Vaccine plus microbicide effective in preventing vaginal SIV transmission in macaques

Abstract

The human immunodeficiency virus epidemic continues in sub-Saharan Africa, and particularly affects adolescent girls and women who have limited access to antiretroviral therapy. Here we report that the risk of vaginal simian immunodeficiency virus (SIV)_mac251 acquisition is reduced by more than 90% using a combination of a vaccine comprising V1-deleted (V2 enhanced) SIV envelope immunogens with topical treatment of the zinc-finger inhibitor SAMT-247. Following 14 weekly intravaginal exposures to the highly pathogenic SIV_mac251, 80% of a cohort of 20 macaques vaccinated and treated with SAMT-247 remained uninfected. In an arm of 18 vaccinated-only animals without microbicide, 40% of macaques remained uninfected. The combined SAMT-247/vaccine regimen was significantly more effective than vaccination alone. By analysing immune correlates of protection, we show that, by increasing zinc availability, SAMT-247 increases natural killer cytotoxicity and monocyte efferocytosis, and decreases T-cell activation to augment vaccine-induced protection.

Fig. 1. Immunization regimen, infection rate and SIV plasma virus.

a, Rhesus macaques were subdivided into four groups: vaccine (n = 18), vaccine + SAMT-247 (n = 20), SAMT-247 (n = 6), and concurrent and historical controls (n = 6 and 31). Thirty-eight animals were primed with ΔV1 DNA-SIVgp160+p57 Gag and boosted with ALVAC-SIV encoding env, gag and pol and ALVAC-SIV + ΔV1 gp120 protein in alum hydroxide at the indicated timepoints. Twelve animals remained naïve until SIV challenge. Beginning at week 17, vaccine efficacy (VE) was assessed by subjecting all animals to up to 14 weekly intravaginal viral exposures (arrows) in the presence or absence of SAMT-247 until infection was confirmed. Animals either received 0.8% SAMT-247 in HEC gel (n = 26) or HEC gel only (n = 24) 4 h before each low-dose SIV_mac251 challenge. b,c, Significant protection in the vaccine group (P = 0.0074) (b) and the vaccine + SAMT-247 group (P < 0.0001) (c) compared with concurrent + historical controls. d, Delayed SIV acquisition in the vaccine + SAMT-247 group compared with the vaccine-only group (P = 0.006). e, No differences in delayed acquisition in the SAMT-247 group were observed compared with the combined concurrent plus historical controls (P = 0.27). f, Viral load (VL) geometric means of all macaque groups over time. Productive infection was qualified by the presence of viral DNA and RNA in mucosa and persistence of viral RNA in plasma over time. Data shown in b–e were analysed with log-rank (Mantel–Cox) test. Source data

Fig. 2. ADCC and NK responses and efferocytosis ex vivo and/or in vitro.

a, Comparison of SAMT-247 non-treated/treated effector cell-mediated ADCC activity in the vaccine (n = 18) and vaccine + SAMT-247 groups (n = 20; P < 0.0001). b, Correlation of SAMT-247-induced ADCC activity with number of intravaginal challenges in the vaccine + SAMT-247 group (n = 20; P = 0.024). c,d, Intracellular Granzyme B, perforin, IFN‐γ and TNF-α in macaque rectal mucosal (n = 9) NKG2A⁺ cells in the presence or absence of different stimuli. e, Macaque rectal mucosal NKp44⁺IL-17⁺ cells in the presence or absence of different stimuli (n = 9). f, Correlation of efferocytosis with number of intravaginal challenges in animals in the vaccine group (n = 18; P = 0.01). g,h, Comparison of percentage of efferocytosis (P < 0.0001) (g) and efferocytosis MFI (P < 0.0001) (h) using week 14 CD14⁺ monocytes in all vaccinated animals (n = 38). i, Correlation of SAMT-247-induced efferocytosis (SAMT-247-untreated efferocytosis subtracted from SAMT-247-treated efferocytosis) with number of intravaginal challenges in the vaccine + SAMT-247 group (n = 20; P = 0.065). Data shown in a, c, d, e, g and h were analysed with the two-tailed Wilcoxon signed-rank test. Data shown in b, f and i were analysed with the two-tailed Spearman correlation test. Horizontal and vertical bars denote mean and standard deviation, respectively. Source data

Fig. 3. T-cell responses ex vivo and in vitro.

a,b, Evaluation of CCR5 and α₄β₇ markers on Th1 and Th2 cells in the absence or presence of stimuli in the vaccine + SAMT-247 group animals (n = 9). c,d, Correlation of gp120 peptide + SAMT-247 stimulated CCR5⁻α₄β₇⁻ Th1 (P = 0.012) and Th2 cells (P = 0.020) with number of intravaginal challenges in the vaccine + SAMT-247 group (n = 9). e,f, IFN‐γ⁺, TNF-α⁺ and IL-10⁺ Th1 and Th2 cells in the rectal mucosa in the absence or presence of stimuli (n = 9). Data shown in a, b, e and f were analysed with the two-tailed Wilcoxon signed-rank test. Data shown in c and d were analysed with the two-tailed Spearman correlation test. Horizontal and vertical bars denote mean and standard deviation, respectively. Source data

Fig. 4. Effect of zinc chelation on NK and monocyte functions.

a, Representative imaging of human NKG2A⁺ cells unstimulated or stimulated with SAMT-247, PMA or PMA + SAMT-247. b, Mean zinc intensity in NKG2A⁺ cells of the healthy human donor in the presence or absence of zinc chelator in different stimulation conditions (n = 8). Fluorescence intensity of each field was measured for zinc expression as indicated by green colour, and the total number of DAPI positive cells were counted to determine the mean intensity of zinc/cells using iMARIS software. The mean of two duplicate fields was evaluated for the calculation. c, Comparison of expressions of NKG2A marker in macaques in the absence or presence of zinc chelator and stimuli in the vaccine + SAMT-247 group (n = 4) and vaccine group (n = 2). d–g, Comparison of expressions of granzyme B, perforin, IFN‐γ and TNF-α by macaque blood NKG2A⁺ cells from week 17 in the absence or presence of different stimulations and zinc chelator in the vaccine + SAMT-247 group (n = 4) and vaccine group (n = 2). h,i, Evaluation of the frequency of CD14⁺ monocytes and CD14⁺IL-10⁺ monocytes in the absence or presence of zinc chelator and stimuli in the vaccine + SAMT-247 group (n = 4) and vaccine group (n = 2). Data shown in b–i were analysed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and standard deviation, respectively. Source data

Fig. 5. Effect of zinc chelation on CCR5^+/– and α₄β₇^+/– Th1 and Th2 cells.

a–d, Comparison of expressions of CCR5 and α₄β₇ markers in Th1 and Th2 memory cells in the absence or presence of zinc chelator and stimuli in the vaccine+SAMT-247 group (n = 4) and vaccine group (n = 2). e–h, Radar plots comparing different expressions of cytokines by different subsets of Th1 and Th2 cells from vaccinated animals at week 17 in the absence or presence of stimulation and zinc chelator (n = 6). Data shown in a–h were analysed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and standard deviation. The radar plot represents the mean percentage value of cytokine responses. Solid lines represent the absence of zinc chelator and dashed lines represent the presence of zinc chelator. Source data

Fig. 6. Model for SAMT-247 modulation of mucosal immune responses.

Vaccination-induced ADCC results in apoptosis of SIV-infected cells, which in turn are cleared by efferocytes to avoid inflammation and preserve tissue homeostasis. Vaccine-induced IL-10 expression in CD14⁺ monocytes further augments efferocytosis. Vaccine-induced NKp44⁺ cells produce the IL-17 cytokine that maintains mucosal epithelium integrity. All of these protective effector responses were enhanced dramatically in the vaccine + SAMT-247 group, increasing protection from SIV_mac251 acquisition. The scheme is adapted from Bissa et al..

Extended Data Fig. 1. Assessment of infection rate in different animal groups.

a) No differences in delayed acquisition in the concurrent control group were observed compared to the historical controls (P = 0.74). Significant protection or trend of protection in b, c) the vaccine group (P = 0.006 and P = 0.18, respectively) and d, e) vaccine+SAMT-247 group (P = 0.0002 and P < 0.0001, respectively) compared with concurrent or historical controls. f, g) No differences in delayed acquisition in the SAMT-247 group were observed compared to concurrent (P = 0.59) or historical controls (P = 0.24). Data shown in (a-g) were analyzed with log-rank (Mantel–Cox) test. Source data

Extended Data Fig. 2. Quantification of humoral responses in plasma of vaccinated rhesus macaques.

a) Plasma antibody titers against ∆V1 gp120 over the course of immunization in the vaccine+SAMT-247 group (n = 20) and vaccine group (n = 18). b) Plasma antibody responses against different peptides encompassing V1 (peptides 15–24) and V2 (peptides 25–29) loop regions of gp120 in the vaccine+SAMT-247 group (n = 20) and vaccine group (n = 18). (c-e) Comparison of c) ADCC titer (P = 0.44), d) V2-specific (NCI05-specific) ADCC activity (P = 0.33) and e) V2-specific (NCI09-specific) ADCC activity (P = 0.13) at week 17 in the vaccine+SAMT-247 group (n = 20) and vaccine group (n = 18). (f-g) Correlation of f) ADCC activity and g) ADCC titer in the vaccine group (n = 18). (h-i) Correlation of h) ADCC activity and i) ADCC titer in the vaccine+SAMT-247 group (n = 20), (j-m) V2-specific (NCI05 and NCI09-specific) ADCC activity with number of intra-vaginal challenges in the vaccine group (n = 18) and vaccine+SAMT-247 group (n = 20). Data shown in (a, c-e) were analyzed with the two-tailed Wilcoxon signed-rank test or two-tailed Mann-Whitney test. Data shown in (f-m) were analyzed with the two-tailed Spearman correlation test. Horizontal and vertical bars denote mean and SD. Source data

Extended Data Fig. 3. Evaluation of NK/ILCs response and efferocytosis in humans and in rhesus macaques.

a, b) Intracellular Granzyme B, perforin, IFN‐γ, and TNF-α in healthy human (n = 6) blood NKG2A⁺ cells in the presence or absence of different stimuli. c) Comparison of Env-specific rectal NKp44⁺IL-17⁺ cells between vaccine+SAMT-247 (n = 20) and vaccine group (n = 18) 1 week post last vaccination (P = 0.43). d) Correlation of rectal mucosal Env-specific NKp44⁺IL-17⁺ cells with number of intra-vaginal challenges in the vaccine group (n = 18). e) Gating of NKG2A⁺NK cells, NKp44⁺ILCs, and NKG2A^–NKp44^–ILCs in rectal mucosal samples in the presence of PMA or PMA + SAMT-247 at 12 hours post stimulation. Gating was done on singlets, live, CD45⁺, CD3⁻, CD20⁻, CD11b⁻ cells. f) Gating of NKp44⁺IL-17⁺ ILCs in the rectal mucosal sample in the presence of PMA or PMA + SAMT-247 at 12 hours post stimulation. g) Correlation of efferocytosis percentage with number of intra-vaginal challenges in the vaccine+SAMT-247 group (n = 20). (h-i) Comparison of h) percentage of efferocytosis (P < 0.0001) and i) efferocytosis MFI (P < 0.0001) using pre CD14⁺ monocytes in all vaccinated animals (n = 38). Data shown in (a, b, h, i) were analyzed with the two-tailed Wilcoxon signed-rank test or two-tailed Mann-Whitney test. Data shown in (d, g) were analyzed with the two-tailed Spearman correlation test. Source data

Extended Data Fig. 4. Determination of ex vivo macaque T-cell responses.

a) Gating strategy of Th1 and Th2 cells, b,c) Comparison of CCR5⁺α₄β₇⁺ (P < 0.0001 and P = 0.01, respectively) and d,e) CCR5⁻α₄β₇^- memory (P < 0.0001 and P < 0.0001, respectively) Th1 and Th2 cells pre vaccination and at 1 week post last vaccination (week 13) in blood (n = 38). Data shown in (b-e) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. Source data

Extended Data Fig. 5. Quantification of activation, proliferation and exhaustion marker on Th1 cells upon stimulation.

Expression of a) OX40, b) CD40L, c) CD69, d) Ki67, e) LAG3, f) CTLA-4, g) PD-1 and h) PDL-1 in Th1 cells upon stimulation in the vaccine+SAMT-247 group (n = 4) and vaccine group (n = 2). Data shown in (a-h) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. Source data

Extended Data Fig. 6. Evaluation of activation, proliferation and exhaustion marker on Th2 cells upon stimulation.

Expression of a) OX40, b) CD40L, c) CD69, d) Ki67, e) LAG3, f) CTLA-4, g) PD-1 and h) PDL-1 in Th2 cells upon stimulation in the vaccine+SAMT-247 group (n = 4) and vaccine group (n = 2). Data shown in (a-h) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. Source data

Extended Data Fig. 7. Quantification of macaque mucosal T-cell responses and human NK cell responses.

a, b) Frequency of Th1 cells and Th2 cells in the rectal mucosa of macaques (n = 9). c-g) Comparison of expressions of granzyme B, perforin, IFN‐γ and TNF-α by NKG2A⁺ cells from healthy humans (n = 6) in the absence or presence of zinc chelator and stimuli (n = 6). Data shown in (a-g) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. The radar plot represents the mean percentage value of cytokine responses. Solid lines represent the absence of zinc chelator and dashed lines represent the presence of zinc chelator. Source data

Extended Data Fig. 8. Comparison of cytokine responses by CCR5⁺α₄β₇⁺ and CCR5⁻α₄β₇^- memory Th1 cells in the presence of stimulation.

a-f) Comparison of intracellular expressions of IFN‐γ, TNF-α, and IL-10 by CCR5⁺α₄β₇⁺ and CCR5⁻α₄β₇^- Th1 memory cells in the absence or presence of zinc chelator and stimuli in the vaccine+SAMT-247 group (n = 4) and vaccine group (n = 2). Data shown in (a-f) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. Source data

Extended Data Fig. 9. Evaluation of cytokine responses by CCR5⁺α₄β₇⁺ and CCR5⁻α₄β₇^- memory Th2 cells in the presence of stimulation.

a-f) Comparison of intracellular IFN‐γ, TNF-α, and IL-10 cytokines by CCR5⁺α₄β₇⁺ and CCR5⁻α₄β₇^- Th2 memory cells in the absence or presence of zinc chelator and stimuli in the vaccine+SAMT-247 group (n = 4) and vaccine group (n = 2). Data shown in (a-f) were analyzed with the two-tailed Wilcoxon signed-rank test. Horizontal and vertical bars denote mean and SD. Source data