A Trivalent Live Vaccine Elicits Cross-Species Protection Against Acute Otitis Media in a Murine Model

Abstract

Background: Acute otitis media (AOM) is a common pediatric infection worldwide and is the primary basis for pediatric primary care visits and antibiotic prescriptions in children. Current licensed vaccines have been incompletely ineffective at reducing the global burden of AOM, underscoring a major unmet medical need. The complex etiology of AOM presents additional challenges for vaccine development, as it can stem from multiple bacterial species including Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis. As such, targeting multiple pathogens simultaneously may be required to significantly impact the overall disease burden. Methods: In this study, we aim to overcome this challenge by engineering a live-attenuated vaccine platform based on an attenuated mutant of S. pneumoniae that expresses H. influenzae and M. catarrhalis surface epitopes to induce protective immunity against all three pathogens. Results: The trivalent live-attenuated vaccine conferred significant protection against all three bacterial otopathogens as measured by seroconversion and the development of AOM, with the inclusion of the additional epitopes providing unexpected synergy and enhanced protection against S. pneumoniae. Conclusions: These data demonstrate a novel mechanism of introducing non-native immunogenic antigens into a live-attenuated vaccine platform to engender protection against AOM from multiple pathogenic species.

Keywords: acute otitis media; live-attenuated vaccine; multiple otopathogens.

Figure 1

LAV is recalcitrant to reversion and is not readily obtained by other pneumococcal serotypes. (a) 19F (circles) and LAV (squares) strains were passaged in vivo via murine nasal colonization. Six mice were intranasally infected, and the lungs, ears, and nasal passages were harvested 3 days post-challenge. The bacterial burden was determined, and the bacterial growth on plates from the nasal passages were recovered (Passage 1). The recovered bacterial population from the nasal passage was used to challenge the mice for the next passage (2); N = 1 per each Passage 1 population. The bacterial population was passaged again using the same method (Passage 3). Each data point represents an individual mouse (lungs and nasal passage) or each ear from individual mice (ear), and the bars represent the median. The dashed line represents the limit of detection. For each strain and tissue, the bacterial burden of the sequential passages was compared via Kruskal–Wallis one-way ANOVA. (b) The competence of the LAV strain was determined by calculating the recombination frequency upon transformation with nontargeted Tn-seq gDNA. Each data point represents an individual biological replicate, and the bars represent the median. The dashed line represents the limit of detection. The recombination frequencies of the wild-type 19F and the LAV strain were compared via a non-parametric Mann–Whitney t-test. (c) The frequency of resistance spread from the LAV strain was determined by transforming strains of different serotypes with the LAV strain’s gDNA (squares, right). As a control, the same strains were transformed with gDNA from a strain harboring the resistance cassette at a neutral location (circles, left). Each data point represents an individual biological replicate, and the bars represent the median. The dashed line represents the limit of detection. For all strains, the recombination frequency of the strains transformed with the gDNA of the LAV strain was compared to that of the strains transformed with the control DNA via a non-parametric Mann–Whitney t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = non-significant.

Figure 2

Prior colonization in mice is not detrimental to the protective efficacy of the LAV strain. (a–e) Mice (N = 15) were colonized with either a homologous strain (19F) expressing its own capsule (Type 19F; squares) or a variant capsule (Type 4; circles), or a heterologous strain (4) expressing its own capsule (Type 4; up triangles) or a variant capsule (Type 19F; down triangles). As a control, mice received a PBS vehicle and were not colonized (none; circles). Three weeks later, all the mice were vaccinated with the LAV strain. Following vaccination, all the mice were challenged with a 19F strain. (a,b) Sera were collected prior to challenge, and IgG seroconversion in the vaccinated mice was determined by ELISA against 19F (a) or 7F (b). Each data point represents an individual mouse, and the bars represent the mean. The immunoreactivity of sera from the mice colonized with each strain was compared to that of the non-colonized mice via an unpaired t-test. (c–e) Post-challenge with 19F, the bacterial burden in the lungs (c), ears (d), and nasal passage (e) of the vaccinated mice was determined. Each data point represents an individual mouse (c,e) or each ear from individual mice (d), and the bars represent the median. The dashed line represents the limit of detection. The bacterial burden in each tissue from the mice colonized with each strain was compared to that of the non-colonized mice via a non-parametric Mann–Whitney t-test. **** p < 0.0001. No significant difference was observed for any comparison without a designated p value.

Figure 3

The protective efficacy of the LAV strain is not diminished with Prevnar-13 vaccination but rather demonstrates a significantly enhanced protective benefit. The mice were vaccinated with either a PBS vehicle control (N = 30; open circles), Prevnar-13 (N = 15; closed circles), LAV strain (N = 30; open squares), or Prevnar-13 followed by vaccination with the LAV strain (N = 30, closed squares). Following the final vaccination, the mice were challenged with a homologous serotype (19F). (a) Sera were collected prior to challenge, and IgG seroconversion was determined by ELISA against 19F. Each data point represents an individual mouse, and the bars represent the mean. The immunoreactivity of sera of the vaccinated mice were compared to that of the mice that received the PBS vehicle control or other vaccines via an unpaired t-test. (b–d) Twenty-four hours post-challenge with 19F, the bacterial burden in the lungs (b), ears (c), and nasal passage (d) of the mice was determined. Each data point represents an individual mouse (b,d) or each ear from individual mice (c), and the bars represent the median. The dashed line represents the limit of detection. The bacterial burden in each tissue of the vaccinated mice was compared to the burden in the tissues of the mice that received the PBS vehicle control or other vaccines via a non-parametric Mann–Whitney t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns = non-significant.

Figure 4

Novel platform for expressing non-native antigenic epitopes in S. pneumoniae. (a) Strategies for generating S. pneumoniae strains expressing antigenic epitopes of H. influenzae and M. catarrhalis. Three strategies were employed to anchor ProteinD to the cell surface, i.e., with C-terminal LPXTG motif (ProteinD-LPXTG), with C-terminal choline-binding motif (ProteinD-CBD), and with N-terminal lipoprotein anchor domain (Lipo-ProteinD). To generate a vaccine against all three otopathogens, both H. influenzae and M. catarrhalis epitopes were incorporated by expressing Lipo-ProteinD with an additional 23-amino-acid UspA epitope (“NNINNIY”) on the C-terminus. (b) The production of ProteinD in the cell lysates of S. pneumoniae expressing each ProteinD variant was measured via Western blot analysis using a polyclonal antibody against ProteinD. The strains included wild-type 19F, strains expressing each ProteinD variant, and the ftsY mutation in 19F (LAV), strain expressing Lipo-ProteinD (LAV-D), and strain expressing Lipo-ProteinD-M (LAV-D-M). The expression of ProteinD-LPXTG (51 kDa) and Protein-CBD (48 kDa) was detected upon longer exposure time than Lipo-ProteinD (42 kDa) and Lipo-ProteinD-M (45 kDa), and the exposure time (s) is listed for comparison. As a loading control, the samples were concurrently run and probed with an antibody against CbpA. (c) The anchoring of ProteinD on the S. pneumoniae cell surface was measured via whole-cell bacterial ELISA using a polyclonal antibody against ProteinD. As a control for differential binding to the plate, ProteinD immunoreactivity was normalized to immunoreactivity against LytA, a cell surface protein. Each data point represents a biological replicate, and the bars represent mean. The immunoreactivity of ProteinD-expressing strains and LAV strains was compared to that of 19F via an unpaired t-test. * p < 0.05, **** p < 0.0001.

Figure 5

Vaccination with strains expressing non-native antigenic epitopes demonstrate seroconversion against multiple respiratory pathogens. (a–c) The mice were vaccinated with the LAV strain (squares), LAV-D strain (down triangles), LAV-D-M strain (up triangles), or PBS vehicle control (circles) and challenged with either S. pneumoniae, H. influenzae, or M. catarrhalis. Sera were collected prior to challenge, and IgG seroconversion in the vaccinated mice was determined by ELISA against 19F (a), H. influenzae (b), or M. catarrhalis (c) and included mice in all the challenge groups. Immunoreactivity against 19F was measured in the sera of the mice vaccinated with all the vaccine strains, against H. influenzae in the sera of the mice vaccinated with LAV-D or LAV-D-M, and against M. catarrhalis in the sera of the mice vaccinated with LAV-D-M; N = 15 for LAV, N = 30 for LAV-D, N = 45 for LAV-D-M and vehicle. Each data point represents an individual mouse, and the bars represent the mean. The immunoreactivity of the sera of the mice vaccinated with vaccine strains were compared to that of the mice that received the PBS vehicle control via an unpaired t-test. ** p < 0.01, **** p < 0.0001; ns = non-significant.

Figure 6

Vaccination with strains expressing non-native antigenic epitopes demonstrate protection against multiple respiratory pathogens. (a–c) The mice were vaccinated with the LAV strain (squares), LAV-D strain (down triangles), LAV-D-M strain (up triangles), or PBS vehicle control (circles) and challenged with either S. pneumoniae, H. influenzae, or M. catarrhalis; N = 15 for each challenge. The bacterial burden in the lungs, ears, and nasal passage of the vaccinated mice was determined for challenge with 19F (a), H. influenzae (b), or M. catarrhalis (c). All the mice challenged with H. influenzae or M. catarrhalis were pre-sensitized with poly (I:C) to enhance translocation to the ear. Each data point represents an individual mouse (lungs and nasal passage) or each ear from individual mice (ear), and the bars represent the median. The dashed line represents the limit of detection. For each challenge, the bacterial burden in each tissue of the vaccinated mice was compared to the burden in the tissues of the mice that received the PBS vehicle control via a non-parametric Mann–Whitney t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. ns = non-significant.