Genetic Diversity of the Pneumococcal Capsule: Implications for Molecular-based Serotyping

Mary Catherine McEllistrem


Future Microbiol. 2009;4(7):857-865. 

In This Article

PCR-based Serotyping Methods

The marked heterogeneity within the capsular operon between serotypes suggests that differentiation of serotypes can be achieved by correlating the size of the PCR product of type-specific genes with the quellung-derived serotype. Indeed, PCR-based serotyping using primers that amplify serotype-specific genes has been shown to be a reliable and accurate method for a few selected serotypes. For example, a single multiplex PCR-based serotyping method published in 2003 could accurately classify clinical strains from six common serotypes and two serogroups (Table 1).[16] While this initial assay could not differentiate serotypes within vaccine serogroups 6 and 18, it seemed likely a more sensitive assay would soon be developed.

Another multiplex PCR assay, which uses seven different reactions, has increased the number of serotypes detected from six to 16. Interestingly, the order of the multiplex reactions can be altered based on the local seroepidemiologic trends to minimize how many reactions are needed to serotype an isolate (Table 1).[17–19] In 2009, this method was expanded to reliably identify 22 different serotypes.[20] While the assay still cannot accurately classify strains from all PCV7 serotypes, the method can identify all serotype 6C strains initially detected with monoclonal antibodies.[20,21] Although the lack of serotype specificity remains a key limitation for PCR-based methods, serotype can sometimes be ascertained directly from the clinical specimen using PCR. Specifically, a single tube, nested PCR reaction followed by dot blotting detected with a labeled, serotype-specific probe can accurately detect serotypes 3, 14, 19F and 23F from respiratory specimens.[22]

PCR-based Reverse Line Blot Hybridization Serotyping Method

To minimize the number of individual PCR reactions required to serotype a strain, multiplex PCR-based reverse line blot hybridization (mPCR/RLB) was developed.[23,24] In brief, amine-labeled oligonucleotide probes are bound to a nylon membrane. Next, biotin-labeled PCR products generated with multiplex PCR are hybridized to the probes. The PCR products are detected with streptavidin-labeled peroxidase and a chemiluminescent substrate. This assay can detect 22 different serotypes using nearly 200 different primers by analyzing the wzy gene and serotype-specific genes.[25,26] Unfortunately, the method cannot accurately differentiate serotypes within serogroups included in the PCV7, PCV13 and the 23-valent pneumococcal polysaccharide (PPS23) vaccines, including 6A/B, 9V/A, 10A/B, 11A/D, 15B/C and 22A/F.[25,26] Therefore, in its current form, this methodology is unlikely to be clinically useful to monitor seroepidemiology trends.

Restriction Fragment Length Polymorphism-based Serotyping Methods

Given serotype-specific PCR assays have been unable to differentiate vaccine-included serotypes within their respective serogroups, others have exploited the genetic diversity of the wzg and wzh (cpsA and cpsB) genes. Lawrence et al. amplified these two genes and then compared restriction fragment length polymorphism (RFLP) patterns after cutting with one of the three following enzymes: AluI, HinfI and RsaI.[27] A set of three banding patterns, or one banding pattern for each enzyme, was correlated with a specific serotype. The band classification scheme was validated for ten strains per serotype, and the results demonstrated that strains from nine serotypes could accurately be determined. However, as was the case with multiplex PCR, the assay could not differentiate 6A from 6B; moreover, the assay could not detect non-PCV7 serotype 19A strains. Importantly, the data suggested that the assay may lose sensitivity with a larger sample size, as some serotype 6B and 9V strains chosen for validation were found to have novel banding patterns (Table 1). These limitations, coupled with the challenge of comparing three sets of banding patterns per strain, suggest that this assay will be unable to reliably classify strains from all 91 serotypes.

Another study which used only one RFLP pattern per serotype was undertaken. Exploiting the presence of the conserved flanking genes dexB and aliA in all serotypes, the capsular operon was amplified and restricted with HinfI. The RFLP-generated bands were compared with Bionumeric-generated dendrograms.[28] As noted with other assays, this method could serogroup, but not serotype, strains from serogroups 6 and 18. Moreover, despite testing a small number of strains, multiple banding patterns were associated with a single serotype. Finally, while Bionumerics is a robust program for clustering analysis, band comparison analyses can be potentially problematic. For example, two strains can be classified as having 'indistinguishable' or 'unrelated' banding patterns depending on where the operator places the line designating the band, the clustering method used and the band tolerance allowed. Taken together, this method may not be a feasible alternative to the quellung reaction.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: