Molecular typing of Bordetella pertussis is routinely performed on bacterial isolates, but not on DNA extracted from nasopharyngeal aspirates or pernasal swabs submitted for diagnostic realtime PCR (qPCR). We investigated whether these DNA extracts were suitable for multilocus variable-number tandem repeat analysis (MLVA) and DNA sequence-based typing. We analysed all the available qPCR-positive samples received by our laboratory from patients ,1 year of age between January 2008 and August 2010. Eighty-one per cent (106/131) of these generated a complete MLVA profile. This rose to 92 % (105/114) if only samples positive for both of the two targets used for the B. pertussis PCR (insertion element IS481 and pertussis toxin promoter ptxP) were analysed. Sequence-based typing of the pertactin, pertussis toxin S1 subunit and pertussis promoter regions (prn, ptxA and ptxP) was attempted on 89 of the DNA extracts that had generated a full MLVA profile. Eighty-three (93 %) of these produced complete sequences for all three targets. Comparison of molecular typing data from the 89 extracts with those from 111 contemporary bacterial isolates showed that the two sources yielded the same picture of the B. pertussis population [dominated by the MLVA-27 prn(2) ptxA(1) ptxP(3) clonal type]. There was no significant difference in MLVA type distribution or diversity between the two sample sets. This suggests that clinical extracts can be used in place of, or to complement, bacterial cultures for typing purposes (at least, in this age group). With small modifications to methodology, generating MLVA and sequence-based typing data from qPCR-positive clinical DNA extracts is likely to generate a complete dataset in the majority of samples from the ,1 year age group. Its success with samples from older subjects remains to be seen. However, our data suggest that it is suitable for inclusion in molecular epidemiological studies of the B. pertussis population or as a tool in outbreak investigations.
INTRODUCTIONThe bacterium Bordetella pertussis causes potentially fatal pertussis disease in humans. Despite vaccination programmes that are highly effective in reducing serious disease and mortality, there has been some concern in developed countries that pertussis disease is on the increase. Various explanations have been proposed, including increased ascertainment due to improved diagnostic methods, poor efficacy of particular batches of vaccine, and genetic variation occurring in the bacterial population (Campbell et al., 2012;Celentano et al., 2005;Mooi et al., 1999;Mooi, 2010;Ntezayabo et al., 2003;Poynten et al., 2004; Tanaka et al., 2003). Protein sequence variation has been observed in many B. pertussis virulence factors, including components of acellular pertussis vaccines such as pertussis toxin (Ptx), pertactin (Prn) and fimbriae (Fim) (Mooi et al., 2007;Mooi, 2010). Experiments using animal models have shown that divergent Ptx, Prn and Fim types can result in immune evasion (Robinson et al., 1989;King et al. 2001;Gzyl et al., 20...