Comparison of contemporary invasive and non-invasive Streptococcus pneumoniae isolates reveals new insights into circulating anti-microbial resistance determinants

Abstract

Streptococcus pneumoniae is a major human pathogen with a high burden of disease. Non-invasive isolates (those found in non-sterile sites) are thought to be a key source of invasive isolates (those found in sterile sites) and a reservoir of anti-microbial resistance (AMR) determinants. Despite this, pneumococcal surveillance has almost exclusively focused on invasive isolates. We aimed to compare contemporaneous invasive and non-invasive isolate populations to understand how they interact and identify differences in AMR gene distribution. We used a combination of whole-genome sequencing and phenotypic anti-microbial susceptibility testing and a data set of invasive (n = 1,288) and non-invasive (n = 186) pneumococcal isolates, collected in Victoria, Australia, between 2018 and 2022. The non-invasive population had increased levels of antibiotic resistance to multiple classes of antibiotics including beta-lactam antibiotics penicillin and ceftriaxone. We identified genomic intersections between the invasive and non-invasive populations and no distinct phylogenetic clustering of the two populations. However, this analysis revealed sub-populations overrepresented in each population. The sub-populations that had high levels of AMR were overrepresented in the non-invasive population. We determined that WamR-Pneumo was the most accurate in silico tool for predicting resistance to the antibiotics tested. This tool was then used to assess the allelic diversity of the penicillin-binding protein genes, which acquire mutations leading to beta-lactam antibiotic resistance, and found that they were highly conserved (≥80% shared) between the two populations. These findings show the potential of non-invasive isolates to serve as reservoirs of AMR determinants.

Keywords: Streptococcus pneumoniae; antibiotic resistance; genomics.

Fig 1

Midpoint-rooted maximum likelihood phylogenetic tree of all study isolates (n = 1,474). Minor serotypes, sequence types (STs), and global pneumococcal sequence cluster (GPSC) were defined as those that contained less than 10 isolates over the study period. The reference was ASM966447v1 (GCA_009664475.1), collected 2014, serotype 19A, ST199.

Fig 2

Summary of the key characteristics of the overrepresented sub-populations. The antibiotics included in the diagram are those that had a significant difference between the two populations (as per Table 1). Trimeth-sulfa refers to trimethoprim-sulfamethoxazole. The overrepresented sub-populations are colored based on which data set had the larger proportion of isolates: red for invasive and blue for non-invasive. The percentages displayed under the overrepresented population type/clusters are the differences in proportion between the two populations. A count of isolates for each typing method is displayed at the top of the bar chart. The proportion of isolates that were susceptible (blue), intermediate (orange), or resistant (red) to the antibiotics listed is also displayed [CLSI (21) breakpoints, meningitis/oral breakpoints used where indicated]. Minor types/clusters are defined as those having <10 isolates over the whole study period.

Fig 3

Validation of the in silico resistance prediction method. Proportion of isolates classified as major or minor errors for each in silico resistance prediction method. Major errors were defined as isolates that were phenotypically resistant or intermediate but labeled as susceptible by genotype. Minor errors were defined as isolates that were phenotypically susceptible but labeled as resistant/intermediate by genotype. Isolates phenotypically classified as intermediate or resistant were grouped to allow uniform comparison between all the methods.

Fig 4

Proportion of pbp alleles that were unique to each data set. The category pbp1a_2x is based on a combined pbp1a and pbp2x allele.

Fig 5

Midpoint rooted maximum likelihood phylogenetic tree of PBP2x. Tree has been built using an amino acid sequence alignment of all unique proteins that had at least one isolate with at least one dilution difference between the phenotypic and genotypic MICs. The number of isolates is based on counts of isolates that had contained the PBP2x alleles. The resistance motifs are the motifs used by WamR-Pneumo to determine the beta-lactam MIC, and the amino acids have been colored based on their side-chain chemistry. The MIC panels [penicillin (oral breakpoints), cefuroxime (meningitis breakpoints) and ceftriaxone (meningitis breakpoints)] show the proportion of isolates with each MIC for each of the PBP2x alleles (blue for susceptible and orange/red for intermediate/resistant). The panels displaying the number of dilutions different between the phenotype and the genotype used WamR-Pneumo to infer the MIC from the genotype.