Regulatory T cells control HIV replication in activated T cells through a cAMP-dependent mechanism

Abstract

We hypothesized that regulatory T cells (Tregs) could play a beneficial role during HIV infection by controlling HIV replication in conventional T cells (Tcons). Purified Tregs and Tcons from healthy donors were activated separately. Tcons were infected with the X4 or R5 HIV strains and cultured with or without autologous Tregs. Coculture of Tcons and Tregs resulted in a dose-dependent inhibition of Tcon infection, which was significant when a 1:1 Treg:Tcon ratio was used. Treg suppression of HIV infection was largely mediated by contact-dependent mechanisms. Blockage of cytotoxic T-lymphocyte-associated antigen-4 did not significantly reduce Treg function. In contrast, Tregs acted through cAMP-dependent mechanisms, because the decrease of cAMP levels in Tregs, the blockade of gap junction formation between Tregs and Tcons, the blockage of CD39 activity, and the blockage of protein kinase A in Tcons all abolished Treg-mediated suppression of HIV replication. Our data suggest a complex role for Tregs during HIV infection. Although Tregs inhibit specific immune responses, their inhibition of HIV replication in Tcons may play a beneficial role, particularly during early HIV infection, when the effector immune cells are not yet activated. Such a protective role of Tregs could have a profound impact on infection outcome.

Figure 1

Suppression of HIV infection in Tcons is mediated by Tregs. Tcons were infected with HIV and cocultured with Tregs for 3 days at a 1:1 ratio. HIV p24^Gag levels were measured by flow cytometry and ELISA. (A) Flow cytometry data from 1 representative experiment. The percentage of HIV-p24^Gag+ cells is indicated in each panel. UI indicates uninfected; INF, infected. (B) Graph showing the mean (± SEM) percentage of HIV-p24^Gag+ Tcons cultured alone or in the presence of Tregs (n = 9). (C) Levels of infection in Tcons cultured alone were considered as 100% infection. The percentage of infection of Tcons cultured with Tregs was then calculated for each donor using the equation (X = percentage of HIVp24^Gag+ Tcons cultured with Tregs × 100; percentage of HIV-p24^Gag+ Tcons cultured alone; n = 9). (D) MFI of HIV-p24^Gag determined in gated HIV-p24^Gag+ Tcons. Graph shows the mean (± SEM) MFI of the HIV-p24^Gag+ Tcons cultured alone and in the presence of Tregs (n = 6). (E-F) HIV-p24^Gag+ levels were measured by ELISA in culture supernatants collected 3 days after infection with the X4 strain (E) or the R5 strain (F) of HIV.

Figure 2

Treg suppression of HIV infection is dose dependent. Tcons were infected with HIV and cocultured at Tcon:Treg ratios of 1:0, 1:0.1, and 1:1. Levels of HIV p24^Gag measured by ELISA in culture supernatants 3 days after infection. One representative experiment for the X4 strain (left panel) and the R5 strain (right panel) of HIV are shown (n = 2 for each virus).

Figure 3

Treg-suppressive activity of HIV infection is mediated by contact. (A) IL-10 levels were measured in 3-day supernatants of cultured Tcons, Tregs, or Tcons + Tregs using the Luminex assay. (B) Tcons were HIV infected and cocultured with Tregs in the presence or absence of anti–IL-10R blocking antibody (10 μg/mL). Graph shows the mean (± SEM) percentage of HIV-p24^Gag+ Tcons (n = 6). (C) Graph showing the mean (± SEM) percentage of HIV-p24^Gag+ Tcons cultured alone, with Tregs, or with Tregs in a Transwell plate (n = 5).

Figure 4

CTLA-4 in Tregs is not the main contact-dependent mechanism. (A) Representative dot plot showing the percentage of infection in Tcons cultured alone (left), with Tregs (middle), or with Tregs in the presence of anti–CTLA-4 blocking antibody (10 μg/mL; right). (B) Graph showing the mean (± SEM) percentage of HIV-p24^Gag+ Tcons in these 3 conditions (n = 6). (C) CFSE MFI measured in Tcons cultured with Tregs in the presence or absence of anti–CTLA-4 blocking antibody (n = 5).

Figure 5

cAMP plays a role in the Treg-suppressive activity of HIV infection. (A) Infected Tcons cultured alone, with untreated Tregs, or with Tregs previously treated with an inhibitor of adenyl cyclase (200uM ddADA). Tregs were loaded with calcein AM and Tcons were stained by CFSE before culture. The formation of contact between Tregs and Tcons was evidenced by the transfer of calcein from Tregs to Tcons. Data represent the percentage of infected cells in gated (CFSE⁺) Tcons after 3 days of coculture. (B) Graph showing the mean (± SEM) percentage of suppression of HIV infection in these conditions (n = 4). (C) Graph showing the MFI of CFSE, Ki67, and cyclin B staining in gated Tcons after their culture with Tregs treated or not with ddADA (n = 4).

Figure 6

Gap junction formation plays a role in the Treg-suppressive activity of HIV infection. (A) Infected Tcons cultured alone, with untreated Tregs, or with Tregs previously treated with the gap junction inhibitor Gap27 (100uM). Tregs were loaded with calcein AM and Tcons were stained by CFSE before culture. The formation of contact between Tregs and Tcons was evidenced by the transfer of calcein from Tregs to Tcons. Data represent the percentage of infected cells in gated (CFSE⁺) Tcons after 3 days of coculture. (B). Graph showing the mean (± SEM) percentage of suppression of HIV infection in these conditions (n = 3). (C) Graph showing the MFI of CFSE, Ki67, and cyclin B staining in gated Tcons after their culture with Tregs treated or not with GAP27 (n = 4).

Figure 7

CD39 activity plays a role in the Treg-suppressive activity of HIV infection. Tregs and Tcons were stained with anti–CD39PE-Cy7 and anti–CD73-APC after 3 days of coculture and analyzed by flow cytometry. (A) CD39 and CD73 expression of 1 representative individual. (B) Graph showing the mean (± SEM) percentage of CD39⁺ (left) and CD73⁺ (right) cells (n = 4). (C) Infected Tcons cultured alone or with Tregs in presence or absence of either an anti-CD39 blocking antibody or a matched isotype control. Tregs were loaded with calcein AM before culture. Graph shows the mean (± SEM) percentage of suppression of HIV infection after the addition of the anti-CD39 blocking antibody or the matched isotype control (n = 3). (D) Infected Tcons cultured in the presence or absence of anti-CD39 blocking antibody (10 μg/mL). Graph shows the mean (± SEM) percentage of HIV-p24^Gag+ Tcons (n = 3). (E) Graph showing the MFI of CFSE, Ki67, and cyclin B staining in gated Tcons after their culture with Tregs treated or not with anti-CD39 (n = 4).

Figure 8

Treg-suppressive activity of HIV infection is PKA mediated. Infected Tcons were cultured alone, in the presence of the PKA inhibitor H89 (10uM), or in the presence of a cAMP analog that specifically activates PKA (6-Bnz-cAMP, 500uM). Tcon were also cultured with Tregs alone or in the presence of 10uM H89. Tregs were loaded with calcein AM and Tcons were stained with CFSE before culture. Data represent the percentage of infected cells in gated (CFSE⁺) Tcons after 3 days. One representative experiment of 2 is shown.