Vectored immunoprophylaxis protects humanized mice from mucosal HIV transmission

Abstract

The vast majority of new HIV infections result from relatively inefficient transmission of the virus across mucosal surfaces during sexual intercourse. A consequence of this inefficiency is that small numbers of transmitted founder viruses initiate most heterosexual infections. This natural bottleneck to transmission has stimulated efforts to develop interventions that are aimed at blocking this step of the infection process. Despite the promise of this strategy, clinical trials of preexposure prophylaxis have had limited degrees of success in humans, in part because of lack of adherence to the recommended preexposure treatment regimens. In contrast, a number of existing vaccines elicit systemic immunity that protects against mucosal infections, such as the vaccines for influenza and human papilloma virus. We recently demonstrated the ability of vectored immunoprophylaxis (VIP) to prevent intravenous transmission of HIV in humanized mice using broadly neutralizing antibodies. Here we demonstrate that VIP is capable of protecting humanized mice from intravenous as well as vaginal challenge with diverse HIV strains despite repeated exposures. Moreover, animals receiving VIP that expresses a modified VRC07 antibody were completely resistant to repetitive intravaginal challenge by a heterosexually transmitted founder HIV strain, suggesting that VIP may be effective in preventing vaginal transmission of HIV between humans.

Figure 1. VIP protects against CD4 cell depletion in humanized mice resulting from challenge with CCR5-tropic or transmitted founder HIV strains

(a) Quantitation of human antibodies in serum one day prior to challenge, 3 weeks after adoptive transfer of human PBMCs and 13 weeks after intramuscular administration of 1×10¹¹ GC of AAV encoding either luciferase, b12, or VRC01-IgG as detected by a gp120-specific ELISA to determine the fraction of human IgG capable of binding HIV (n=21). (b) CD4 cells as a percentage of CD3 positive T-lymphocytes in the peripheral blood of PBMC-NSG humanized mice expressing luciferase (top), b12 (center) or VRC01 (bottom) following intravenous challenge with either JR-CSF (left) or REJO.c (right) strains of HIV as measured by flow cytometry. (c) Highlighter plot of envelope sequences amplified from control JR-CSF plasmid, or spleen genomic DNA from infected luciferase- or b12-expressing humanized mice as compared to parental JR-CSF sequence. Colored marks along each line indicate the position of missense mutations, silent mutations or gaps in alignment in red, green or grey respectively. Yellow highlight denotes mutations commonly identified among sequences isolated from antibody but not luciferase-expressing animals. (d) (top) Alignment of envelope sequences from wildtype compared to M373R, or V372E mutant JR-CSF demonstrating the location of mutations relative to the CD4 binding loop. (bottom) Relative frequency of the two observed escape mutations obtained from four independent b12-expressing mice. (e) In vitro neutralization assays performed using the TZM-bl cell line infected with either wildtype or the indicated mutant strain of JR-CSF in the presence of serial dilutions of either b12 (left) or VRC01 (right) (n=3). Each plot contains data generated from one individual experiment, and each experiment was performed once with the indicated number of mice.

Figure 2. VIP prevents mucosal transmission of CCR5-tropic HIV following repetitive intravaginal challenge

(a) Quantitation of gp120-binding human antibody in BLT humanized mice at the indicated times following administration of 1×10¹¹ GC of AAV encoding either luciferase or VRC01-IgG prior to challenge as measured by ELISA (n=9–10). Limit of detection = 70 ng ml⁻¹. (b) Quantitation of gp120-binding human antibody in vaginal wash samples taken from BLT humanized mice at the indicated times following administration of 1×10¹¹ GC of AAV encoding either luciferase or VRC01-IgG prior to challenge as measured by ELISA (n=9–10). Limit of detection = 1.3 ng ml⁻¹. (c) CD4 cells as a percentage of CD3 positive T-lymphocytes in the peripheral blood of BLT humanized mice expressing luciferase (white) or VRC01 (red) throughout weekly intravaginal challenges with JR-CSF as measured by flow cytometry. (d) (left) HIV p24 detection by immunohistochemical (IHC) staining of representative sections taken from spleens of challenged animals. Scale bar represents 40 micrometers. (right) Quantitation of IHC staining of spleen denoting the relative frequency of p24-expressing cells in spleens of challenged animals. (e) CD4 cells as a percentage of CD3 positive T-lymphocytes in mouse spleen (left) or vaginal lamina propria (right) tissues after repetitively challenge as detected by flow cytometry. (f) HIV viral load detected in plasma of mice at the time of sacrifice of repetitively challenged BLT mice as detected by the Abbott RealTime HIV-1 Viral Load Assay. Limit of detection = 200 copies ml⁻¹. (g) HIV viral load in plasma throughout weekly intravaginal challenge of BLT mice as detected by an in-house viral load assay. Limit of detection = 1500 copies ml⁻¹. (h) Fraction of uninfected mice over the course of repetitive intravaginal challenge. Positive infection defined by two consecutive viral load measurements above the limit of detection (n=9–10). Statistics for grouped comparisons calculated using either a one or two-tailed t-test. Statistics for Kaplan-Meier analysis calculated by log rank test. Samples from mice exhibiting fewer than 20 CD3 positive cells were excluded from this analysis. Each plot contains data generated from one individual experiment, and each experiment was performed once with the indicated number of mice.

Figure 3. VIP prevents mucosal transmission of transmitted founder HIV following repetitive intravaginal challenge

(a) Quantitation of gp120-binding human antibody in serum taken from BLT humanized mice at the indicated times following administration of 1×10¹¹ GC of AAV encoding either luciferase or VRC07G54W-IgG prior to challenge as measured by ELISA (n=12–13). Limit of detection = 70 ng ml⁻¹. (b) Quantitation of gp120-binding human antibody in vaginal wash samples taken from BLT humanized mice at the indicated times following administration of 1×10¹¹ GC of AAV encoding either luciferase or VRC07G54W-IgG prior to challenge as measured by ELISA (n=12–13). Limit of detection = 1.3 ng ml⁻¹. (c) CD4 cells as a percentage of CD3 positive T-lymphocytes in the peripheral blood of BLT humanized mice expressing luciferase (white) or VRC07G54W (red) throughout weekly intravaginal challenges with REJO.c as measured by flow cytometry. (d) HIV p24 detection by immunohistochemical (IHC) staining of a representative section taken from vaginal tissue of a REJO.c infected, luciferase expressing animal. Dashed line represents interface between epithelium (left) and lamina propria (right) demonstrating infected lamina propria lymphocytes. Scale bar represents 40 micrometers. (e) CD4 cells as a percentage of CD3 positive T-lymphocytes in mouse spleen, gut intraepithelial, gut lamina propria or vagina lamina propria lymphocytes tissues after repetitive challenge animals as detected by flow cytometry. (f) HIV viral load detected in plasma of mice at the time of sacrifice of repetitively challenged BLT mice as detected by the Abbott RealTime HIV-1 Viral Load Assay. Limit of detection = 200 copies ml⁻¹. (g) HIV viral load detected in plasma throughout weekly intravaginal challenge of BLT mice as detected by an in-house viral load assay. Limit of detection = 1000 copies ml⁻¹. (h) Fraction of uninfected mice over the course of repetitive intravaginal challenge. Positive infection defined by two consecutive viral load measurements above the limit of detection (n=12–13). Statistics for grouped comparisons calculated using either a one or two-tailed t-test. Statistics for Kaplan-Meier analysis calculated by log rank test. Each plot contains data generated from one individual experiment, and each experiment was performed once with the indicated number of mice.