Non-neutralizing antibodies elicited by recombinant Lassa-Rabies vaccine are critical for protection against Lassa fever

Abstract

Lassa fever (LF), caused by Lassa virus (LASV), is a viral hemorrhagic fever for which no approved vaccine or potent antiviral treatment is available. LF is a WHO priority disease and, together with rabies, a major health burden in West Africa. Here we present the development and characterization of an inactivated recombinant LASV and rabies vaccine candidate (LASSARAB) that expresses a codon-optimized LASV glycoprotein (coGPC) and is adjuvanted by a TLR-4 agonist (GLA-SE). LASSARAB elicits lasting humoral response against LASV and RABV in both mouse and guinea pig models, and it protects both guinea pigs and mice against LF. We also demonstrate a previously unexplored role for non-neutralizing LASV GPC-specific antibodies as a major mechanism of protection by LASSARAB against LF through antibody-dependent cellular functions. Overall, these findings demonstrate an effective inactivated LF vaccine and elucidate a novel humoral correlate of protection for LF.

Fig. 1

Diagram of vaccine constructs and controls. BNSP333 is the parental vector and FILORAB1, the control used, is based on BNSP333 with a codon-optimized Zaire Ebola Virus Glycoprotein (EBOV GP) inserted between N and P through the BsiWI and NheI restriction digest sites. LASSARAB was generated in a similar manner as FILORAB1, from BNSP333 by cloning a codon-optimized version of Lassa virus glycoprotein (LASV GPC) in the BsiWI and NheI restriction digest sites. LASSARAB-ΔG was further generated from LASSARAB by removing the native rabies glycoprotein (G) by using the restriction digest sites SmaI and PacI. rVSV-GPC was generated by replacing the native VSV glycoprotein (G) by LASV GPC at the same sites. rVSV-GPC was created to be used as a control vector and as a scaffold to produce a native LASV GPC antigen for ELISAs (see Methods section)

Fig. 2

Evaluation of LASSARAB and LASSARAB-ΔG vectors in cell culture and inactivated virion characterization. a LASSARAB, FILORAB1 and uninfected VERO cells were probed for LASV-GPC and RABV-G expression with 37.7H anti-LASV human mAb and 1C5 anti-RABV G mouse mAb and analyzed by flow cytometry 48 h post infection. b VERO cells were infected at a MOI of 0.1 with 4 viruses: FILORAB1, LASSARAB, LASSARAB-ΔG, and rVSV-coGPC. 48 h later (24 h for VSV based vectors) cell surface expression of LASV glycoprotein (GPC), in red, and RABV glycoprotein (G), in green, was probed by a α-LASV GPC rabbit polyclonal and a α-RABV G human 4C12 monoclonal, respectively. In LASSARAB infected cells, yellow is observed as the superimposition of LASV GPC surface expression with RABV G. The bar indicates 12 μm. c VERO CCL-81 cells were infected with a MOI of 0.01 and media supernatant was collected at 0, 24, 48, 72, and 96 h. Virus titers were measured through foci-forming assay (in Y-axis) and plotted through time (X-axis). d, e LASSARAB and FILORAB1 virions were concentrated through TFF and sucrose purified through ultra-centrifugation. Pellets were resuspended in PBS, BPL inactivated at 1:2000 for 24 h, and 2 μg of each was loaded in a denaturing 10% SDS-PAGE gel. In d SYPRO Ruby staining was used. e, f LASV GPC incorporation in LASSARAB particles was confirmed by western Blot with either an anti-LASV GP2 rabbit polyclonal (upper panel) and anti-GPC/GP1/GP2 guinea pig survivor serum (lower panel). Uncropped versions are available in supplementary figures

Fig. 3

Evaluation of LASSARAB, LASSARAB-ΔG, and rVSV-GPC pathogenicity. a Weight curves of 6- to 8-week-old female Swiss Webster mice that were inoculated intranasally with 10⁵ ffu of either LASSARAB, LASSARAB-ΔG, or rVSV-GPC. As controls, mice were inoculated with the same dosage of either BSNP parent vector (Rabies) without the 333 mutation in the Rabies G, FILORAB1, rVSV-EGFP, or Mock (PBS). Weight is standardized as percentage of weight loss or gain in comparison with first day of exposure. Rabies virus-infected animals developed clinical signs on day 8 with further weight loss until day 11 when endpoint criteria were reached. In LASSARAB-ΔG one mouse died at day 14 without displaying any signs or weight loss. In rVSV-GPC, three mice displayed signs of neurological deficit with two succumbing and one surviving. All other mice showed no signs of pathology. b Survival curves of BALB/c or SCID mice that were subjected to intracranially (IC) exposure with either LASSARAB or BNSP333. No signs of disease nor death were observed post- exposure. c IC exposure of Swiss Webster suckling mice with either LASSARAB or BNSP333. Suckling mice started succumbing to infection by day 7 in BNSP333 group and survived as long as day 12 in LASSARAB group with none surviving by the end of the study

Fig. 4

Analysis of the humoral response towards Lassa virus glycoprotein. C57BL/6 mice were immunized IM in the gastrocnemius muscle with either 10 μg of β-propiolactone inactivated viral particles in PBS or adjuvanted with 5 μg of GLA, a TLR-4 agonist formulated in 2% of stable emulsion (SE); LASSARAB+GLA-SE, LASSARAB, FILORAB1 groups) and boosted two times with the same amount on day 7 and 28 (a). Immunizations with replication-competent viruses were executed with a single time inoculation of 10⁶ ffu or pfu virus IM in the gastrocnemius (rc-LASSARAB; rc-FILORAB1 groups and rVSV-GPC). b The EC₅₀ values (obtained from the 4PL regression ELISA curve) of the total IgG titers against LASV GPC are plotted since day 0 until day 42. Error bars are representative of the standard error mean (SEM) and is calculated from 15 mice per group. Statistical significance was calculated by using 2-way ANOVA–post-hoc Tukey’s Honest Significant Difference Test. c ELISA of total IgG against LASV GPC of all day 42 groups are shown for all immunized groups. ELISA curves are generated from 4PL regression. Error bars are representative of the SEM of OD 490 values (five mice per group, in triplicates). d Day 35 EC₅₀ antibody titer of IgG sub-isotype (IgG2c and IgG1) against LASV GPC of sera from LASSARAB+GLA-SE and LASSARAB group was analyzed. Error bars are the SEM of a total of five mice per group and statistical significance by 2-way ANOVA (post-hoc Tukey’s Honest Significant Difference Test). e The ratios of the respective EC₅₀ antibody titers IgG1/IgG2c are plotted and the F test was applied to check for variance difference (p < 0.001). (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05)

Fig. 5

Virus neutralization antibody titers. Day 42 sera from immunized mice was incubated with pseudotyped rVSV-ΔG-NL-GFP. a rVSV-ΔG-NL-GFP was pseudotyped with either RABV-G, LASV-GPC, or EBOV-GP to assay for RABV-G, LASV-GPC, or EBOV-GP NAb titers, respectively. 12.1F, 37.7H, and 25.10c are LASV-GPC neutralizing antibodies used as a positive control for LASV-GPC neutralization. Y-axis in b represents 50% of inhibitory serum dilution (IC₅₀) titers obtained based on the antibody dilution that has 50% infection percentage of infected cells curves obtained in a. All groups achieved high neutralizing titers against RABV-G except for the groups immunized with virus lacking RABV-G: rc-LASSARAB-ΔG and rVSV-coGPC, as expected. Regarding LASV-GPC pseudotyped VSVs, no immunization achieved appreciable amounts of neutralizing antibodies. Neutralization of LASV GPC pseudotyped viruses with 12.1F, 37.7H, and 25.10C had an average IC50 of 1546 ng/ml, 375 ng/ml, and 69 ng/ml, respectively. c Rabies neutralizing titers were calculated by using the IC₅₀ values of the WHO sera standard (2 IU/ml) serial diluted with rVSV-ΔG-NL-GFP pseudotyped with RABV G. WHO international units/ml (IU/ml) were then calculated using the following formula: (sample IC₅₀ titer)/(WHO standard IC₅₀ titer) × 2.0 (WHO IU/ml standard starting dilution). IU/ml from test sera is plotted Y-axis. All error bars represented are the SEM of triplicate values of 5 mice per group. (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05)

Fig. 6

LASV challenge of outbred Hartley guinea pigs immunized with several vaccine candidates and control. a Guinea pigs were immunized with either two replication competent vaccines: rVSV-GPC (positive control for survival) and LASSARAB replication-competent at 10⁶ ffu by intraperitoneal injection (IP); or inactivated LASSARAB+GLA-SE with different immunization schedules: day −58 (LASSARAB+GLA-SE (1)), day −58, day −51 (LASSARAB+GLA-SE (2)) and day −58, day −51, and day −30 (LASSARAB+GLA-SE (3)). RabAvert was used as mock immunization (negative control). b Survival curves post IP exposure with 10⁴ pfu guinea pig-adapted LASV Josiah strain. Statistical significance is compared against Rabvert group using log-rank (Mantel–Cox) test. c Heat plot representing the clinical score information. X-axis represents days' post-challenge and Y-axis represents the individual animal number. d Terminal viremia was plotted using LASV RNA copies/ml in Y-axis. Statistical significance was calculated using Kruskal–Wallis one-way ANOVA (not significant). e LASV neutralizing antibody titers is reported as the IC₅₀ (half maximal inhibitory concentration) of serum dilution. The human mAbs 25.10C, 12.1F, and 37.7H^, were used as positive LASV neutralization controls. f Pre-challenge titers of LASV GPC specific IgG were performed on sera collected on day −15 prior to challenge by ELISA with LASV GPC antigen and the EC₅₀ (50% effective concentration) of serum dilution was plotted in the Y-axis. Statistical significance (compared to the RabAvert group) was calculated by using one-way ANOVA (post-hoc test Tukey Honest Significant Difference Test). g Post-challenge titers of LASV GPC-specific IgG was performed on sera collected on terminal bleeding of both succumbed animals and survivors (day 50 post challenge) and the EC₅₀ of serum dilution is plotted on the Y-axis. Statistical significance reported between survivors and succumbed in e, g was determined by using two-way ANOVA. All error bars represented are the standard error mean (SEM) of 10 animals per group (in triplicates). (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05)

Fig. 7

Evaluation of antibody effector cell functions mediated by murine NK and macrophage cells against 3T3 expressing LASV GPC. Day 42 sera from immunized mice was incubated with 3T3-LASV cells and 30 min either murine NK or macrophage cells were added and results were analyzed 4 h later by either flow cytometry (Supplementary Fig. 3 and a–d) or confocal microscopy (e). Purified murine C57BL/6 NK cells (a) or IC-21 macrophages (d) were added at different Target:Effector cell ratios (T:E) with target cells incubated with either LASSARAB sera (yellow), FILORAB1 sera (gray) or no sera (pink). The Y-axis represents the percentage of cellular cytotoxicity based on GFP⁺/PI⁺ cells (gating strategy and flow plots in Supplementary Fig. 3). b To determine which antibody isotype class is important for ADCC, NK cells were added at 1:5 T:E and incubated with either unprocessed sera (sera 1:100 condition), purified IgG (20 μg/ml), or IgG impoverished sera (1:5 dilution) from LASSARAB and FILORAB1 immunized mice. The Y axis represents cytotoxicity fold change of LASSARAB sera or IgG compared to FILORAB1 sera or IgG with same respective conditions. Anti-CD16/32 (Fcγ-RII/III) was also added at 25 μg/ml to confirm that FcγR blockade reduces ADCC activity. c, e, f To analyze ADCP J774.A1 macrophages were added at 1:5T:E or 1:1T:E (confocal) to 3T3-LASV cells incubated with either LASSARAB sera (c, e) or FILORAB1 sera (c, f). In c anti-CD16.2 (Fcγ-RIV), anti-CD16/32 (Fcγ-RII/III), and anti-CD64 (Fcγ-RI) were added at 25 μg/ml to check the effect of different FcγR blockade on ADCP activity. All error bars are the SEM of at least three independent experiments executed with duplicates. All statistical significance represented was performed through either a one- or two-way ANOVA and using a post-hoc analysis Tukey Honest Significant Difference test. (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05). The bar in e indicates 6 μm (80×) or 12 μm (40×)

Fig. 8

Evaluation of in vivo relevance of non-NAbs LASV GPC specific antibodies induced by LASSARAB + GLA-SE vaccination. a 8- to 10-week-old Balb/c (WT) or Balb/c with Fcγ chain KO (Fcγ^−/−) female mice were immunized with 10 µg of inactivated particles of either LASSARAB or FILORAB1 (mock control) on day 0 and boosted on day 28. All four groups in total were adjuvanted with 5 µg of GLA in a 2% SE with each vaccination. One day before exposure (day 41) animals were injected with 1.25 mg of anti-Ifnar mAb (MAR1-5A3, Leinco technologies) through intra-peritoneal injection (IP). On day 42, mice were exposed to 10⁴ rVSV-GPC virus IP and general health (weights and clinical observation) was recorded until endpoint criteria were reached or end of study (supplemental). b Survival curves post-exposure of rVSV-GPC. Significance is compared between the WT LASSARAB vaccinated and the Fcγ^−/− vaccinated using the log-rank (Mantel–Cox) test. c Pre-exposure total IgG titers anti LASV GPC were measured by ELISA on day 35 post-prime and ELISA curves were plotted according to OD490 reading value (Y-axis) and serum dilution (X-axis). On the right, EC₅₀ (half maximal effective concentration) of serum dilution of both LASSARAB groups (WT and Fcγ^−/−) is plotted on Y-axis on a log scale; statistical significance was calculated using one-way ANOVA. d Virus neutralization assay using pseudotyped VSV-GFP-NanoLuc with LASV GPC. On the right, percentage of cells infected is plotted against the serum dilution (survivors on day 14 post-exposure) of each respective group. On the right, the IC₅₀ (half maximal inhibitory concentration) of serum dilution is plotted individually and significance was calculated using one-way ANOVA. Error bars represent Standard Error Mean (SEM) and include all mice (n = 10 per group [WT and KO] in LASSARAB and n = 5 per group [WT and KO] in FILORAB1 control) in pre-challenge and survivor mice in post-challenge. (****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05)