Unintegrated HIV-1 provides an inducible and functional reservoir in untreated and highly active antiretroviral therapy-treated patients

Abstract

Background: The presence of HIV-1 preintegration reservoir was assessed in an in vitro experimental model of latent HIV-1 infection, and in patients treated or not with highly active antiretroviral therapy (HAART).

Results: In resting CD4+ T lymphocytes latently infected in vitro with HIV-1, we demonstrated that the polyclonal activation induced a HIV-1 replication, which could be prevented by the use of an HIV-1 integrase inhibitor. We also showed that this reservoir was labile since the rescuable HIV-1-antigens production from unintegrated HIV-1 genomes declined over time. These data confirm that our experimental approach allows the characterization of a functional unintegrated HIV-1 reservoir. We then explored the preintegration reservoir in HIV-1-infected patients. This reservoir was detected in 11 of 12 untreated patients, in 4 of 10 sustained responders to HAART, and in one incomplete responder. This reservoir was also inducible, labile, and anti-HIV-1 integrase drug inhibited its induction. Finally, this reservoir was associated with the presence of spontaneous HIV-1 antigens producing CD4+ T cells in blood from 3 of 3 untreated patients and 2 of 2 sustained responders to HAART harboring a preintegration reservoir.

Conclusion: This preintegration phase of HIV-1 latency could be a consequence of the ongoing viral replication in untreated patients and of a residual viral replication in treated patients.

Figure 1

Experimental protocol and culture conditions. A. In order to study the mobilization of the functional preintegration reservoir, resting CD4⁺ T cells were activated and cultured with the HIV-1 integrase inhibitor L-731,988 at the final concentration of 40 μM. B. To assess the correlation between the unintegrated HIV-1 DNA decay in vitro and the decline of rescuable viral production, infected resting CD4⁺T cells were preincubated 1 or 2 days before polyclonal stimulation. In both cases, in order to prevent infection of others cells by de novo-synthesized HIV-1, 1 μg/ml of the viral entry inhibitor T20 was also added in culture medium.

Figure 2

In vitro model of latently infected resting CD4⁺ T cells. A. The experimental approach was validated using in vitro latently infected resting CD4⁺ T cells that were unstimulated and directly polyclonaly activated in four experiments (nos. 1, 2, 3, and 4) or directly polyclonaly activated and cultured with L-731,988 in two other assays (nos. 3 and 4). B. In vitro latently infected resting CD4⁺ T cells were directly polyclonaly activated or preincubated 2 days before polyclonal activation in two experiments (nos. 3 and 4).

Figure 3

Mobilization of the functional preintegration reservoir. Resting CD4⁺ T lymphocytes secreting HIV-1 viral proteins in untreated (A) and HAART-treated patients (B). The CD4⁺ T lymphocytes were polyclonally activated and cultured with 1 μg/ml of enfuvirtide and with or without 40 μM of the HIV-1 integrase inhibitor L-731,988. The median values are shown as black bars. Comparison of results was done by the Wilcoxon signed-rank test.

Figure 4

Functional preintegration reservoir decay over the time in untreated patients (A) and in HAART-treated patients harboring a functional preintegration reservoir (B). Resting CD4⁺ T lymphocytes were polyclonally activated (J0) or preincubated 1 (J1) and 2 days (J2) before stimulation. HIV-1-Ag-SCs were enumerated at the end of culture. The median values are shown as black bars.

Figure 5

Safeguarding of CD4⁺ lymphocytes viability. Representative flow cytometry histograms (patient no. 9) characterizing viability of CD4⁺ T cell subset at the end of culture when cells were preincubated one or two days before their polyclonal activation. A gate A was set on the forward-scatter vs side-scatter histogram. As shown on different histograms, gate A corresponded to CD69⁺ CD4⁺ T lymphocytes. The analysis of the 7AAD level expression demonstrated that activated CD4⁺ T lymphocytes were viable cells.

Figure 6

Characterization of preintegrated HIV-1 DNA using Alu-LTR real-time PCR. The level of integrated HIV-1 DNA copies was assessed in CD4⁺ T lymphocytes from the in vitro model of infection and from four patients (nos. 4, 12, 13 and 14) using Alu-LTR real-time PCR experiments. CD4⁺ T cells that were directly stimulated, preincubated 1 and 2 days before polyclonal activation, and directly stimulated and cultured with L-731,988 were recovered from ELISpot assays and tested in PCR experiments.