Reply to “Cross-Protective Immunity against Heterologous Streptococcus pneumoniae”

REPLY

In the foregoing letter responding to our article in Infection and Immunity, Moens et al. provide further evidence that immunization with an inactivated strain of Streptococcus pneumoniae despite being highly protective against infection with the homologous strain is only weakly protective against infection with a heterologous strain. Previously these data may have been interpreted as demonstrating that adaptive protective immunity to S. pneumoniae is mainly mediated by antibody to capsular polysaccharide. However, our data (1) demonstrate that this is not the case due to the following observations. (i) No significant antibody to capsular polysaccharide was detected in mice immunized with the live attenuated TIGR4Δpab strain. (ii) Immunization with the live attenuated capsular serotype 4 TIGR4Δpab strain was highly protective against subsequent infection with a capsular-locus-switched TIGR4 strain expressing a serotype 2 capsule but not against a serotype 2 D39 strain. Furthermore, colonization-induced protection against invasive S. pneumoniae infection does not require antibody to the capsule (2, 3). Instead, after vaccination with a Δpab strain, strong protective immunity against the homologous strain was associated with specific IgG responses to a range of well-recognized protein antigens, including pneumolysin, PspA, PpmA, PsaA, and CbpD (1). Which antigens were recognized by IgG in sera from mice immunized with TIGR4Δpab varied with the dose of bacteria given and the route of inoculation but always included several antigens like PpmA and PsaA that are highly conserved in S. pneumoniae strains. Some of these antigens have been shown to induce cross-protective responses to different S. pneumoniae strains when used as purified protein antigen vaccines (4). Hence, the failure of cross-protection when mice are immunized with an attenuated S. pneumoniae strain is surprising. These results are important as they suggest a protein antigen-based vaccine for S. pneumoniae may not necessarily provide much advantage in inducing greater cross-protective immunity than the conjugated vaccine.

Our data obtained using capsular-locus-switched TIGR4 strains demonstrate that variable access to subcapsular antigens between different capsular serotypes is not the cause of the failure of cross-protection. In addition, flow cytometry did show that antibody in the sera obtained from vaccinated mice recognized some but not all heterologous strains (1). These data would be consistent with protection after vaccination with attenuated S. pneumoniae depending on only some of the anti-protein antigen antibody responses, with the protective antigens varying significantly either in structure or relative expression between strains. For example, antibodies to choline-binding protein(s) such as PspA and CbpD, which have significant allelic variation in strains, may be mediating protection, whereas antibodies to conserved surface antigens like PpmA and PsaA are not functionally important. Alternatively, protection could be independent of IgG responses, or there could be steep dose-response curves with the protective responses induced after the short-lived infection with attenuated bacteria not adequate to protect against heterologous strains despite strong protection against homologous strains. In order to ensure the maximum efficacy of potential future vaccines based on protein antigens, considerably more research is required to further explore the reasons for lack of cross-protection induced by immunization with attenuated S. pneumoniae.