Evaluation in cynomolgus macaques challenged with SHIV-SF162-p3 of the protective efficacy of a measles-based SHIV vaccine combined with a peptide-based vaccine targeting 3 retroviral functions – ARTT-HIVAC

In the first NHP study (Nzounza et al., 2021), the measles virus (MV) encoding SIV Gag and HIV-1 Env (full-length envelope protein encompassing gp120 and gp41 subunits) MV-Gag/Env, was administered. The present study uses a novel measles virus encoding SIV Gag and HIV-1 gp41 subunit (MV-Gag/gp41). In both studies, the same measles virus encoding the SIV Nef protein was used.

In preliminary work in mice, the humoral and cellular immunities induced by MV-Gag/gp41 were compared to those induced by MV-Gag/Env. Furthermore, immune responses to the MV-Gag/gp41 virus were evaluated in the presence or absence of a boost using two protein fragments of the gp41 subunit adjuvanted with Alum (fragments corresponding to the HR1 and HR2 heptad repeat domains of gp41).

During the present study conducted in the NHP, female cynomolgus macaques were divided into two groups: "vaccine" and "control".

Animals in the vaccine group were subjected to the following protocol:

- Phase 1: Vaccination with MV-Gag/gp41 and MV-Nef viruses.

- Phase 2: Administration of 50-70 amino acid domains of the gp41 and Nef proteins in the presence of Alum, first intramuscularly and then intranasally (mucosal route). This involves the use of the HR1 and HR2 subdomains of gp41, comprising three targeted regions: 1) the immunosuppressive domain, 2) the "3S" region, and 3) the MPER (Membrane-Proximal External Region); and a 67 amino acid structured subdomain of the SIV Nef protein.

- Phase 3: Viral pre-challenge, vaccination with recombinant MV vaccines, and protein boosts.

Animals in the control group received unmodified measles virus (phase 1), Alum adjuvant alone (phase 2), and finally unmodified measles virus in the presence of Alum (phase 3).

Antibodies against MV, Nef, and gp41 were measured by ELISA 15 days post-immunization.

Cellular responses against MV, Gag, Nef, and gp41 were measured 7 days post-immunization by ELISPOT.

Intracellular staining of T lymphocytes was performed by cytometry post-phase 1, post-phase 3, and post-viral challenge.

Following the immunization phase, animals in both groups received 0.5 AID50 (Animal Infectious Dose 50%, 25% infection rate per exposure) of SHIV162p3 virus, a difficult-to-neutralize clade B tier-2 virus, intravaginally once a week once a week for 10 weeks (i.e., repeated low-dose challenges). Viral challenges were stopped after two successive detections of lentiviral RNA in the blood by qRT-PCR. The challenge phase of this study differs from the first NHP study, which consisted of administering 0.5 AID50 of SHIV16p3 by the rectal route to male monkeys.

Preliminary studies in mice indicated that MV-Gag/gp41 induced humoral and cellular immune responses comparable to or greater than those induced by MV-Gag/Env. Similarly, strong immune responses were obtained in mice vaccinated with MV-Gag/gp41 (prime) and then with gp41 fragments in the presence of Alum (boosts).

Vaccination of NHP with MV-Gag/gp41 and MV-Nef (phase 1), and comparable to what we obtained in the first NHP study, led to moderate humoral and cellular responses. Anti-gp41 and Nef immune responses were significantly enhanced by the intramuscular administration of protein boosts in the presence of Alum (phase 2). Intranasal administration of protein boosts did not induce mucosal IgA or IgG antibodies (phase 2).

Overall, MV-Gag/gp41 and MV-Nef immunization combined with gp41 and Nef protein boosts did not induce antibodies capable of neutralizing viral entry (phases 1-3), as verified in in vitro tests against tier-1 HIV-1 viruses, which are highly susceptible to neutralization (work carried out in C. Moog's laboratory).

Six weeks after the last vaccination (phase 3), the NHPs were submitted to a viral challenge via vaginal administration of 0.5 AID50 of SHIV162p3 virus. Overall, the animals in the control and vaccine groups were infected at a rate consistent with the SHIV162p3 dose used. Of the 8 monkeys in the vaccine group, 1 animal remained uninfected despite 10 exposures to the virus. Two animals exhibited a very low blood viral load of 10^2 viruses per ml, rapidly undetectable, corresponding to the limit of virus quantification with the qRT-PCR method used. Conversely, in the control group, 8 out of 8 animals showed plasma viral load of >10^4 viruses/ml. Interestingly, this study confirms the results obtained in the first NHP study, namely the induction of immunity capable of significantly reducing viral load (area under the curve, vaccine vs. control, p<0.05, non-parametric Mann-Whitney test). Furthermore, the study of viruses integrated after infection in the most exposed compartments, namely the lymph nodes, showed no trace of provirus in 3 out of 8 monkeys in the vaccine group (3/8, or 37.5% total protection). One monkey from this group had an extremely low proviral load of 8 proviruses per million cells. In the control group, all animals had proviruses in their lymph nodes (8/8, no protection). The other reservoirs studied —PBMCs and the spleen— confirmed the results obtained for the lymph nodes. In our first study conducted on NHP, among the animals in the "vaccine" group, 3 out of 16 had undetectable proviral loads in their lymph nodes (3/16, 18.75% protection), compared to 0/8 in the control group. The protection obtained in the present study constitutes a crucial step in the development of a vaccine against HIV-1 delivered by the measles virus. However, the small number of animals used in the "vaccine" group (8 NHP) does not allow for an accurate determination of the percentage of animals protected against infection.

The vaccination protocol for this study includes the administration of the measles virus as a vector for HIV gp41 and SIV Gag/Nef (MV-SHIV), followed by the injection of adjuvanted gp41 and Nef peptides (peptide boosts). The results of the present study reinforce those obtained in a first NHP study which was based solely on the administration of the measles virus as a vector for the full-length HIV Env and SIV Gag-Nef vaccine antigens (Nzounza et al., NPJ vaccine, 2021). The first NHP study was conducted by several partners who are also involved in the present study, namely UMR9196, Pasteur Institute, IDMIT and Viroxis SAS.

The present study shows that MV-SHIV vaccination followed by peptide boosts induces (i) protection against infection in 37.5% of animals (3/8 with no trace of provirus in reservoir cells despite measurable blood viremia in 2 of them, indicating a transient and fully controlled viremia), (ii) a decrease in blood viremia, with 6/8 exhibiting viremia <10^4 viruses per ml compared to 0/8 in the control group. In the first NHP study, comparable results were obtained: (i) protection against infection in 3 out of 16 animals (18.75%, absence of reservoir cell infection) and reservoir cell infection at the limit of detection in 5/16 (10 proviral copies per 10^6 cells), (ii) control of blood viremia: 12/16 (75%) had viremia <10^4 viruses per ml compared to 1/8 in the control group (12.5%).

A direct next step for this study will be to seek new funding with Viroxis —the industrial partner in this study— to measure the effects of this SIV vaccination approach in non-human primates—in a prophylactic and therapeutic vaccination context. If this vaccination strategy has a significant effect on protection against infection, control of plasma viremia, and the occurrence of acquired immunodeficiency, a clinical trial in a small number of HIV-positive patients will be considered.

The repeated failures of previous clinical HIV-1 vaccine candidates and the moderate efficacy of the RV144 vaccine (31% at 42 months) have emphasized the need of new vectors inducing new immune functions and the setup of vaccine regimens combining several pre-existing immunogens/vaccine strategies. Measles virus (MV) vector priming combined with protein boosts could fulfill these conditions. In fact, in an initial study we demonstrated that vaccination with MV vectors expressing Gag, Env and Nef simian-human immunodeficiency virus immunogens (MV-SHIV) controlled the SHIVSF162p3 challenge virus in cynomolgus macaques. Indeed, the peak viral load median of the vaccinated monkeys was nearly 2 logs lower compared to the placebo group, and plasma virus load was strongly reduced within a week (p=0.0001, Wilcoxon test). Moreover in contrast to the control monkeys, the vaccinated monkeys maintained plasma CD4+ T-cell counts >1000 cells/µl following challenges. Consequently, the MV-SHIV vaccine markedly reduced the reservoir size in PBMCs, spleen, axillary and inguinal lymph nodes and rectum, as evidenced by 50% of the animals exhibiting = 10 proviral DNA copies per million of cells. Interestingly, the control of SHIV162p3 found in the vaccinated monkeys was correlated with the Gag-specific cellular immune responses. However MV-SHIV vaccine alone was not able to delay SHIV acquisition after repeated intrarectal challenges, which is crucial for a HIV-vaccine to prevent virus integration. This could be attributed to the lack of induction of plasma neutralizing IgG (against tier-2 SHIV162p3) and mucosal IgA (in rectal secretions), which are known to be associated in vivo with a sterilizing protection against SHIV and SIV challenges, or at least with a delay of acquisition. Thus, we propose in this study to combine the MV-SHIV vaccine with protein boosts of the external region of the HIV gp41 envelope subunit. That gp41 polypeptide will include 3 highly conserved functional domains: the immunosuppressive domain ISD (the target of IgA antibodies in “Exposed Uninfected” patients), the “3S-motif” (anti-3S antibodies inhibit NK activity and cytotoxicity) and the membrane-proximal external region MPER (recognized by neutralizing and anti-HIV transcytosis antibodies in macaques challenged by the SHIVSF162p3). Chemically synthesized gp41 polypeptide in the presence of PLGA-based nanoparticles and TLR4 and TLR7/8 agonists adjuvants, or recombinant gp41 expressed by measles virus vector will be first assessed in mice to define the protein-boost regimen yielding the highest levels of circulating and mucosal antibodies. Then, cynomolgus macaques will be primed with the previously assessed MV-SHIV vaccine and boosted with the best protein-boost regimen before intravaginal challenges with the SHIVSF162p3 strain. We aim to provide the proof of concept of the protective efficacy for this new vaccine regimen/combination in the perspective of a first-in-man Phase I clinical trial. The final objectives of this project are the evaluation in patients of this vaccine candidate as a prophylactic and therapeutic vaccine.