|Home | About | Journals | Submit | Contact Us | Français|
One hundred fifty-two nonculturable lyophilized carriage pneumococcal isolates from a vaccine trial were subjected to PCR for serotyping, and 149 (98%) were successfully classified as vaccine or nonvaccine types, which were similar to viable isolates. The methodology will be useful for analysis of this and other studies where stored pneumococcal isolates fail to grow.
Streptococcus pneumoniae, the leading cause of childhood pneumonia and meningitis, is responsible for 700,000 to 1 million child deaths per year worldwide (9–11). However, the organism remains underreported, specifically in Asia (1), because many laboratories are unable to isolate this fastidious organism. Gathering evidence on the pneumococcal disease burden for vaccination policy can be difficult, since the organism is fragile and has >90 different serotypes. Invasive disease is caused by only a limited number of serotypes, but they vary geographically and with time. Furthermore, serotyping of pneumococcus is labor-intensive and expensive and is done by only a limited number of laboratories globally. Thus, isolates are often preserved and shipped in batches to reference laboratories for serotyping. The process of shipping the isolates is also challenging, since the preservation of these strains for long time periods is difficult (8). The strains can be preserved either at ultralow temperature (−70°C to −80°C) or by lyophilization. The latter is more convenient, especially with the air freight regulations for shipping with dry ice. However, there are reports that lyophilized pneumococcal isolates survive longer at 4°C than at room temperature (25 to 30°C), 68 months versus 4 months, respectively (8).
Recently, we worked in collaboration with International Center for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) and Cincinnati Children's Hospital Medical Center on serotyping of nasal-carriage pneumococcal strains isolated from the infants and mothers enrolled in the Mother's Gift study. This randomized study was designed to measure the effect of maternal antenatal 23-valent pneumococcal polysaccharide (23vPPS) vaccine on the infant response to 7-valent conjugate pneumococcal vaccine (PCV7) and also to assess the safety and immunogenicity of PCV7 in Bangladesh (12). The serotype information for pneumococcal isolates from this study will be useful in understanding the impact on pneumococcal carriage among infants born to mothers receiving antenatal 23vPPS vaccine with or without infant immunization with PCV.
This report describes the challenges we faced during the serotyping of these valuable isolates using quellung reactions.
We received preserved isolates which had been lyophilized and stored at room temperature since 2005 from the Mother's Gift study to assess their serotypes. In brief, the lyophilization was done using 107 to 109 CFU of pneumococcus suspended in 10% skim milk. The suspension was frozen at −70° and then dried at −50°C with 100-mtorr vacuum pressure. The laboratory was blinded regarding the identities of the subjects and the study vaccine group. On receiving 446 lyophilized pneumococcal carriage isolates in 2009, we attempted to revive them by reconstituting the lyophilized powder in brain heart infusion broth. An aliquot of the reconstituted broth was directly subcultured on Mueller-Hinton (MH) blood agar. A second aliquot was enriched by incubation in Tryptone soya broth (TSB) with 7% sterile defibrinated sheep blood for 18 h at 37°C and then cultured on MH blood agar. Two hundred ninety-four pneumococcal (66%) isolates were recovered following the procedures of direct (n = 238/294; 81%) and enriched subculture (n = 56/294, 19%), and 6 of them could not be confirmed as pneumococcus (Fig. 1). The pneumococcal isolates were serotyped by the capsular swelling method as described earlier (7).
Considering the importance of these isolates in interpreting the data generated from this unique vaccine study, we carried out further investigation of lyophilized material (n = 152) from which pneumococcus could not be isolated.
As the first step, we wanted to confirm the presence of pneumococcus in these specimens and then assayed for serotypes by PCR. In brief, DNA was extracted from the culture-negative lyophilized powder specimens using the QIAamp DNA minikit (Qiagen, Hilden, Germany) and subjected to PCR for the pneumococcus ply gene (3). Of 152 culture negative-specimens, 149 (98%) were positive for pneumococcus. The extracted DNA from these pneumococcal PCR-positive nonculturable strains was subjected to sequential multiplex PCR to identify serotype-specific genes, as described earlier for culture-negative cerebrospinal fluid (CSF) specimens (4–6). For this study, serotype detection was limited to the 23 vaccine types (except 9N, since we do not have a primer against this), considering the final analysis plan, outcome of this, and its implication in policy decisions. The primers were sequentially grouped into 6 reactions (reaction 1, primers for serotypes 6, 15B, 19F, and 23F; reaction 2, primers for serotypes 9V, 17F, 19A, and 33F; reaction 3, primers for serotypes 3, 11A, 14, and 18; reaction 4, primers for serotypes 4, 5, 7F, and 20; reaction 5, primers for serotypes 1, 8, 10A, and 12F; reaction 6, primers for serotypes 2 and 22F) based on the prevalent carriage serotypes in Bangladeshi children (2; S. K. Saha, unpublished data). Each multiplex reaction included the pneumococcal isolates of the serotypes of corresponding primers as positive controls. The PCR mixtures without template DNA served as negative controls. In addition, every 5th culture-positive specimen (n = 59) was subjected to direct PCR, following the same methodology as for culture negatives. All culture-positive cases revealed serotypes identical to those of the quellung reaction.
Sequential multiplex PCR revealed one of the vaccine types from 63 of 149 specimens. The PCR-positive specimens were classified as vaccine type, and the negatives were classified as nonvaccine type. Overall, the recovery of stored study specimens by PCR increased serotype data by 33%. There was no remarkable difference in the distribution of vaccine and nonvaccine types in culture positives and culture negatives detected by PCR (P = 0.361). Similarly, no significant difference was observed when the prevalence of individual serotypes in two groups (culture positives and negatives) were compared as determined by the Pearson chi-square test (Fig. 2).
One limitation of this study was that we did not have the primer for 9N and also lacked the specific primers to differentiate the types of serogroup 18. However, serotypes against these primers are not common. Despite that, detection of serotypes of the culture-negative lyophilized pneumococcal isolates will be useful for the Mother's Gift study because it will enable the researchers to measure the impact of maternal and infant immunization, independently and in combination, on serotype distribution of carriage. We suspect that such a loss of isolates is not uncommon and that it may occur on many other similar occasions and with other studies and researchers, where the fastidious and autolytic pneumococcus strains turn out to be nonculturable and unavailable for serotyping by classical microbiological techniques. The findings of this study are of importance for similar situations, now and in the future, where PCR techniques will improve the generation of pneumococcal identification and serotyping data at a lower cost (6) than classical microbiological techniques.
We are thankful to Mrittika Chowdhury for her help in conducting microbiological procedures.
The study was partially supported by Cincinnati Children's Hospital Medical Center and Child Health Research Foundation.
Published ahead of print 1 August 2012