Molecular serotyping of Streptococcus pneumoniae

Abstract

Streptococcus pneumoniae, known as the pneumococcus, remains the leading cause of mortality and morbidity worldwide in children <5 years due to pneumonia, with the highest burden of disease in developing countries. The pneumococcus is responsible for a variety of diseases, from localised infections such as otitis media and sinusitis, to life-threatening diseases such as meningitis and pneumonia. The polysaccharide capsule, exterior to the cell wall, is the major virulence factor of the organism. To date, 93 different capsular types, called serotypes, have been described. These differ in their chemical structures, and <15% of serotypes are responsible for majority of pneumococcal disease worldwide. Some serotypes have a higher invasive potential than others, while some are associated with more severe outcome than others. There is also a geographic and age-specific variation in the distribution of serotypes. Pneumococcal diseases are largely vaccine preventable. The current vaccine formulations contain serotypes that are most prevalent across regions. The licensed polysaccharide and polysaccharide conjugate vaccines (PCV) contain at least 23 and 10 or 13 of most prevalent serotypes, respectively. Serotypes that were historically not commonly associated with invasive pneumococcal disease (IPD) are increasing in prevalence due to the reduction of vaccine serotypes. Thus, it is important to monitor serotype distribution for future vaccine formulations and to determine the vaccine effectiveness in the region. The “gold standard” for serotyping, the Quellung reaction, is dependent on the viability of the organism, and hence it is problematic in settings where a culture cannot be obtained due to insensitivity of blood cultures, incorrect culturing techniques, initiation of antibiotic therapy before the specimen is taken or autolysis of the pneumococcus. Although molecular methods have been established to identify pneumococcus as the disease-causative agent, serotyping was not possible on culture-negative samples until recently. Sequencing of the capsular biosynthetic loci of >90 serotypes revealed conserved regions and serotype-specific genes, which serve as targets for the PCR-based assays. In this study, PCR-based serotyping assays were established in South Africa (SA) for culture-negative samples initially by using isolates of known serotypes. The conventional PCR assay (C-PCR) showed 100% sensitivity (801/801) and specificity (29/29) for serotypes included in the assay when compared with the Quellung reaction. However, specificity of the real-time PCR serotyping assay (RT-PCR) was 95% (89/94) compared to the Quellung reaction and the C-PCR assay, due to cross-reactions between genetically related serotypes/serogroups. The prevalent disease-causing serotypes were detected in the first three reactions of the C-PCR assay and first four reactions of the RT-PCR serotyping assay, allowing for the majority of isolates to be assigned a serotype in a time-efficient manner. While both C-PCR and RT-PCR serotyping assays had comparable sensitivities on samples with sufficient DNA concentration, RT-PCR had an increased sensitivity in assigning serotypes for clinical specimens and non-viable transport medium samples (NVTMs). These samples often had a lytA cycle threshold (Ct) values of >26, indicative of low bacterial loads/DNA concentration. In 2010, 4201 cases of IPD were reported to the national IPD surveillance laboratory in SA. Viable isolates were not available for 10% (n=440) of all cases reported and 395 were available for serotyping. Of these, 80% (173/217) of NVTMs and 91% (162/178) of culture-negative clinical specimens could be assigned a serotype using the PCR-based methods. The PCR-based assays added 8% (335/4201) of serotyping data to the national surveillance database for 2010. The newly discovered serotype 6D was not identified in SA in 2010. The establishment of PCR-based serotyping assays was useful in the monitoring of serotype-specific disease burden. The prevalent serotypes in all SARI (severe acute respiratory illness) cases were 19A, 1 and 6A/B. This data is useful for determining the effectiveness of the 7-valent pneumococcal conjugate vaccine (PCV7) effectiveness in children with known vaccination status. The PCR-based serotyping assays could assign a serotype to 95% (18/19) of the pleural fluids positive for lytA received as part of an empyema study. Serotypes 6A/B (n=8, 44%), 19A (n=3, 17%), 23F (n=3, 17%), 14 (n=2, 11%) and 1 (n=2, 11%) were prevalent in these specimens. PCV13 serotypes 19A and 1 were prevalent in cultures and culture-negative samples included in this study, suggesting the potential increased positive public health impact offered by introduction of 13-valent PCV in SA in 2011. PCR provided important serotyping data on a significant proportion of culture-negative samples. A number of culture-negative samples were positive for PCV7 serotypes, and therefore these data are useful for monitoring vaccine impact on serotype distribution. Thus, the burden of disease and the serotype distribution can be monitored effectively for vaccine failures and future vaccine developments, in a cost-effective and timeous manner.