Typhoid fever remains one of the world’s most consequential waterborne diseases. Salmonella enterica serovar Typhi, a Gram-negative bacterium restricted to humans, causes a systemic febrile illness accompanied by diarrhoea and abdominal cramps, and without prompt antibiotic treatment it can lead to severe complications and death. Global estimates attribute more than ten million illnesses and roughly one hundred thousand deaths each year to the disease, with the overwhelming burden falling on low- and middle-income countries where sanitation infrastructure is inadequate and contaminated water sustains transmission. South Africa is considered a low-endemicity setting, yet outbreaks continue to occur, particularly among vulnerable communities in informal settlements and densely populated areas with limited access to safe water. One of the country’s largest documented outbreaks, in Delmas in 1993, produced more than a thousand cases and was traced epidemiologically to the municipal water supply, although the bacterium was never actually isolated from the water itself.

That persistent gap between epidemiological suspicion and environmental proof is precisely what the new study set out to address. South Africa’s national surveillance system, run through the Group for Enteric, Respiratory and Meningeal Disease Surveillance in South Africa at the National Institute for Communicable Diseases, relies heavily on clinical case detection supported by whole-genome sequencing. Five confirmed typhoid outbreaks have been detected in the past five years, including one between January and October 2025 that produced 147 laboratory-confirmed cases across eight provinces. Yet no definitive source has ever been identified for any of these outbreak-associated cases. Because S. Typhi is shed in the faeces of infected individuals, wastewater networks and urban rivers can act as sensitive indicators of community-level transmission, an approach already demonstrated in Blantyre, Malawi, and Vellore, India, where environmental monitoring detected pathogen circulation independently of fluctuations in clinical reporting. Comparable data from South Africa had been essentially absent.

The isolates at the centre of the study came from the Jukskei River, a major waterway running through Johannesburg in Gauteng Province. As part of a broader twelve-month surveillance study, researchers sampled water monthly at five sites along the river during 2023, collecting sixty samples in total. Four S. Typhi isolates were recovered, a recovery rate of 6.7 percent, and all four came from March 2023 at two sites roughly 3.35 kilometres apart: one within a business district and the other inside a recreational park. The DNA was extracted, sequenced on an Illumina platform at approximately one-hundred-fold coverage, and analysed through a bioinformatics pipeline incorporating quality control, species confirmation, assembly, annotation, serovar prediction, multilocus sequence typing, and screening for antimicrobial resistance determinants, plasmids, virulence factors, and Salmonella pathogenicity islands.

The genomic results revealed a picture of unexpected diversity. Genome sizes ranged from 4.67 to 4.73 megabase pairs, all with the characteristic 52.2 percent GC content of Salmonella enterica. Multilocus sequence typing assigned two isolates to sequence type 1 and two to sequence type 2, both of which are among the most commonly reported S. Typhi types worldwide. Critically, the two ST1 isolates belonged to genotype 4.3.1, better known as the H58 lineage, with one further resolved as the East African I sublineage. H58 emerged in Asia before spreading through repeated intercontinental introductions into Africa, where sustained intra-continental transmission established it across countries including Tanzania, Kenya, Malawi, Zambia, Zimbabwe, and South Africa. It is the lineage most strongly associated with multidrug resistance, and it has driven three major enteric fever outbreaks in South Africa between 2020 and 2021. Until now, however, no whole-genome-confirmed H58 isolate had ever been reported from an environmental source such as a river anywhere in the country.

The resistance profiles of the two H58 isolates were alarming. Both carried a suite of genes conferring resistance to aminoglycosides, beta-lactams, phenicols, sulfonamides, and trimethoprim: aph(6)-Id, aph(3″)-Ib, blaTEM-1, catA1, sul1, sul2, and dfrA7. One isolate additionally carried a point mutation in the gyrA gene, S83F, within the quinolone resistance-determining region, a change classically associated with reduced fluoroquinolone susceptibility. Mobile genetic element analysis identified an IncQ1 plasmid replicon co-located on the same contig as several resistance genes, alongside the insertion sequence IS26, which is widely recognised as a key driver of genomic plasticity and has been implicated in the majority of antimicrobial resistance gene transfer events among plasmids. IncQ1 plasmids are highly mobilisable with a broad host range, and surface waters are considered hotspots for plasmid-mediated horizontal gene transfer, raising the possibility that the river itself could serve as an arena for resistance dissemination, although the authors caution that complete plasmid reconstruction and conjugation experiments were not performed.

Perhaps the most novel molecular finding was the detection of qacEΔ1, a gene associated with reduced susceptibility to quaternary ammonium compounds, a class of disinfectants widely used in food processing facilities. The gene had previously been reported in clinical multidrug-resistant S. Typhi isolates from paediatric patients, but never before in S. Typhi recovered from an urban river. In both H58 isolates, qacEΔ1 was co-localised on the same contig with sul1 and dfrA7, a pattern consistent with the well-documented tendency of disinfectant resistance and antibiotic resistance determinants to travel together on mobile elements such as integrons. No plasmids or insertion sequences were identified on those particular contigs, however, and phenotypic susceptibility to quaternary ammonium compounds was not tested, so the practical implications of the gene’s presence remain an open question for future investigation.

The virulence repertoire of the isolates was equally complete. All four carried genes for adhesion, iron acquisition, and toxin production, including sinH, iroB, iroC, and cdtB, along with the full complement of Salmonella pathogenicity islands from SPI-1 through SPI-10 in the ST1 isolates, while the ST2 isolates lacked only SPI-4. Mercury resistance genes were consistently present in all four genomes. Pairwise SNP comparisons showed the four river isolates were genetically distinct from one another, differing by between 50 and 382 SNPs, which suggests they originated from separate transmission sources rather than a single contamination event, despite being recovered in the same month from sites less than three and a half kilometres apart.

The epidemiological significance emerged when the team expanded the analysis to 128 S. Typhi genomes from South Africa collected in 2023, combining the four river isolates with 124 clinical isolates from the EnteroBase database and applying a core genome MLST hierarchical clustering scheme. The analysis identified twelve clusters of at least three isolates differing by five or fewer alleles, and four of those clusters each contained one of the Jukskei River isolates alongside two or more clinical isolates. The largest cluster comprised four clinical isolates and one environmental isolate differing by no more than two alleles. The clinical isolates in these clusters were collected between February and April 2023, overlapping precisely with the March recovery of the river isolates. Because national surveillance recorded no outbreaks during that period, with 2023 case numbers at baseline levels, the clustering points to small, localised transmission events rather than widespread dissemination, likely sustained by chronic sewage contamination of the river and exposure among communities along its banks.

Genomic relatedness alone cannot resolve the direction of transmission, whether bacteria moved from contaminated water into people or from infected individuals into the river, and the authors are careful not to overstate the claim. But the study delivers the first genomic evidence that a polluted South African urban waterway harbours the same high-risk, multidrug-resistant H58 genotypes circulating in human populations, and it argues persuasively for extending the country’s already advanced genomic surveillance framework to include routine environmental monitoring of high-risk rivers. Within a One Health approach, such integrated monitoring could reveal contamination sources earlier, sharpen outbreak investigations that have historically ended without a confirmed source, and support more timely public health interventions before resistant typhoid strains find new pathways from water to community.

Subject of Research: Genomic surveillance of multidrug-resistant H58 Salmonella Typhi in an urban South African river and its phylogenetic links to clinical typhoid cases

Article Title: Genomic Detection of Multidrug‐Resistant H58 ST1 Salmonella Typhi in a South African Urban Waterway Reveals Epidemiological Links With Clinical Isolates

Article References: Duze, S. T., Marimani, M., & Patel, M. (2026). Genomic Detection of Multidrug‐Resistant H58 ST1 Salmonella Typhi in a South African Urban Waterway Reveals Epidemiological Links With Clinical Isolates. MicrobiologyOpen, 15(5), Article e70408. https://doi.org/10.1002/mbo3.70408

Image Credits: AI Generated

DOI: 10.1002/mbo3.70408

Keywords: Salmonella Typhi, typhoid fever, H58 lineage, whole-genome sequencing, antimicrobial resistance, environmental surveillance, Jukskei River, Johannesburg, wastewater, IncQ1 plasmid, qacEΔ1, One Health