Urine samples were collected between February 2016 and June 2016 from a total of 96 patients presenting to their general practitioner with symptoms of UTI. The practice for general medicine is situated in a suburban community in the region of Zürich, Switzerland and has a catchment area of 10,000 people of all age groups, levels of education and professions. The patient collective is therefore representative for the average Swiss primary care patient. Informed consent was obtained from the participating patients and the study was approved by the local ethics committee of Zürich (BASEC-Nr.Req-2016-00374). Medical records were reviewed to obtain demographic and clinical data. Comorbidities were defined as one or more coexisting medical conditions that were additional to the diagnosis of UTI. Dipstick urinalysis was performed using Combur® urine test sticks (Roche, Rotkreuz, Switzerland). Specimens giving an elevated white cell count were collected using sterile screw-cap collection tubes containing boric acid, sodium formate and sodium borate as a preservative (Becton Dickinson, Allschwil, Switzerland). Thereafter, samples were diluted 1:1,000 in sterile 0.9% NaCl and 100 μl were streaked on UTI Brilliance agar (Oxoid, Pratteln, Switzerland) and on MacConkey agar (Becton Dickinson, Allschwil, Switzerland) used as growth controls. From plates corresponding to samples with a viable cell count of ≥10⁴ colony forming units (cfu)/mL, single colonies with morphological Gram-negative characteristics were subcultured on MacConkey agar following standard laboratory procedures guidelines. Isolates were subjected to species identification using API ID 32 E (bioMérieux). A total of 44 non-duplicate clinical isolates of E. coli from 44 patients with UTI were obtained for the study.

DNA from E. coli isolates were subjected to triplex PCR targeting the chuA gene, the yjaA gene and an unspecified DNA fragment termed TspE4.C2, as described previously (Clermont et al., 2000). Isolates were classified as belonging to one of the four phylogenetic groups A, B1, B2, or D.

For multilocus sequence typing of E. coli isolates, internal fragments of the seven housekeeping genes (adk, fumC, gyrB, icd, mdh, purA, and recA) were amplified by PCR from DNA, as described by Wirth et al. (2006). Sequencing of the amplification products was performed by Microsynth (Balgach). Sequences were imported into the E. coli MLST database website (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli) to determine MLST types. Alleles and STs that had not been previously described were designated new ST, but not assigned numerical designations, since whole-genome sequencing was not performed.

All ST131 isolates were further analyzed using CH typing based on fumC and fimH allele combination (Weissman et al., 2012). Isolates belonging to the ST131-CTX-M-15-H30-Rx sublineage were identified by detection of the specific amino acid alteration Gly723 → Ala within the allantoin transporter-encoding gene, ybbW, using PCR primers and conditions described previously (Banerjee et al., 2013b). Isolates typed ST10, ST69, ST73, ST140 (CC95), ST127, and ST131 were regarded as belonging to major UPEC clones.

All 44 isolates were screened for genetic markers of virulence associated with ExPEC by conventional PCR using primers and conditions described previously for afa, papAH, papC, papEF, sfaS, fyuA, hlyA, iutA, KpsMII, PAI, and traT (Johnson and Stell, 2000; Johnson et al., 2015), vat and yfcV (Spurbeck et al., 2012). The aggregate VF score was defined as the number of unique VF detected for each isolate, counting the PAI marker as one.

Strains that were negative for KpsMII were tested by PCR for the presence of the group 2 capsule subgroup K15 as described by Johnson et al. (2015).

Isolates that were positive for ≥2 of the five markers afa, papAH, and/or papC, sfa, KpsM II, or iutA were presumed to be ExPEC, and isolates that tested positive for ≥3 of the four genes chuA, fyuA, vat, or yfcV were considered UPEC, as described by Johnson et al. (2015).

Strains that tested negative for all VFs were screened by PCR for the presence of aggR, which encodes is a transcriptional regulator of enteroaggregative E. coli (EAEC) (Boisen et al., 2012).

Antimicrobial susceptibility testing was performed using the disk-diffusion method and the antibiotics ampicillin (AM), amoxicillin-clavulanic acid (AMC), cefotaxime (CTX), nalidixic acid (NA), ciprofloxacin (CIP), gentamicin (GM), kanamycin (K), streptomycin (S), sulfamethoxazole (SMZ), trimethoprim (TMP) tetracycline (T), and chloramphenicol (C) (Becton Dickinson, Heidelberg, Germany). Results were interpreted according to Clinical and Laboratory Standards Institute (CLSI) performance standards (Clinical Laboratory Standards Institute, 2016). For sulfamethoxazole, for which breakpoints are not listed separately from trimethoprim, an inhibition zone of ≤10 mm was interpreted as resistant. Isolates displaying resistance to three or more classes of antimicrobials (counting ß-lactams as one class) were defined as multidrug-resistant (MDR). Synergistic effects between AMC and CTX were regarded as an indication of the presence of an ESBL producer (Kaur et al., 2013). Putative ESBL producers were grown on Brilliance™ ESBL agar (Oxoid, Hampshire, UK). The presence of bla_ESBLs was confirmed by PCR and amplicons were sequenced as described previously (Pitout et al., 1998; Woodford et al., 2006; Geser et al., 2012; Zurfluh et al., 2015). Quinolone-resistant strains were examined for mutations in quinolone resistance-determining regions (QRDRs) of gyrA and parC, using previously described PCR and sequencing primers (Zurfluh et al., 2014). Screening for the plasmid-mediated fluoroquinolone resistance genes aac(6′)-Ib-cr, qnrA, qnrB, qnrC, qnrD, qnrS, and qepA genes was carried out as described previously (Park et al., 2006; Robicsek et al., 2006; Cattoir et al., 2007; Cavaco et al., 2009; Kim et al., 2009; Wang et al., 2009; Karczmarczyk et al., 2010; Zurfluh et al., 2014). Screening for the plasmid-mediated colistin resistance genes mcr-1, mcr-2 and the plasmid-mediated azithromycin resistance gene mph(A) was performed using previously published primers (Ojo et al., 2004; Liassine et al., 2016; Liu et al., 2016) using appropriate positive controls (Nüesch-Inderbinen et al., 2016; Zurfluh et al., 2017).

Primers used in this study for targeting virulence and antimicrobial resistance genes are listed in Supplementary Table S1. Synthesis of primers and DNA custom sequencing was carried out by Microsynth (Balgach, Switzerland) and nucleotide sequences were analyzed with CLC Main Workbench 6.6.1. For database searches the BLASTN program of NCBI (http://www.ncbi.nlm.nih.gov/blast/) was used.

Comparisons of proportions of E. coli STs, proportions of virulence genes and ExPEC/UPEC status within patient comorbidity status were performed by Fisher's exact test in a series of individual pairwise comparisons using 2 × 2 tables where each characteristic was determined as present or absent. The significance criterion was set at p ≤ 0.05. Calculations were performed using the VassarStats website for statistical computation (http://www.vassarstats.net).

The 44 strains analyzed in this study were isolated from urine samples of 44 patients with a female/male ratio of 39/5, and a median age of 53. Ages ranged from 16 to 87 and were allocated to either group 1, 16–45 years (n = 18), or group 2, 46–87 years (n = 26), respectively (Table 1). A history of hospitalization during the 4 months before participation in the study was noted for 2 (4.5%) of the patients (Supplementary Table S2). None of the patients had a history of antimicrobial therapy in the 4 months before sample collection. Comorbidities were recorded for 17 (38.6%) of the patients (five in age group 1 and 12 in age group 2) and are summarized in Table 1 and detailed in Supplementary Table S2. Treatment following clinical diagnosis of UTI included fosfomycin (19/43.2% of the patients), norfloxacin (12/27.3%), ciprofloxacin (4/9%), or sulfamethoxazole/trimethoprim (3/6.8%). Six (13.6%) of the patients received no treatment (Supplementary Table S2).

The majority (52.3%) of the 44 UPEC isolates belonged to phylogenetic group B2, followed by group D (22.7%), group A (13.6%) and B1 (11.4%). Twenty-seven different ST were identified, the four most common represented by ST131 (n = 6/13.6% of the isolates), ST69 (n = 6/13.6%), ST141 (n = 5/11.4%) and ST73 (n = 3/6.8%). Overall, 19 (43.2%) showed STs associated with major UPEC clonal groups (ST10, ST69, ST73, ST140, ST127, or ST131). Two (4.5%) belonged to ST12, and ST14, which are found occasionally among UPEC (Banerjee et al., 2013a; Nichols et al., 2016). Twenty (45.5%) of the isolates belonged to STs that occurred only once, whereof three were new STs. The new STs were not assigned numerical designations by the E. coli MLST database (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli), however, the allelic profiles were related to those of ST23 (n = 2) and ST165 (n = 1) (Supplementary Table S3). Subtyping of the ST131 isolates identified five isolates with the fumC/fimH type 40-30, and one isolate with the fumC/fimH type 40–41 (Supplementary Table S2). Accordingly, the isolates were assigned to the subclones ST131-H30 and ST131-H41, respectively. All five ST131-H30 isolates were fluoroquinolone resistant and thus considered to belong to the subclone H30-R as determined by Price et al. (2013). Furthermore, both ST131-CTX-M-15 isolates were identified as belonging to the H30-Rx sublineage (Price et al., 2013).

The distribution of STs among the age groups, genders and comorbidities of the patients is shown in Table 1. ST distribution was invariable for age groups and gender, but varied by health status. The majority (n = 14/73.7%) of the 19 isolates that belonged to major UPEC lineages were isolated from patients without comorbidities, without reaching statistical significance (p = 0.2133). By contrast, the majority (n = 11/64.7%) of the 20 isolates that belonged to singly occurring STs were from patients for whom comorbidities were noted, approximating, but not reaching statistical significance (p = 0.0633).

Overall, 90.9% of the isolates tested positive for one or more of the VF genes (Table 2). Among the 44 isolates, the prevalence of individual VF genes ranged from 13.6% (afa) to 70.5% (fyuA). None of the KpsMII negative isolates were positive for K15 (data not shown). The overall median aggregate VF scores (and ranges) were lowest for isolates belonging to ST10 (median VF score 3.5, range 3–4) and highest for ST127 (median VF score 9, range 9) and ST141 (median VF score 11, range 5–12). Among the different STs, PAI, yfcv and hlyA were distinguished by their absence from ST69 (Table 2).

Applying the criteria for ExPEC and UPEC described above, 27 (61.4%) of the isolates qualified as ExPEC, and 22 (50%) as UPEC (Supplementary Table 2).

No statistically significant associations were observed for STs, VFs, or ExPEC/UPEC status with comorbidities (data not shown).

Notably, four (9%) of the isolates (belonging to a new ST, to ST40, ST841, and ST1001, respectively), tested negative for all ExPEC VF (Supplementary Table S2). Screening of these isolates for aggR remained negative throughout (data not shown).

Antimicrobial susceptibility rates for the E. coli isolates are summarized in Table 2 and shown in detail in Supplementary Table S2. Overall, 22 (50%) of the isolates remained susceptible to all antimicrobials tested. None of the isolates were resistant to fosfomycin or to nitrofurantoin. Isolates belonging to ST131 were the most extensively resistant, exhibiting a high prevalence of 83.3% of MDR, ciprofloxacin resistance, and ampicillin resistance. ST69 was characterized by a prevalence of 66.7% of MDR, trimethoprim and ampicillin resistance as well as a high rate of 83.3% resistance to sulfamethoxazole (Table 2). Overall, 10 (22.7%) of the isolates were resistant to nalidixic acid, and seven thereof were ciprofloxacin resistant (Table 2). Among the isolates displaying ciprofloxacin resistance, all revealed chromosomal mutations leading to amino acid substitutions in GyrA and ParC (Table 3). Only one of the ciprofloxacin resistant ST131 strains harbored the aac(6′)-Ib-cr gene. Further, one (16.7% of all E. coli ST69) and three (50%) of the ST131 isolates were ESBL producers and resistant to cefotaxime. These ESBLs were CTX-M-14 (n = 1) and CTX-M-15 (n = 3) (Table 2). Of note was the association of the plasmid-mediated azithromycin resistance gene mph(A) with ST69 and ST131.