Rarely Occurring 19A-Like cps Locus from a Serotype 19F Pneumococcal Isolate Indicates Continued Need of Serology-Based Quality Control for PCR-Based Serotype Determinations

Abstract

Determination of the DNA sequences corresponding to all 91 pneumococcal capsular biosynthetic (cps) loci (1, 3, 6, 9) has allowed serotype determinations from pneumococcal isolates and clinical specimens using conventional PCR assays (2, 4, 7, 8, 10). We developed sequential, multiplexed PCR schemes that resolve 33 serologic specificities, including 20 serotypes and 13 serotype subsets (4, 7, 8). We recently expanded the scheme to 40 specificities (www.cdc.gov/ncidod/biotech/strep/pcr.htm). PCR primers targeted serotype-specific open reading frames within central portions of known cps loci (4, 7, 8). Subsequent information (1, 3, 6) revealed that we primarily targeted serotype-specific oligosaccharide repeat unit polymerases (wzy encoded), glycosyl transferases, and flippases (wzx encoded). Our serotype 19A primers, however, targeted the mnaA gene that encodes UDP-N-acetylglucosamine-2-epimerase (8). Although the mnaA sequence differs markedly among the 15 cps loci that harbor it (serotypes 19A, -B, -C, -F; 12A, -B, -F; 9A, -N, -L, -V; 4; 36; 44; 46), due to potential exchanges of heterologous mnaA genes, we recently changed the 19A PCR assay to target wzy (our unpublished data). The wzy19A and wzy19F genes putatively provide the basis of the structural difference between the related 19A and 19F capsules (1). Recently, we evaluated Streptococcus pneumoniae strain 2584-08(19F) that was typed as 19F using the serology-based reaction and found that it yielded a false-positive 19A result with our mnaA-based assay as well as a false-negative 19F PCR result using our wzy19F-specific assay. We found that this strain had an apparently rare multilocus sequence genotype 3040 (ST3040) that is not related to other known genotypes. Strain 2584-08(19F) reactivity with seven different monoclonal antibodies targeting 19A and/or 19F (5, 11) was consistent with results from previously characterized 19F strains (M. H. Nahm, unpublished data). We retrospectively tested 71 diverse 19F isolates using the 19A mnaA assay and found no false positives. We also found no discrepant results among 100 mnaA PCR-based 19A isolates also determined to be 19A based upon quellung reaction results. Unexpectedly, upon testing strain 2584-08(19F) using our newer 19A assay based upon the wzy19A gene sequence (completely conserved within the three available cps19A loci), we again obtained a false-positive 19A result. Using primers 19AFwzyF (5′TTGAAGTTGATGAAAGAAAACGAGG) and 19AFwzyR (5′ATAAAGTTGCCAGTAACTGTACTC), we subsequently amplified and sequenced the 1,335-bp wzy structural gene from 2584-08(19F) and found that the sequence (GenBank accession no. {type:entrez-nucleotide,attrs:{text:FJ829071,term_id:229559728,term_text:FJ829071}}FJ829071) shared 88% sequence identity with wzy19A from known 19A cps loci (GenBank accession no {type:entrez-nucleotide,attrs:{text:CR931675,term_id:68643359,term_text:CR931675}}CR931675 is representative) and only 78% identity with wzy19F from known 19F cps loci (GenBank accession no. {type:entrez-nucleotide,attrs:{text:CR931678,term_id:68643444,term_text:CR931678}}CR931678 is representative). As expected, we found near identity to the wzy19A-based primers and marked divergence from our standard wzy19F-based typing primers. Subsequently, we formulated new serotype 19A identification primers (19AF3 [5′ GAGAGATTCATAATCTTGCACTTAGCCA] and 19AR3 [5′ CATAATAGCTACAAATGACTCATCGCC]) based upon differences between wzy19A and all other wzy genes, including the 2584-08(19F) wzy gene. The current wzy19A-based PCR assay tested positive against 42 serotype 19A isolates and was negative against 31 diverse 19F isolates, including 2584-08(19F). Due to the high homology of the wzy gene of 2584-08(19F) with the serotype 19A wzy gene, we are unable to design a new wzy-specific primer set that would be specific for both 2594-08(19F) and commonly recovered type 19F isolates without giving false-positive PCR results for 19A isolates. We place more priority upon eliminating rare false-positive 19A results than illuminating rarely encountered false nontypeable results for serotype 19F. Based upon cumulative results from concurrent serology- and PCR-based testing of diverse isolate sets, we are confident that the sequential PCR assay is accurate for the vast majority of isolates. However, in view of this newly identified potential for false-positive 19A results, we will continue quellung serotyping in parallel with sequential PCR-based serotype determination until biologically based sequence signatures are identified for each serotype.