Cooperation Between Systemic and Mucosal Antibodies Induced by Virosomal Vaccines Targeting HIV-1 Env: Protection of Indian Rhesus Macaques Against Low-Dose Intravaginal SHIV Challenges

Abstract

A virosomal vaccine inducing systemic/mucosal anti-HIV-1 gp41 IgG/IgA had previously protected Chinese-origin rhesus macaques (RMs) against vaginal SHIV_SF162P3 challenges. Here, we assessed its efficacy in Indian-origin RMs by intramuscular priming/intranasal boosting (n=12/group). Group K received virosome-P1-peptide alone (harboring the Membrane Proximal External Region), Group L combined virosome-rgp41 plus virosome-P1, and Group M placebo virosomes. Vaccination induced plasma binding but no neutralizing antibodies. Five weeks after boosting, all RMs were challenged intravaginally with low-dose SHIV_SF162P3 until persistent systemic infection developed. After SHIV challenge #7, six controls were persistently infected versus only one Group L animal (vaccine efficacy 87%; P=0.0319); Group K was not protected. After a 50% SHIV dose increase starting with challenge #8, protection in Group L was lost. Plasmas/sera were analyzed for IgG phenotypes and effector functions; the former revealed that protection in Group L was significantly associated with increased binding to FcγR2/3(A/B) across several time-points, as were some IgG measurements. Vaginal washes contained low-level anti-gp41 IgGs and IgAs, representing a 1-to-5-fold excess over the SHIV inoculum's gp41 content, possibly explaining loss of protection after the increase in challenge-virus dose. Virosomal gp41-vaccine efficacy was confirmed during the initial seven SHIV challenges in Indian-origin RMs when the SHIV inoculum had at least 100-fold more HIV RNA than acutely infected men's semen. Vaccine protection by virosome-induced IgG and IgA parallels the cooperation between systemically administered IgG1 and mucosally applied dimeric IgA2 monoclonal antibodies that as single-agents provided no/low protection - but when combined, prevented mucosal SHIV transmission in all passively immunized RMs.

Keywords: HIV-1 gp41; Indian-origin rhesus macaque model; SHIV; intramuscular prime/intranasal boost vaccination; intravaginal challenge; mucosal immunity; virosomal vaccine; virosomes.

Figure 1

HIV gp41-derived antigens used with virosomes. (A) Scheme illustrating the various regions of HIV gp120 and gp41 with the signal peptide (SP), variable regions #1 to #5 (V1-V5), conserved domains #1 to #5 (C1-C5), gp41 fusion peptide (FP), the two helix regions 1 and 2 (HR1 and HR2), the membrane-proximal external region (MPER), and the transmembrane domain (TM). The numbers indicated correspond to the amino acid position in the HxBc2 gp160 sequence, with some of the key neutralizing gp41 epitope sequences (QARILAVERY, 2F5, 4E10, 10E8) to show their location. (B) The recombinant gp41 antigen (rpg41) covers the amino acid sequence 540-664 with a deletion from 593-617, followed by leucine and glutamic residues from the cloning site and a 4-histidine tag sequence ending with a C-terminal cysteine for lipidation. The last 16 rgp41 residues on the C-terminal end (residues 649-664) overlap with the first 16 amino terminal residues of the P1 peptide. Of note, the originally described P1 sequence covers the gp41 residues 649-683, and it was subsequently modified by adding the natural leucine residue (649-684) on the C-terminal end, followed by serine and cysteine residues to improve peptide solubility, stability, upscaling as well as allowing lipidation. This modified P1 sequence and the rgp41, respectively, were anchored onto separate virosomes and used for nonhuman primate studies.

Figure 2

Production schema to generate HIV-1 gp41 virosomes. The production of unadjuvanted placebo virosomes (MYM-VP01), as well as unadjuvanted virosome-P1 (MYM-V101) and virosome-rgp41 (MYM-V102) is based on components derived from influenza virus membranes. Step 1, inactivated influenza A/H1N1 is solubilized with detergent; Step 2, nucleocapsids are discarded; Step 3, the viral membrane lipids with the native influenza hemagglutinin (HA) and neuraminidase (NA) are recovered and used as carriers for vaccinal antigens. The following steps are specific to each virosomal vaccine: Step 4a for placebo virosomes; only synthetic phospholipids are mixed with the influenza virus-derived components, while the antigen P1 (Step 4b, blue rods) or rgp41 (Step 4c, pink rods) is mixed with synthetic phospholipids. During gradual removal of the detergent in Steps 5a-5c, placebo virosomes (6a), virosome-P1 (6b), and virosome-rgp41 (6c) are gradually assembled in vitro. To generate the HIV-1 liquid virosomal vaccine, virosome-P1 (MYM-V101) and virosome-rgp41 (MYM-V102) are then diluted, combined, and mixed to achieve the target antigen concentration of the final HIV-1 vaccine MYM-V201. Quality control is performed to verify that values for particle size, particle population homogeneity, antigen and HA content are within the predefined specificities.

Figure 3

Study design and timeline for vaccine administration and virus challenges. For three groups of 12 Indian rhesus macaques, inactivated influenza virus was given intramuscularly 4 weeks (week -4) before the first intramuscular vaccine dose (week 0), which was followed by the second intramuscular vaccine dose at week 7. The third and fourth vaccine doses were given intranasally at weeks 14 and 24, respectively. Intravaginal SHIV_SF162P3 virus challenges were performed from week 29 to week 51. The challenges were stopped after a given animal became persistently viremic (>10,000 copies/ml) or after 22 challenges. During the vaccination phase, the study was performed blinded. For Groups K and L, blinding was maintained throughout the entire study; Group M animals were recognized as controls after immunogenicity analyses for anti-P1 peptide antibody reactivity prior to starting the SHIV_SF162P3 challenges.

Figure 4

Plasma viral RNA (vRNA) loads after repeated low-dose intravaginal challenges with the tier 2 R5 SHIV_SF162P3. (A–C) Challenge Phase I, comprising the first seven intravaginal challenges with plasma vRNA loads up to day 32, just before challenge #8. (D–F) Challenge Phases I (days 0-32) and II (day 32 to end of study). For challenge Phase II, the SHIV_SF162P3 dose had to be increased by 50% to follow the same experimental strategy used in the initial study in Chinese macaques (14). Horizontal dotted line (A–F), limit of detection 50 vRNA copies/ml (25). Vertical dotted line on day 32 in all panels, start of Challenge Phase II at the increased virus dose. Red ticks, day for each intravaginal SHIV challenge in all panels. ^†animal 30648 (M8) died of unrelated causes on day 46.

Figure 5

Time to event endpoints, by study day, for SHIV Challenge Phase I (A–C). This included the first 7 challenges and the time up to but not including challenge #8, when the challenge virus dose was increased. Endpoints include (A) time to first viremia; (B) time to peak viremia; and (C) time to persistent systemic infection (PSI). For interval-censored endpoints (time to first viremia, time to PSI), blocks in figures indicate periods with no plasma samples taken, over which the time to event curve is linearly interpolated. There were no significant differences among Groups K, L, and M when the entire study was analyzed (Challenges Phases I and II; for a total of 22 SHIV challenges; data not shown). Statistical analyses were performed by Dr. Chris Gast.

Figure 6

Sudden loss of protection in Group L one week after challenge #8 when the virus dose was increased by 50%. On day 38, five Group L animals lost protection (green box). We hypothesize that vaccine-induced mucosal antibodies were quite protective during Challenge Phase I, when only two breakthrough infections occurred. However, the additional antigen load in the 50% higher challenge virus dose overwhelmed vaccine protection in a large fraction of Group L animals immediately after the dose escalation – akin to flood water overtopping the dam, leading to flash floods and inundation. The implication of this “dam” hypothesis is that the immunogenicity of the virosome platform may need to be increased to boost both mucosal and systemic immune defenses for facing the virus doses administered during experimental intravaginal SHIV challenges in primate models estimated to exceed the HIV-1 inocula passed between infected men to their female partners (please see text).

Figure 7

Vaccine-induced anti-HIV-1 gp41 plasma IgG responses in Group L measured by ELISA with rgp41. The timeline for the intramuscular (i.m.) and intranasal (i.n.) immunizations is given in Figure 3. Plasma samples collected at the time points indicated were assessed for each individual vaccinee in Group L. Data for the two animals with early breakthrough infection during Challenge Phase I are indicated in red symbols; data for vaccinees that remained aviremic during the first 15 challenges or longer are indicated in blue symbols. No clear pattern emerged for either subset of vaccine recipients.

Figure 8

Antigen-specific IgG concentration in plasma samples and viral RNA (vRNA) loads over time. Each panel represents data for one rhesus macaque of Group L that had been vaccinated with the combination of virosome-P1 + virosome-rgp41. Red circles in green squares, RMs with breakthrough infection 6 days after increasing the SHIV challenge dose by 50% on day 32 (challenge #8; please see ). Red dotted line, limit of detection for vRNA [50 copies/ml (25)]. Black dotted line, ≥ 10⁴ plasma vRNA copies/ml, threshold for Persistent Systemic Infection (PSI). Red ticks, time points at which each animal underwent SHIV_SF162P3 challenges. Once vRNA levels were ≥10⁴ copies/ml, no further virus challenges were administered. The number of SHIV_SF162P3 challenges thus varied for each animal. Immediately following challenge #8 at the 50% higher virus dose, many Group L animals had breakthrough infections, which were not linked to decreases in anti-P1 or anti-rgp41 plasma IgG levels.

Figure 9

IgG Binding Antibody Multiplex Assay (BAMA) response magnitude measured by Mean Fluorescence Intensity (MFI), for the vaccine groups L and M, by time point and antigen (Methods). Dots are colored by response status with boxplots displaying distributions among positive responders. Baseline (pre) and negative samples are shown in shades of grey color; blue and black dots show positive samples for Groups L (immunized with the combination of virosome-P1 + virosome-rgp41) and M (given placebo virosomes), respectively. Response rates [percent positive, and ratio of number of positive samples/total number of samples] are shown for each combination. Lines join the same animal over time. High background reactivity against some gp41 antigens is notable for some animals.

Figure 10

Volcano plots depicting fold change and significance of differences in Fc Array antibody features between immunization Groups L (top, given virosome-P1 + virosome-rgp41) and K (bottom given single-agent virosome-P1) and the control arm (Group M given placebo virosomes). Symbol shapes indicate Fc specificity and color indicates Fc characteristic. Horizontal lines indicate statistical significance by Mann-Whitney U test, with p = 0.05 indicted by the black dotted line, and a Bonferroni-adjusted significance threshold indicated in the red dotted line. Group comparisons of antibody profiles at time points including pre-immunization (pre), one or two weeks after the 2^nd intranasal boost, the day of the 1^st SHIV challenge, and the day of the 7^th SHIV challenge are reported; please see Figure 3.

Figure 11

Levels of FcγR2a-4 binding (left) and total IgG (right) specific for the P1 peptide used as an immunogen (top) and recombinant gp41 from the HxBc2 strain (bottom) over time. Immunization groups are indicated in color; green, Group K; blue, Group L; black, Group M. Time points include pre-immunization (pre), one and two weeks after the 2^nd intranasal boost, the day of the 1^st SHIV challenge, and the day of the 7^th SHIV challenge (please see Figure 3). Median fluorescent intensities (MFI) are reported for each antibody characteristic; group medians are represented by the bar with Tukey’s box and whiskers.

Figure 12

Antibody-dependent Cellular Phagocytosis (ADCP) (top) and Antibody-dependent Neutrophil Phagocytosis (ADNP) (bottom) were assessed against gp140 SHIV_SF162p3 (left) and the recombinant Mymetics rgp41 (right) antigens. Immunization groups are indicated in color; green, Group K; blue, Group L; black, Group M. Time points include pre-immunization (pre), one and two weeks after the 2^nd intranasal boost, the day of the 1^st SHIV challenge, and the day of the 7^th SHIV challenge (please see Figure 3). Phagocytic scores are reported for each assay; group medians are represented by the bar with Tukey’s box and whiskers.

Figure 13

Antigen-specific antibody concentrations (ng/ml) measured in vaginal washes with the Imperacer^® assay (Methods). Group K (green circles) was immunized with single-agent virosome-P1 (P1 alone), Group L (blue circles) received the combination of virosome-P1 plus virosome-rgp41 (P1+rgp41), whereas Group M was given placebo virosomes (black symbols). (A–D) The sum of antigen-specific IgG plus IgA is given at the weeks after intranasal (i.n.) boosts indicated; (A, B) anti-P1 responses, and (C, D) anti-rgp41 responses.