Genotypic diversity and serotype distribution of group B streptococcus isolated from women before and after delivery

Abstract

Background: Most studies of the dynamics of maternal group B Streptococcus (GBS) colonization have relied on capsular serotyping to define GBS acquisition or loss. Newer molecular methods that distinguish GBS clones may expand our knowledge and influence vaccination strategies. We used multilocus sequence typing (MLST) and GBS capsular gene cluster (cps) genotyping to investigate the dynamics of perinatal GBS colonization.

Methods: A total of 338 GBS isolates obtained from 212 colonized women who were enrolled in a prior prospective cohort study were serotyped and genotyped by MLST and cps typing before (visit 1) and 6 weeks after (visit 2) delivery.

Results: Of the 212 women, 126 were colonized at both visits, whereas 66 lost and 20 acquired GBS by visit 2. MLST of the 338 strains identified 29 sequence types marking distinct bacterial clones. A change in sequence type or cps and serotype occurred in 23 (18.3%) of the 126 women who were colonized at both visits. Specific sequence types were associated with GBS loss and persistence. Older maternal age and exclusive intrapartum antibiotic use were associated with persistent colonization.

Conclusions: Although most GBS-positive pregnant women were stably colonized during the peripartum period, we detected changes in capsule expression and recolonization with antigenically distinct GBS clones over time by applying MLST. Combining the epidemiologic and molecular typing data revealed host factors and clones associated with persistent colonization, as well as a clone that was more readily lost. This knowledge is useful for the development of prevention and intervention strategies to reduce the likelihood of maternal GBS colonization.

Figure 1.

Distribution and frequency of group B Streptococcus strains, representing 29 distinct multilocus sequence types (STs), by capsular (cps) genotype. Dendrogram is a consensus of 500 bootstrap trees generated with the neighbor-joining algorithm with use of sequence data for 7 genes comprising 1152 codons. STs followed by a “v” represent ST variants.

Figure 2.

Distribution of multilocus sequence types (STs) and frequency of capsule (cps) genotypes among 338 group B Streptococcus strains from 212 women who were colonized during pregnancy (visit 1) and/or at 6–8 weeks after delivery (visit 2). The STs are ranked in order of decreasing overall frequency, and the proportion of some STs differed between visits. STs followed by a “v” represent ST variants; STs 23v and 1v are pooled together because of the high degree of similarity (bootstrap >96% and >79%, respectively) in the phylogenetic analyses. The overall frequency of cps genotypes is noted, because there were no differences in the cps distribution by visit; 2 ST-1 strains had cps5 variant genotypes but were included in the cps5 category.

Figure 3.

Distribution of multilocus sequence types, stratified by capsule (cps) genotypes, among 212 pregnant women who were colonized with group B Streptococcus (GBS) before (visit 1) and/or after (visit 2) delivery, by colonization dynamic. The sequence types are ranked in order of decreasing overall frequency, with the “others” category comprising all nonpredominant sequence types grouped together. Two women who had persistent colonization with the same GBS sequence type were colonized with different cps genotypes at visits 1 and 2; these cases were included in the lost and acquired distributions because of the cps change.