HIV-1 tat promotes integrin-mediated HIV transmission to dendritic cells by binding Env spikes and competes neutralization by anti-HIV antibodies

Abstract

Use of Env in HIV vaccine development has been disappointing. Here we show that, in the presence of a biologically active Tat subunit vaccine, a trimeric Env protein prevents in monkeys virus spread from the portal of entry to regional lymph nodes. This appears to be due to specific interactions between Tat and Env spikes that form a novel virus entry complex favoring R5 or X4 virus entry and productive infection of dendritic cells (DCs) via an integrin-mediated pathway. These Tat effects do not require Tat-transactivation activity and are blocked by anti-integrin antibodies (Abs). Productive DC infection promoted by Tat is associated with a highly efficient virus transmission to T cells. In the Tat/Env complex the cysteine-rich region of Tat engages the Env V3 loop, whereas the Tat RGD sequence remains free and directs the virus to integrins present on DCs. V2 loop deletion, which unshields the CCR5 binding region of Env, increases Tat/Env complex stability. Of note, binding of Tat to Env abolishes neutralization of Env entry or infection of DCs by anti-HIV sera lacking anti-Tat Abs, which are seldom present in natural infection. This is reversed, and neutralization further enhanced, by HIV sera containing anti-Tat Abs such as those from asymptomatic or Tat-vaccinated patients, or by sera from the Tat/Env vaccinated monkeys. Thus, both anti-Tat and anti-Env Abs are required for efficient HIV neutralization. These data suggest that the Tat/Env interaction increases HIV acquisition and spreading, as a mechanism evolved by the virus to escape anti-Env neutralizing Abs. This may explain the low effectiveness of Env-based vaccines, which are also unlikely to elicit Abs against new Env epitopes exposed by the Tat/Env interaction. As Tat also binds Envs from different clades, new vaccine strategies should exploit the Tat/Env interaction for both preventative and therapeutic interventions.

Conflict of interest statement

Competing Interests: IKS, VAS, EK, JBU, RR and SWB are employees of Novartis Vaccines & Diagnostics, Inc. Novartis provided the recombinant HIV-1 Env proteins as a partner in several collaborative studies (the European projects AVIP and MUVAPRED, the “Joint Program ISS/Novartis (Chiron) for the development of a combined vaccine against HIV/AIDS.” There is one patent to declare: Novel Tat complexes, and vaccines comprising them (Inventor: Ensoli B.; Applicant: ISS (100%); priority date March 11, 2004; N. PCT/EP2005/003043, Publication N. WO2005/090391). There are no products in development or marketed products to declare. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors.

Figures

Figure 1. Virological outcome in Tat/Env-vaccinated or control monkeys after intrarectal challenge with the SHIVSF162P4cy (70 MID50).

Box plots of (A) Viral RNA, (B) proviral DNA in blood at 2, 3, 4 and 5 weeks after challenge, respectively; and (C) proviral DNA at week 4 after challenge in rectal tissue (RT) and inguinal lymph nodes (LN). Statistical analysis was performed by the one-sided Wilcoxon rank sum test. Red: monkeys vaccinated with Tat/Env (n = 6); blue: control animals (n = 6).

Figure 2. Enhancement by soluble Tat of HIV-1 infection in MDDCs.

Peak infection fold increase reached 3–8 days after infection with the R5 pSF162LUC in MDDCs from 10 different donors. The virus was pre-incubated with 1 µM cys22 Tat or with PBS-0.1% BSA (control buffer). Fold increase were calculated as the ratio between the RLU from MDDCs infected with virus pre-incubated with cys22 Tat and RLU from MDDCs infected with virus pre-incubated with control buffer (baseline RLU: 298.95) (both values had been previously subtracted of the uninfected MDDCs RLU background).

Figure 3. Tat-mediated entry of VLPs expressing R5 or X4 Env into MDDCs.

(A) and (B): entry in immature MDDCs of null-VLPs (control) or (A) VLPs expressing R5-Env (HIV-1 BaL) (VLP-R5Env) or (B) VLPs expressing X4-Env (HIV-1 HXBc2) (VLP-X4Env), in the absence or presence of increasing concentrations of cys22 Tat after 4 h of incubation, evaluated by flow cytometry. Results are shown as either dot-plot (A) or histograms (B). SSC: side scatter. Numbers in the insets represent the percentage of GFP positive cells. (C) Confocal microscopy analysis of VLP-R5Env entry into MDDCs in the presence of cys22 Tat (60 nM) after 4 h of incubation (0.2 µm optical sections form the apical to the basal cell side).

Figure 4. Tat released by producing T cells increases entry and infection of a Tat-independent replication-incompetent SF162 pseudovirus.

CEMss cells were either infected (CEMss-Tat) or not infected (CEMss) with VSV-G/HIV. After 24 h cells were co-cultured for 4 days with TZM-bl cells in the presence (TZM-bl/VP-SF162) or absence (TZM-bl) of a Tat-independent GFP-expressing single-cycle HIV-1 SF162 virus (VP/SF162), and then GFP expression evaluated by flow cytometry. Results are expressed as MFI fold increase with respect to the TZM-bl + CEMss co-culture (baseline MFI: 1.9).

Figure 5. Adherent Tat binds HIV particles and increases MDDC productive infection and transmission to T cells.

(A) Immuno-scanning electron microscopic analysis of HIV bound to Tat-coated latex beads. Images obtained with secondary electrons (left panels) or with backscattered electron (right panels) are shown. Latex beads were incubated with control buffer and HIV (A, B); with Tat and medium (C, D); or with Tat and HIV (E, F). Bar represents 5 µm. (B) MDDC infection with HIV SF162 on adherent wt Tat at different time points after infection. Infection of MDDCs from 2 different donors was carried out on plates coated with wt Tat (0.01 µM for donor 1, 0.1 µM for donor 2) or control buffer. Infection fold increase was calculated as the ratio between the content of p24 antigen detected in the supernatants of MDDCs seeded on adherent Tat and that detected in the supernatants of MDDCs seeded on control buffer (baseline, for donor 1, p24 was 154 and 119 pg/mL at day 7 and 10, respectively; for donor 2, p24 was 38 and 159 pg/mL at day 7 and 10, respectively). (C) Infection of PHA-activated CD4 T cell blasts with supernatants from MDDCs infected with HIV SF162 on plates coated with Tat wt (0.1 µM) or with control buffer (BSA). Infection was evaluated at the indicated time points by measuring p24 antigen in the culture supernatants in duplicate wells.

Figure 6. Anti-integrin antibodies block the enhancement of MDDC infection induced by soluble or adherent Tat.

(A) pSF162LUC infection of MDDCs for 8 days in the presence of soluble cys22 Tat (0.1 or 1 µM) and a combination of mAbs directed against the α5β1, αvβ3 and αvβ5 integrins (10 µg/mL each) or the control isotype mAb (30 µg/mL). For each Tat concentration (Tat 0.1 µM, blu bar; Tat 1 µM, red bar) infection fold changes were calculated as the ratio between RLU from MDDCs pre-incubated with the control isotype mAb in the presence and in the absence of Tat (baseline RLU: 97) or as the ratio between RLU from MDDCs pre-incubated with anti-integrin mAbs in the presence and in the absence of Tat (baseline RLU: 112) (all values had been previously subtracted of the uninfected MDDC RLU background). Experiments were performed in duplicate. (B) Infection of MDDCs at different time points in the presence of coated wt Tat (0.01 µM for donor 1, 0.1 µM for donor 2) and of mAbs directed against the α5β1, αvβ3 and αvβ5 integrins (10 µg/mL each) or the control isotype mAb (30 µg/mL). Fold reduction of infection was calculated as the ratio between the content of p24 antigen detected in the supernatants of MDDCs seeded on adherent Tat in the presence of anti-integrin mAbs and that detected in the supernatants of MDDCs seeded on adherent Tat in the presence of the control isotype mAb (baseline, for donor 1, p24 was 186 and 479 pg/mL at day 7 and 10, respectively; for donor 2, p24 was 135 and 538 pg/mL at day 7 and 10, respectively). Experiments were performed in duplicate.

Figure 7. Entry of Tat or R5 Env-VLPs in MDDCs and block by anti-integrin antibodies.

(A) Entry of wt Tat into MDDCs from 3–10 different donors (depending on the anti-integrin mAbs used), in the presence of mAbs against the indicated integrins alone or combined, a control isotype mAb, or nil (buffer). (B) Entry of cys22 Tat into MDDCs from 3 different donors and block by the combined anti-integrin mAbs versus an Ig control isotype mAb. The percentages of Tat positive cells +/− standard deviations are shown. (C) Entry of VLP-R5Env (BaL) in MDDCs in the presence of Tat and block by anti-integrin mAbs or an Ig control isotype mAb. The percentages of fluorescent cells are shown. A representative experiment out of 4 performed is shown.

Figure 8. Tat-mediated entry of Env molecules from different clades in MDDCs and block by anti-integrins antibodies.

(A) Clade B trimeric wt Env (twt Env), trimeric ΔV2 Env (tΔV2 Env), monomeric wt Env (mwt Env), or monomeric ΔV2 Env (mΔV2 Env) and (B) clade A and C trimeric wt Env (twt Env) molecules were incubated with control buffer or increasing concentrations of Tat, and then added to MDDCs at 1∶100 final dilution. Cells were then stained for intracellular Env. Open circles: control isotype mAb; filled circles: anti-integrin mAbs directed against the α5β1, αvβ3 and αvβ5 integrins (10 µg/mL each). Data from the same donor out of 3–8 donors tested are shown. Results are expressed as the percentage of Env-positive cells.

Figure 9. Tat/Env binding kinetics and thermodynamics.

(A) Binding kinetics of trimeric wt Env (tw Env), trimeric ΔV2 Env (tΔV2 Env), monomeric wt Env (mwt Env) or monomeric ΔV2 Env (mΔV2 Env) to Tat bound to the Biacore chip. The range of concentrations of Env molecules is indicated. (B) Tat/Env binding thermodynamics by ITC. Tat was titrated with trimeric wt Env. The top panel shows calorimetric data versus time; the lower panel shows integrated areas normalized to the number of moles of Env subunits injected at each injection step.

Figure 10. Tat/Env complex and ternary Tat/Env/αvβ3 complex by modeling-docking calculations.

(A) ribbon representation of the Tat/Env complex showing that the Env CD4 binding site and the RGD domain of Tat are both exposed. Color code: ΔV1-2 Env: blue; Tat: red; Tat-RGD: yellow. (B) Surface representation of the ternary Tat/Env/αvβ3 complex. Color code: ΔV1-2 Env: green; Tat: purple; αvβ3 integrin: grey. See experimental procedures for details.

Figure 11. Tat/V3 loop interaction by ELISA and ITC

(A). Tat interaction with cyclic (open circles) or linear (filled circles) V3 peptides by ELISA. Data are expressed as O.D. corrected for wells coated with the Tat buffer. (B) Tat interaction with cyclic (left panels) or linear (right panels) V3 peptides by ITC. Upper panels: calorimetry data versus time; lower panels: integrated areas normalized to the number of moles of V3 loop injected at each injection step.

Figure 12. Uptake of trimeric ΔV3 Env by MDDCs and block by anti-integrins mAbs.

Uptake of ΔV2 Env or ΔV3 Env pre-incubated with buffer or increasing concentrations of Tat, in the presence or absence of anti-integrin mAbs (10 µg/mL each) or a control isotype mAb (30 µg/mL). Trimeric ΔV2 Env: circles; trimeric ΔV3 Env: squares. Control isotype Ab: open symbols; anti-integrin blocking mAbs combined: filled symbols.

Figure 13. Neutralization of Tat/Env complex entry in MDDCs by sera from HIV-infected individuals.

(A) Neutralization of trimeric ΔV2 Env entry in MDDCs by sera from HIV-infected subjects in the presence or absence of Tat in anti-Tat Ab negative (n = 8) and anti-Tat Ab positive (n = 8) subjects. The bars represent the percentage of entry of Env alone incubated in buffer (in blue) or with Tat (in red). The percentage of Env-positive cells is shown. Data are expressed as the mean with standard deviation of experiments performed in duplicate. The codes of the anti-Tat Ab negative or positive sera are indicated at the bottom of the bars. (B) Geometric mean (GM) of the ratio, with 95% confidence interval (CI) of the percentage of MDDCs internalizing Env in the absence (blue bar) or in the presence (red bar) of Tat in anti-Tat negative (n = 8) and anti-Tat positive (n = 8) subjects. Statistical analysis was performed by the two-tailed Student’s t-test.

Figure 14. Neutralization of Tat/Env complex entry or infection in MDDCs by sera from Tat-vaccinated HIV-infected individuals.

(A) Percentage of MDDCs internalizing Env in the presence of sera from trial subjects (n = 7) at baseline (week 0, anti-Tat Ab negative, in blue) and after Tat vaccination (week 48, anti-Tat Ab positive, in red) in the absence (empty bars) or in the presence (filled bars) of Tat. The codes of the sera are indicated at the bottom of the bars. (B) Geometric mean (GM) of the ratio (week 48 vs baseline), with 95% confidence interval (CI) of the percentage of MDDCs internalizing Env in the presence or in the absence of Tat in trial subjects (n = 7). Statistical analysis was performed by the two-tailed Student’s t-test. (C) MDDC infection performed in duplicate for 48 h with pSF162LUC in the presence of coated cys22 Tat (0.01 µM) or BSA (control) with sera from a representative trial subject before (baseline, anti-Tat Ab negative) and after Tat vaccination (anti-Tat positive). Sera prior to or after vaccination contained the same anti-Env Ab titers. Data are expressed as RLU.

Figure 15. Neutralization of Tat/Env complex entry in MDDCs by sera from Tat/Env-vaccinated monkeys.

(A) Percentage of MDDCs internalizing Env in the presence of plasma from Tat/Env vaccinated or control monkeys (6 animals/group) before (week 0, anti-Tat Ab negative, in blue) or after vaccination (week 38, anti-Tat Ab positive, in red) in the absence (empty bars) or in the presence (filled bars) of Tat. The codes of the plasma are indicated at the bottom of the bars: MU1-6, vaccinated monkeys; MU7-12, control monkeys. (B) Geometric mean (GM) of the ratio (week 38 vs baseline), with 95% confidence interval (CI) of the percentage of MDDCs internalizing Env in the presence or in the absence of Tat in vaccinated monkeys (n = 6). Statistical analysis was performed by the two-tailed Student’s t-test.

Figure 16. Outcome of DC infection in the absence or presence of Tat, anti-Env and/or anti-Tat antibodies.

Tat redirects HIV to RGD-binding integrins evading neutralization by anti-Env Abs and both anti-Env and anti-Tat Abs are required to block infection. Extracellular Tat released by infected neighbor cells binds to trimeric Env on HIV, decreasing recognition of C-type lectin receptors and promoting engagement of RGD-binding integrins, which are expressed by inflammatory DCs, macrophages and endothelial cells (ECs) present at the site of infection. As a result, virions escape neutralization by anti-Env Abs directed against high mannose determinants and enters target cells upon binding to RGD-binding integrins. Anti-Tat Abs neutralize this binding, preventing virus entry through RGD-binding integrins. DC-SIGN: Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing Non-integrin; SIGN-R: homologue of DC-SIGN present on ECs; MR: mannose receptor.