Limits...
Selective and genetic constraints on pneumococcal serotype switching.

Croucher NJ, Kagedan L, Thompson CM, Parkhill J, Bentley SD, Finkelstein JA, Lipsitch M, Hanage WP - PLoS Genet. (2015)

Bottom Line: Eleven of these were within-serogroup switches, representing a highly significant (p < 0.0001) enrichment based on the observed serotype distribution.Whereas the recombinations resulting in between-serogroup switches all spanned the entire cps locus, some of those that caused within-serogroup switches did not.Exclusion of these genetic and physiological hypotheses suggested future work should focus on alternative mechanisms, such as host immunity spanning multiple serotypes within the same serogroup, which might explain the observed pattern.

View Article: PubMed Central - PubMed

Affiliation: Department of Infectious Disease Epidemiology, Imperial College London, London, United Kingdom; Center for Communicable Disease Dynamics, Department of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, Massachusetts, United States of America.

ABSTRACT
Streptococcus pneumoniae isolates typically express one of over 90 immunologically distinguishable polysaccharide capsules (serotypes), which can be classified into "serogroups" based on cross-reactivity with certain antibodies. Pneumococci can alter their serotype through recombinations affecting the capsule polysaccharide synthesis (cps) locus. Twenty such "serotype switching" events were fully characterised using a collection of 616 whole genome sequences from systematic surveys of pneumococcal carriage. Eleven of these were within-serogroup switches, representing a highly significant (p < 0.0001) enrichment based on the observed serotype distribution. Whereas the recombinations resulting in between-serogroup switches all spanned the entire cps locus, some of those that caused within-serogroup switches did not. However, higher rates of within-serogroup switching could not be fully explained by either more frequent, shorter recombinations, nor by genetic linkage to genes involved in β-lactam resistance. This suggested the observed pattern was a consequence of selection for preserving serogroup. Phenotyping of strains constructed to express different serotypes in common genetic backgrounds was used to test whether genotypes were physiologically adapted to particular serogroups. These data were consistent with epistatic interactions between the cps locus and the rest of the genome that were specific to serotype, but not serogroup, meaning they were unlikely to account for the observed distribution of capsule types. Exclusion of these genetic and physiological hypotheses suggested future work should focus on alternative mechanisms, such as host immunity spanning multiple serotypes within the same serogroup, which might explain the observed pattern.

Show MeSH

Related in: MedlinePlus

Extent of recombinations affecting the cps locus.(A) Positions of recombinations affecting the cps locus. Each of the recombinations identified in Fig. 3 is represented by a horizontal bar coloured according to the inferred impact of the recombination. Recombinations causing a change in serotype are uniquely annotated to correspond with Fig. 3. The grey column represents the extent of the cps locus; the width displayed is that of the longest cps locus across the different sequence clusters, and each recombination is scaled relative to the positions of this cps locus’ boundaries. (B) Extent of recombinations affecting the cps locus. The degree to which the recombinations displayed in (A) impinge on the flanking regions is summarised, with the dashed line representing the falling numbers of recombinations extending to bases further removed from the cps locus edges (displayed on the same scale as (A)). The red lines represent the fit of exponential decay curves to these trends. The tick marks above the plots represent 10 kb intervals.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4380333&req=5

pgen.1005095.g004: Extent of recombinations affecting the cps locus.(A) Positions of recombinations affecting the cps locus. Each of the recombinations identified in Fig. 3 is represented by a horizontal bar coloured according to the inferred impact of the recombination. Recombinations causing a change in serotype are uniquely annotated to correspond with Fig. 3. The grey column represents the extent of the cps locus; the width displayed is that of the longest cps locus across the different sequence clusters, and each recombination is scaled relative to the positions of this cps locus’ boundaries. (B) Extent of recombinations affecting the cps locus. The degree to which the recombinations displayed in (A) impinge on the flanking regions is summarised, with the dashed line representing the falling numbers of recombinations extending to bases further removed from the cps locus edges (displayed on the same scale as (A)). The red lines represent the fit of exponential decay curves to these trends. The tick marks above the plots represent 10 kb intervals.

Mentions: Including only the switch from serotype 6B to 6C from SC6 (see S1 Text), the analysis of whole genome alignments used to generate Fig. 3 identified 50 putative homologous recombination events that overlapped the cps loci of the relevant reference genome sequences (Fig. 4). The nine recombinations that resulted in a change in serogroup had a median length of 42.7 kb, and all spanned the entire cps locus. The eleven recombinations that led to a within-serogroup switch had a shorter median length, at 30.4 kb (Fig. 5A). These almost all spanned the 5’ region of the cps locus, but in some cases did not extend so far as the 3’ end. In two cases, this could be ascribed to the presence of the rml rhamnose synthesis operon at the 3’ end of the cps locus. Both the acquisition of the 6B capsule in SC13, and the 6C capsule in SC6, terminated within this gene cluster; however, the rml operon is found in several serogroups [2], and recombinations causing between-serogroup switches have previously been observed to end within it [29]. Only three recombination events were observed where the 3’ recombination breakpoint occurred in a region that might be considered serogroup-specific. Two of these were switches from 23A to 23F within SC9 (Fig. 4). In both cases, the wzy polymerase gene that distinguished these cps loci was replaced; however, the recombinations were far more extensive than this minimal alteration, as they extended to or beyond the 5’ boundary of the cps locus. The switch to 6C within SC13 that ended before the 3’ boundary of the locus was also far more extensive than simply encompassing the wciN gene. The overall difference in the distribution of lengths between recombinations causing within- and between-serogroup switches was not significant (Fig. 5; Wilcoxon rank sum test, W = 35, p value = 0.29). By contrast, the 30 recombinations inferred to overlap with the cps loci that did not affect serotype had a median length of 11.7 kb and were significantly shorter than both the recombinations that alter serogroup (Wilcoxon rank sum test, W = 22, p value = 3.4x10-5) and those that caused within-serogroup switching (Wilcoxon rank sum test, W = 61, p value = 0.0016). As the majority of within-serogroup switches were caused by recombinations long enough to cause changes in serogroup, restrictions on transformation event length cannot fully explain the observed pattern of switching.


Selective and genetic constraints on pneumococcal serotype switching.

Croucher NJ, Kagedan L, Thompson CM, Parkhill J, Bentley SD, Finkelstein JA, Lipsitch M, Hanage WP - PLoS Genet. (2015)

Extent of recombinations affecting the cps locus.(A) Positions of recombinations affecting the cps locus. Each of the recombinations identified in Fig. 3 is represented by a horizontal bar coloured according to the inferred impact of the recombination. Recombinations causing a change in serotype are uniquely annotated to correspond with Fig. 3. The grey column represents the extent of the cps locus; the width displayed is that of the longest cps locus across the different sequence clusters, and each recombination is scaled relative to the positions of this cps locus’ boundaries. (B) Extent of recombinations affecting the cps locus. The degree to which the recombinations displayed in (A) impinge on the flanking regions is summarised, with the dashed line representing the falling numbers of recombinations extending to bases further removed from the cps locus edges (displayed on the same scale as (A)). The red lines represent the fit of exponential decay curves to these trends. The tick marks above the plots represent 10 kb intervals.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4380333&req=5

pgen.1005095.g004: Extent of recombinations affecting the cps locus.(A) Positions of recombinations affecting the cps locus. Each of the recombinations identified in Fig. 3 is represented by a horizontal bar coloured according to the inferred impact of the recombination. Recombinations causing a change in serotype are uniquely annotated to correspond with Fig. 3. The grey column represents the extent of the cps locus; the width displayed is that of the longest cps locus across the different sequence clusters, and each recombination is scaled relative to the positions of this cps locus’ boundaries. (B) Extent of recombinations affecting the cps locus. The degree to which the recombinations displayed in (A) impinge on the flanking regions is summarised, with the dashed line representing the falling numbers of recombinations extending to bases further removed from the cps locus edges (displayed on the same scale as (A)). The red lines represent the fit of exponential decay curves to these trends. The tick marks above the plots represent 10 kb intervals.
Mentions: Including only the switch from serotype 6B to 6C from SC6 (see S1 Text), the analysis of whole genome alignments used to generate Fig. 3 identified 50 putative homologous recombination events that overlapped the cps loci of the relevant reference genome sequences (Fig. 4). The nine recombinations that resulted in a change in serogroup had a median length of 42.7 kb, and all spanned the entire cps locus. The eleven recombinations that led to a within-serogroup switch had a shorter median length, at 30.4 kb (Fig. 5A). These almost all spanned the 5’ region of the cps locus, but in some cases did not extend so far as the 3’ end. In two cases, this could be ascribed to the presence of the rml rhamnose synthesis operon at the 3’ end of the cps locus. Both the acquisition of the 6B capsule in SC13, and the 6C capsule in SC6, terminated within this gene cluster; however, the rml operon is found in several serogroups [2], and recombinations causing between-serogroup switches have previously been observed to end within it [29]. Only three recombination events were observed where the 3’ recombination breakpoint occurred in a region that might be considered serogroup-specific. Two of these were switches from 23A to 23F within SC9 (Fig. 4). In both cases, the wzy polymerase gene that distinguished these cps loci was replaced; however, the recombinations were far more extensive than this minimal alteration, as they extended to or beyond the 5’ boundary of the cps locus. The switch to 6C within SC13 that ended before the 3’ boundary of the locus was also far more extensive than simply encompassing the wciN gene. The overall difference in the distribution of lengths between recombinations causing within- and between-serogroup switches was not significant (Fig. 5; Wilcoxon rank sum test, W = 35, p value = 0.29). By contrast, the 30 recombinations inferred to overlap with the cps loci that did not affect serotype had a median length of 11.7 kb and were significantly shorter than both the recombinations that alter serogroup (Wilcoxon rank sum test, W = 22, p value = 3.4x10-5) and those that caused within-serogroup switching (Wilcoxon rank sum test, W = 61, p value = 0.0016). As the majority of within-serogroup switches were caused by recombinations long enough to cause changes in serogroup, restrictions on transformation event length cannot fully explain the observed pattern of switching.

Bottom Line: Eleven of these were within-serogroup switches, representing a highly significant (p < 0.0001) enrichment based on the observed serotype distribution.Whereas the recombinations resulting in between-serogroup switches all spanned the entire cps locus, some of those that caused within-serogroup switches did not.Exclusion of these genetic and physiological hypotheses suggested future work should focus on alternative mechanisms, such as host immunity spanning multiple serotypes within the same serogroup, which might explain the observed pattern.

View Article: PubMed Central - PubMed

Affiliation: Department of Infectious Disease Epidemiology, Imperial College London, London, United Kingdom; Center for Communicable Disease Dynamics, Department of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, Massachusetts, United States of America.

ABSTRACT
Streptococcus pneumoniae isolates typically express one of over 90 immunologically distinguishable polysaccharide capsules (serotypes), which can be classified into "serogroups" based on cross-reactivity with certain antibodies. Pneumococci can alter their serotype through recombinations affecting the capsule polysaccharide synthesis (cps) locus. Twenty such "serotype switching" events were fully characterised using a collection of 616 whole genome sequences from systematic surveys of pneumococcal carriage. Eleven of these were within-serogroup switches, representing a highly significant (p < 0.0001) enrichment based on the observed serotype distribution. Whereas the recombinations resulting in between-serogroup switches all spanned the entire cps locus, some of those that caused within-serogroup switches did not. However, higher rates of within-serogroup switching could not be fully explained by either more frequent, shorter recombinations, nor by genetic linkage to genes involved in β-lactam resistance. This suggested the observed pattern was a consequence of selection for preserving serogroup. Phenotyping of strains constructed to express different serotypes in common genetic backgrounds was used to test whether genotypes were physiologically adapted to particular serogroups. These data were consistent with epistatic interactions between the cps locus and the rest of the genome that were specific to serotype, but not serogroup, meaning they were unlikely to account for the observed distribution of capsule types. Exclusion of these genetic and physiological hypotheses suggested future work should focus on alternative mechanisms, such as host immunity spanning multiple serotypes within the same serogroup, which might explain the observed pattern.

Show MeSH
Related in: MedlinePlus