Preclinical development of a replication-competent vesicular stomatitis virus-based Lassa virus vaccine candidate advanced into human clinical trials

Abstract

Background: Lassa fever (LF) is a zoonotic haemorrhagic disease caused by Lassa virus (LASV), which is endemic in West African countries. The multimammate rat is the main animal reservoir and its geographic range is expected to expand due to influences like climate change and land usage, and this will place larger parts of Africa at risk. We conducted preclinical development on a promising experimental vaccine that allowed its advancement into human trials.

Methods: The LF vaccine is based on a vesicular stomatitis virus (VSV) vector in which the VSV glycoprotein (G) was replaced with the LASV glycoprotein complex (GPC). Earlier studies showed that this vaccine (VSVΔG-LASV-GPC) was efficacious in macaques, thus we regenerated VSVΔG-LASV-GPC using laboratory and documentation practices required to support vaccine manufacturing and human trials. The efficacy of the clinical vaccine candidate was assessed in cynomolgus macaques and more extensive immunologic analysis was performed than previously to investigate immune responses associated with protection.

Findings: A single VSVΔG-LASV-GPC vaccination elicited innate, humoural and cellular immune responses, prevented development of substantial LASV viraemia, and protected animals from disease. Vaccinated macaques developed polyfunctional antibodies and serum was shown to neutralize virus expressing GPCs representative of geographically diverse LASV lineages.

Interpretation: The VSVΔG-LASV-GPC clinical candidate elicited immunity that protected 10 of 10 vaccinated macaques from disease supporting its use in a clinical development program, which recently progressed to phase 2 clinical trials. Moreover, immunologic analysis showed that virus-neutralizing serum antibodies likely played a role in preventing LASV disease in vaccinated macaques.

Funding: This work was supported by the Coalition for Epidemic Preparedness Innovations (CEPI), The National Institute of Allergy and Infectious Diseases (NIAID)/National Institutes of Health (NIH), The Bill and Melinda Gates Global Vaccine Accelerator Program, the Burroughs Wellcome Fund, and financial gifts and support by Nancy Zimmerman, Mark and Lisa Schwartz, and Terry and Susan Ragon.

Keywords: Clinical vaccine candidate; Lassa virus; Neutralizing antibodies; Vesicular stomatitis virus vector.

Fig. 1

VSVΔG-LASV-GPC chimeric virus vaccine. (a) Illustration of the enveloped VSV particle and its (−)ssRNA genome. The genes encoding the N, P, M, G and L proteins are expressed from a single promoter at the 3’ end. Genes closest to the promoter are transcribed more abundantly. (b) Schematic of the replication-competent VSVΔG-LASV-GPC vaccine used in these studies.^, (c) Map of the GPC precursor protein and its proteolytic processing. SSP, stable signal peptide. (d) Nanoflow virometry was used to analyse purified VSVΔG-LASV-GPC vaccine material. Particle counts (ordinate) and large-angle light scatter (abscissa) are shown. (e) Expression and antigenicity of GPC on the cell surface was evaluated by flow cytometry using infected Vero cells and three different GPC-specific mAbs or a negative control antibody specific for HIV-1 Env (PGT-145⁴¹). Intracellular VSV N was stained with mAb 10G4. (f) Purified vaccine material was analysed by Western blot to detect GP1 (mAb 3.3B, lanes 1–2) and GP2 (mAb 22.5D, lanes 3–4). VSV N was detected in lanes 5 and 6 with a polyclonal rabbit antibody. Lanes 2, 4 and 6 contained VSVΔG-LASV-GPC; lanes 1, 3 and 5 contained VSVΔG-MARV-GP as a control. (g) Electron micrograph composition of negative-stained VSVΔG-LASV-GPC particles and reference-free 2D class averages of GPC coating the viral surface. This analysis was conducted with the vaccine material used in this study. (h) 3D reconstruction of GPC on the surface of VSVΔG-LASV-GPC with C3 symmetry applied (left panel), and with the full-length atomic model of GPC (PDB: 7PUY) docked into the reconstruction (right panel).

Fig. 2

VSVΔG-LASV-GPC vaccination and LASV challenge. Cynomolgus macaques were divided into three groups including unvaccinated control macaques that were injected with buffer (n = 3; PBS containing 5% trehalose) and animals injected with 2 × 10⁷ PFU or 2 × 10⁵ PFU of VSVΔG-LASV-GPC (n = 5 per group). Vaccination was performed with a single IM injection, after which samples were collected through day 27 for immunological analysis. On day 28, animals were challenged with a single IM injection of 3.5 × 10³ PFU LASV (Josiah strain, lineage IV) and were monitored for an additional 28 days for signs of LASV disease. Animals exhibiting disease symptoms were euthanized based on predetermined health criteria. (a) Timeline of study activities. (b) Group descriptions. (c) Kaplan–Meier survival curves. One control animal became moribund on day 11 following challenge, the remaining two on day 13. The 10 animals vaccinated with VSVΔG-LASV-GPC survived until the completion of the study. ∗∗, p < 0.01 by log-rank test. (d) Infectious LASV in blood of challenged animals was quantified by plaque assay. LOD, limit of detection.

Fig. 3

Health monitoring and blood chemistry following LASV challenge. (a) Clinical observation scores through day 28 after LASV challenge. (b–f) Key analytes in the blood of animals were quantified from the day of challenge to the end of the study: (b) alanine aminotransferase (ALT); (c) aspartate aminotransferase (AST); (d) albumin (ALB); (e) C-reactive protein (CRP); and (f) calcium (CA). n = 3–5/group.

Fig. 4

Quantification of adaptive immune responses. (a) Seroconversion was monitored by direct ELISA using plates coated with recombinant, covalently linked GP1–GP2 (GPC). (b–d) The presence of peripheral blood T cells specific for LASV GPC or VSV N was quantified by IFN-γ ELISpot assay. PBMCs from vaccinated animals were stimulated in vitro with (b) overlapping peptides spanning the complete sequence of GPC; (c) the same recombinant linked GP1–GP2 protein (GPC) used for ELISA in (a); or (d) overlapping peptides spanning the complete sequence of VSV N. Results are expressed as spot-forming cells (SFC) per one million PBMCs. Box plots indicate the interquartile range and median. Data from viraemic/RNAemic animal are indicated by their identifier (4 and 13). ∗∗, p < 0.01; ∗, p < 0.05 by Kruskal–Wallis test with Dunn’s post-hoc multiple comparisons test. n = 3–5/group.

Fig. 5

Quantification of neutralizing serum antibodies. VSVΔG-LASV-GPC chimeras expressing GPC from LASV lineages I–V and VII or MARV GP (Musoke; VSVΔG-MARV-GP) were used in a PRNT to quantify serum nAbs. Genbank accession numbers for the GPC sequences used in development of the VSVΔG-LASV-GPC chimeras are included in Supplementary Table S2. The ability of serum from day 27 after vaccination to neutralize these virions was evaluated. (a) Schematic representing the replication-competent viruses used in this assay. (b) Serum neutralizing titres. Values plotted are the serum dilutions that reduced the number of plaques by 50% (NT50). Boxplots indicate the interquartile range and median. Data from viraemic/RNAemic animals are indicated by their identifier (4 and 13). This assay may provide a more sensitive assessment of neutralization than single-cycle infection (see Discussion). Results from intraclade Kruskal–Wallis test with Dunn’s post-hoc multiple comparisons test are indicated. ∗, p < 0.05. n = 3–5/group.

Fig. 6

Effector properties of vaccine-elicited antibodies in serum. Serum harvested on day 0, 10 and 27 after vaccination was used to quantify different non-neutralizing effector functions of vaccine-induced serum antibodies. (a) Antibody-dependent cellular phagocytosis (ADCP); (b) antibody-dependent neutrophil phagocytosis (ADNP); (c) antibody-dependent complement deposition (ADCD); and (d) antibody-dependent NK-cell activation as measured by MIP-1β expression (ADNKA). Assays were performed with lineage II or lineage IV GP antigens including a soluble GP that adopts a native-like prefusion conformation (GP prefusion) or a soluble GP1–GP2 fusion protein (GP-link). (e) Polar plots of the median percentile rank of each antibody function, isotype, and FcγR binding titre against LIV GP-link at day 27. Legend is shared with panel (g) and shown on the right indicating rank score (upper right) and colour scheme (lower right). (f) A PCA was built using all available data on antibody features. Colouring corresponds to the detection of viraemia (cf.Fig. 2d). (g) Polar plots of the median percentile rank for all antibody features for animals binned by segregation along PC2 or PC1, as indicated. The legend is shared with (e). n = 3–5/group.

Fig. 7

Effect of IM injection with VSVΔG-LASV-GPC on the whole-blood transcriptome. Gene expression prior to and on days 1 and 3 after vaccination was analysed by RNA-seq on whole-blood samples from the high-dose (2 × 10ˆ7 PFU) and low-dose (2 × 10ˆ5 PFU) groups. (a) Heatmap showing normalized expression values of all differentially expressed genes (DEGs) with unsupervised hierarchical clustering. (b and c) Volcano plots showing changes in expression and the associated statistical significance in the (b) high-dose and (c) low-dose group, respectively, on day 1 after vaccination compared to baseline. Symbol shape indicates whether a given gene was differentially expressed on day 3 after vaccination. (d) Venn diagram showing the overlap of DEGs between the two time points and the two dose groups. (e) Venn diagram of the overlap between VSVΔG-LASV-GPC-induced DEGs and a framework of representative genes identified previously as differentially expressed following vaccination with VSVΔG-ZEBOV-GP in macaques and humans.^,, (f) Transcription factor motifs enriched in the promoter region of VSVΔG-LASV-GPC-induced DEGs, identified by Homer. (g) Heatmap showing mean log₂ fold-changes in expression of genes within gene sets enriched for VSVΔG-LASV-GPC-induced DEGs. Gene sets in the heatmap include those that showed significant enrichment for DEGs in at least one study group at one or both post-vaccination timepoints. Dots in heatmap cells indicate significant enrichment for DEGs. “DC”, “LI” and “HALLMARK” prefixes on gene sets indicate the published collection from which they originated,^,, respectively. n = 3–5/group.