CD4⁺ T cells in our in vitro experiments. We suggest that this transfer is due to the different relative binding affinities of ICs to CD4⁺ T cells or monocytes rather than on an active mechanism. Because the binding affinities between the IC and the Fc receptors on monocytes is weak compared with the high-affinity CD4–gp120 interaction (13), we speculate that ICs are captured by CD4⁺ T cells from the monocytes in a stochastic process in which the ICs eventually localize on the cells to which they exhibit increased binding affinity. This transfer was dependent on the CD4 receptor and consequently did not occur for the control IC or in the presence of a CD4 receptor–blocking Ab, and therefore we excluded that ICs bound to the CD4⁺ T cells in an unspecific manner. Accordingly, we suggest that ICs trapped on host cells via Fc or complement receptors as well as soluble gp120–anti-gp120 ICs might be effectively adsorbed on innocent bystander CD4⁺ T cells. This could provoke unresponsiveness to subsequent TCR-induced activation and thereby add to CD4⁺ T cell dysfunction during HIV-1 infection. Importantly, gp120–anti-gp120 ICs generated by the addition of gp120 to plasma from HIV-1 patients showed that the repertoire of in vivo circulating polyclonal gp120-specific Abs have the potential to generate gp120–anti-gp120 ICs that can be adsorbed by CD4⁺ T cells. Abs that target the evolutionary conserved CD4BS (99–102) might prevent gp120-containing ICs from binding to CD4 receptors and hence could block gp120-related functional impairment of CD4⁺ T cells similar to our control IC. Our data suggest, however, that the presence of high CD4BS Ab titers in polyclonal plasma from HIV-1⁺ donors does not prevent gp120–anti-gp120 IC formation upon gp120 addition and that the thereby generated gp120-containing ICs bind to CD4⁺ T cells and negatively affect anti-human CD3 mAb-induced CD4⁺ T cell activation. This might be explained by the effective ratio of CD4BS versus non-CD4BS anti-gp120 Abs in the circulation, with a low ratio still allowing the formation of gp120-containing IC capable of binding to CD4 receptors. Of note, plasma in which anti-gp120 Ab titers were below the detection limit did not support the formation of gp120-containing ICs on CD4⁺ T cells upon addition of gp120 (data not shown).

It should be considered in further studies whether gp120–anti-gp120 IC-coated CD4⁺ T cells not only contribute to CD4⁺ T cell dysfunction by means of impairing their TCR-induced activation but also by enhancing Ab-dependent cellular phagocytosis or Ab-dependent cellular cytotoxicity.

Overall, we demonstrate a central role of gp120–CD4 receptor interaction in impairing CD4⁺ T cell function and consequently potential hindrance of establishing functional immune responses as one factor in the developing progressive immunodeficiency during HIV-1 infection.

The level of unresponsiveness toward TCR-induced stimulation likely depends on the extent of IC deposition on CD4⁺ T cells. Although it is conceivable that the amounts of ICs we used in our in vitro assays might be at the upper limit or even higher compared with the average load of ICs detected on CD4⁺ T cells isolated ex vivo in peripheral blood, the levels of gp120 and gp120-containing ICs are higher in lymphoid tissues (23, 24), and consequently we would assume that the levels of ICs on CD4⁺ T cells might also be higher in these anatomic compartments. Beyond dispute, immune dysfunction that occurs in HIV-1–infected hosts is likely to be caused by multiple additional factors. Only very few studies compared the activation potential of CD4⁺ T cells isolated from HIV-1 patients to CD4⁺ T cells isolated from healthy controls, and a few of them indicated decreased in vitro lymphocyte proliferation/activation of CD4⁺ T cells (56, 103, 104). Daniel et al. (56) revealed that this could be attributed to the presence of gp120-containing ICs on the surface of CD4⁺ T cells. Along these lines, we demonstrate that HIV-1 gp120 directly influences CD4⁺ T cell activation, whereby the geometry of CD4 receptor cross-linking by gp120 dictates the functional outcome of TCR-mediated CD4⁺ T cell activation by differentially affecting IS formation. Although spatially and temporally linked engagement of CD4 receptors and TCR promote CD4⁺ T cell activation, dispersed CD4 receptor engagement in relation to TCR triggering signals results in the opposite outcome, namely reduced propagation of TCR signaling and hence CD4⁺ T cell activation. Our data thereby provide an explanation for the conflicting reports on gp120–CD4 receptor interaction on TCR-induced CD4⁺ T cell activation and suggest a relevant role for the HIV-1–derived gp120 protein in mediating general downmodulation of CD4⁺ T cell responsiveness in HIV-1–infected individuals aside the prolonged, nonphysiological activation of immune cells occurring in the context of HIV-1 infection (4, 105).

We are thankful to our patients for their commitment and specially thank C. Grube for patient care, N. Oetiker for experimental support as well as P. Rusert for providing us with anti-gp120 Abs and plasma samples from HIV-1 patients.

The authors have no financial conflicts of interest.