NEW YORK (GenomeWeb) – A team led by investigators at Houston Methodist Research Institute and Weill Cornell Medical College have explored the relationships between Streptococcus pyogenes strains using a combination of population genomic, transcriptomic, and virulence profiling.
The researchers did whole-genome sequencing on the emm28 "M Protein" cell surface protein-coding gene type from 2,101 invasive S. pyogenes strains collected in the US, Canada, Denmark, Finland, Norway, and Iceland between 1991 and 2016, selecting nearly 500 strains for transcriptome sequencing and 50 strains for virulence testing with the help of phylogenetic information. Their findings, published online yesterday in Nature Genetics, offered a look at the interplay between phylogenetics, transcriptome variation, and virulence in S. pyogenes, a representative from the invasive infection-causing collection of "group A Streptococcus" (GAS).
"GAS has been used as a model organism for studying the relationships among strain type and disease phenotype, and epidemics," senior author James Musser, a pathology and genomic medicine researcher at Houston Methodist's Center for Molecular and Translational Human Infectious Diseases Research, and his colleagues wrote. "[E]mm28 strains are among the top five emm types associated with invasive GAS infections in the United States and many European countries."
The researchers performed Illumina short-read genome sequencing on the 2,101 emm28 S. pyogenes isolates, selecting 24 strains for further long-read sequencing on the Oxford Nanopore Minion from a set that had been RNA sequenced three times during the mid-exponential or early stationary growth phases. To that, they added RNAtag-seq-based transcriptomic analyses on another 442 replication-free S. pyogenes strains, including wild-type strains and strains with mutations in regulatory genes such as covRS, and 50 strains assessed with virulence testing in mouse models of necrotizing myositis.
"Despite important advances in the genomics of selected organisms, little is known about the nature and extent of transcriptome diversity among clonally related progeny of bacterial strains that have shared a recent common ancestor," the authors explained. "Data on this issue are critical for enhanced understanding of bacterial evolution in natural populations, phenotypic diversification, and microbial epidemics."
The team narrowed in on a small insertion or deletion involving a single nucleotide in a non-coding site between the Spy1336/R28 and Spy1337 genes that influenced virulence and transcriptional profiles in S. pyogenes, including the expression of a secreted virulence factor called R28 that is encoded by Spy1336/R28. Subsequent mouse model experiments also highlighted a sub-clade known as SC2A that included S. pyogenes strains that were particularly virulent.
The researchers went on to analyze the genomic, transcriptomic, and virulence data on hand with genome-wide association, expression quantitative trait locus, machine learning, and other approaches, identifying enhanced virulence in strains with the suspicious indel or in the SC2A subclade. In the mouse experiments, for example, the investigators found that SC2A strains were significantly more virulent than those in the SC1A or SC1B sub-clades.
"We found that, compared with non-SC2A strains from the United States, a significantly higher proportion of SC2A strains from the United States was associated with puerperal sepsis, neonatal infections, and female genital tract infections," the authors reported, noting that the SC2A strains "were also significantly more virulent in the mouse necrotizing myositis experiments."
The investigators concluded that their overall research strategy "is generally applicable to any microbe — pathogenic or otherwise — and may lead to new therapeutics," noting that "integration of the three different types of data resulted in a more enhanced understanding of the molecular genetics of a pathogen than the study of any one or two of the three types of data."