Scientific Publications

Combination vaccine formulations contain distinct components targeting multiple strains of a single pathogen or multiple pathogens. By minimizing the number of separate vaccine administrations required, combination vaccines have been critical in allowing the broad expansion of the number and range of diseases that can now be prevented by immunization. Recent advances in vaccine development and our understanding of the immune system now make it possible to envision how new combination vaccines could play a major role in helping immunization programs address a much wider range of emerging or still problematic pathogens. However, few combinations are currently in the pipeline, in part due to their inherently increased complexity and cost of development compared to standalone formulations. This complexity, in turn, is partly driven by the regulatory requirements surrounding the clinical study program for the combination vaccine, especially the primary clinical endpoints and the required degree of precision around those endpoints, as these ultimately determine the sample size, cost, and duration of the study. As part of a larger effort to facilitate combination vaccine development, vaccine experts at the World Health Organization and PATH coordinated a one-day meeting in March 2025 gathering current and former national regulatory agency staff from a dozen countries, together with vaccine developers, representatives from funding and procurement agencies, and public health and policy officials. The convened participants held spirited discussions on how multiple immune markers and controlled human infection models (CHIM) might contribute to the demonstration of vaccine efficacy. In addition, participants considered the possibility of relying on clinical endpoints when the vaccine components are directed against pathogens causing the same disease syndrome but etiological determination of each component's contribution is not feasible. Regulators welcomed scientifically sound, creative proposals for demonstration of efficacy, and agreed that the benefit-risk of the combination vaccine as a whole should be the primary focus.

Circulating antibodies from previous immune encounters impact subsequent humoral responses. Here, we investigated how local epitope-specific competition shapes ongoing germinal center (GC) responses by delivering an mRNA-LNP-encoded membrane-bound immunogen displaying three conserved HIV-1 envelope (Env) epitopes to mouse models bearing B cell receptors (BCRs) of defined affinities. High-affinity B cells exhibited shorter GC residency than lower-affinity counterparts. B cells engaged GC reactions at equivalent rates in the presence or absence of clonal lineages binding the same epitope with similar affinities; however, higher-affinity clones suppressed lower-affinity counterparts targeting the same epitope. Spatial transcriptomics revealed plasma-like cells within and adjacent to the GC, and early immunoglobulin G (IgG) was detectable in draining lymph nodes. Our findings suggest that a self-modulating local antibody feedback loop limits epitope-specific recognition-dampening selection for higher-affinity B cells and facilitating epitope spreading by redirecting the response toward alternative epitopes.

Current immunization strategies to elicit broadly neutralizing antibodies (bnAbs) against HIV-1 generally propose complex, multiboost regimens. In rhesus macaques, simian-human immunodeficiency virus (SHIV) infection rapidly drives the development of some bnAb classes sharing structural similarities with those in humans. Here, we generated a knockin (KI) mouse model with B cells bearing the unmutated common ancestor of a V2 apex-targeted bnAb lineage, V033-a. A single immunization with a germline-targeting native-like trimer, Q23-APEX-GT1, recapitulated the ontogeny of the mature rhesus bnAb in KI mice, including rare, disfavored somatic mutations. Resulting antibodies exhibited potent neutralization against a broad panel of heterologous HIV-1 strains. Boosting with Env escape mutant trimers further improved breadth and potency, and cryo-electron microscopy analysis revealed the structural basis for heterologous neutralization breadth. Nonhuman primate and mouse models combined with structure can serve as a platform for identifying and validating immunogens that streamline HIV vaccination regimens.

Transmembrane glycoproteins of enveloped viruses are targets of neutralizing antibodies and essential vaccine antigens. mRNA-LNP technology allows in vivo expression of transmembrane glycoproteins, but in vitro biophysical characterization of transmembrane antigens and analysis of post-immunization antibody responses typically rely on soluble proteins. Here, we present a platform for assembling transmembrane glycoprotein vaccine candidates into lipid nanodiscs. We demonstrate the utility of nanodiscs in HIV membrane proximal external region (MPER)-targeting vaccine development by binding assays using surface plasmon resonance (SPR), ex vivo B cell sorting with fluorescence-activated cell sorting (FACS), and by determining the structure of a prototypical HIV MPER-targeting immunogen nanodisc in complex with three broadly neutralizing antibodies (bnAbs), including MPER bnAb 10E8, to 3.5 Å by cryogenic electron microscopy (cryo-EM), providing a template for structure-based immunogen design. To demonstrate general applicability we characterize Ebola virus glycoprotein nanodiscs. Overall, the platform offers a tool for accelerating development of next-generation vaccines.

Background/Objectives: The rapid development of safe and efficacious vaccines is often hindered by extensive, mandated non-clinical safety evaluations in animals. With the aim to provide scientific evidence supporting a "vaccine platform approach", here we present the complete non-clinical studies for two investigational vaccines, GRAd-COV2 and GRAdHIVNE1, based on GRAd, a gorilla-derived group C adenoviral vector. Methods: The biodistribution of GRAd genomes following the intramuscular administration of the vaccines was assessed in rats by a sensitive qPCR method. Local tolerance and systemic toxic effects were evaluated in single- and repeated-dose toxicity studies in rabbits. Results: GRAd-COV2 and GRAdHIVNE1 were well-tolerated. Distribution was highly confined to the injection site and draining lymph nodes, and toxicity profile consisted of transient, non-adverse inflammatory responses, while the expected immune responses to the encoded antigens were successfully induced. Notably, both vaccines demonstrated a consistent safety profile despite transgene and backbone differences, comparable to other replication-defective adenoviral vectors. Conclusions: The established non-clinical safety profile of the GRAd platform provides a robust foundation for a more efficient and streamlined regulatory pathway. By leveraging this prior knowledge, future GRAd-based vaccines can achieve accelerated clinical development while fully adhering to the ethical principles of replacement, reduction, and refinement of animal use in research.

Priming rare subdominant precursor B cells in germinal centers (GCs) is a central goal of vaccination to generate broadly neutralizing antibodies (bnAbs) against HIV. Multivalent immunogen display on protein nanoparticle scaffolds can promote such responses, but it also generates scaffold-specific B cells that could theoretically limit bnAb precursor expansion in GCs. We rationally designed DNA origami-based virus-like particles (DNA-VLPs) displaying a germline-targeting HIV envelope protein immunogen, which elicited no scaffold-specific antibody responses. Compared with a state-of-the-art clinical protein nanoparticle, these DNA-VLPs increased the expansion of epitope-specific GC B cells relative to off-target B cells and enhanced expansion of bnAb-lineage B cells in a humanized mouse model of CD4 binding site priming. Thus, minimizing off-target responses enhances bnAb priming and indicates that DNA-VLPs are a promising vaccine platform.

Introduction: HIV testing is at the cornerstone for interventions aimed at controlling the number of new HIV infections in Sub-Saharan Africa. The current HIV RDTs in use are not able to identify acute HIV infection. Methods: We collaborated with 10 health centers in Uganda, offering HIV voluntary counselling and testing services to obtain 'left-over' EDTA blood samples after HIV RDT. We investigated for HIV RNA using the GeneXpert HIV-1 Viral Load assay among individuals who had non-reactive Determine (HIV negative) or an inconclusive test result (Indeterminate). We used a formula from Brookmeyer and Quinn to estimate HIV incidence. The study started on the 17th February 2021 and ended on the 24th December 2021. Study activities were suspended between June and September 2021, due to travel restrictions instituted by government of Uganda aimed at reducing levels of connectivity between and within communities, to curb the rising number of COVID-19 infections. Results: Fourteen of 4,941 'left-over' plasma samples retested were reactive with a mean pVL of 1.5X106 copies per ml. The overall HIV incidence per 100 person years was 7.4 (95% CI: 4.3-12.5): higher for females at 9.4 (95% CI: 5.2-16.9) compared to 4.2 (95% CI: 1.3-12.9) for males. The HIV Incidence per 100 person years was 7.8 (95% CI: 3.7-16.4) before travel restrictions were instituted and 7.0 (95% CI: 3.3-14.7) after they were lifted. Conclusion: We demonstrated that incident HIV infection can be determined without assembling a longitudinal cohort. GeneXpert HIV-1 Viral Load assay can detect acute infection. There is no evidence that COVID-19 travel restrictions was associated with increased HIV Incidence rates in our communities.

Applying cryoelectron microscopy (cryo-EM) to small protein complexes is usually challenging due to their lack of features for particle alignment. Here, we characterized antibody responses to 21 kDa human immunodeficiency virus (HIV) membrane-proximal external region germline-targeting (MPER-GT) immunogens through cryo-EM by complexing them with 10E8 or Fabs derived from MPER-GT-immunized animals. Distinct antibody-antigen interactions were analyzed using atomic models generated from cryo-EM maps. Mutagenesis screening revealed that off-target monoclonal antibodies (mAbs), which do not compete with 10E8, bind non-MPER epitopes, and the binding of two most dominant epitopes were verified by cryo-EM. The structures of 10E8-class on-target Fabs showed binding patterns that resemble the YxFW motif in the 10E8 heavy chain complementarity-determining region 3 (HCDR3) loop. Additionally, we demonstrate that high-resolution maps can be generated from heterogeneous samples with pooled competing Fabs. Overall, our findings will facilitate the optimization of MPER-GT antigens and push the size limit for cryo-EM-based epitope mapping with smaller antigens and heterogeneous antibody mixes.

Background: Monoclonal antibodies (mAbs) are an increasingly essential class of medicines across many disease areas. In the human body, there are five antibody isotypes, each with potential therapeutic benefits for different disease indications. However, 97% of all clinically approved mAbs are produced as the IgG isotype, largely due to challenges associated with recombinantly producing non-IgG isotypes like IgM or IgA, which have additional N-linked glycan sites and can present as multivalent oligomers. One potential way to circumvent this challenge is to express mAbs in situ using mRNA encapsulated in lipid nanoparticles (LNP), bypassing the need for recombinant protein production. Objective: Here, we demonstrate the feasibility of expressing a mAb as both IgG and IgA in non-human primates (NHPs) using mRNA-LNPs. Methods: We express ePGDM1400v9, a broadly neutralizing mAb targeting human immunodeficiency virus (HIV), in both IgG1 and IgA2 formats by infusing NHPs with LNPs containing the appropriate mRNAs. Results: Though IgA2 expression levels were low, both formats were detectable in serum within one day of LNP infusion in all NHPs, and both were detectable in mucosal secretions of most animals. Importantly, serum mRNA-produced IgG1 and IgA2 retained HIV-neutralizing function. Furthermore, mass spectrometry analysis confirmed that mAbs of either isotype produced in situ exhibited glycosylation patterns highly similar to that of native antibody, which is likely to confer therapeutic advantages. Conclusion: Altogether, this work demonstrates that mRNA-LNPs can be used to express native like mAbs of non-IgG isotypes in primates at detectable levels and enables further development and optimization of non-IgG mAb constructs.

Rhesus macaques (RMs) are a vital model for studying human disease and are invaluable to preclinical vaccine research, particularly for the study of broadly neutralizing antibody responses. Such studies require robust genetic resources for antibody-encoding genes within the immunoglobulin (IG) loci. The complexity of the IG loci has historically made them challenging to characterize accurately. To address this, we developed experimental and computational methodologies to generate a collection of integrated antibody repertoire and long-read genomic sequencing data in 106 Indian-origin RMs. We created a resource of IG heavy- and light-chain variable (V), diversity (D), and joining (J) alleles, as well as leader, intronic, and recombination signal sequences (RSSs). This includes the curation of 1,095 previously unidentified alleles, unveiling tremendous diversity and expanding existing IG allele sets by 40%. This publicly available, continually updated resource (https://vdjbase.org/reference_book/Rhesus_Macaque) provides the foundation for advancing RM immunogenomics, vaccine discovery, and translational research.

RNA-based vaccines have emerged as a highly effective delivery platform. However, this approach eliminates the possibility for immunogen purification, common in manufacturing of recombinant immunogens. In HIV-1 vaccine design, this is of particular importance because non-native epitopes can compromise the desired immune response, and native immunogen assembly is important for presentation of glycan-based epitopes targeted by broadly neutralizing antibodies. Here, we investigate the assembly and glycosylation of the archetypal trimeric HIV-1 immunogen, BG505, in the soluble single-chain format (NFL.664) that bypasses the need of maturation by furin cleavage. We have investigated the presence of the trimer-associated mannose-patch as oligomannose-type structures at these N-linked glycosylation sites are indicative of native-like glycoprotein structure. Despite the presence of features of native-like glycosylation, both electron microscopy and glycopeptide analysis indicated the presence of a sub-population of non-native material. We also investigated the glycosylation of material derived from cell-types that likely produce immunogens near the site of intramuscular RNA injection. We show that replicon-transformed dendritic and muscle cell lines generate immunogens displaying similar oligomannose-type glycan content, whereas sites presenting complex-type glycosylation differed substantially in the levels of glycan processing. Overall, the control of the immunogen assembly by protein engineering is sufficient to drive native-like glycosylation at the majority of glycosylation sites independent of producer cells. Furthermore, we explored the engineering of RNA immunogens to improve glycan site occupancy. Controlling immunogen assembly at the nucleotide level offers a route to enhanced RNA-based immunogens.

Background: Human immunodeficiency virus type 1 (HIV-1) is one of the fastest-evolving human pathogens. Understanding HIV-1 transmission, within-host adaptation, and evolutionary dynamics is pivotal for development of interventions and vaccines. HIV-1 infection is generally caused by a single transmitted founder virus (TFV), and TFV sequences are typically obtained using single genome amplification (SGA). However, suboptimal sample quality can cause sequencing failures, representing considerable losses considering the scarcity of acute HIV-1 infection (AHI) samples. Sequencing failures may be mitigated by molecular cloning (MC), which can be less vulnerable to sample quality but more susceptible to polymerase chain reaction (PCR) errors. Here, we explore the feasibility of supplementing SGA with MC data using samples from clinical and research cohorts to determine whether sequence diversity and evolutionary rate estimates are comparable between the techniques. Methods: Plasma samples were selected from participants with documented AHI from an East African research cohort (the International AIDS Vaccine Initiative, 2006-2011) and a clinical cohort from Sweden (1983-2011). SGA and MC sequencing were done on the HIV-1 env V1-V3 region (~940 base pairs). Within-host sequence diversity was determined from maximum likelihood phylogenetic trees, and evolutionary rate by Bayesian phylogenetic analysis. Highlighter plots, Hamming distances, and assessment of star phylogenies were used to quantify TFVs. Results: One hundred participants (median age 30.3 years, 15% female), contributing 350 samples from four longitudinal time points, 10-540 days post-infection, met the inclusion criteria. SGA succeeded on 90% of research cohort and 48% of clinical cohort samples. Comparative analysis of linked SGA and MC data from 10 samples indicated that approximately eight sequences were necessary for diversity estimates. Consistently higher sequence diversity was observed among MC relative to SGA sequences (median [IQR]: 0.009 [0.003, 0.015] and 0.004 [0.001, 0.012] substitutions/site, P = .002), whereas evolutionary rates were comparable between the two methods (0.016 [0.012, 0.019] and 0.011 [0.008, 0.020] substitutions/site/year, P = .232). Five participants with samples obtained within 45 days post-infection were eligible for TFV quantification, and all found to have one TFV using both techniques. Conclusion: MC data is a suitable supplement for SGA-based HIV-1 studies to preserve the value of precious samples for analysis of evolutionary rate, but not for sequence diversity.