Risk assessment of Enterococcus faecium, Klebsiella pneumoniae, and Pseudomonas aeruginosa in environmental water sources: Development of surrogate models for antibiotic resistance genes

The presence of Enterococcus faecium (E. faecium), Klebsiella pneumoniae (K. pneumoniae), Pseudomonas aeruginosa (P. aeruginosa), and the aminoglycoside resistance genes, aac(6′)-Ib and aac(6′)-aph(2″), was investigated in environmental water sources obtained from informal settlements in the Western Cape (South Africa). Using ethidium monoazide bromide quantitative polymerase chain reaction (EMA-qPCR) analysis, E. faecium, K. pneumoniae, and P. aeruginosa were detected in 88.9 %, 100 %, and 93.3 % of the samples (n = 45), respectively, with a significantly higher mean concentration recorded for K. pneumoniae (7.83 × 10⁴ cells/100 mL) over the sampling period. The aac(6′)-Ib gene was detected in 95.6 % (43/45) of the environmental water samples [mean concentration of 7.07 × 10⁶ gene copies (GC)/100 mL], while the aac(6′)-aph(2″) gene was detected in 100 % (n = 45) of the samples [mean concentration of 6.68 × 10⁵ GC/100 mL]. Quantitative microbial risk assessment (QMRA) subsequently indicated that the risks posed by K. pneumoniae and P. aeruginosa were linked to intentional drinking, washing/bathing, cleaning of the home, and swimming, in the samples collected from the various sampling sites. Surrogate risk assessment models were then designed and applied for Gram-positive [aac(6′)-aph(2″) gene] and Gram-negative [aac(6′)-Ib gene] pathogens that may exhibit aminoglycoside resistance. The results indicated that only the Gram-negative pathogens posed a risk (>10⁻⁴) in all the samples for cleaning of the home and intentional drinking, as well as for washing laundry by hand, garden hosing, garden work, washing/bathing, accidental consumption, and swimming at the stream and marsh sites. Thus, while environmental waters may pose a health risk of exposure to pathogenic bacteria, the results obtained indicate that screening for antibiotic resistant genes, associated with multiple genera/species, could serve as a surrogate model for estimating risks with the target group under investigation.