Effect of Oral Streptococci Expressing Pneumococcus-like Cross-Reactive Capsule Types on World Health Organization Recommended Pneumococcal Carriage Detection Procedure.

Abstract

Carriage studies are fundamental to assessing the effects of pneumococcal vaccines. Because a large proportion of oral streptococci carry homologues of pneumococcal genes, non-culture-based detection and serotyping of upper respiratory tract (URT) samples can be problematic. In the current study, we investigated whether culture-free molecular methods could differentiate pneumococci from oral streptococci carried by adults in the URT. Paired nasopharyngeal (NP) and oropharyngeal (OP) samples were collected from 100 older adults twice a month for 1 year. Extracts from the combined NP + OP samples (n = 2400) were subjected to lytA real-time polymerase chain reaction (PCR). Positive samples were subjected to pure culture isolation, followed by species confirmation using multiple approaches. Multibead assays and whole-genome sequencing were used for serotyping. In 20 of 301 combined NP + OP extracts with positive lytA PCR results, probable pneumococcus-like colonies grew, based on colony morphology and biochemical tests. Multiple approaches confirmed that 4 isolates were Streptococcus pneumoniae, 3 were Streptococcus pseudopneumoniae, 12 were Streptococcus mitis, and 1 were Streptococcus oralis. Eight nonpneumococcal strains carried pneumococcus-like cps loci (approximate size, 18-25 kb) that showed >70% nucleotide identity with their pneumococcal counterparts. While investigating the antigenic profile, we found that some S. mitis strains (P066 and P107) reacted with both serotype-specific polyclonal (type 39 and FS17b) and monoclonal (Hyp10AG1 and Hyp17FM1) antisera, whereas some strains (P063 and P074) reacted only with polyclonal antisera (type 5 and FS35a). The extensive capsular overlap suggests that pneumococcal vaccines could reduce carriage of oral streptococci expressing cross-reactive capsules. Furthermore, direct use of culture-free PCR-based methods in URT samples has limited usefulness for carriage studies.