The inability to disrupt the immunological synapse between infected human T cells and APCs distinguishes HIV-1 from most other primate lentiviruses

Abstract

Viruses that infect T cells, including those of the lentivirus genus, such as HIV-1, modulate the responsiveness of infected T cells to stimulation by interacting APCs in a manner that renders the T cells more permissive for viral replication. HIV-1 and other primate lentiviruses use their Nef proteins to manipulate the T cell/APC contact zone, the immunological synapse (IS). It is known that primate lentiviral Nef proteins differ substantially in their ability to modulate cell surface expression of the TCR-CD3 and CD28 receptors critical for the formation and function of the IS. However, the impact of these differences in Nef function on the interaction and communication between virally infected T cells and primary APCs has not been investigated. Here we have used primary human cells to show that Nef proteins encoded by HIV-2 and most SIVs, which downmodulate cell surface expression of TCR-CD3, disrupt formation of the IS between infected T cells and Ag-presenting macrophages or DCs. In contrast, nef alleles from HIV-1 and its simian precursor SIVcpz failed to suppress synapse formation and events downstream of TCR signaling. Our data suggest that most primate lentiviruses disrupt communication between virally infected CD4+ Th cells and APCs, whereas HIV-1 and its SIV precursor have largely lost this capability. The resulting differences in the levels of T cell activation and apoptosis may play a role in the pathogenesis of AIDS.

Figure 1. Nef alleles that downmodulate TCR-CD3 impair the ability of virally infected primary T cells to form complexes with APCs.

(A and B) PBLs were infected with NL4-3–IRES–eGFP viruses encoding different Nefs, and complex formation with autologous SEE-pulsed DCs and MDMs was assessed 4 days after infection by fluorescence microscopy (A) and flow cytometry (B). (A) Representative acquisition frames of cocultures between infected PBLs and MDMs or DCs, with phase-contrast (Nomarski), eGFP, and PKH26 superimposed. Scale bars: 10 μm. The proportion of infected (eGFP⁺) cells contacting labeled autologous MDMs or DCs was scored randomly by single-blind method with n = 60 per sample per experiment. Results are mean ± SD of 4–5 independent experiments. In all experiments, 25%–55% of PBLs infected with the nef-defective control HIV-1 construct (nef–) formed complexes with autologous APCs; nef– values are set as 100%. (B) Representative flow cytometry acquisitions for complex formation between infected PBLs and MDMs or DCs. Percentages denote the number of eGFP⁺ cells in the DC gating that are PKH26⁺ (as a percentage of all eGFP⁺ cells). Graphs show mean ± SD of 3 and 4 experiments in DCs and MDMs, respectively. In all experiments, 3%–13% of PBLs infected with nef– formed complexes with autologous APCs; nef– values are set as 100%. Red and green bars denote nef alleles of group 1 and group 2, respectively. Combined results for group 1 (G1) and group 2 (G2) Nefs represent average values ± SD. *P < 0.05; ^#P < 0.01; ^§P < 0.001.

Figure 2. Effect of primate lentiviral Nefs on TCR-CD3 cell surface expression and clustering at the IS.

Primary CD4⁺ T cells were not infected (n.i.) or were infected with HIV-1 constructs expressing the indicated nef alleles. At 3 days after infection, the T cells were incubated with autologous sAg-pulsed DCs, fixed, and stained with anti-CD3. Images show representative confocal acquisitions from cocultures between infected CD4⁺ T cells and autologous sAg-pulsed DCs labeled with DAPI. Shown are CD3 Ab labeling alone and merged images of eGFP, DAPI, and CD3. Arrowheads point to zones of close contact between infected T cells and DCs. Close contact was rarely observed in T cell/APC cultures infected with viral constructs expressing group 1 Nefs, and these cells did not express CD3 at their surface. Scale bars: 5 μm.

Figure 3. Nef-mediated downmodulation of TCR-CD3 prevents Lck polarization at the IS.

(A) CD4⁺ T cells infected with HIV-1 constructs expressing the indicated nef alleles were incubated with autologous sAg-pulsed DCs, fixed, and stained with anti-Lck. Images show representative confocal acquisitions from cocultures between infected CD4⁺ T cells and autologous sAg-pulsed DCs labeled with DAPI. Shown are Lck Ab labeling alone and merged images of eGFP, DAPI, and Lck. Arrowheads point to zones of close contact between infected T cells and DCs. Scale bars: 5 μm. (B) Number of HIV-1–infected eGFP⁺ T cells engaged in complex formation with APCs, showing accumulation of Lck at the IS. Results are mean ± SD of 2 independent experiments. *P < 0.05; ***P < 0.001.

Figure 4. Nef alleles that downmodulate CD3 inhibit talin clustering at the IS.

(A) CD4⁺ T cells infected with HIV-1 constructs expressing the indicated nef alleles were incubated with autologous sAg-pulsed DCs, fixed, and stained with anti-talin. Images show representative confocal acquisitions from cocultures between infected CD4⁺ T cells and autologous sAg-pulsed DCs labeled with DAPI. Shown are talin Ab labeling alone and merged images of eGFP, DAPI, and talin. Arrowheads point to zones of close contact between infected T cells and DCs. Scale bars: 5 μm. (B) Number of HIV-1–infected eGFP⁺ T cells engaged in complex formation with APCs, showing polarization of talin at the zone of contact. Results are mean ± SD of 2 independent experiments (except for NA7, tNef, and SIVblu Nefs, which were derived from a single experiment). ***P < 0.001.

Figure 5. Nef alleles that downmodulate TCR-CD3 impair CD43 exclusion at the IS.

Images show representative confocal acquisitions from cocultures between infected PBLs and autologous SEE-pulsed DCs. Shown are CD43 Ab labeling alone and merged images of eGFP, PKH26, and CD43. Scale bars: 5 μm. Close contact was rarely observed in PBL/APC cultures infected with viral constructs expressing HIV-2 BEN or SIVblu Nefs, and occasional cell-cell complexes did not show CD43 exclusion, as determined by microscopic examination and analysis of the CD43 signal intensity along the plane of the cell-cell interface. The x axis represents distance along the contact zone; the y axis denotes pixel intensity. The location of 3–4 representative points per construct, and their corresponding intensities, are shown by the numbered arrowheads.

Figure 6. Downmodulation of IS formation perturbs early TCR-CD3 signaling events after coculture with APCs.

PBLs were allowed to return to a less-activated state after infection by removal of all cytokines in the growth medium for 3 days. PBLs were stimulated with CD3/CD28 beads for the indicated times or left unstimulated, and then labeled for intracellular phospho–ZAP-70 kinase. (A) Percentage of pZAP-70⁺ infected cells, shown as mean ± SD of 3 independent experiments. In all experiments, 5%–20% of PBLs infected with nef– were pZAP-70⁺ after 15 minutes’ stimulation with CD3/CD28 beads; nef– values are set as 100%. *P < 0.05; ***P < 0.001. (B) Kinetic of ZAP-70 phosphorylation in PBL cultures infected with HIV-1 constructs expressing the indicated nef alleles. Values represent percent ZAP-70⁺ cells and are representative of 2 independent experiments performed.

Figure 7. TCR-CD3 downmodulation leads to decreased IL-2 production after dynamic synapse interaction with primary APCs.

Infected PBLs were cocultured with autologous SEE-pulsed DCs or MDMs for 16 hours and then labeled for intracellular IL-2. Shown is the percentage of IL-2⁺ cells in the infected PBL population; results are mean ± SD of 3 independent experiments. In all experiments, 2%–7% of PBLs infected with nef– stained positive for intracellular IL-2; nef– values are set as 100%. *P < 0.05; ***P < 0.001.