NiV vaccine candidates are immunogenic in mice

NiV vaccine candidates (adjuvanted NiV sG protein, adjuvanted NiV mcsF protein, and ChAdOx1 NiV G) were first evaluated for immunogenicity in BALB/c mice. Responses were compared to those elicited by immunisation with adjuvanted HeV sG protein (the antigen used in the licensed equine HeV vaccine). Serum antibody and splenocyte responses were assessed 21 days after prime and booster immunisations (Fig. 1). A single immunisation with all four vaccines induced antigen-binding antibody titres. Immunisation with NiV sG produced significantly higher NiV-G specific antibody titres at day 42 (determined by ELISA) when compared to serum from ChAdOx1 NiV G immunised mice (p < 0.01) (Fig. 1A). Immunisation with NiV mcsF induced high titres of F-binding antibodies, comparable to the G-binding titres induced by NiV sG. All vaccines induced NiV neutralising antibody titres after a single immunisation, and titres were elevated following the booster immunisation, with only the HeV sG group showing a significant increase in titres following the boost. Unexpectedly, the highest neutralising titres were observed in day 42 sera from HeV sG immunised mice (Fig. 1B). ChAdOx1 NiV G was the only vaccine to show substantial frequency of CD8+ T-cells producing IFN-γ and TNF-α, with a significantly higher frequency observed in the mice culled at day 21 (Fig. 1C). NiV G-specific CD4+ T-cell responses were also most prominent in the ChAdOx1 NiV G immunised mice, also with a significant higher frequency observed in the mice culled at day 21, although they were lower than the CD8+ T-cells (Fig. 1D). Mice immunised with NiV mcsF showed a significant CD8+ T-cell response, although the magnitude was lower than observed in ChAdOx1 NiV G immunised mice (Fig. 1E). The limited impact of booster immunisation on antibody titres and T cell responses were similar to that observed previously in mice immunised with a COVID-19 vaccine candidate24, and suggests that the vaccine dosages used may have saturated the immune response and obscured the ability to discern the impact of the boost.

Fig. 1: Immunisation of mice with NiV sG, ChAdOx1 NiV G or NiV mcsF induces varied antibody and T-cell responses. Mice were immunised on day 0 and 21 by intramuscular inoculation of 5 µg NiV sG, NiV mcsF, or HeV sG proteins in adjuvant, or 1 × 108 IU ChAdOx1 NiV G. A NiV mcsF and sG binding antibody titres (EPT) in serum on day 21 and 42; B Virus neutralising titres in day 21 and 42 serum as assessed by NiV M VNT; C, D CD8+ and CD4+ T-cell cytokine responses after stimulation of splenocytes with NiV G peptides, respectively; E, F CD8+ and CD4+ T-cell cytokine responses after stimulation of splenocytes with NiV F peptides, respectively. Splenocytes from all groups were restimulated with NiV G peptides, whereas splenocytes from the NiV sG, NiV mcsF and adjuvant only groups were restimulated with NiV F peptides. Datapoints represent individual mice, with the bars showing the group mean and error bars represent the standard deviation (SD). Significant differences were determined using two-way ANOVA and signified with the following letter: a—significant difference to adjuvant; b—significant difference to NiV sG; c—significant difference to ChAdOx1 NiV G; d—significant difference to NiV mcsF; e—significant difference to HeV sG. Full size image

NiV vaccine candidates are immunogenic in pigs

Since all three vaccine candidates were immunogenic in mice, they were next evaluated for immunogenicity in pigs (Fig. 2). Responses were once again benchmarked against the HeV sG antigen in adjuvant. All vaccines appeared to show no reactogenicity, with no significant elevation in rectal temperatures nor other clinical signs being observed (Supplementary Fig. 1). After receiving a homologous booster immunisation on day 21, a significant increase in NiV G binding antibody titres was detected in day 42 serum from pigs immunised with NiV sG and ChAdOx1 NiV G (p < 0.001 and <0.01, respectively) (Fig. 2A). NiV G antibody titres in pigs immunised with NiV sG were significantly higher on day 42 compared to pigs immunised with HeV sG (p < 0.0001) (Fig. 2A). The kinetic of this antibody response is shown in Supplementary Fig. 2. All vaccines induced neutralising antibody titres (Fig. 2B). One-week post-boost, there was a significant increase in neutralising titres in NiV sG immunised pigs (p < 0.0001), although all vaccines showed a trend for increased titres following the boost. On day 42, pigs immunised with NiV sG had significantly higher neutralising titres compared to all other vaccine candidates (HeV sG p < 0.01; ChAdOx1 NiV G p < 0.0001 and NiV mcsF p < 0.0001). The ability of sera to reduce NiV glycoprotein mediated cell-to-cell fusion was assessed using the mFIT assay (Fig. 2C). Sera from pigs immunised with NiV sG and NiV mcsF inhibited fusion significantly more than pigs immunised with ChAdOx1 NiV G (p < 0.0001) and HeV sG (p < 0.01 and <0.001, respectively). Like the mouse data, ChAdOx1 NiV G was the only vaccine to show significant production of CD8+ T-cells producing IFN-γ and TNF-α at day 28 (7 days post-boost; p < 0.0001) (Fig. 2D). T-cells producing single cytokines (IFN-γ or TNF-α) are shown in Supplementary Fig. 3. A moderate increase in CD4+ T cells producing IFN-γ and TNF-α was detected after boosting with ChAdOx1 NiV G (Fig. 2E). Pigs immunised with NiV mcsF showed a significant increase in CD8+ T-cells producing IFN-γ and TNF-α at day 14 when compared to day 7 (p < 0.0001) but no boosting effect was observed (Fig. 2F). No notable NiV F specific CD4+ T cell response could be detected (Fig. 2G). The peak in IFN-γ responses at 28 dpv was also shown using an IFN-γ ELISpot assay (Fig. 2H) where, on day 28, the number of IFN-γ secreting cells in PBMC from ChAdOx1 NiV G immunised pigs was significantly higher than all other groups (p < 0.0001). Of note, pigs immunised with NiV sG showed spontaneous secretion of both IFN-γ and TNF-α at all timepoints post-immunisation in both the flow cytometric and IFN-γ ELISpot assays, for reasons that could not be elucidated (Supplementary Figs. 4 and 5).

Fig. 2: Immunisation of pigs with NiV sG, ChAdOx1 NiV G, NiV mcsF similarly induces varied antibody and T-cell responses. Pigs were immunised on day 0 and 21 by intramuscular inoculation of 100 µg NiV sG, NiV mcsF, or HeV sG proteins in adjuvant, or 1 × 109 IU ChAdOx1 NiV G. A NiV mcsF and sG binding antibody titres (EPT) on day 21 and 42; B Virus neutralising titres as assessed by NiV M VNT; C Inhibition of NiV glycoprotein-mediated cell-cell fusion evaluated using day 42 sera; D, E CD8+ and CD4+ T-cell cytokine responses after stimulation of PBMC with NiV G peptide pool, respectively; F, G CD8+ and CD4+ T-cell cytokine responses after stimulation of PBMC with NiV F peptide pool, respectively. H, I IFN-γ ELISpot assay to assess PBMC responses to NiV G and F peptide stimulation, respectively. For A and C, datapoints represent individual pigs, with the bars showing the group mean and error bars represent the SD. For the other panels, datapoints represent the group mean and error bars represent the SD. Significant differences were determined using two-way ANOVA and signified with the following letter: a—significant difference to adjuvant; b—significant difference to NiV sG; c—significant difference from ChAdOx1 NiV G; d—significant difference to NiV mcsF; e—significant difference to HeV sG. Full size image

Prime-boost immunisation of pigs with NiV vaccine candidates significantly reduces virus shedding and loads post-challenge

After determining immunogenicity profiles in pigs, which were unique to each vaccine administered, three vaccine candidates were evaluated in an efficacy study, using the same prime-boost immunisation regimen (the adjuvanted HeV sG protein was not included in this and subsequent studies). Pigs were then challenged on day 42 by oronasal inoculation of NiV M . Since experimental infection of pigs with NiV M did not induce clinical disease, protection was determined by reduction in viral shedding and viral loads in tissues harvested after 6 days; which has been described as the peak of infection15 (Fig. 3). Neutralising antibody titres at days 21 and day 42 (Fig. 3A) were comparable to those seen in the previous immunogenicity study (Fig. 2B). At the point of challenge at day 42 post-immunisation, only pigs receiving NiV sG had a significantly higher level of neutralising antibodies to the unvaccinated control pigs. At 6 days post-challenge, pigs immunised with NiV sG had significantly higher neutralising antibodies than those immunised with ChAdOx1 NiV G (p < 0.05) and NiV mcsF (p < 0.01) and the unvaccinated controls (p < 0.001) (Fig. 3A). At 4 days post-challenge, unvaccinated pigs had significantly higher copies of NiV RNA in both nasal (p < 0.0001) and oral (p < 0.05) swab samples than animals in the three vaccine groups (Fig. 3B, C, respectively). At 6 days post-challenge, significantly lower NiV RNA copies were detected in the nasal swabs of all vaccinated pigs when compared to unvaccinated controls (p < 0.0001) (Fig. 3B), whereas only pigs immunised with NiV sG or ChAdOx1 NiV G had significantly lower (p < 0.0001) NiV RNA in oral swabs compared to the unvaccinated controls (Fig. 3C). Infectious NiV could only be detected in oral swabs from unvaccinated pigs (p < 0.0001) (Fig. 3E). At post-mortem, significantly higher amount of NiV RNA copies were found in all tissues sampled from unvaccinated pigs (p < 0.0001), except the thymus, which was not significantly different from the vaccinated pigs (Fig. 3F). Infectious NiV could not be isolated from tissues of vaccinated pigs but could from unvaccinated controls; with prescapular (p < 0.01) and submandibular lymph nodes (p < 0.0001), tonsil (p < 0.0001), trigeminal ganglion (p < 0.05) and nasal turbinates (p < 0.05) being significantly different from the vaccinated groups (Fig. 3G).

Fig. 3: NiV vaccine candidates are protective in pigs following a prime-boost immunisation regimen. Pigs were immunised on day 0 and 21 by intramuscular inoculation of 100 µg NiV sG or NiV mcsF proteins in adjuvant, or 1 × 109 IU ChAdOx1 NiV G. On day 42, all pigs were challenged by oronasal inoculation with 1 × 105 PFU NiV M . A Neutralising antibody titres as assessed by NiV M VNT; B, C Viral loads in nasal and oral swabs detected post-challenge by RT-qPCR, respectively. D, E Level of infectious NiV from nasal and oral swabs detected post-challenge by virus isolation, respectively. F, G Viral loads from tissue samples at 6 days post-challenge detected by RT-qPCR and virus isolation, respectively. Postmortem tissues collected: PSLN prescapular lymph node, RPLN retropharyngeal lymph node, SMLN submandibular lymph node, TBLN tracheobronchial lymph nodes, Olf bulb olfactory bulb, Trig ganglion trigeminal ganglion. Significant differences were determined using two-way ANOVA and signified with the following letter: a—significant difference to unvaccinated; b—significant difference to NiV sG; c—significant difference from ChAdOx1 NiV G; d—significant difference to NiV mcsF. NS not significant. Full size image

NiV vaccine candidates are not effective in pigs following a single immunisation

After demonstrating efficacy in pigs with all three vaccine candidates after a prime-boost regimen, efficacy after a single immunisation was determined by challenging pigs with NiV M on day 21, 3 weeks post-prime (Fig. 4). Neutralising antibody titres at day 21 (Fig. 4A) were comparable to those seen in the previous immunogenicity study (Fig. 2B) and efficacy study (Fig. 3A). At 6 days post-challenge, neutralising antibody titres from pigs immunised with ChAdOx1 NiV G were significantly higher than the unvaccinated controls (p < 0.01), However, no significant difference in viral loads could be detected between any vaccinated groups when compared to the unvaccinated controls in either nasal (Fig. 4B) or oral swabs (Fig. 4C), or postmortem tissue samples (Fig. 4F). Infectious NiV could only be detected in nasal swabs from unvaccinated pigs on day 4 and 6 post challenge and pigs immunised with NiV mcsF on day 4 post challenge (Fig. 4D). At post-mortem, NiV RNA copies were found in all tissues sampled from unvaccinated pigs (Fig. 4F). Infectious NiV was assessed in samples which gave a CT value of ≤34. Infectious virus was isolated from tissues of pigs vaccinated with NiV sG (olfactory bulb and trigeminal ganglion) but no significance between groups was found. Infectious virus was isolated from all samples from unvaccinated controls except the thymus, liver, and kidney (Fig. 4G).

Fig. 4: NiV vaccine candidates are not effective in pigs following a prime-only immunisation regimen. Pigs were immunised on day 0 by intramuscular inoculation of 100 µg NiV sG or NiV mcsF proteins in adjuvant, or 1 × 109 IU ChAdOx1 NiV G. On day 21, all pigs were challenged by oronasal inoculation of 1 × 105 PFU NiV M . A Neutralising antibody titres as assessed by NiV M VNT; B, C Viral loads in nasal and oral swabs detected post-challenge by RT-qPCR, respectively. D, E Level of infectious NiV from nasal and oral swabs detected post-challenge by virus isolation, respectively. F, G Viral loads from tissue samples at 6 days post-challenge detected by RT-qPCR and virus isolation, respectively. Postmortem tissues collected: PSLN prescapular lymph node, RPLN retropharyngeal lymph node, SMLN submandibular lymph node, TBLN tracheobronchial lymph nodes, Olf bulb olfactory bulb, Trig ganglion trigeminal ganglion. Significant differences were determined using two-way ANOVA and signified with the following letter: a—significant difference to unvaccinated. Full size image

Comparison of vaccine candidate immunogenicity following prime only and prime-boost immunisation regimens

After showing a reduced efficacy from the vaccine candidates 21 days after a single immunisation, the three vaccine candidates were evaluated in pigs in a second immunogenicity study to determine immunological differences between the prime-only and prime-boost regimen over a 4-month time period, which is comparable to the time required to rear a weaned pig to slaughter weight (Fig. 5). No significant differences were detected in NiV G or F specific antibody titres between pigs receiving a single shot or prime boost regimen of either NiV sG, NiV mcsF or ChAdOx1 NiV G at day 21, 42 or 112 (Fig. 5A). Longitudinal NiV G and F binding antibody data is shown in Supplementary Fig. 6. However, at 112 dpv, significantly higher (p < 0.05) neutralising antibody titres were seen in pigs receiving a prime-only regimen of NiV sG compared to a prime-boost regimen of ChAdOx1 NiV G (Fig. 5B). Pigs receiving a prime-only regimen of NiV sG also had significantly higher (p < 0.01) neutralising antibody titter compared to pigs receiving a single dose of NiV mcsF (Fig. 5B). A similar trend was seen in neutralising titres using the pseudoVNT, but no significant differences were detected (Supplementary Fig. 7). All groups showed an upward trend or sustained level in neutralising titres throughout the study period. It is of interest that by day 112, pigs receiving a single immunisation with NiV sG or ChAdOx1 NiV G had similar titres to their matching vaccine group receiving a prime-boost regimen (Fig. 5B). However, there was a clear effect of the booster immunisation with NiV mcsF to augment neutralising antibody titres. When determining day 112 serum inhibition of cell-to-cell fusion, again there was no significant difference between pigs receiving the prime-only compared to prime-boost regimens in any of the vaccine groups (Fig. 5C). However, pigs receiving ChAdOx1 NiV G had significantly less ability to inhibit fusion when compared to pigs receiving NiV mcsF under both regimens (p < 0.05 and <0.01). Pigs immunised with ChAdOx1 NiV G again showed the highest level of IFN-γ producing cells which appeared sustained. Pigs receiving the prime-boost regimen had the highest levels of IFN-γ producing cells; significantly greater than other groups (p < 0.01) apart from the prime-only ChAdOx1 NiV G group (Fig. 5D). NiV mcsF prime-only vaccinated pigs had the highest, and sustained, F specific T-cell responses, which at 112 dpv was significantly (p < 0.0001) higher than all other groups (Fig. 5E). As seen in the first pig immunogenicity study (Supplementary Figs. 4 and 5), pigs receiving NiV sG displayed spontaneous T-cell IFN-γ responses (Supplementary Fig. 8).

Fig. 5: Comparison of the immunogenicity of NiV sG, ChAdOx1 NiV G, and NiV mcsF following prime only and prime-boost immunisation regimens. All pigs were immunised on day 0 and 3 out of the 6 groups were boosted on day 21 by intramuscular inoculation of 100 µg NiV sG or NiV mcsF proteins in adjuvant, or 1 × 109 IU ChAdOx1 NiV G. A NiV sG and mcsF binding antibody titres (EPT) on day 21, 42 and 112; B Virus neutralising titres as assessed by NiV M VNT; C Inhibition of NiV glycoprotein-mediated cell-cell fusion evaluated using day 112 sera; D, E IFN-γ ELISpot assay to assess PBMC responses to NiV G and F peptide stimulation, respectively. For A and C, datapoints represent individual pigs, with the bars showing the group mean and error bars represent the SD. For the other panels, datapoints represent the group mean and error bars represent the SD. Significant differences were determined using two-way ANOVA and signified with the following letter: b—significant difference to NiV sG prime-boost; c—significant difference to ChAdOx1 NiV G prime-boost; d—significant difference to NiV mcsF prime-boost; f—significant difference to NiV sG prime only; g—significant difference to ChAdOX1 NiV G prime only; h—significant difference to NiV mcsF prime only. Full size image

Evaluation of vaccine candidate immunogenicity in pigs under field conditions in the Nipah virus endemic region

A final assessment of the vaccine candidates was undertaken in a NiV endemic region of Bangladesh, immunising indigenous breed pigs under field conditions (Fig. 6). An initial pre-screen for antibody reactivity to NiV sG and mcsF was undertaken using sera from 60 pigs (Supplementary Fig. 9). Five pigs with an OD value higher than twice the negative control were excluded from the study. Immunisation with the three vaccine candidates produced NiV G or F specific antibody titres (Fig. 6A), comparable to those elicited in the European breed pigs under controlled conditions (Figs. 2A and 5A). The indigenous backyard pigs immunised with NiV mcsF produced higher neutralising antibody titres (Fig. 6B) compared to the previous studies (Figs. 2B, 3A, 4A and 5B) with pigs immunised with ChAdOx1 NiV G producing a reduced neutralising titter compared to NiV sG immunised pigs (p < 0.01). These data were also replicated in the pseudoVNT assay (Fig. 6C). When determining the ability of vaccinated pig sera to inhibit cell-to-cell fusion, at 42 dpv, pigs receiving the ChAdOx1 NiV G had a significantly reduced ability to do so compared to pigs immunised with NiV sG and NiV mcsF (p < 0.0001) (Fig. 6D).