Rapid seeding of the viral reservoir prior to SIV viraemia in rhesus monkeys

Abstract

The viral reservoir represents a critical challenge for human immunodeficiency virus type 1 (HIV-1) eradication strategies. However, it remains unclear when and where the viral reservoir is seeded during acute infection and the extent to which it is susceptible to early antiretroviral therapy (ART). Here we show that the viral reservoir is seeded rapidly after mucosal simian immunodeficiency virus (SIV) infection of rhesus monkeys and before systemic viraemia. We initiated suppressive ART in groups of monkeys on days 3, 7, 10 and 14 after intrarectal SIVMAC251 infection. Treatment with ART on day 3 blocked the emergence of viral RNA and proviral DNA in peripheral blood and also substantially reduced levels of proviral DNA in lymph nodes and gastrointestinal mucosa as compared with treatment at later time points. In addition, treatment on day 3 abrogated the induction of SIV-specific humoral and cellular immune responses. Nevertheless, after discontinuation of ART following 24 weeks of fully suppressive therapy, virus rebounded in all animals, although the monkeys that were treated on day 3 exhibited a delayed viral rebound as compared with those treated on days 7, 10 and 14. The time to viral rebound correlated with total viraemia during acute infection and with proviral DNA at the time of ART discontinuation. These data demonstrate that the viral reservoir is seeded rapidly after intrarectal SIV infection of rhesus monkeys, during the 'eclipse' phase, and before detectable viraemia. This strikingly early seeding of the refractory viral reservoir raises important new challenges for HIV-1 eradication strategies.

Extended Data Figure 1

Viral dynamics modeling of initial viral growth and decay following ART initiation. Red lines indicate fitted values from the model in monkeys that initiated ART on days 3, 7, 10, and 14 of infection and were utilized for AUC VL calculations. The red asterisk indicates the time of treatment initiation.

Extended Data Figure 2

Ultrasensitive plasma viral loads in monkeys during ART. Log plasma viral RNA (copies/ml) at week 20 in rhesus monkeys infected with SIVmac251 and following initiation of ART on days 3, 7, 10, and 14 of infection. Assay sensitivity is 6 RNA copies/ml.

Extended Data Figure 3

Intracellular cytokine staining raw data of Gag-specific CD8+ and CD4+ T cells. Representative data for the magnitude of Gag-specific IFN-γ+ CD8+ and CD4+ T cell responses at week 20 in monkeys that initiated ART on days 3, 7, 10, and 14 of infection or with no ART.

Extended Data Figure 4

Intracellular cytokine staining of Gag-specific CD8+ and CD4+ T cells. Summary data for the magnitude of Gag-specific IFN-γ+ CD8+ and CD4+ T cell responses and Ki67 expression at week 4 and week 20 in monkeys that initiated ART on days 3, 7, 10, and 14 of infection or with no ART (N=4 animals per group).

Extended Data Figure 5

Proviral DNA in CD4+ T cell subpopulations during ART. Log proviral DNA (copies/10⁶ CD4 T cells) in sorted naïve (N), transitional effector memory (TM), and central memory (CM) CD4 T cell subpopulations from PBMC and from genital, inguinal, iliac, para-aortic, axillary, and/or mesenteric lymph nodes obtained from two animals necropsied on day 3 and two animals necropsied on day 7 following mucosal SIVmac251 infection.

Extended Data Figure 6

Viral dynamics modeling of viral rebound following ART discontinuation. Red lines indicate fitted values from the model in monkeys that initiated ART on days 3, 7, 10, and 14 of infection.

Extended Data Figure 7

Viral kinetics and setpoint viral loads following ART discontinuation. Log initial viral loads, exponential viral growth rate, and log setpoint viral loads following viral rebound derived from the model fits in groups of monkeys that initiated ART on days 3, 7, 10, and 14 of infection (N=4 animals per group).

Extended Data Figure 8

Early ART impacts AUC VL and time to viral rebound. Log AUC VL and interpolated time of viral rebound derived from the model fits are shown in monkeys that initiated ART on days 3, 7, 10, and 14 of infection (N=4 animals per group).

Figure 1. Viral decay kinetics after treatment with ART

Log plasma viral RNA (copies/ml) in rhesus monkeys infected with SIVmac251 and following initiation of ART on days 3, 7, 10, and 14 of infection (a) or with no ART (b). Assay sensitivity is 50 RNA copies/ml. Red arrows indicate initiation of ART. Black dots below x-axis indicate sampling timepoints.

Figure 2. SIV-specific humoral and cellular immune responses during ART

Env-specific ELISA antibody titers at weeks 0, 4, 10, and 24 (a) and Env-, Pol-, and Gag-specific IFN-γ ELISPOT responses at weeks 0, 4, 10, and 20 in SIV-infected monkeys that initiated ART on days 3, 7, 10, and 14 of infection or with no ART (b). Mean responses are shown (N=4 animals per group). Error bars reflect standard errors.

Figure 3. Proviral DNA during ART

Log proviral DNA (copies/10⁶ cells) in peripheral blood mononuclear cells (PBMC), lymph node mononuclear cells (LNMC), and gastrointestinal mucosa mononuclear cells (GMMC) in monkeys that initiated ART on days 3 (a), 7 (b), 10 (c), and 14 (d) of infection or with no ART (e). Comparisons of mean levels of log proviral DNA/10⁶ cells at the time of ART discontinuation (week 24) in ART treated and untreated (NT) monkeys in PBMC (f), LNMC (g), and GMMC (h) are also shown (N=4 animals/group). Assay sensitivity is 3 DNA copies/10⁶ cells. P-values reflect one-sided t-tests. Error bars reflect standard errors.

Figure 4. Viral rebound kinetics after ART discontinuation

Log plasma viral RNA (copies/ml) in animals following discontinuation of ART at week 24 in monkeys that initiated ART on days 3, 7, 10, and 14 of infection (a). Assay sensitivity is 50 RNA copies/ml. Median times to viral rebound, defined as the first timepoint at which plasma viral RNA was >50 copies/ml, is also shown by the red bars (b). Median setpoint viral loads following viral rebound, defined as day 56–112 following ART discontinuation, are also shown by the red bars in ART treated monkeys compared with untreated monkeys (c). P-values reflect one-sided t-tests.

Figure 5. Viral dynamics and correlations

Correlations of area under the curve for pre-ART viral loads (AUC VL) with proviral DNA in PBMC (a) and LNMC (b) at the time of ART discontinuation are shown. Correlations of AUC VL (c) and proviral DNA in PBMC prior to ART discontinuation (d) with the interpolated time to viral rebound are also shown. R² and P-values were calculated from correlation analyses, and trendlines were calculated using total least squares regression. Animals with undetectable proviral DNA were plotted at the detection limit of 3 DNA copies (0.48 log)/10⁶ cells.