Mapping the emergence of a highly invasive lineage reveals a crucial step in the evolution of Salmonella strains.
Charting the evolutionary trajectories of bacterial pathogens allows us to understand the history of acquisition of virulence traits. Scientists are increasingly implementing phylogenetic approaches to classify bacterial strains, understand the evolution of antibiotic resistance and monitor outbreaks. Writing in this issue of Nature Microbiology, Pulford et al. reconstruct the evolutionary history of Salmonella Typhimurium in Africa and uncover the recent emergence of a highly invasive lineage in Malawi1.
Salmonella species are capable of causing disease with varying degrees of severity. Most often, humans that become infected with Salmonella develop uncomplicated symptoms associated with food poisoning. However, Salmonella can also produce severe bloodstream infections that require antibiotic treatment2. Salmonella-associated bloodstream infections are particularly problematic in Africa, where case fatality rates range from approximately 25% in children to 50% in adults, resulting in almost 50,000 deaths each year3.
Interestingly, the vast majority of severe Salmonella infections are caused by a single sequence type known as ST313 (ref. 4). Since the initial publication of a representative ST313 genome sequence in 2009 (ref. 3), researchers have accumulated evidence suggesting that ST313 has undergone adaptation to support an invasive, extra-intestinal lifestyle in humans5,6. Genome sequencing of Salmonella isolates has provided a ‘fossil record’ of sorts, revealing signatures of gene loss in ST313 genomes that indicate mutational inactivation of dispensable virulence factors.
It was previously shown that the ST313 clade was dominated by two clonal lineages (L1 and L2) that are distinctly clustered3. The two lineages were thought to have diverged after L2 developed resistance to chloramphenicol, which was the first-line treatment for septic infections in Malawi in 2001 (ref. 7). After this deviation event, L2 became the primary cause of ST313-associated bloodstream infections, continuing to acquire antibiotic resistance elements and spreading throughout the African continent, the United Kingdom and Brazil8. Today, most ST313 isolates are resistant to not only chloramphenicol but also to ampicillin, trimethoprim and sulfamethoxazole; some have acquired resistance to beta-lactam antibiotics as well9.
Despite analyses of key ST313 isolates using comparative genomic, transcriptomic and proteomic studies, a comprehensive timeline of the major genetic events that shaped ST313 evolution has been lacking. In their study, Pulford et al. tracked the stepwise evolution of ST313 for the first time, broadly defining its phylogenetic structure and key phenotypes (Fig. 1). By examining 680 isolates from Salmonella-associated bloodstream infections between 1996 and 2018 — the largest collection ever studied — the authors constructed a maximum likelihood phylogeny to reveal three distinct ST313 clusters. While two clusters corresponded to the previously defined L1 and L2 lineages, an intriguing third cluster represented a new lineage (L3). Phylogenetic prediction suggested that L3 was an intermediate of the L1 and L2 lineages, appearing to cluster with isolates from the United Kingdom and Brazil, rather than those from Africa. The authors predicted that the clonal expansion of L3 was a fairly recent evolutionary event, with its most recent common ancestor dated to about 2007 by their estimates.
With the increased phylogenetic resolution that their study provides, Pulford et al. revealed an evolutionary blueprint in ST313 that seemed to be shaped by clinical practice. For example, changes in antibiotic policy in Malawi led to the withdrawal of chloramphenicol from widespread clinical use in 2001, which was followed by a ~24% reduction in chloramphenicol-resistant isolates between 2005 and 2018. Further, while isolates present within L1 and L2 displayed varying degrees of antibiotic resistance, L3 was genotypically and phenotypically susceptible to all antibiotics tested, aligning it with ST313 isolated from the United Kingdom. With these data in hand, it is tempting to speculate that perhaps the current presence of L3 in Malawi may be linked to an international transmission event.
To explore the potential for invasiveness in ST313 L3, Pulford et al. applied a machine-learning algorithm10 initially developed by Wheeler et al. This approach invokes a random forest classifier to identify mutational signatures of adaptation to an invasive pathogenic lifestyle — one that involves exit from the intestinal tract into extra-intestinal, systemic sites. Herein lies the most foreboding finding from the study by Pulford et al.: benchmarked against 196 extra-intestinal predictor genes (those associated with immunogenicity, metabolic functioning, motility and other phenotypes), L3 has a significantly higher potential for invasiveness relative to all other ST313 lineages from the United Kingdom and Malawi. Genetic drivers of this invasive potential seem to stem from further degradation of genes involved in stress resistance, biofilm formation and gut-specific metabolism in L3 compared to other lineages. Consistent with similar findings for other human-adapted Salmonella serovars (that is, Enteritidis, Typhi and Paratyphi), L3 appears to have moved yet another step away from the types of Salmonella that cause gastrointestinal illness, undergoing genetic changes that optimize it for systemic virulence at the expense of a typical gut lifestyle.
The combination of genomics and epidemiology presented by Pulford et al. to reconstruct the evolutionary history of ST313 reveals how this clade is expanding. Connecting this evolutionary trajectory to clinical practice regarding antibiotic use also serves to reinforce the importance of antibiotic stewardship. However, the story of ST313 is not over and the current work begs further investigation. While in silico estimates of invasiveness are one way to connect sequencing data with infection-relevant phenotypes, verifying these predictions in vivo will be an important next step to understand the wider public health implications of L3-associated infections. The application of cellular and animal models of Salmonella infection will be a useful first step, but new models may need to be developed to fully understand the idiosyncrasies that set invasive Salmonella infections apart from the rest. Previous studies have failed to differentiate between ST313 lineages in existing infection models for Salmonella, suggesting the inability of these mouse models to capture human-specific virulence phenotypes. More investigation may also be warranted to better understand the role of antibiotic resistance in accelerating the diversification of not only ST313, but also other pathogens. In this case, antibiotic usage in Malawi between 2002 and 2015 — perhaps the phased removal of chloramphenicol from clinical practice — likely created a window of opportunity for L3 to emerge.
As work continues apace to shore up eroding antibiotic pipelines, further investigation into how antibiotic administration shapes pathogen evolution is also urgently needed.
Pulford, C. V. et al. Nat. Microbiol. https://doi.org/10.1038/s41564-020-00836-1 (2020).
Stanaway, J. D. et al. Lancet Infect. Dis. 19, 1312–1314 (2019).
Kingsley, R. A. et al. Genome Res. 19, 2279–2287 (2009).
Feasey, N. A., Dougan, G., Kingsley, R. A., Heyderman, R. S. & Gordon, M. A. Lancet 379, 2489–2499 (2012).
Canals, R. et al. PLoS Biol. 17, e3000059 (2019).
Canals, R. et al. PLoS Pathog. 15, e1007948 (2019).
Okoro, C. K. et al. Nat. Genet. 44, 1215–1221 (2012).
Ashton, P. M. et al. Genome Med. 9, 92 (2017).
Van Puyvelde, S. et al. Nat. Commun. 10, 4280 (2019).
Wheeler, N. E., Gardner, P. P. & Barquist, L. PLoS Genet. 14, e1007333 (2018).
The authors declare no competing interests.
About this article
Cite this article
Tsai, C.N., Coombes, B.K. Emergence of invasive Salmonella in Africa. Nat Microbiol 6, 273–274 (2021). https://doi.org/10.1038/s41564-021-00864-5