The quest for an HIV-1 vaccine: will mRNA deliver us from evil?

KEYWORDS:

• HIV-1
• Vaccine
• mRNA
• Virus-like particles
• Primate models
• Neutralizing antibodies
• Protection
• Heterologous tier-2 challenge

Forty-two years have gone by since the first report of five unusual cases of Pneumocystis carinii pneumonia in previously healthy young men in Los Angeles [1], which heralded the official appearance of AIDS on the world stage, but the quest for an HIV-1 vaccine is still on. The obstacles have been enormous. HIV-1 remains unrivaled in its extraordinary armamentarium of immune-evasive tactics, spanning from antigenic variation to glycan-mediated shielding to conformational camouflage. Most importantly, the bar to achieve protection from HIV/AIDS is extremely high because, in the case of HIV-1, protection equals ‘sterilizing’ immunity. The reason is that, as a retrovirus, HIV-1 integrates its genome into the host DNA where it can survive indefinitely, outside the reach of current therapeutic strategies. For most other infectious agents, the major goal of a vaccine is prevention of the acute illness that occurs during primary infection. For HIV-1 this objective is of no relevance to protection because the real disease is caused by a progressive corruption of the adaptive immune system that occurs over the course of several years of chronic infection. Once the virus gets its foot in the door, the clock starts. And the disease will almost invariably progress unless antiretroviral treatment is initiated and continued indefinitely. Thus, the sole way to immunize against HIV-1 disease is to lock the virus out of the body from square one: ‘sterilizing’ immunity.

As high as the bar may be, we now have convincing evidence that a protective HIV-1 vaccine is feasible. This reassuring assertion is based on two fundamental lines of evidence: first, as documented by antibody cloning from HIV-infected subjects, the human immune system has the ability to produce broadly neutralizing antibodies (bNAbs), which are widely believed to be the Holy Grail for a protective vaccine; second, as documented by passive infusion studies in nonhuman primate models, bNAbs can confer complete protection from chimeric viruses (SHIV) displaying the HIV-1 envelope on their surface [2]. But how can we educate the immune system to produce bNAbs? This remains the million-dollar question that has not, hitherto, found an answer.

1. The promise of mRNA

mRNA technology has been a cornerstone in the largely victorious campaign against COVID-19. But if one had asked people on the street ‘what is mRNA?’ in the early days of the pandemic, nobody would have even heard of it. Before 2010, even the majority of scientists would have been skeptical that mRNA transcripts could be safely and efficiently transduced into living cells in vivo and inform the endogenous production of an efficacious therapeutic protein or vaccine. But it happened. Thanks to the faith and vision of a few pioneers who pursued what initially appeared to be a ‘mission impossible’ (reviewed in [3]). Indeed, when the first cases of an unknown severe pulmonary infection first emerged in 2019 in the city of Wuhan, China [4], mRNA technology was already well established as it was being developed for a vaccine against one of SARS-CoV-2's first cousins: MERS-CoV [5]. Owing to the exceptional versatility of the mRNA platform, the transfer of the new pathogen genetic sequences, the vaccine design and production, and the clinical experimentation all happened within the span of a few months [6]. The rest is in the history books.

2. Can mRNA solve the HIV vaccine riddle?

I am often asked this question, and my default answer is ‘no, but … .’ No, because mRNA per se cannot provide a solution to this unprecedented challenge. But... yes, because mRNA can and will help tremendously in the path to the final victory. Indeed, mRNA is endowed with a unique set of qualities that can be of great benefit for an HIV-1 vaccine. Besides the general attributes of mRNA that make it an exceptional platform for any vaccine (i.e. safety, efficiency of expression, lack of immunogenicity, intrinsic adjuvanticity, speed of development, universal applicability, and relatively low cost), there are some properties specifically useful for an HIV-1 vaccine: i) endogenous protein expression by host cells, which thereby become the actual vaccine ‘factory.’ This is essential for the production of the HIV-1 envelope (Env) in its membrane-anchored form, which is analogous to the native Env presented on the surface of infectious virus; ii) decoration of HIV-1 Env glycoproteins with native N-linked glycosylation. This is essential because the quality of glycosylation has profound effects on the correct structuring and antigenicity of the Env trimer; iii) potential expression of multiple Env forms, including Envs from difficult-to-grow strains, as well as engineered variants with altered biologic or antigenic properties. This is essential for employing optimized immunogens for bNAb elicitation; and iv) potential expression of multiple HIV-1 proteins simultaneously. This is essential for inducing immune responses against several viral components and for promoting the in-vivo production of native-like virus-like particles (VLPs) when Env and Gag are simultaneously co-expressed.

Despite these favorable qualities, however, mRNA remains a vehicle, and we cannot expect a vehicle to solve the HIV-1 vaccine riddle. The medium is not the message. Or better, in this case: The messenger is not the message. To elicit bNAbs, we must devise appropriately designed immunogens and complex immunization protocols for stimulating and expanding rare breeds of B cells expressing germline precursors of such antibodies, also referred to as unmutated common ancestors (UCA), and then for pushing their affinity maturation through a very narrow path until they acquire improbable mutations that eventually will confer the ability to recognize tightly protected sites of vulnerability on the surface of the native HIV-1 Env trimer.

3. A VLP-forming env-gag mRNA vaccine platform

In 2016, well before mRNA was dragged to the center stage by the COVID-19 pandemic, we started to discuss with Moderna the possibility of utilizing mRNA for an HIV-1 vaccine. By virtue of the unique properties listed above, mRNA appeared to be an ideal technology to try and overcome some of the hurdles posed by an HIV-1 vaccine. Our main guiding principles were the following: i) to express the HIV-1 Env through endogenous pathways via mRNA in order to achieve native N-linked glycosylation; ii) to express only membrane-anchored Env since this form adopts the native antigenic configuration and does not expose immune-distractive epitopes; iii) to promote the in vivo formation of VLPs via Gag co-expression (Figure 1), as VLPs are believed to be more efficient immunogens compared to subunit proteins; iv) to perform the initial priming with a UCA-engaging Env in order to jump-start the expansion and maturation of bNAb lineages; and v) to perform multiple booster immunizations with accurately selected heterologous tier-2 Envs from different HIV-1 clades in order to focus antibody responses on the shared epitopes among such strains, i.e. bNAb epitopes.

Figure 1. Production of mature virus-like particles by LNP-encapsulated co-formulated env-gag-gagpol mRNA.

A VLP-forming, multiclade env-gag mRNA vaccine, encapsulated in lipid nanoparticles (LNPs) for in-vivo delivery, was tested both in mice and in Rhesus macaques. And mRNA lived up to the expectations. Not only did the macaques develop strong humoral and cellular immune responses, but also, after repeated immunizations, we started to observe the emergence of broad-spectrum tier-2 neutralizing antibodies, albeit at relatively low titers. Most importantly, the animals were significantly protected from mucosal infection with a difficult-to-neutralize heterologous virus, SHIV-AD8. The calculated risk reduction was 79% per-exposure [7]. These results were recently replicated in a second, more robust, macaque study, which also documented the beneficial effects of including the viral protease as a third component of the vaccine formulation (Figure 1) in order to achieve complete Gag processing and, thereby, production of mature VLPs (P. Zhang et al., unpublished data). A VLP-forming env-gag mRNA platform expressing the VRC01-UCA engager Env 426c-deglyco3 has also proven to be highly effective in inducing specific B-cell activation in knock-in mice expressing both heavy and light chains of bNAb VRC01-UCA (L. Stamatatos et al., unpublished data). These promising results have prompted the design of a Phase-I clinical trial based on the VLP-forming env-gag mRNA platform (HVTN 310), which is scheduled to launch in early 2024.

4. The future has begun

The above results raise the hope that an mRNA-based vaccine for HIV-1 might be an important step in the right direction. At the same time, other promising approaches based on mRNA are in the pipeline, most notably protocols aimed at the efficient engagement of bNAb UCAs. Similar to the VLP-forming env-gag mRNA vaccine platform described above, two of the most interesting such approaches are designed to express self-assembling nanoparticles, corroborating the current trend to favor particulate vaccines displaying repetitive antigen arrays for optimal immune stimulation. The first, which employs mRNAs expressing self-assembling ferritin nanoparticles bearing a stabilized clade-C Env that engages the UCA of the V3-glycan-specific bNAb DH270, has been tested in knock-in mice [8]. The second, which employs mRNA expressing self-assembling lumazine nanoparticles displaying on their surface an engineered core protein derived from the gp120 outer domain (eOD-GT8) that engages the UCA of bNAb VRC01, is being evaluated in a human trial. A vaccine using eOD-GT8 protein (IAVI G001) has already provided the proof-of-principle that this molecule can effectively prime VRC01 precursors in vivo and promote their expansion and initial somatic hypermutation [9]. Based on these encouraging results, Phase-I clinical experimentation is currently under way with the mRNA version of eOD-GT8 (IAVI G002 and G003) [10,11]. Although these studies are designed to achieve only the very first step of bNAb development, i.e. the recruitment and initial expansion of germline antibody precursors, they extend this important proof-of-principle to the mRNA realm, paving the way for future, more complex and ambitious trials. Another preparatory human trial (HVTN 302) is being conducted with mRNA encoding a stabilized membrane-anchored versus soluble Env trimer derived from a stabilized form of the reference clade-A Env BG505. Since all 3 scheduled immunizations in this trial will be performed with the same Env, which is not a UCA engager, the vaccine is not expected to elicit bNAb development, but it will provide helpful preliminary safety and immunogenicity data to guide the design of future trials.

Are we close to a solution of the HIV-1 vaccine conundrum? Probably not, but we may be closer. Now that mRNA has established itself as a legitimate carrier for an HIV-1 vaccine, the field is open to testing new ideas for further optimization. Approaches that will undoubtedly be explored include mucosal delivery, slow delivery, multiplex antigen display, addition of specific adjuvants, and combination of mRNA with other vaccine platforms. Nanoparticle-displayed protein immunogens and viral vectors expressing multiple HIV-1 proteins are among the most promising potential partners of mRNA for future combination approaches.

The next few years will see an explosion of studies employing mRNA for the delivery of HIV-1 vaccines. Although mRNA is only one element of a vaccine, it is a fundamental one. Especially for a virus like HIV-1, for which the bar for protection is so high, the unique set of features of mRNA may be fundamental to tilt the balance in our favor. The future of HIV-1 vaccines has begun, and mRNA is destined to play a protagonist role in this quest.

Declaration of interest

P Lusso is listed as a co-inventor in a patent application related to VLP-forming mRNA vaccines. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or material discussed in the manuscript apart from that disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This work was supported by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases (NIAID), Bethesda, Maryland.