Single-dose VSV-based vaccine protects cynomolgus macaques from disease after Taï Forest virus infection

ABSTRACT

Taï Forest virus (TAFV) is a lesser-known ebolavirus that causes lethal infections in chimpanzees and is responsible for a single human case. Limited research has been done on this human pathogen; however, with the recent emergence of filoviruses in West Africa, further investigation and countermeasure development against this virus is warranted. We developed a vesicular stomatitis virus (VSV)-based vaccine expressing the TAFV glycoprotein as the viral antigen and assessed it for protective efficacy in nonhuman primates (NHPs). Following a single high-dose vaccination, NHPs developed antigen-specific binding and neutralizing antibodies as well as modest T cell responses. Importantly, all vaccinated NHPs were uniformly protected from disease after lethal TAFV challenge while the naïve control group succumbed to the disease. Histopathologic lesions consistent with filovirus disease were present in control NHPs but were not observed in vaccinated NHPs. Transcriptional analysis of whole blood samples obtained after vaccination and challenge was performed to gain insight into molecular underpinnings conferring protection. Differentially expressed genes (DEG) detected 7 days post-vaccination were enriched to processes associated with innate immunity and antiviral responses. Only a small number of DEG was detected in vaccinated NHPs post-challenge while over 1,000 DEG were detected in control NHPs at end-stage disease which mapped to gene ontology terms indicative of defense responses and inflammation. Taken together, this data demonstrates the effective single-dose protection of the VSV-TAFV vaccine, and its potential for use in outbreaks.

KEYWORDS:

• Filovirus
• Taï forest ebolavirus
• vesicular stomatitis virus
• live-attenuated vaccine
• nonhuman primate
• transcriptomics

Introduction

The Filoviridae family in the Mononegavirales order consists of six genera each represented by one virus: Zaire ebolavirus (represented by Ebola virus; EBOV), Sudan ebolavirus (Sudan virus; SUDV), Reston ebolavirus (Reston virus; RESTV), Taï Forest ebolavirus (Taï Forest virus; TAFV), Bundibugyo ebolavirus (Bundibugyo virus; BDBV), and Bombali ebolavirus (Bombali virus; BOMV). These are negative-sense, single-stranded RNA viruses that can cause Ebola disease (EBOD) [1]. EBOV, SUDV, BDBV, and TAFV have caused recorded disease outbreaks in humans and nonhuman primates (NHPs) while RESTV is considered to be apathogenic in humans [1]. EBOD symptoms include fever, myalgia, diarrhoea, and vomiting resulting in severe dehydration. Patients either recover or deteriorate further with hemorrhages, neurologic manifestations, and multi-organ failure [2].

TAFV is a lesser-known ebolavirus first identified in 1994 in the Parc National de Taï of Côte d’Ivoire following several outbreaks in chimpanzees [3]. There has only been a single non-lethal human case of a TAFV infection following the 1994 chimpanzee outbreak. A 34-year-old female ethologist was exposed to TAFV during a wild chimpanzee necropsy and presented 8 days later with classical EBOD symptoms including a macular rash, headache, diarrhea, vomiting, and weight loss [4,5]. This individual experienced temporary hair loss and was fully recovered after six weeks. It is unclear how the chimpanzees were originally infected with TAFV; however, it is obvious that TAFV can be transmitted to humans. Only limited research has been performed on this human-pathogenic filovirus. With the recent emergence of filoviruses in West Africa we sought to address existing knowledge gaps on the pathogenicity of this virus to inform outbreak response preparedness.

The U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) have approved two vaccines for EBOV for human use: the single-dose Ervebo vaccine and the prime/boost Zabdeno/Mvabea vaccine, respectively [6]. The Ervebo vaccine is based on vesicular stomatitis virus (VSV) expressing the EBOV glycoprotein (GP) (rVSV-ZEBOV or VSV-EBOV) which is a live-attenuated vaccine. Ervebo is a single dose, fast-acting vaccine that elicits a protective humoral immune response against EBOV [6,7].

We developed a TAFV-specific vaccine based on the VSV platform with the TAFV GP as the viral antigen [8,9] and investigated the protective efficacy of a single high-dose of VSV-TAFV. The crab-eating macaque (Macaca fascicularis or Cynomolgus macaque) is susceptible to TAFV infection [10], therefore, this NHP species was used here to assess TAFV disease and pathogenesis. NHPs were either vaccinated with VSV-TAFV one month prior to challenge with a lethal dose of TAFV or remained as naïve controls. Our results demonstrate that VSV-TAFV is a viable, single-dose vaccine and ideal for use in emergency situations including outbreaks.

Materials and methods

Ethics statement

All infectious work with TAFV was performed following standard operating procedures (SOPs) approved by the Rocky Mountain Laboratories (RML) Institutional Biosafety Committee (IBC) in the biosafety level 4 (BSL4) laboratory at RML, Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health. Animal work was performed in accordance with the recommendations described in the Guide for the Care and Use of Laboratory Animals of the National Institute of Health, the Office of Animal Welfare and the United States Department of Agriculture and was approved by the RML Animal Care and Use Committee (ACUC; protocol #2021-018-E). Procedures were conducted in animals anesthetized with ketamine by trained personnel under the supervision of veterinary staff. Animals were housed in adjoining individual primate cages that enabled social interactions, under controlled conditions of humidity, temperature, and light (12 h light:12 h dark cycles). Food and water were available ad libitum. Animals were monitored and fed commercial monkey chow, treats, and fruit at least twice a day by trained personnel. Environmental enrichment consisted of commercial toys, music, and video. Endpoint criteria as specified by RML ACUC-approved clinical score parameters (as previously described [11]) were used to determine when animals were humanely euthanized.

Study design

Eight Macaca fascicularis (cynomolgus macaques, 3 female and 5 male) of Cambodian origin, 3–4 years old and 2.5–5.0 kg in weight were included in this study. The NHPs were herpes B virus positive. A group of 4 NHPs (2 female and 2 male) were vaccinated with 1 × 10⁷ PFU of VSV-TAFV IM on −28 DPC and 4 NHPs remained unvaccinated as naïve controls (1 female and 3 male). On 0 DPC, all 8 NHPs were challenged IM with a target dose of 1,000 PFU of TAFV (1,125 TCID₅₀/mL) [12]. Physical examinations and blood draws were performed on −28, −21, −14, −7, 0, 1, 3, 6, 8, 11, 14, 21, 28, 35, and 42 DPC. The animals were observed at least twice daily for clinical signs of disease according to a RML ACUC-approved scoring sheet and humanely euthanized when they reached endpoint criteria. The study end point was 42 DPC when all surviving NHPs were humanely euthanized.

Cell lines, challenge viruses and VSV-based vaccines

VeroE6 cells were grown at 37°C and 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) (Sigma-Aldrich, St. Louis, MO) containing 10% fetal bovine serum (FBS) (Wisent Inc., St. Bruno, Canada), 2 mM L-glutamine (Thermo Fisher Scientific, Waltham, MA), 50 U/mL penicillin (Thermo Fisher Scientific), and 50 μg/mL streptomycin (Thermo Fisher Scientific). BHK-T7 (baby hamster kidney) cells expressing T7 polymerase were grown at 37°C and 5% CO₂ in minimum essential medium (MEM) (Thermo Fisher Scientific) containing 10% tryptose phosphate broth (Thermo Fisher Scientific), 5% FBS, 2 mM L-glutamine, 50 U/mL penicillin, and 50 μg/mL streptomycin. The TAFV stock was isolated from a sample obtained from the human case from the Republic of Côte d’Ivoire in 1994 [13] and a passage 2 virus preparation was used in this study [12]. Deep sequencing revealed no contaminants; however, 3 base pair changes (2 of them coding) were noted when compared to reference sequence NCBI KU182910 (Table S1). The VSV-TAFV vaccine expressing TAFV GP was previously described [8]. VSV-TAFV-GFP was generated by adding the GFP gene into the VSV genome downstream of the TAFV GP and recovered following standard protocols [8].

Heamatology and serum biochemistry

Blood samples were collected in tubes containing EDTA where the total cell counts were determined via Idexx ProCyte DX analyzer (Idexx Laboratories, Westbrook, ME). Serum biochemistry was analysed on a Vetscan 2 using Preventive care profile disks (Abaxis, Union City, CA).

RNA extraction and RT-qPCR

Whole blood (EDTA) and swab (oral, nasal, rectal, urogenital) samples were extracted with QIAmp Viral RNA Mini Kit (Qiagen, Hilden, Germany) and tissues (maximum of 30 mg/tissue) were extracted with RNeasy Mini Kit (Qiagen) according to manufacturer specifications. One-step RT-qPCR was performed with QuantiFast Probe RT–PCR + ROX Vial Kit (Qiagen) with specific TAFV-L primer-probe set as described previously [14] on the Rotor-Gene Q (Qiagen). RNA from the TAFV viral stocks were extracted via QIAmp Viral RNA Mini Kit (Qiagen) and used alongside samples as standards with known median tissue culture infectious dose (TCID₅₀) concentrations.

Enzyme-linked immunosorbent assay

Serum was inactivated by gamma-irradiation (4 MRad) [15] and removed from the maximum containment laboratory according to SOPs approved by the RML IBC. TAFV GP IgG levels were measured in serum samples at a 1:1000 dilution using the human anti-TAFV GP IgG ELISA kit (Alpha Diagnostic International, San Antonio, TX) according to manufacturer specifications. The kit was modified by using the anti-monkey IgG HRP secondary antibody (Alpha Diagnostic International) instead of anti-human IgG HRP. IFNα levels were measured in undiluted serum samples using the VeriKine Cynomolgus/Rhesus IFN Alpha Serum ELISA kit (PBL, Piscataway, NJ) according to manufacturer specifications.

Virus neutralization assay

Neutralization of irradiated and heat-inactivated serum samples were assessed in Vero UNC cells. Neutralization assays with VSV-TAFV-GFP were performed as previously described [16]. Samples were run on the FACSymphony A5 Cell Analyzer (BD Biosciences, Mississauga, ON, Canada) where the GFP-positive cell count was measured and data analysed on FlowJo V10.

Serum cytokine and chemokine levels

Serum samples were assessed using the Milliplex MAP Non-Human Primate Cytokine Magnetic Bead kit (Millipore, Burlington, MA) by diluting in serum matrix (1:2) according to manufacturer specifications. Samples were read on the Bio-Plex 200 system (Bio-Rad, Hercules, CA).

Cellular immune responses

Peripheral blood mononuclear cells (PBMCs) were isolated from EDTA whole blood which was overlaid on a Histopaque®−1077 density gradient and separated according to manufacturers’ instructions. Isolated PBMCs were resuspended in FBS with 10% DMSO and frozen at −80°C until analysis. For cellular immune response analysis, cells were in duplicate stimulated with 5μg/ml TAFV GP peptide pool, media, cell stimulation cocktail (containing PMA-Ionomycin, Biolegend), or SARS-CoV-2 nucleocapsid peptide pool (negative control stimulation) together with 5μg/ml Brefeldin A (Biolegend). Cells were surface stained with Live/Dead-UV450, CD45-BV786, CD3-FITC, CD4-PerCP Cy5.5, CD8-PeTexas Red, CD69-AF700, CCR7-BV605, CD45-RA-APC, and CD11b-BUV660. Cells were fixed with 4% paraformaldehyde and stained intracellularly with IFN-γ-BV421 and TNFα-PE diluted in Perm-Wash buffer (Biolegend). Sample acquisition was performed on a Cytoflex-S (Beckman Coulter) and data analysed in FlowJo V10. The gating strategy is shown in Figure S1.

Histology and immunohistochemistry

Necropsies and tissue sampling were performed according to IBC-approved SOPs. Harvested tissues were fixed for eight days in 10% neutral-buffered formalin, embedded in paraffin, processed using a VIP-6 Tissue Tek (Sakura Finetek, USA) tissue processor, and embedded in Ultraffin paraffin polymer (Cancer Diagnostics, Durham, NC). Samples were sectioned at 5 μm, dried overnight at 42°C, and resulting slides were stained with heamatoxylin and eosin. Specific anti-VP40 immunoreactivity was detected using a previously described cross-reactive anti-EBOV VP40 antibody (a generous gift by Dr. Yoshihiro Kawaoka, University of Wisconsin–Madison) at a 1:1000 dilution [17,18]. The immunohistochemical assay was carried out on a Discovery ULTRA automated staining instrument (Roche Tissue Diagnostics) with a Discovery ChromoMap DAB (Ventana Medical Systems) kit. All tissue slides were evaluated by a board-certified veterinary pathologist.

Library preparation and sequencing

Libraries were generated via the NEBNext Ultra II Directional RNA Library Prep Kit per manufactures instructions with rRNA depletion by the NEBNext rRNA Depletion kit. An Agilent 2100 Bioanalyzer was used to assess cDNA library quality and concentration before multiplexing and sequencing on the Illumina NextSeq 2000 platfrorm.

Bioinformatics

The Trim Galore package was used to trim raw sequences to 70 bp and an average Phred score of 30. Trimmed sequences were aligned to the Macaca fascicularis genome “Macaca_fascicularis.Macaca_fascicularis_6.0.dna.toplevel.fa” using hisat. Genes were annotated with Ensembl file “Macaca_fascicularis.Macaca_fascicularis_6.0.106.gtf.” Raw gene counts were identified by the summarizeOverlaps package before being converted to transcripts per kilobase million (TPM).

Gene expression analysis was performed using three approaches: (1) EdgeR, (2) STEM, and (3) MaSigPro [19–21]. The EdgeR package identifies differentially expressed genes (DEGs) in one condition relative to another. All DEGs were identified relative to either −28 DPC (Figure S1) or 0 DPC (Figures 4 and 5, S5). The −28 and 0 DPC heatmap data were included as a reference for gene comparisons. DEGs were then filtered for only those with an FDR ≤ 0.05, fold change ≤ −1 or ≥ 1, and encoded for human protein-coding homologs. The Short Time Series Expression Miner (STEM) software was used to identify significant patterns of longitudinal gene expression change [19] within each condition. The MaSigPro package was used to identify significant longitudinal gene expression profile changes between conditions using a two-way forward regression strategy. The 11 DPC data set from the control group includes all 4 NHPs at time of euthanasia. Gene ontology (GO) terms were identified using Metascape [22]. All heatmaps, bar plots, bubble plots, and Venn diagrams were made using R package ggplot2.

Statistical analysis

Log-rank (Mantel–Cox) test was performed for survival rates between vaccinated and control groups. Kruskal–Wallis test with a Dunn’s multiple comparison test was performed between groups for the neutralization assay. Friedman test with a Dunn’s multiple comparison test was used for the longitudinal cellular immune response comparisons. Multiple Mann–Whitney test was used between groups for 0 DPC comparison of cellular immune responses. Analysis was performed using GraphPad Prism Software (v. 9.3.1).

Results

VSV-TAFV protects NHPs from lethal TAFV infection

We conducted a VSV-TAFV efficacy study by intramuscularly (IM) vaccinating four NHPs with 1 × 10⁷ plaque-forming units (PFU) of VSV-TAFV 28 days before challenge (−28 days post-challenge; DPC) with four additional naïve NHPs serving as controls. The vaccination did not result in adverse events in NHPs as evidenced by normal values resulting from serum chemistry analysis (Figure S2). To gain a deeper understanding of the molecular underpinnings of the vaccine-elicited responses prior to TAFV challenge, longitudinal transcriptional changes in whole blood samples were assessed in vaccinated NHPs by RNA sequencing (RNA-Seq) at −21 and 0 DPC relative to −28 DPC (Figure S3A). Functional enrichment analysis mapped to gene ontology (GO) terms associated with “inflammatory response” and “mononuclear cell migration” (Figure S3B) as well as upregulated differentially expressed genes (DEG) associated with innate immunity and antiviral responses (AQP9, LTF, S100A12, RNASE6) (Figure S3C).

On 0 DPC, all 8 NHPs were IM-challenged with 1,000 PFU of TAFV. After TAFV challenge, all vaccinated NHPs were uniformly protected from disease, however, the control NHPs met endpoint criteria between 8 and 11 DPC and were euthanized (Figure 1(A) and 1(B). These control NHPs presented with characteristic EBOD-like disease including fever (Figure S2A), viremia (Figure 1(C)), thrombocytopenia (Figure 1(D)), and elevated liver (Figure 1(E) and 1(F)) and kidney (Figure 1(G)) enzyme levels. At the time of euthanasia, high levels of TAFV RNA were detected in all tissues collected from these NHPs (Figure S4G). Since filovirus transmission occurs through bodily fluids, oral, nasal, rectal, and urogenital swabs were collected throughout the acute phase of the disease. Detectable amounts of TAFV RNA in most swab samples were only present in control NHPs beginning 8 DPC while no TAFV RNA was detected in any of the samples collected from vaccinated NHPs (Figure S4C–F). Finally, only control NHPs had increased inflammatory chemokine and cytokine responses indicative of the cytokine storm associated with EBOD (Figure S5)[23].

Figure 1. Clinical and serological findings in cynomolgus macaques after TAFV infection. Cynomolgus macaques (n = 4) were vaccinated with 1 × 10⁷ PFU of VSV-TAFV 28 days before challenge. All NHPs (n = 8) were challenged with 1 × 10³ PFU of TAFV on 0 DPC. (A) Survival curve, (B) clinical score, (C) TAFV RNA in the blood, (D) platelet counts, and (E-F) liver and (G) kidney enzyme levels for the first 2 weeks after challenge are depicted. (H) TAFV GP-specific IgG levels in serum throughout the study. (I) Serum neutralization presented as 50% reduction of GFP-positive cells (FRNT₅₀) at the time of vaccination (−28 DPC), challenge (0 DPC) and euthanasia (42 DPC; study end). ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen. Statistical significance calculated by Mantel-Cox test or Kruskal-Wallis test with a Dunn’s multiple comparisons are indicated as **p < 0.01.

Histopathologic evaluation after TAFV infection

No significant histopathologic changes attributable to EBOD were observed in the vaccinated NHPs at 42 DPC (Figure 2(A) and 2(C)). In contrast, a detailed investigation of liver, spleen, and peripheral lymph nodes of the control NHPs collected at end-stage disease revealed histologic lesions consistent with EBOD. Lesions included minimal to mild, widely disseminated, random foci of hepatocellular necrosis accompanied with histiocytic infiltration throughout the liver (Figure 2(E)). Lymphoid necrosis, sinus histiocytosis and rarely, erythrophagocytosis were observed in lymphoid tissues including the spleen of all control NHPs (Figure 2(G)). Immunoreactivity to VP40 was not observed in any cell type in any vaccinated NHP (Figure 2(B) and 2(D)), whereas samples from control NHPs showed immunoreactivity associated with cells primarily of histiocytic lineage in the liver (Kupffer cells) (Figure 2(F)) and spleen (Figure 2(H)).

Figure 2. Histopathology and immunohistochemistry (IHC) in NHPs after TAFV infection. Cynomolgus macaques (n = 4) were vaccinated with 1 × 10⁷ PFU of VSV-TAFV whereas control NHPs remained naïve. All NHPs (n = 8) were challenged with 1 × 10³ PFU of TAFV 28 days after vaccination. Tissue samples were collected at the time of euthanasia (8-11 or 42 DPC for control and vaccinated NHPs, respectively) and stained with heamatoxylin and eosin (H&E) or TAFV antigen by IHC. (A,B) liver and (C-D) spleen samples of vaccinated NHPs. (E-F) liver and (G-H) spleen tissues of control NHPs.

VSV-TAFV induces robust humoral and CD4 T helper responses

The primary protective mechanism of VSV-based filovirus vaccines is the humoral response [24–26]; therefore, we measured binding and neutralizing antibody responses following vaccination and challenge. The vaccinated NHPs developed a robust TAFV GP-specific antibody response as measured in serum samples collected throughout the study. The IgG binding antibody levels peaked 14 days post vaccination (−14 DPC) and remained constant throughout the duration of the study (Figure 1(H)). Neutralization activity was assessed using a VSV-TAFV-GFP assay. We measured a significant increase of neutralizing activity at 42 DPC but not at challenge (0 DPC) compared to the time of vaccination for the vaccinated NHPs (−28 DPC). As expected, control NHPs had low neutralizing activity on 0 DPC indicating a lack of neutralizing antibodies (Figure 1(I)).

Next, we sought to investigate the contribution of T cell responses by flow cytometry analysis of PBMCs. Antigen-specific T cell responses were evaluated at 0 DPC comparing vaccinated and control NHPs by measuring expression of activation markers and cytokine production following stimulation with overlapping peptide libraries covering TAFV GP (Figure S6). A significantly higher frequency of central and effector memory TNFα⁺ CD4⁺ T cells was observed in the vaccinated NHPs compared to the control NHPs (Figure S6 A–D); however, no CD8⁺ T cell responses were detected at this time (Figure S6 E–H). To gain a better understanding of antigen-specific T cell responses from the vaccinated NHP group, PBMCs were evaluated longitudinally at the time of vaccination (−28 DPC), challenge (0 DPC), and on 14 and 28 DPC (Figure 3). The peak cellular response was detected at 0 DPC, with a skew towards antigen-specific CD4⁺ T cell involvement (Figure 3(A)–(D)). A significant increase of activated (AIM+) naïve and central memory (CM) CD4⁺ T cells were detected at this time point (Figure 3(A)–(D)). No CD8⁺ T cell responses were detected at 0 DPC, but CD8⁺ T cell responses were evident after challenge (Figure 3(E)–(H)). EM-RE CD8⁺ responses were mostly present at 14 DPC (Figure 3(H)) while EM CD8⁺ responses were highest at 28 DPC (Figure 3(G)).

Figure 3. Longitudinal analysis of cellular immune responses in vaccinated NHPs. Cynomolgus macaques (n = 4) were vaccinated with 1 × 10⁷ PFU of VSV-TAFV 28 days before challenge with 1 × 10³ PFU of TAFV. Whole blood samples were collected at the time of vaccination (−28 DPC), at challenge (0 DPC), and 14 and 28 DPC. (A) Naive CD4+ T cells, (B) central memory (CM) CD4+ T cells, (C) effector memory (EM) CD4+ T cells, (D) Effector memory re-expressing (EM-RE) CD4+ T cells, (E) naive CD8+ T cells, (F) CM CD8+ T cells, (G) EM CD8+ T cells, and (H) EM-RE CD8+ T cells in PBMC preparations. Statistical significance calculated by Friedman test with a Dunn’s multiple comparison test is indicated as follows *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. “*” represents single timepoint comparisons, and “+” comparisons across multiple timepoints.

We then performed Short Time-series Expression Miner (STEM) analysis to gain a more detailed insight into these longitudinal changes in the vaccinated NHP group. STEM analysis of transcriptional changes during the acute disease phase identified 3 clusters. The 33 genes in cluster 1 were drastically upregulated between 8 and 11 DPC (Figure S7A) and mapped to GO terms such as “metabolic process” and “immune system process” (Figure S7B). Genes within clusters 2 and 3 showed modest gene expression level changes over the course of the acute disease phase (Figure S7A). Functional enrichment analysis indicated that DEG within clusters 2 and 3 mapped to GO terms such as “regulation of defense response” (TLR2, TLR4, NLRP12, SIGLEC10, CD300A, C1RL, IL6R), “myeloid leukocyte activation” (MYD88, TLR4, TLR2), “cellular response to cytokine stimulus” (IL1RAP, IFNGR1, IL6R, IL4R) and “defense response to virus” (S100A8, S100A9, MX2, APOBEC3H, TRIM5, GBP3, IFI6, IFIT2, ISG15) (Figure S7 B,C). Interestingly, expression of genes associated with the interferon (IFN) response (IFIT2, IFIT3, ISG15, OAS2, OAS3, OASL) peaked at 8 DPC and then subsided at 11 DPC which correlates with the increased T cell-mediated responses on 11 DPC (IFNGR1, IFNGR2, IL-16, STAT1, STAT6) (Figure S7C).

TAFV infection leads to a heightened inflammatory response in unvaccinated NHPs

Differences between the vaccinated and control NHPs after TAFV challenge were assessed by transcriptomic analysis relative to 0 DPC for whole blood samples during the acute phase of the disease. The number of DEG in the control group increased dramatically from 3 to 11 DPC, reaching over 1,000 DEG (Figure 4(A)). They were upregulated genes primarily related to inflammatory processes, such as IFN-stimulated genes (ISG; TRIM5, TRIM25, IFIT2, MX1, HERC5, ISG15, DDX60L, STAT1), pattern recognition receptors (PRR; JAK2, MYD88, TLR3, TLR4, STING1), and immune mediators (MYD88, TNF, ISG15, IFNG, CCL2, CXCL10, CCR5, STAT1, STING1, IFI44L, DDX60) (Figure 4(C) and 4(D)). In contrast, a small number of DEG were only identified at 11 DPC in the vaccinated group (Figure 4(A)) which were related to roles in pathogen sensing (RNASE6, TLR8, TLR1, TLR6, TLR8) (Figure 4(C) and 4(D)). The 27 DEG shared between the vaccinated and control groups on 11 DPC (Figure 4(B)) were associated with leukocyte migration and inflammation (Figure 4(C) and 4(D)).

Figure 4. Transcriptional response in vaccinated and control NHPs after TAFV challenge. (A) Bar graph of the number of DEG identified in vaccinated and control NHPs at 3, 8, and 11 DPC relative to 0 DPC. (B) Venn diagram of DEGs between 11 DPC relative to 0 DPC for vaccinated and control animals. (C) Bubbleplot representing GO terms for genes from B. Color indicates – log10(q) and size indicates the number of genes within the GO term. (D) Heatmap of average TPM values for selected genes mapping to GO terms in Panel C.

We used the MaSigPro software to identify longitudinal gene expression signatures to distinguish the vaccinated and control groups (Figure 5). This analysis identified 2 clusters with temporal expression changes that were extremely different between the vaccinated and control groups starting around 8 DPC (Figure 5(A)). Expression of genes within cluster 1 greatly increased in control NHPs, but modestly increased in vaccinated NHPs over time (Figure 5(A)). Enrichment analysis showed that cluster 1 DEG mapped to antiviral response and host defense processes such as “response to virus” (STAT1, STAT2,OAS1, ISG15, MX1, IFIH1, DDX60, IFI44L, TRIM22, CD68, CD81), “regulation of defense response” (BCL6, CD163, CD274, CD300LF, CD300A), “cellular response to cytokine stimulus” (TNFRSF1B, IL4R, IL1RAP, ISG15), “transcription factor binding” (STAT1, STAT3, NFKBIA, RELA, JAK2, JAK3, IRF3), and “inflammatory response” (S100A8, S100A9, MYD88, TNPFAIP6, IL1B, NFKBIB) (Figure 5(B) and 5(C)). In contrast, expression of genes in cluster 2 decreased in control NHPs but increased in vaccinated NHPs (Figure 5(A)). Genes in cluster 2 mapped to processes related to protein folding and trafficking (FKBP8), phosphatase activity (PHOSPHO1), amino acid transport and erythropoiesis (SLC25A39), and cytokinesis (TUBB6, CENPV, SPECC1) (Figure 5(D)).

Figure 5. MaSigPro analysis of transcriptional changes after challenge. (A) Average TPM of the two gene expression clusters identified by MaSigPro using two-way forward regression analysis. (B) Bubbleplot representing GO terms for genes from cluster 1 in panel A. Color indicates – log10(q) and size indicates the number of genes within the GO term. (C) Heatmap of average TPM values for selected genes mapping to GO terms in Panel B. Color is based on scaled and centered TPM values. (D) Heatmap of average TPM values for 7 genes in cluster 2. Color is based on scaled and centered TPM values.

Discussion

While only a single case of TAFV disease was reported in humans [4], TAFV is responsible for lethal outbreaks in chimpanzees in Côte d’Ivoire [3]. Given the continued human encroachment into wild areas, it is likely that increased contact between potentially TAFV-infected animals and humans will occur. However, there are currently no approved TAFV-specific countermeasures and the FDA-approved VSV-based EBOV vaccine has limited to no cross-protective efficacy against other ebolavirus species [27]. Therefore, we designed a TAFV-specific VSV-based vaccine of which a single dose resulted in uniform protection against TAFV infection in NHPs. While the vaccinated NHPs did not show any signs of disease, the unvaccinated control NHPs developed severe signs of disease after TAFV challenge and were humanely euthanized. We did not observe any differences indicative of a sex-based difference to vaccination or challenge in this study, however, with this small number of NHPs the data is limited.

Clinical findings and histopathologic analysis of tissue samples collected at the time of euthanasia in control NHPs revealed characteristic findings of EBOD including thrombocytopenia, elevated liver and kidney parameters, as well as microscopic lesions in the liver and spleen. Viral involvement in hepatic and splenic damage was confirmed by immunohistochemistry and molecular evaluation for viral RNA. A previous study by Geisbert et al. confirms disease in NHPs after TAFV infection where they had three out of five NHPs succumb to the disease [12]. The difference in lethality between their study and the current study could be related to differences in TAFV viral stocks. These findings contrast with TAFV infection of ferrets which developed only mild signs of disease [14] and survived the challenge, thus emphasizing the value of the NHP model to study TAFV disease.

The humoral response has been described to be the primary mechanism of protection by VSV-based filovirus vaccines [24,26] and the VSV-TAFV-specific humoral response followed the pattern observed for VSV-EBOV [28] with a peak binding antibody response −14 DPC. In addition, vaccinated NHPs also developed TAFV-specific neutralizing antibodies albeit at low titres at the time of challenge that increased by the end of the study. In contrast to VSV-MARV and VSV-EBOV vaccination and respective challenges, the TAFV challenge did not appear to boost the TAFV GP-specific IgG response [29,30]. These findings are similar to what occurs in unvaccinated Ebola virus disease survivors where neutralizing antibodies only gradually increase up to one year after infection and may be linked to accumulated somatic hypermutation over time [31]. Together with the lack of viral RNA detected at any time point during the study in samples from vaccinated NHPs, these observations suggest protection from disease and possibly even infection. Moreover, the development of the antibody response was accompanied by the detection of activated central and effector CD4 T cell responses, critical for the development of an effective humoral responses. TNFα- and IFNγ-producing CD4 T cells also have the potential to clear infected cells [32–36]. As observed in previous studies [37–41], limited antigen-specific CD8 T cells were induced by vaccination. These results are supported by our previous findings where we showed that VSV-EBOV-elicited CD8 T cell immunity plays a limited role in mediating protection against EBOV challenge [25]. Interestingly, we measured some increased levels of T cell activation at −28 DPC which can be attributed to cross-reactivity of the stimulated peptides with herpes B virus. We aligned the peptide pool sequences to the herpes B virus protein sequence and looked for homology of at least 8 amino acid overlap since this smaller peptide size is needed for efficient TCR binding [42]. We identified 17 peptides with overlapping reads; 7 of which were associated with the capsid protein. We attributed the reactivity of these peptides to the observed T cell activation in the −28 DPC samples.

The development of the antibody response correlated with the induction of genes important for innate and humoral immune responses at −21 DPC, including genes involved in leukocyte migration, lymphocyte activation, antigen processing and presentation, as well as antibody production. These findings are in line with those obtained with the VSV-EBOV and VSV-MARV vaccines [26,29,43]. The number of DEG and the duration of VSV-TAFV-induced transcriptional changes are similar to the ones elicited by VSV-EBOV, which are fewer DEG for a shorter period of time than VSV-MARV [26]. In addition, both VSV-TAFV and VSV-EBOV vaccines induced expression of genes involved in antiviral immunity, PRR signalling, and inflammation [44]. Furthermore, the rapid activation of innate immune genes and of DEG involved in antiviral responses suggest that VSV-TAFV could be a fast-acting prophylactic vaccine similar to VSV-EBOV which elicited partial protection when delivered 3 days prior to lethal viral challenge [30].

Following TAFV challenge, vaccinated animals elicited a tightly regulated immune response. GO terms were associated with “inflammatory response,” “PRR signaling pathway,” and “response to virus.” These were associated with genes specific to the innate immune response and T cell activation. Specifically, CSF1R which is the receptor for colony stimulating factor 1 that controls production, differentiation, and function of macrophages. NLRP6 which increases the activation of NFκB; a critical regulator of the response to viral infections. CR1L helps with the regulation of complement activation while IL-16 is a modulator of T cell activation and GPX4 protects cells from oxidative damage. All of these genes were upregulated within the vaccinated NHPs and downregulated in the control NHPs. When assessing longitudinal changes in vaccinated NHPs, there was an increase of IFN-related genes on 8 DPC and an increase of T cell-mediated genes on 11 DPC which is interesting since T cell responses are associated with IFN signalling. Therefore, VSV-TAFV-vaccinated NHPs were able to mount a protective and regulated immune response against the lethal TAFV challenge.

In contrast, transcriptional analysis of blood samples in control NHPs had substantial upregulation of genes associated with inflammatory processes, antiviral innate immune responses, leukocyte migration, and myeloid cell activation. Control NHPs had upregulated JAK2 and STAT1 which correlates to the increased IL-10 in the serum since IL-10 is related to regulation of the JAK-STAT pathway and NFκB activity [45]. There was also an upregulation of S100A8 and S100A9 in control NHPs. Interestingly, these genes have been associated with increased MIP-1β and IL-1β levels [46] which we also measured in the serum of the control NHPs in the current study. The increased expression of these genes is in line with the observed uncontrolled inflammation and cytokine storm after other lethal filovirus challenges [26,29]. Several DEG induced by TAFV infection are also modulated by EBOV infection including genes involved in antiviral (STAT1, MX1, IFIH1, DDX60, ISG15, HERC5) and inflammatory (IL6, TLR4, TNF, S100A8) responses. Additional identified common genes with EBOV are PYCARD (apoptosis), CD40LG and SELP (innate leukocyte immunity) [44]. Concurrently, genes involved in T and B cell activation as well as antigen processing were downregulated in TAFV-challenged control NHPs indicative of a disrupted adaptive immune response. Interestingly, there were several upregulated genes associated with the innate immune activation of PAMPs for infectious agents (TLR1 and TLR6) within both vaccinated and control NHP groups.

In conclusion, this study shows that a single dose of the VSV-TAFV vaccine provides complete protection in NHPs against lethal TAFV challenge. While the duration of the protection induced by VSV-TAFV has yet to be established, VSV-based filovirus vaccines have been shown to elicit long-term humoral immunity as demonstrated by VSV-EBOV with single-dose vaccination eliciting antibodies that persist up to 2 years after vaccination [47]. Future work will include protection provided by a pan-filovirus VSV-based vaccine against multiple filovirus species building upon the work by Geisbert et al. who showed that a blended VSV-EBOV, VSV-SUDV, and VSV-MARV vaccine cross-protected 3/3 NHPs from TAFV challenge [12]. However, protection after a single dose is an important feature for vaccines intended to be used to mitigate disease outbreaks. Therefore, this data demonstrating efficacious single-dose protection of the VSV-TAFV vaccine highlights its potential for use in outbreaks.

Competing interests

T.W.G. claims intellectual property regarding recombinant VSV-based vaccines for the prevention and treatment of filovirus infections. All other authors declare that they have no competing interests.

Acknowledgments

We thank members of the Rocky Mountain Veterinary Branch for supporting the NHP studies and members of the Rocky Mountain Research Technology Branch for assistance with sequencing the viral stocks.

Data availability

Data underlying the transcriptome analysis are available under project number PRJNA914302 at https://dataview.ncbi.nlm.nih.gov/object/PRJNA914302?reviewer = h0ggieqj0sd3hfmvslsnnncsp0. All additional data are available in the main text or the supplementary materials.

Disclosure statement

T.W.G. claims intellectual property regarding recombinant VSV-based vaccines for the prevention and treatment of filovirus infections. All other authors declare that they have no competing interests.

Additional information

Funding

This work was funded by the Intramural Research Program, NIAID, NIH. Transcriptome analysis was funded by the University of Kentucky, College of Medicine internal funding awarded to IM.