Rapid identification of Streptococcus pneumoniae serotypes by cpsB gene-based sequetyping combined with multiplex PCR

Background/purpose: Streptococcus pneumoniae is an important human pathogen that causes invasive infections in adults and children. Accurate serotyping is important to study its epidemiological distribution and to assess vaccine efficacy.Methods: Invasive S. pneumoniae isolates (n Z 300) from 27 teaching hospitals in China were studied. The Quellung reaction was used as the gold standard to identify the S. pneumoniae serotypes. Subsequently, multiplex PCR and cpsB gene-based sequetyping methods were used to identify the serotypes.Results: Based on the Quellung reaction, 299 S. pneumoniae isolates were accurately identified to the serotype level and 40 different serotypes were detected. Only one strain was non -typeable, and five most common serotypes were identified: 23F (43, 14.3%), 19A (41, 13.7%), 19F (41, 13.7%), 3 (31, 10.3%), and 14 (27, 9.0%). Overall, the multiplex PCR method identified 73.3 and 20.7% of the isolates to the serotype and cluster levels, respectively, with 1.7% of the isolates misidentified. In contrast, the cpsB sequetyping method identified 59.0 and 30.3% of the isolates to the serotype and cluster levels, respectively, and 7% were misidentified.Conclusions: The cpsB gene sequetyping method combined with multiplex PCR, can greatly improve the accuracy and efficiency of serotyping, besides reducing the associated costs.Copyright 2022, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

In this study, 300 invasive S. pneumoniae isolates from 27 teaching hospitals in China were analyzed using the Quellung reaction, multiplex PCR, and cpsB gene-based sequetyping methods for serotyping. The multiplex PCR method showed higher accuracy in identifying serotypes and clusters, while the cpsB gene-based sequetyping method exhibited better performance in cluster level identification despite a certain misidentification rate.