Use of Pyrosequencing To Differentiate Streptococcus pneumoniae Serotypes 6A and 6B

Abstract

Accurate serotyping of Streptococcus pneumoniae remains important to monitor the changes in seroepidemiology of the organism over time. Though several PCR-based systems have been developed for this purpose, the cross-reactivity within serogroups often limits discrimination between types. All serogroup 6 isolates can be identified using a multiplex PCR system; however, due to the high sequence homology between the cps-6B and cps-6A loci, serotypes 6A and 6B cannot be differentiated by this method. We describe the use of pyrosequencing to reliably differentiate between serotypes 6A and 6B using a previously described single nucleotide polymorphism at codon 195 of the cps locus wciP gene. We observed complete concordance between capsular serotyping results and wciP pyrosequencing among 210 isolates examined, indicating that pyrosequencing is a rapid and accurate technique for deducing serotypes 6A and 6B.

Streptococcus pneumoniae is a major cause of morbidity and mortality worldwide, leading to ∼1.2 million deaths among young children, mostly in developing countries (6). Immunity to pneumococcal infection is largely type specific with 90 distinct immunological types being described (4). However, not all known serotypes cause serious infections; ∼15 serotypes cause the majority of invasive pneumococcal disease worldwide (3). Although this increases the probability of developing effective vaccines targeting the most frequent types, the distribution of serotypes can vary with age, geography, and time, posing greater challenges for vaccine development. Therefore, monitoring the changes in seroepidemiology of this organism is important, especially to evaluate the effect of newer vaccines.

Conventionally, serotyping is performed using the standard capsular reaction test (12), although the high cost of antisera and the technical expertise required are limitations of the system. The development of PCR-based serotyping systems has overcome some of the difficulties associated with serologic testing (1, 9), although most of the PCR systems described are limited to identifying serogroups that include cross-reacting serotypes (e.g., they can identify serotypes 6A and 6B only as serogroup 6), leading to incomplete serotype identification. The availability of the sequences of the capsular biosynthetic loci (cps) from all 90 pneumococcal serotypes at the website www.sanger.ac.uk and development of newer techniques such as pyrosequencing for accurate determination of single nucleotide polymorphisms (SNPs) may resolve this issue. Pyrosequencing is a real-time DNA sequencing technique for generating short reads rapidly and inexpensively (13). The method employs coupled enzymatic reactions to detect inorganic pyrophosphate (PP_i) released as a result of nucleotide incorporation by DNA polymerase. The released PP_i is converted to ATP by ATP sulfurylase, which provides the energy for luciferase to oxidize luciferin and generate light that is detected by a charge-coupled device camera and is seen as a peak in the pyrogram (7, 8, 13).

Recently, the difference between serotypes 6A and 6B has been putatively correlated with a single nonsynonymous substitution in the putative rhamnosyl transferase gene (wciP) (11). We utilized this information to test a pyrosequencing assay to reliably differentiate types 6A and 6B among isolates initially deduced as serogroup 6 from a PCR assay.

Primers targeting a wciP gene segment of types 6A and 6B, which encompasses the putative key polymorphism at codon 195 (see GenBank accession AF246897), were designed using the pyrosequencing assay design software (Biotage AB). wciP-f (5′ AATTTGTATTTTATTCATGCCTATATCTGG) (biotinylated at 5′ end) and wciP-r (5′ TTAGCGGAGATAATTTAAAATGATGACTA) were used to amplify a 250-bp product. PCR was done in 25-μl volumes with 1× PCR buffer (Promega Inc., WI), 2.5 mM of MgCl₂, and 200 μM of each deoxynucleoside triphosphate (New England Biolabs, MA) with 250 nM of each primer and 2.0 U of Taq DNA polymerase (Promega Inc., WI). Thermal cycling was performed with the following conditions: 94°C for 4 min and 30 amplification cycles of 94°C for 45 s, 55°C for 45 s, and 65°C for 2 min. Products were detected on 2% NuSieve agarose gels (Cambrex Bio Science, Inc., Rockland, ME). Since multiplex PCR-based serotyping may become a viable alternative to conventional serotyping (1, 9), we incorporated primers wciP-f and wciP-r into a multiplex PCR also containing primers targeting cps loci of three of the most predominant serotypes seen in the Centers for Disease Control and Prevention's (CDC's) Active Bacterial Core surveillance (see http://www.cdc.gov/ncidod/dbmd/abcs/) during 2002 and 2003 (types 19A, 3, and 22F). The primer sequences, product sizes, and concentrations used for the reaction are listed in Table 1. All 250-bp PCR products obtained using the multiplex reaction were identified as serogroup 6 and were stored at 4°C until further use in pyrosequencing.

TABLE 1.

List of oligonucleotide primers used in the multiplex PCR

Primer	Sanger's strain no. or GenBank accession no.b	Primer sequence (5′-3′)	Nucleotide positionc	Product size (bp)	Primer product concn (μM)
3-f	Z47210	ATG GTG TGA TTT CTC CTA GAT TGG AAA GTA G	9179	371	1.5
3-r		CTT CTC CAA TTG CTT ACC AAG TGC AAT AAC G	9519
6A/B-f (biotin)	AF316640	AAT TTG TAT TTT ATT CAT GCC TAT ATC TGG	8656	250	0.5
6A/B-r		TTA GCG GAG ATA ATT TAA AAT GAT GAC TA	8877
19A-fa	AF094575	GTT AGT CCT GTT TTA GAT TTA TTT GGT GAT GT	12118	478	1.0
19A-r		GAG CAG TCA ATA AGA TGA GAC GAT AGT TAG	12566
22F-f	1772/40 (da)	GAG TAT AGC CAG ATT ATG GCA GTT TTA TTG TC	11055	643	1.5
22F-r		CTC CAG CAC TTG CGC TGG AAA CAA CAG ACA AC	11666

^aPrimer previously published as in reference 1.

^bPrimers were designed using the sequence from either GenBank (accession numbers provided) or www.sanger.ac.uk (strain numbers provided).

^cStart position of each primer.

Pyrosequencing was performed with the Pyro Gold kit (Biotage AB, Uppsala, Sweden), and primer wciP-s (5′ CCATACTCTACTGCAAAC) was used to obtain a 25-bp sequence. Briefly, 14 μl of biotinylated PCR product was bound to streptavidin-coated Sepharose beads and denatured. Using a vacuum manifold device (Biotage AB, Uppsala, Sweden) the PCR product was denatured and washed, leaving the biotinylated sense strand for pyrosequencing. After cleanup, the products were transferred to a 96-well plate containing 333 nM sequencing primer in 45 μl of annealing buffer. The reaction cartridge was then loaded with deoxynucleoside triphosphates, enzyme, and substrate supplied with the kit, and the cartridge and the plate were placed in the instrument (Biotage AB) for analysis. Since each nucleotide addition takes 1 minute for completion, only 26 min was required to complete all 96 reactions.

Two hundred ten serogroup 6 isolates obtained through Active Bacterial Core surveillance were included for analysis (2). These included sterile-site isolates from both adults and children obtained during 2002 and 2003. In the first phase of this study, 175 isolates were amplified in PCRs containing wciP-f and wciP-r and subsequently pyrosequenced with wciP-s. Of the 175 products tested by pyrosequencing, 86 were identified as serotype 6B (Fig. 1) and 88 were serotype 6A (Fig. 1), correlating perfectly with conventional serotyping results. Though the genetic diversity among the isolates in this set was not determined, the strains of serotypes 6A and 6B have been previously found to be genetically diverse (5), and the assay can therefore be used to reliably differentiate serotypes even among strains with high genetic diversity. Only one isolate among our isolate set was consistently identified as serotype 6A by pyrosequencing and not 6B as originally recorded by serotyping. Repeat serotyping revealed that the original results were in error and that the isolate was actually type 6A.

FIG. 1.

The pyrograms indicate the two polymorphic sites that were consistently associated with serotypes 6A and 6B. The regions above each graph correspond to the specific SNPs used for differentiation. The two 25-base wciP coding sequence segments from the type 6A and 6B isolates examined during this work are depicted. The key residue within rhamnosyl transferase that putatively determines the key structural difference between the 6B and 6A serotypes is underlined (8).

In the second phase of this work, primers wciP-f and wciP-r were also incorporated as part of a multiplex PCR system used in part to identify serogroup 6 isolates, which also contained primers to identify serotypes 19A, 3, and 22F (CDC unpublished data), and 35 strains of serogroup 6 identified by the multiplex PCR were further subjected to pyrosequencing. There was no cross-reactivity between the primers in the multiplex PCR, and the specificity of each primer was checked against pneumococci of 59 different serogroups/types during initial standardization. Since the pyrosequencing includes a cleanup procedure before the sequencing reaction, the high primer concentration in the multiplex reaction mixture did not interfere with the pyrosequencing. All of the 35 isolates identified as serogroup 6 through multiplex PCR were differentiated as 6A or 6B by pyrosequencing, achieving an additional level of differentiation that we cannot currently achieve by conventional PCR. In addition to the SNP at residue 195, another nonsynonymous substitution that differentiated all 6A and 6B isolates tested was observed at codon 192, corresponding to an alanine-to-serine alteration (Fig. 1). It must be noted, however, that codon 192 was not entirely reliable in differentiating serotype 6A and 6B isolates (11).

Accurate determination of pneumococcal serotypes is a continuing priority, especially since the introduction of the heptavalent pneumococcal conjugate vaccine (PCV7). Monitoring of serotypes subsequent to introduction of the highly effective PCV7 is necessary to record the decline in vaccine types while simultaneously detecting possible serotype replacements with nonvaccine types. Though capsular testing has long been the preferred method for serotyping, the subjectivity in interpretation and the technical expertise required make alternate approaches attractive (1). The development of multiplex PCR systems that are rapid, reliable, and cost-effective has the capacity to make serotyping more affordable in resource-limited settings.

Pyrosequencing assays might generally allow distinguishing serotypes within strains that can currently only be serogrouped using PCR-based techniques. It is possible that similar assays can be developed to deduce the serotypes within other cross-reacting serogroups (e.g., serogroups 12, 18, 22, and 7). The assay described here rapidly and reliably detected the key polymorphism believed to be the genetic basis of the structural difference between the serotype 6A and 6B capsules (10). Though the initial setup costs for pyrosequencing are considerable, the cost per test is low at less than $1 (United States). With pyrosequencing finding greater application for diagnosis of infectious and noninfectious diseases (10, 14), the cost may drop with time, making it more affordable. This is the first report on the use of pyrosequencing for deducing serotypes of S. pneumoniae and highlights the possibility of developing multiplex pyrosequencing systems for rapid and reliable determination of pneumococcal serotypes.