A prophylactic multivalent vaccine against different filovirus species is immunogenic and provides protection from lethal infections with Ebolavirus and Marburgvirus species in non-human primates

Abstract

The search for a universal filovirus vaccine that provides protection against multiple filovirus species has been prompted by sporadic but highly lethal outbreaks of Ebolavirus and Marburgvirus infections. A good prophylactic vaccine should be able to provide protection to all known filovirus species and as an upside potentially protect from newly emerging virus strains. We investigated the immunogenicity and protection elicited by multivalent vaccines expressing glycoproteins (GP) from Ebola virus (EBOV), Sudan virus (SUDV), Taï Forest virus (TAFV) and Marburg virus (MARV). Immune responses against filovirus GP have been associated with protection from disease. The GP antigens were expressed by adenovirus serotypes 26 and 35 (Ad26 and Ad35) and modified Vaccinia virus Ankara (MVA) vectors, all selected for their strong immunogenicity and good safety profile. Using fully lethal NHP intramuscular challenge models, we assessed different vaccination regimens for immunogenicity and protection from filovirus disease. Heterologous multivalent Ad26-Ad35 prime-boost vaccination regimens could give full protection against MARV (range 75%-100% protection) and EBOV (range 50% to 100%) challenge, and partial protection (75%) against SUDV challenge. Heterologous multivalent Ad26-MVA prime-boost immunization gave full protection against EBOV challenge in a small cohort study. The use of such multivalent vaccines did not show overt immune interference in comparison with monovalent vaccines. Multivalent vaccines induced GP-specific antibody responses and cellular IFNγ responses to each GP expressed by the vaccine, and cross-reactivity to TAFV GP was detected in a trivalent vaccine expressing GP from EBOV, SUDV and MARV. In the EBOV challenge studies, higher humoral EBOV GP-specific immune responses (p = 0.0004) were associated with survival from EBOV challenge and less so for cellular immune responses (p = 0.0320). These results demonstrate that it is feasible to generate a multivalent filovirus vaccine that can protect against lethal infection by multiple members of the filovirus family.

Fig 1. Immunity to filovirus glycoprotein using a tetravalent vaccine in a heterologous (Ad26-Ad35) regimen.

Cynomolgus macaques were immunized with heterologous Ad26-prime at week 0 and Ad35-boost at week 4 with a tetravalent vaccine (total dose 8x10¹⁰ vp) or empty Ad vectors as controls. (A) Cellular immune response 2 weeks after boost immunization using IFNγ ELISpot after stimulation with the indicated filovirus GP peptide pools. (B-E) Humoral immune responses after prime and boost immunizations as measured by ELISA for the indicated filovirus GP. Bars designate the mean response and each circle represents an individual animal. The black dotted line represents the lower limit of detection. (F-I) Humoral immune responses over 85 weeks measured by ELISA for the indicated filovirus GP. Each green line represents an individual animal, the red dotted line is the average response after priming, the black dotted line represents the lower limit of detection.

Fig 2. Immunogenicity of heterologous (Ad26-Ad35) tetravalent and homologous (Ad26) trivalent vaccine regimens and protection from MARV Angola challenge.

(A-D) Cynomolgus macaques were immunized with heterologous Ad26 prime at week 0 and Ad35 boost at week 4 with a tetravalent vaccine, or a monovalent Ad26 MARV GP vaccine, or empty Ad vectors, at the doses indicated. (E-H) Cynomolgus macaques were immunized with homologous trivalent Ad26 prime at week 0 and Ad26 boost at week 4, or a monovalent Ad5 MARV GP vaccine (prime only at week 4), or empty Ad vectors, at the doses indicated. A challenge with 1000 pfu MARV Angola was given at week 8. (A+E) Kaplan-Meier representation of survival. (B+F) Clinical scoring of individual animals after lethal challenge. (C+G) Humoral immune response over time measured by MARV GP-specific ELISA. Horizontal dotted line represents the lower limit of detection. Solid lines indicate the group mean. (D+H) Cellular immune response to MARV GP peptide pool by IFNγ ELISpot. Vertical dotted lines indicate the time of boost immunization. Solid lines indicate the group mean.

Fig 3. Immunogenicity of heterologous (Ad26-Ad35) and homologous (Ad26-Ad26) trivalent vaccine regimen and protection from SUDV Gulu challenge.

Cynomolgus macaques were immunized with heterologous (Ad26-Ad35) or homologous (Ad26-Ad26) prime at week 0 and boost at week 4 with a trivalent vaccine, or a monovalent Ad5 SUDV GP vaccine (prime only at week 4), or Ad empty vectors at the doses indicated. A challenge with 1000 pfu SUDV Gulu was given at week 8. (A) Kaplan-Meier representation of survival. (B) Clinical scoring of individual animals after lethal challenge. (C) Humoral immune response over time measured by SUDV GP-specific ELISA. Solid lines indicate the group mean. (D) Cellular immune response to SUDV GP peptide pool measured by IFNγ ELISpot. Vertical dotted lines indicate the time of boost immunization. Solid lines indicate the group mean.

Fig 4. Immunogenicity of tetravalent, trivalent, and monovalent vaccines and protection from EBOV Kikwit challenge.

(A-D) Cynomolgus macaques were immunized with heterologous Ad26-prime at week 0 and Ad35 boost at week 4 with a trivalent vaccine, or a monovalent vaccine (Ad.ZEBOV), or empty Ad vectors, at the doses indicated. A challenge with 100 pfu EBOV Kikwit was given at week 8. (E-H) Cynomolgus macaques were immunized with a heterologous (Ad26-Ad35) or homologous (Ad26-Ad26) prime at week 0 and boost at week 4 with a trivalent vaccine, or tetravalent vaccine, or empty Ad vectors, or a bivalent Ad5 EBOV GP + SUDV GP vaccine (prime only at week 4), at the doses indicated. A lethal challenge with 1000 pfu EBOV Kikwit was given at week 8. (A+E) Kaplan-Meier representation of survival. (B+F) Clinical scoring of individual animals after lethal challenge. Clinical score criteria for euthanasia was 15 in (B) and 9 in (F). (C+G) Humoral immune response over time measured by EBOV GP-specific ELISA. Horizontal dotted line represents the lower limit of detection. Solid lines indicate the group mean (D+H) Cellular immune response to EBOV GP peptide pool by IFNγ ELISpot. Vertical dotted lines indicate the time of boost immunization. Solid lines indicate the group mean.

Fig 5. Immunogenicity of heterologous (Ad26-Ad35, Ad26-MVA-BN-Filo or MVA-BN-Filo-Ad26) trivalent, and Ad26-Ad35 monovalent vaccines and protection from EBOV Kikwit challenge.

Heterologous (Ad26-Ad35, Ad26-MVA or MVA-Ad26) prime at week 0 and boost at week 8 with a trivalent vaccine, or empty Ad26-Ad35 vectors, at the doses indicated. One group received a heterologous Ad26-Ad35 prime at week 4 and boost at week 8 of monovalent vaccine (Ad.ZEBOV). A challenge with 100 pfu EBOV Kikwit was given at week 12. (A) Kaplan-Meier representation of survival. (B) Clinical scoring of individual animals after lethal challenge. (C) Cellular immune response to EBOV GP peptide pool by IFNγ ELISpot. (D) Humoral immune response over time measured by EBOV GP-specific ELISA. Horizontal dotted line represents the lower limit of detection. (E) Neutralizing antibody response over time measured by pseudovirion neutralization assay. (C-E) Solid lines indicate the group mean. (F) Correlation of pre-boost (black symbols) and post-boost (blue symbols) ELISA titers with virus neutralization antibody titers. Vertical dotted lines indicate the time of boost immunization.

Fig 6. Comparison of EBOV GP-specific cellular and antibody responses 1 week prior to EBOV challenge in challenge survivors and non-survivors.

Summary of data from lethal challenge experiments with EBOV Kikwit shown in Fig 4 and Fig 5. Animals receiving vectors with EBOV GP in mono- or multivalent regimens are shown. Dots: 100 pfu challenge, squares: 1000 pfu challenge; blue: 0–4 week regimen, red: 0–8 week regimen. (A) Cellular immune response to EBOV GP peptides measured by IFNγ ELISpot. (B) Humoral immune response measured by EBOV GP-specific ELISA. p-values were calculated with the exact Wilcoxon rank sum test, and adjusted for multiplicity for each of the 2 pairs of tests (ELISpot and ELISA) using the Bonferroni adjustment.