Conjugate vaccines successfully target specific serotypes but also lead to serotype replacement. Alternative strategies include bacterial proteins. Selection of protein antigens would benefit from knowledge of those most abundantly expressed during stages of human pneumococcal pathogenesis. Furthermore, vaccine protein candidates must also be expressed in mice for testing of candidates.
We designed a Nanostring codeset and analyzed 200 genes encoding for bacterial surface-exposed proteins. We evaluated transcriptomic expression in mouse colonization, pneumonia and sepsis models, as well as clinical CSF samples and human controlled infection.
The 30 genes most highly expressed in each system (mouse model, human samples) were identified. There was excellent correlation between mouse colonization, pneumonia and blood specimens (R>0.85). Correlations between human colonization and meningitis were moderate (R=0.56), as were those between mouse and human transcriptomic profiles. We identified several genes highly expressed in all mouse and human conditions. Two in particular encoded for proteins that we show conferred protection against colonization and induced opsonic antibodies.
We present a novel approach to identify putative protective pneumococcal antigens, based on transcriptomic analyses of pneumococcal RNA harvested in different conditions. We identified two novel potential candidates, which also highlights the utility of studying pneumococcal gene expression from human samples.