Comparative distribution of extended-spectrum beta-lactamase-producing Escherichia coli from urine infections and environmental waters

Abstract

Extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli have been reported in natural environments, and may be released through wastewater. In this study, the genetic relationship between ESBL-producing E. coli collected from patient urine samples (n = 45, both hospitalized patients and out-patients) and from environmental water (n = 82, from five locations), during the same time period, was investigated. Three independent water samples were collected from the municipal wastewater treatment plant, both incoming water and treated effluent water; the receiving river and lake; and a bird sanctuary near the lake, on two different occasions. The water was filtered and cultured on selective chromID ESBL agar plates in order to detect and isolate ESBL-producing E. coli. Illumina whole genome sequencing was performed on all bacterial isolates (n = 127). Phylogenetic group B2 was more common among the clinical isolates than the environmental isolates (44.4% vs. 17.1%, p < 0.01) due to a significantly higher prevalence of sequence type (ST) 131 (33.3% vs. 13.4%, p < 0.01). ST131 was, however, one of the most prevalent STs among the environmental isolates. There was no significant difference in diversity between the clinical isolates (DI 0.872 (0.790-0.953)) and the environmental isolates (DI 0.947 (0.920-0.969)). The distribution of ESBL genes was similar: blaCTX-M-15 dominated, followed by blaCTX-M-14 and blaCTX-M-27 in both the clinical (60.0%, 8.9%, and 6.7%) and the environmental isolates (62.2%, 12.2%, and 8.5%). Core genome multi-locus sequence typing showed that five environmental isolates, from incoming wastewater, treated wastewater, Svartån river and Hjälmaren lake, were indistinguishable or closely related (≤10 allele differences) to clinical isolates. Isolates of ST131, serotype O25:H4 and fimtype H30, from the environment were as closely related to the clinical isolates as the isolates from different patients were. This study confirms that ESBL-producing E. coli are common in the aquatic environment even in low-endemic regions and suggests that wastewater discharge is an important route for the release of ESBL-producing E. coli into the aquatic environment.

Fig 1. Schematic map of the study area.

The red crosses indicate the sampling locations at the wastewater treatment plant (WWTP) and in Svartån River, Hjälmaren Lake, and Oset ponds. Six water samples were collected from each sampling site (three in June and three in October). Presented in the figure are also the number of isolates included from each site and the sequence types (ST) found in the respective locations., STs in bold were also found among the clinical isolates.

Fig 2. Neighbor–joining (NJ) tree based on core genome multi–locus sequence typing (cgMLST) analysis.

The NJ tree is based on allele differences in 1957 core genes found in 126 ESBL–producing E. coli isolates included in the cgMLST analysis. Isolate ID, ST, and ESBL type are written in red for the clinical isolates (n = 45), and in black for the environmental isolates (n = 81). The colored areas represent the different phylogenetic groups.

Fig 3. Minimum spanning tree (MST) of the ST131 isolates.

The MST shows all ST131 isolates found in patients and the environment (n = 26) and is based on allele differences in 2682 core genes. Each circle represents one isolate. Isolates marked with * carried bla_CTX–M–15 and isolates marked with # carried bla_{CTX–M–27.} The different colors show the sampling locations and the numbers next to the lines connecting the circles show the number of allele differences between the isolates.

Fig 4. The monthly variation of ST131 and ST38.

(A) The number of clinical ESBL–producing E. coli belonging to ST38, ST131, and other STs per month, 2013. (B) The number of ESBL–producing E. coli from water environments belonging to ST38, ST131, and other STs in June and October 2013.