Effects of therapeutic vaccination on the control of SIV in rhesus macaques with variable responsiveness to antiretroviral drugs

Abstract

A therapeutic vaccine that induces lasting control of HIV infection could eliminate the need for lifelong adherence to antiretroviral therapy. This study investigated a therapeutic DNA vaccine delivered with a single adjuvant or a novel combination of adjuvants to augment T cell immunity in the blood and gut-associated lymphoid tissue in SIV-infected rhesus macaques. Animals that received DNA vaccines expressing SIV proteins, combined with plasmids expressing adjuvants designed to increase peripheral and mucosal T cell responses, including the catalytic subunit of the E. coli heat-labile enterotoxin, IL-12, IL-33, retinaldehyde dehydrogenase 2, soluble PD-1 and soluble CD80, were compared to mock-vaccinated controls. Following treatment interruption, macaques exhibited variable levels of viral rebound, with four animals from the vaccinated groups and one animal from the control group controlling virus at median levels of 103 RNA copies/ml or lower (controllers) and nine animals, among all groups, exhibiting immediate viral rebound and median viral loads greater than 103 RNA copies/ml (non-controllers). Although there was no significant difference between the vaccinated and control groups in protection from viral rebound, the variable virological outcomes during treatment interruption enabled an examination of immune correlates of viral replication in controllers versus non-controllers regardless of vaccination status. Lower viral burden in controllers correlated with increased polyfunctional SIV-specific CD8+ T cells in mesenteric lymph nodes and blood prior to and during treatment interruption. Notably, higher frequencies of colonic CD4+ T cells and lower Th17/Treg ratios prior to infection in controllers correlated with improved responses to ART and control of viral rebound. These results indicate that mucosal immune responses, present prior to infection, can influence efficacy of antiretroviral therapy and the outcome of immunotherapeutic vaccination, suggesting that therapies capable of modulating host mucosal responses may be needed to achieve HIV cure.

Fig 1. Therapeutic vaccine study design.

Indian origin rhesus macaques were infected with SIVΔB670 at week 0 (red arrow) and were treated with ART starting at 6 weeks post-infection (wpi). Purple arrows indicate a series of 5 DNA immunizations spaced 1 month apart, occurring between 32 wpi and 48 wpi. At week 55, ART was interrupted to assess the efficacy of the therapeutic vaccine on viral control. Animals were necropsied at 80 wpi or earlier in the presence of AIDS-defining conditions. Red triangles indicate blood draws for PBMC isolation and brown circles indicate MLN biopsies to measure systemic and gut-associated immune responses. Prior to administering therapeutic immunizations, macaques were stratified so that each group had comparable viral loads and CD4 counts prior to and during ART.

Fig 2. DNA vaccination increases Env-specific antibody responses and total SIV-specific IFN-γ T cell responses during ART.

(A) The magnitude of the SIV Env-specific IgG response in the plasma was measured by ELISA, using SIV gp130 as the capture antigen. Shown are medians with individual animals’ data layered over. (B-C) PBMCs were stimulated with Gag, Env, Pol, Vif, Vpr, Rev, Nef, and Tat peptides to quantify the SIV-specific IFN-γ response. Samples were considered positive if peptide-specific responses were at least twice that of the negative control plus at least 0.01% after background (DMSO) subtraction. (B) Shown are medians of the cumulative (sum of response against all peptides) IFN-γ response, with data from individual animals layered over. (C) The cumulative SIV-specific IFN-γ response is shown after 5 vaccinations (50 wpi). Shown are medians and interquartile ranges with individual responses layered over each bar. (D) The breadth of the SIV-specific IFN-γ response is the number of peptide pools with a positive IFN-γ response. The cumulative breadth of the SIV-specific IFN-γ response post-vaccination (50 wpi) is shown, with medians and interquartile ranges of each group depicted, with individual animals’ data layered over each bar. (E) The cumulative breadth of the SIV-specific IFN-γ response post-vaccination (50 wpi) is shown as medians and interquartile ranges with individual responses layered over each bar. (C, E) A Dunn’s multiple comparisons test was used when making multiple comparisons between vaccine groups and the mock group. Results are considered significant if P ≤ 0.05. Shown are the medians and individual responses layered over each timepoint. (F) The percent of the IFN-γ response specific for Gag, Env and accessory proteins was calculated from the cumulative IFN-γ response at peak breadth (34 wpi for MAG + LT, 42 wpi for MAG + AC and Mock).

Fig 3. Three out of five animals in the MAG + AC group control virus replication during ATI.

(A-C) Plasma viral RNA levels were quantified using RT-q-PCR, with a limit of detection of 30 viral RNA copies per 1 mL of plasma, as indicated by the dashed line. (D) Shown are the median viral load and interquartile ranges for each treatment group. The dotted line indicates the threshold for control of virus replication, based on previous studies using SIVΔB670.

Fig 4. Five controllers maintained significantly lower viral burden during ATI compared to nine non-controllers.

(A) Plasma viral RNA levels were quantified using RT-q-PCR, with a limit of detection of 30 viral RNA copies per 1 mL of plasma, as indicated by the dashed line. The dotted line denotes the threshold for control of virus replication, based on previous studies using SIVΔB670. Controllers were defined as animals that maintained a median viremia at or below 1000 viral RNA copies per 1 mL of plasma for 5 months post-ART. Non-controllers were defined as animals with a median viremia that exceeded 1000 viral RNA copies per 1mL of plasma for 5 months post-ART. (B) Viral burden during ATI was calculated as the area under the curve of each animal’s viral load from 55 wpi to 76 wpi, shown are the median viral burden and interquartile ranges. Statistics were calculated using a Mann-Whitney t-test. Results are considered significant if P ≤ 0.05.

Fig 5. Controllers have higher SIV Gag-specific CD8⁺ T cell responses in PBMC and MLN post-vaccination and during ATI.

(A-C) PBMCs and MLNs were thawed and stimulated with Gag peptides, and expression of cytokines was quantified using intracellular cytokine staining. Shown are the medians and interquartile ranges of the SIV Gag-specific CD8⁺ T cell responses of controllers and non-controllers, with individual responses layered over each bar at a post-vaccination timepoint (50 wpi) and during ATI (62 wpi for PBMC and 66 wpi for MLN). Statistical differences between controllers and non-controllers at each timepoint were calculated using a Mann-Whitney t-test. Benjamini-Hochberg adjusted P values are shown, results are considered significant if P ≤ 0.05. (D) The SIV Gag-specific TNFα CD8⁺ T cell responses in PBMC at 50 wpi negatively correlated with the viral burden measured as area under the curve (AUC) during ATI. (E) The SIV Gag-specific IFN-γ CD8⁺ T cell responses in PBMC at 50 wpi negatively correlated with the viral burden measured as area under the curve (AUC) during ATI. (F) The SIV Gag-specific IL-2 CD8⁺ T cell responses in MLN at 62 wpi negatively correlated with the viral burden measured as area under the curve (AUC) during ATI. The P and r values shown were calculated using a Spearman rank correlation test. Benjamini-Hochberg adjusted P values are shown, results are considered significant if P ≤ 0.05.

Fig 6. Increased frequencies of polyfunctional CD8⁺ T cells in MLN and PBMC of controllers.

(A-D) The polyfunctional SIV-specific CD8⁺ T cell responses in MLN and PBMC post-vaccination and during ATI negatively correlated with ATI viral burden. The P and r values shown were calculated using a Spearman rank correlation test. Benjamini-Hochberg adjusted P values are shown, and results are considered significant if P ≤ 0.05.

Fig 7. Lower viral replication during acute infection correlates with improved ART responsiveness that in turn is associated with better control of viral rebound during ATI.

Improved virological response to ART and better control of viral rebound are associated with the relative populations of mucosal CD4⁺ T cells and Th17/Treg cells before infection. (A-C) Viral loads were measured via RT-q-PCR and viral burden was calculated as the area under the curve of animals’ viral loads. (A) Statistics were calculated using a Mann-Whitney t-test. Benjamini-Hochberg adjusted P values are shown. Results are considered significant if P ≤ 0.05. (B-C) Acute log viral burden (0–6 wpi) correlated with log viral burden during ART treatment (6–32 wpi). ART log viral burden in turn correlated with log viral burden during ATI (55 wpi– 76 wpi). (D) The frequency of CD4⁺ T cells in the colon at baseline (0 wpi) correlated with viral burden during ART. (E) The ratio of Th17 and Treg cells at baseline correlated with viral burden during ART. (F) The frequency of CD4⁺ T cells in the colon at baseline (0 wpi) correlated with viral burden during ATI. (G) The ratio of Th17 and Treg cells at baseline correlated with viral burden during ATI. (B-G) A Spearman rank correlation test was used to determine P and r values. Shown are Benjamini-Hochberg adjusted P values, results are considered significant if P ≤ 0.05.