Single-Nucleotide Polymorphisms within the cps Loci: Another Potential Source of Clinically Important Genetic Variation for Streptococcus pneumoniae?

ABSTRACT

The Streptococcus pneumoniae capsule is essential for disease pathogenesis, suggesting that even minor genetic changes within the cps locus could potentially have important consequences. Arends et al. (D. W. Arends, W. R. Miellet, J. D. Langereis, T. H. A. Ederveen, et al., Infect Immun 89:e00246-21, 2021, https://doi.org/10.1128/IAI.00246-21) have identified 79 different nonsynonymous single-nucleotide polymorphisms (SNPs) in the cps locus of 338 19A serotype strains and shown significant variations between strains in nucleotide sugar content and capsule shedding. Further work is required to characterize whether any of these changes have important functional consequences on capsule-host interactions.

COMMENTARY

The virulence of the Gram-positive Streptococcus pneumoniae pathogen is critically dependent on its polysaccharide capsule. The capsule assists initial colonization of the nasopharynx by preventing S. pneumoniae from being entrapped by mucus (1), and it is essential for invasive infection as it inhibits both innate and adaptive immune clearance by inhibiting bacterial opsonization with complement and antibody (2). The S. pneumoniae capsule consists of repeating units of oligosaccharides linked to the bacterial cell wall, and the considerable variation in the monosaccharides, their order and type of chemical bonds within the repeating unit, and the presence or absence of side chains results in multiple chemical structures that are antigenically distinct. Around 100 S. pneumoniae capsular serotypes have been described, and this biochemical and antigen diversity has major implications for disease. Some capsular serotypes are much more likely to cause invasive disease (e.g., serotypes 4, 7F, and 14), and an invasive phenotype correlates closely with resistance to complement-mediated neutrophil phagocytosis (3–5). Other capsular serotypes (e.g., 6A, 23F, and 19F) are much less invasive and are more sensitive to complement-mediated immunity but tend to have longer duration of colonization (3). For most serotypes, the genes necessary for capsule synthesis are contained within a single cps locus consisting of between 10 to 20 genes. As a consequence, S. pneumoniae strains can change capsular serotype by replacing their existing cps locus by recombination with DNA from another S. pneumoniae serotype (6), which could have large effects on resistance to opsonophagocytosis. In addition, relatively minor genetic changes can alter the capsule chemical structure enough to change serotype and alter sensitivity to complement (7). The importance of the capsule for S. pneumoniae biology has been further emphasized by the widespread use of vaccination of both children and adults using capsular polysaccharide as the target antigen. Vaccination of children largely eradicates the vaccine serotypes as nasopharyngeal commensals, resulting in major changes in S. pneumoniae ecology due to compensatory expansion in the prevalence of nonvaccine serotypes.

As well as the profound effects of capsular serotype on S. pneumoniae biology and interactions with the host, there are additional effects of the capsule that are independent of its chemical structure and serotype. The best described is phase variation, with opaque-phase S. pneumoniae cells having a relatively thick capsule layer and an invasive phenotype, whereas transparent-phase S. pneumoniae cells have thinner capsule layers and are associated with colonization and biofilm formation (8, 9). In addition, sensitivity to complement-mediated phagocytosis differs between S. pneumoniae strains expressing the same capsular serotypes (10). More recently, genetic and biochemical data have demonstrated that other commensal streptococci (mainly Streptococcus mitis, the closest genetic relative of S. pneumoniae) carry cps loci very similar to some S. pneumoniae cps loci and express serologically identical capsules (11). However, the thickness and functional effects of S. mitis capsules can differ compared to the same serotype expressed by S. pneumoniae, generally resulting in weaker protection against host innate immunity (12). These data demonstrate that serotype-independent factors can influence phenotypes associated with capsule expression and raise the question of whether genetic variation between cps loci of S. pneumoniae strains expressing the same capsular serotype can have functional consequences that are important for disease. For example, can minor genetic changes in the cps locus alter capsule width or cause subtle changes in the capsule’s biochemical structure that alter the physical properties and thereby affect interactions with the host?

In this issue of Infection and Immunity, the paper by Arends et al. (13) starts to address this question. The authors present data on genetic variation within the cps loci for 338 serotype 19A strains, a serotype which increased in prevalence as a cause of invasive disease after the introduction of routine vaccination of children with Prevnar, which targets seven other serotypes. The authors identify a considerable amount of genetic variation in the cps locus between these serotype 19A strains. This genetic variation divides the 19A strains into 8 subtypes based on a PCR method for assessing cps subtypes. However, the number of 19A subtypes increases substantially to 100 when cps locus genome data were used to divide strains into multilocus sequence types (termed cpsMLST). In total, they identified 79 different nonsynonymous single-nucleotide polymorphisms (SNPs), mainly concentrated in the following three genes: rmlB and rmlD, both required for synthesis of rhamnose, which is one of the three monosaccharides present in the 19A capsule repeating unit; and wzg, which has a poorly understood function but is involved in capsule expression and perhaps shedding. The SNPs resulted in 22 different RmlB, RmlD, or Wzg proteins in total, and similar rml gene polymorphisms tended to the cluster within each 19A subtype. Furthermore, the cpsMLST phylogenetic tree divided strains containing the SNPs into two groups consisting of SNP1 to SNP29 plus SNP49 and SNP30 to SNP48. Additional genetic changes in the cps locus identified by Arends et al. (13) include two subtypes of the cps locus promoter. The concentration of mutations in three out of 16 genes within the 19A cps locus is interesting and could suggest potential evolutionary advantages for mutations in these genes. However, it could also reflect the opposite effect, with a lack of functional effects allowing these genes to tolerate mutations more readily than other cps loci genes.

Does this genetic variation within the 19A cps locus have any functional consequences and, therefore, could it be biologically important, or do the SNPs have little consequence for S. pneumoniae biology? The answer to this question remains unclear. The genome data indicate some clustering of the mutations to specific sites, but for rmlB and rmlD, the amino acid changes were predicted to affect sites in the protein that are less likely to affect function directly. To answer whether the genetic changes affect the capsule function requires comparing suitable phenotypes between the 19A strains showing genetic differences in their cps loci. The authors evaluated two phenotypes that could be affected by these mutations. One was capsule shedding, which has been linked to the function of Wzg. However, capsule shedding differed markedly between strains but did not correlate with any specific wzg SNPs. Arends et al. also used mass spectrometry to measure the levels of nucleotide sugars in the bacterial cell (13). Mass spectrometry demonstrated marked variations between strains in the total sugar content and relative ratios of different monosaccharides. This is an important observation but does require confirmation of the reproducibility of the differences between strains over time and under different culture conditions. The monosaccharides affected included those utilized in the 19A capsule, such as rhamnose, and this provides a potentially link to the SNPs affecting rmalB and rmalD. However, the differences between strains in specific nucleotide sugars did not correlate with specific SNPs, although the subgroup consisting of SNP1 to SNP29 and SNP49 had higher levels overall of multiple monosaccharides compared to those in the SNP30 to SNP48 subgroup. As the supply of substrates is thought to affect capsule quantity (14), this observation may have functional relevance that needs further investigation.

The most obvious question about the biological relevance of the SNPs that needs to be answered is whether they affect the width of the capsule layer. This was not addressed by the authors, as it is in fact not straightforward to measure differences in capsule width between strains, especially when investigating a large number of strains. The most common way of measuring capsule width uses fluorescein isothiocyanate (FITC)‐dextran exclusion and conventional microscopy, but this methodology lacks the sensitivity needed to identify subtle changes in capsule size (15). Perhaps the most sensitive technique is electron microscopy using capsule sparing preparation techniques, but this is a low-throughput technique that can only be applied to a limited number of strains (12). Atomic force microscopy has recently been used to directly assess the physical properties of the capsule and has the potential to identify subtle differences between strains (12). However, this technique is highly time-consuming and even more low-throughput than electron microscopy. Both electron and atomic force microscopy could only be used to compare a few strains after prior downselection by higher-throughput methods. Higher-throughput methods suitable for assessing smaller differences in capsule quantity between multiple strains include enzyme-limited immunosorbent assay (ELISA) or flow cytometry using anticapsular antibody (12, 16). Alternatively, flow cytometry assays can rapidly compare complement resistance or antibody binding to subcapsular protein antigens for dozens of strains and therefore identify important phenotypes that differ with capsule thickness or chemical composition (4, 5, 10). These assays give an indirect indication of changes in capsule width. When preparing S. pneumoniae bacteria for assays measuring capsule content, culture under stress conditions specifically affecting carbohydrate availability or metabolism could be important, as this might accentuate subtle effects of cps locus mutations on capsule synthesis. Those strains showing significant differences with the higher-throughput assays could then be investigated using electron and/or atomic force microscopy.

Whether any differences in the capsule phenotype translate into effects on virulence can be assessed using competitive infection experiment in mouse models of infection. These are highly sensitive at identifying relatively small effects on virulence and can assess interactions between bacterial genetic factors and host immune effectors (17). For all phenotype analyses, any differences identified between strains will be confounded by genetic variation outside the cps loci. Hence, linking any specific phenotypes directly to genetic change in the cps locus will require reconstructing the mutations in another strain and repeating the phenotype analyses. Although SNPs that cause changes in S. pneumoniae resistance to immune effectors and in virulence in mice could be biologically important, this ultimately needs to be demonstrated by epidemiology studies showing that they associate with differences in colonization or invasive disease in humans. The authors found no associations in the 19A SNPs with whether a strain was isolated from the nasopharynx or from invasive infection. However, it is likely only a small number of the SNPs (or combinations of certain SNPs) could have serious functional consequences. To identify these will require longitudinal studies that can identify whether specific serotype subtypes rapidly become dominant in humans compared to other subtypes of the same capsular serotype.

The capsule is vital for virulence and disease caused by multiple pathogens; so, do the findings by Arends et al. (13) have any broader relevance for nonpneumococcal pathogens? In fact, SNPs have already been shown to affect the quantity of capsule expressed by Streptococcus agalactiae (16). Hence, minor genetic changes affecting capsule genes could indeed have important functional consequences for other pathogens. Compared to S. pneumoniae, most pathogens have very limited variation between strains in the chemical structure of their capsule and therefore limited numbers of capsular serotypes. This perhaps might make minor genetic variation in capsule genes more important for other pathogens than it is for S. pneumoniae, as any effects on disease potential will be less obscured by variations in virulence between serotypes.

The description by Arends et al. (13) of genetic variation within the cps loci of S. pneumoniae strains expressing the same capsular serotype potentially adds another layer of complexity to the investigation of the consequences of genetic and epigenetic variation between strains on disease pathogenesis. Further detailed phenotyping of multiple strains will be needed to ascertain whether genetic variation within the cps locus does result in significant biological effects independent of capsular serotype.

The views expressed in this Commentary do not necessarily reflect the views of the journal or of ASM.