Australian researchers have uncovered new genetic insights into Staphylococcus aureus, revealing what makes the bacterium so dangerous when it enters the blood.
While common, with up to a 20% mortality and increasing incidence rates, Staphylococcus aureus infections are a key challenge to human health. Golden staph is notorious for its ability to become resistant to antibiotics, making it hard to treat, which can lead to adverse health outcomes for patients infected with a drug resistant form of the bacteria.
In one of the most comprehensive studies of its kind, published 12 September in Cell Reports, researchers led by the Peter Doherty Institute for Infection and Immunity (Doherty Institute), analysed the unique genetic profiles of more than 1,300 Golden staph strains.
By combining this data with patient and antibiotic information, the researchers found that while patient factors were critical in determining mortality risks, specific genes were linked to antibiotic resistance, along with the bacteria’s ability to linger in the blood, evading antibiotics, and the immune system.
University of Melbourne Dr Stefano Giulieri, a Clinician-Researcher at the Doherty Institute and lead author, explained that a key concept underlying bacterial phenotype-genotype studies was pathoadaptation, where bacteria acquire mutations that enhance their ability to survive in the host, and that antibiotic resistance was an adaptive phenotype with high heritability and strong selective pressure, where convergent evolution was expected to be apparent.
“Upon infection, colonizing S. aureus strains are repeatedly subjected to similar selective pressures; thus, mutations promoting survival in the host during infection are anticipated to be preferentially maintained and become evident when examining S. aureus microevolution,” he said.
“Bacterial adaptation can be frequently detected in clinical infections, leading to antibiotic resistance, antibiotic tolerance, or immune evasion, and potentially driving treatment failure.”
Genome wide association studies (GWASs) scan the genes of a large collection of bacteria to look for tiny changes (mutations) that show up more often in strains with a certain characteristic, such as antibiotic resistance, and mutations with a strong statistical link are precious clues to figure out how bacteria acquire attributes that are important for patient outcomes.
“However, it was unclear whether clinical outcomes of bacterial infections were dependent on bacterial genotypes, as shown by bacterial genome wide association studies (GWASs) of clinical outcomes, which have been either negative or have not been reproduced across independent cohorts,” Dr Giulieri said.
“This is due to the considerable genetic diversity of bacterial pathogens and the strong clonal structure of bacterial populations that complicate GWASs. In addition, many bacterial GWASs addressing clinical outcomes face specific issues like small sample size, retrospective data collection, and unbalanced datasets.
“Here, we propose a statistical genomics framework that incorporates convergent evolution explicitly and tracks molecular signatures of adaptation across three independent SAB patient cohorts to build a model of bacterial genotype-phenotype associations for treatment outcomes.”
Given the complexity of identifying genetic correlates of clinical outcomes, the team progressed their analysis from clinically relevant bacterial phenotypes with high expected role of adaptation (vancomycin MIC) to intermediate outcomes (duration of bacteremia), to final endpoints with strong multi-factorial pathogenesis (infection mortality).
University of Melbourne’s Professor Ben Howden, the Director of the Microbiological Diagnostic Unit Public Health Laboratory at the Doherty Institute and co- author of the paper, said that consistent with the heritability estimate, the team identified several mutations that were significantly associated with vancomycin MIC.
“While this indicates inflation (despite correction for population structure in the linear mixed model) and confirms the critical role of lineage effects in antibiotic resistance in S. aureus, the analysis was strongly biased by the population structure,” he said.
“We next assessed genetic correlates of duration of bacteremia and as expected, the distribution of this phenotype was skewed, with 59% of strains being associated with a duration of bacteremia of 1 day or less.
“Using automated normalization optimization (based on Pearson P statistics), we found that none of the most common available strategies could improve the distribution of this variable, and just one mutation (non-synonymous substitution in the isomerase galE) barely achieved genome-wide significance, in accordance with a lower heritability and an expected polygenic signal associated with this complex trait.”
However, loci that were associated with duration of bacteremia were identified using a truncation burden test, with the strongest association found in protein truncations in murQ, an esterase that is part of the peptidoglycan recycling machinery, and metN2, a methionine transporter in the phospholipid synthesis pathway.
“Mapping of murQ truncations showed that they were independently acquired multiples times across the phylogeny, suggesting convergent evolution,” Professor Howden said.
“We found two significant associations linked to virulence factor expression and resistance to antimicrobial peptides, but only in cohort C could we detect a genome-wide significant association with rare mutations in four putative metabolic genes (SAUSA300_1088, hxlA, glcU, ctaB).”
This suggested that single locus effects did not have a strong impact on mortality or that this impact could not be detected with the size of the datasets – the attempt of applying bacterial GWASs to infection outcomes (such as mortality) was less successful than GWASs on antibiotic resistance mechanisms.
“We found that the bacterial genotype has limited impact on these outcomes, in particular mortality, as demonstrated by low heritability, poor predictive performance in our classification model of SAB, and a limited number of significant findings in the bacterial GWAS,” Dr Giulieri said.
“By revealing the genes responsible for antibiotic resistance in Golden staph, our GWAS is pointing the scientific community to clearer targets for the development of effective solutions to treat Golden staph bloodstream infections.
“This knowledge has the potential to shape and enhance our ability to tackle these persistent infections. As bacterial genomes become increasingly available in the clinical routine, we inch closer to customised therapeutic strategies, where treatments will be tailored to the unique genetic makeup of the infecting strain, rather than treating everyone in the same way.”