Antimicrobial resistance profile and molecular analysis of virulence among Escherichia coli isolated from pregnant women with asymptomatic bacteriuria in Gorgan, Iran

Abstract

Urinary tract infections (UTIs) are common among pregnant women and are frequently caused by uropathogenic Escherichia coli (UPEC). This cross-sectional study investigated the distribution of key virulence genes, antimicrobial resistance patterns, and their relationship with phylogenetic groups in UPEC isolates obtained from pregnant women with asymptomatic bacteriuria in Gorgan, Iran. Out of 913 urine samples collected between 2018 and 2023, 360 non-duplicate UPEC isolates were confirmed using standard biochemical methods. Nineteen virulence genes, including neuC, fimH, papC, sfaS, sfa/focDE, focG, kpsMTII, ecpA, iutA, ibeA, rfc, fyuA, traT, iroN, papGII, hlyA, cnf1, cdtB, and cvaC, were detected by PCR, while phylogenetic grouping was performed through triplex PCR targeting chuA, yjaA, and tspE4.C2. Antimicrobial susceptibility testing showed that imipenem (96.3%) and nitrofurantoin (93.0%). The capsule gene kpsMTII was the most prevalent, and fimH was the most frequent adhesin gene (89.8%). Phylogenetic group B2 exhibited the highest diversity and frequency of virulence genes. Multidrug resistance (MDR) was observed in 51.4% of isolates, and although 58.3% of B2 strains were MDR, the association between phylogeny and MDR was not statistically significant (p > 0.05). Strong biofilm formation correlated significantly with resistance to ciprofloxacin and ampicillin (p < 0.01). This study documents a high prevalence of antimicrobial resistance and ESBL production among UPEC isolates from pregnant women with asymptomatic bacteriuria. The coexistence of multiple virulence traits with MDR and strong biofilm formation suggests potential biological interactions that warrant further investigation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-026-08765-3.

Introduction

Urinary tract infections (UTIs) are among the most common bacterial infections in women and occur more frequently during pregnancy due to physiological and hormonal changes that promote urinary stasis and vesicoureteral reflux [1, 2]. A notable proportion of cases, approximately 2–10% of pregnancies, involve asymptomatic bacteriuria (ASB), which, if left untreated, can lead to pyelonephritis, preterm labor, low birth weight, and other adverse outcomes [3, 4]. Despite the clinical relevance of ASB, data on the molecular characteristics of uropathogenic Escherichia coli (UPEC) in pregnant women remain limited in northern Iran. This study addresses this knowledge gap by characterizing virulence profiles and phylogroups of UPEC isolates from pregnant women with ASB in Gorgan.

The UPEC is the primary etiological agent of ASB and symptomatic UTIs [5]. These strains possess a wide array of virulence factors that enable colonization, immune evasion, and persistence within the urinary tract [6]. The key virulence determinants include adhesins, toxins, and iron-acquisition systems. For example, fimH is involved in adhesion to urothelial cells, neuC contributes to sialic acid capsule synthesis and immune evasion, and papC mediates adhesion to renal cells. The inclusion of neuC in this panel was motivated by its potential role in capsule-mediated persistence during pregnancy-related UTI. The hlyA gene encodes a hemolysin toxin that disrupts host cell membranes [7]. Other genes, such as fyuA and iroN, are crucial for iron acquisition, allowing UPEC to thrive in nutrient-limited environments like the urinary tract. In addition, genes such as kpsMTII and cnf1 contribute to immune evasion and tissue damage, respectively. The ecpA gene is associated with biofilm formation, while sfaS and ibeA play roles in adhesion to epithelial cells and invasion of host tissues. Finally, traT helps UPEC evade the host immune system, further promoting persistent infections and resistance to antimicrobial treatment [8, 9].

The treatment of UTIs caused by UPEC is increasingly complicated by the rise of MDR strains. MDR UPEC isolates frequently harbor extended-spectrum β-lactamase (ESBL) genes such as bla_CTX−M and bla_TEM, rendering many first-line antibiotics ineffective [10]. Resistance to fluoroquinolones, trimethoprim-sulfamethoxazole, and β-lactams is particularly concerning. Resistance rates range from 50% to over 75% in several countries across the Middle East and South Asia, including Iran. Recent studies have reported ciprofloxacin and trimethoprim-sulfamethoxazole resistance exceeding 65%, and β-lactam resistance reaching up to 77.5% in UPEC isolates [11]. A multicenter study in Iran reported ESBL production in over 34.6% of UPEC isolates from community-acquired infections [12], while WHO regional reports have highlighted similar trends in antimicrobial resistance in Southeast Asia and the Eastern Mediterranean [13]. Additionally, the widespread use of antibiotics has led to the emergence of carbapenem-resistant UPEC, further limiting therapeutic options and highlighting the need for novel treatments [14].

The interplay between virulence factors and antibiotic resistance presents a dual challenge in managing UPEC infections. Many resistance genes are co-located with virulence determinants on mobile genetic elements such as plasmids and pathogenicity islands, facilitating their horizontal transfer [15]. For instance, the co-occurrence of resistance genes like blaCTX-M with adhesion-related virulence genes such as fimH has been reported in MDR UPEC isolates. This genomic association enhances both antimicrobial resistance and colonization capacity [16]. It not only increases the pathogenic potential of UPEC strains but also complicates eradication efforts by limiting antibiotic efficacy and promoting persistence. The ability to form biofilms further strengthens this survival strategy, providing a protective niche against antibiotics and host immune responses [12].

Biofilms serve as protective barriers that enhance bacterial survival under hostile conditions, including antibiotic exposure and host immune responses [15, 17]. The ability of UPEC to form biofilms is often associated with the expression of adhesins, such as fimH, and other surface molecules that mediate attachment to epithelial and abiotic surfaces [18]. This biofilm-forming capacity not only contributes to the chronicity of infections but also facilitates the development and dissemination of antibiotic resistance, complicating treatment outcomes. Recent studies have demonstrated the role of biofilm formation in the persistence and antibiotic resistance of UPEC strains. In this context, biofilm not only enhances bacterial resilience by reducing susceptibility to antibiotics and immune clearance, but also serves as a compensatory mechanism in less resistant strains. In clinical isolates, strong biofilm formation is often coupled with high antimicrobial resistance, highlighting its central contribution to both survival and pathogenicity [15, 19, 20].

Addressing the global burden of UPEC-mediated UTIs requires a multifaceted approach, including improved surveillance, the development of targeted therapies, and enhanced understanding of the genetic mechanisms underlying resistance and virulence. Comprehensive analyses of virulence factors and resistance profiles can inform effective strategies for prevention and treatment [21, 22]. It should be noted that the triplex Clermont scheme used in this study does not differentiate the expanded phylogroups (C, E, F) introduced in the updated 2013–2015 revisions, and these groups may therefore be incorporated into the traditional A or D categories.

This study aims to elucidate the prevalence and distribution of key virulence factors - including fimH, papC, sfa/focDE, neuC, iroN, hlyA, cnf1, sfaS, focG, kpsMTII, ecpA, iutA, ibeA, rfc, fyuA, traT, papGII, cdtB, and cvaC - among UPEC isolates from pregnant women with asymptomatic bacteriuria. Using PCR for virulence gene detection and phenotypic antimicrobial susceptibility testing in accordance with CLSI guidelines, we also explore how these virulence determinants and resistance profiles correlate with phylogenetic groups, aiming to better understand UPEC pathogenicity and inform more effective management strategies.

Materials and methods

Study design and population

This cross-sectional study was conducted on pregnant women aged 18 to 43 years who were suspected of having ASB. The study period was from January 2018 to February 2023. Sample size calculation was not performed prospectively. All eligible pregnant women with suspected ASB who attended the participating laboratories during the study period were consecutively approached for inclusion, yielding a final sample of 913 participants. This study was framed as an exploratory cross-sectional investigation, and the achieved sample size reflected routine clinical case load rather than a priori power calculations. Samples were collected from the central and private diagnostic laboratories located in Gorgan, Iran. ASB suspicion was based on routine prenatal screening, and participants were recruited upon detection of significant bacteriuria without urinary symptoms. Inclusion criteria required participants to be within the second and third trimesters of pregnancy, as these periods are associated with significant urinary stasis due to ureteral dilation. Recruitment was performed consecutively throughout the study period, ensuring that all eligible women were included without selective sampling. The diagnosis of ASB was based on the presence of ≥ 10⁵ CFU/mL of a single uropathogen in a midstream urine specimen, in accordance with standard ASB diagnostic criteria in pregnant women. Women with a history of urinary tract infections were excluded to avoid confounding effects of recurrent infections on virulence gene distribution. Additional exclusion criteria included kidney diseases, urinary stones, congenital or acquired urological anomalies, pre-existing diabetes mellitus, or ongoing antibiotic therapy at the time of recruitment. These conditions were excluded due to their potential to affect bacterial colonization or host immune responses.

While all participants met the diagnostic criteria for asymptomatic bacteriuria at the time of enrollment, a small number later reported developing mild urinary symptoms - such as dysuria or flank pain - either during clinical follow-up or when asked retrospectively. Rather than excluding these cases, we retained them to allow for exploratory analysis of potential links between symptom emergence and virulence gene profiles. To evaluate whether their inclusion introduced bias, we performed a sensitivity analysis in which these post-enrollment symptomatic cases were excluded; the direction and significance of all key findings remained unchanged.

Demographic and clinical data were collected using a structured questionnaire. The questionnaire gathered detailed information about age, gravidity, prior UTI history, urinary symptoms, and criteria for exclusion. These data were carefully reviewed to ensure appropriate stratification and inclusion of participants. Informed written consent was obtained from all participants prior to sample collection. This study was approved by the Ethics Committee of Golestan University of Medical Sciences (Approval Code: IR.GOUMS.REC.1395.2).

Sample collection and isolation of bacterial strains

Midstream clean-catch urine samples were obtained under sterile conditions from each participant. Urine samples were inoculated on MacConkey agar and blood agar plates (Oxoid, UK) and incubated at 37^°C for 24 h. Colony-forming units (CFUs) exceeding 10⁵ per milliliter were classified as indicative of significant bacteriuria. This threshold was based on standard ASB diagnostic criteria in pregnant women [23]. Suspected E. coli colonies were identified through standard biochemical tests, including indole production, citrate utilization, and lactose fermentation on MacConkey agar. While phenotypic identification is widely used in routine microbiology, a small degree of misclassification with other Enterobacterales cannot be fully excluded. Confirmed isolates were subsequently stored at -70^°C in Luria-Bertani broth containing 30% glycerol for downstream analyses.

Antibiotic susceptibility testing

The antibiotic susceptibility profiles of the isolates were determined using the Kirby-Bauer disk diffusion method on Mueller-Hinton agar. The antibiotics tested, along with their respective concentrations, included ampicillin (10 µg), ciprofloxacin (5 µg), trimethoprim-sulfamethoxazole (1.25/23.75 µg), cefotaxime (30 µg), ceftriaxone (30 µg), amoxicillin-clavulanic acid (20/10 µg), nitrofurantoin (300 µg), piperacillin (100 µg), piperacillin-tazobactam (100/10 µg), gentamicin (10 µg), nalidixic acid (30 µg), aztreonam (30 µg), ceftazidime (30 µg), and imipenem (10 µg). Antibiotic discs were obtained from Liofilchem Company (Italy). The panel was selected based on clinical guidelines for UTI treatment in pregnant women and prevalent resistance patterns in UPEC. Results were interpreted following CLSI guidelines, 30th edition (2020), using E. coli ATCC 25,922 as the control strain [24]. Quality control strains were used for each batch of testing. MDR isolates were defined as those resistant to at least one agent in three or more antimicrobial categories, based on standardized criteria.

Phenotypic detection of ESBLs

ESBLs production was detected phenotypically following CLSI guidelines. Initially, Ceftazidime (30 µg) and Cefotaxime (30 µg) discs were used. An inhibition zone size of ≤ 22 mm for Ceftazidime and/or ≤ 27 mm for Cefotaxime indicated potential ESBL producers. Confirmatory testing was conducted using the combination disc test (CDT), as recommended by CLSI guidelines. Discs containing Ceftazidime (30 µg), Cefotaxime (30 µg), Ceftazidime-Clavulanic acid (30/10 µg), and Cefotaxime-Clavulanic acid (30/10 µg) were placed 15 mm apart on a Muller-Hinton agar plate inoculated with a bacterial suspension standardized to 0.5 McFarland turbidity. Plates were incubated at 37^°C for 18–24 h. An increase in the inhibition zone diameter of ≥ 5 mm with the combination disc, compared to the individual antibiotic discs, confirmed ESBL production. This approach followed the CLSI Performance Standards for Antimicrobial Susceptibility Testing, M100, 30th edition (2020) [24], together with the confirmatory protocol described by Sahni et al. [25].

Molecular detection of virulence genes

DNA extraction was performed on all E. coli isolates using the High Pure DNA Template Preparation Kit (Roche, Germany) in accordance with the manufacturer’s guidelines. PCR was employed to detect the presence of key virulence and resistance genes. A comprehensive list of the oligonucleotide primers used for amplification of virulence, adhesion genes, along with their nucleotide sequences, is provided in Table 1. The primers were adopted from previously validated studies, including Iranian investigations confirming their specificity for local UPEC isolates. Virulence genes analyzed included neuC, fimH, papC, sfaS, sfa/focDE, focG, kpsMTII, ecpA, iutA, ibeA, rfc, fyuA, traT, iroN, papGII, hlyA, cnf1, cdtB, and cvaC. PCR conditions were optimized based on previous study [26, 27]: Thermal cycling conditions consisted of an initial denaturation at 95^°C for 5 min, followed by 30 amplification cycles comprising 1 min of denaturation at 94^°C, 1 min of primer-specific annealing, and a 30-second extension at 72^°C. A final elongation step was performed at 72^°C for 5 min. Amplified PCR products were separated by electrophoresis on a 1.5% agarose gel stained with SYBR Safe DNA Gel Stain (Invitrogen), and bands were visualized under a UV trans-illumination imaging system. Negative controls (PCR-grade water) were included in each reaction to monitor contamination. Amplicon identity was confirmed by the presence of single, correctly sized bands for each gene target, verified against molecular weight markers; only isolates with reproducible bands were included in the analysis.

Table 1.

Primer sequences for UPEC virulence and phylogenetic genes

Gene	Primer sequence (5′–3′)	Product size (bp)	Gene Description	Reference
fimH	TGCAGAACGGATAAGCCGTGG GCAGTCACCTGCCCTCCGGTA	508	Involved in adhesion to urothelial cells, facilitating bacterial attachment in the urinary tract.	[28]
papC	GTGGCAGTATGAGTAATGACCGTTA ATATCCTTTCTGCAGGGATGCAATA	200	Responsible for the formation of P fimbriae, aiding adhesion to renal and urothelial cells.	[28]
sfaS	GTGGATACGACGATTACTGTG CCGCCAGCATTCCCTGTATTC	240	Contributes to adhesion to epithelial cells, enhancing the ability to colonize the urinary tract.	[29]
kpsMTII	GCGCATTTGCTGATACTGTTG CATCCAGACGATAAGCATGAGCA	272	Encodes a capsular polysaccharide important for immune evasion and protection from phagocytosis.	[29]
cnf1	AAGATGGAGTTTCCTATGCAGGAG CATTCAGAGTCCTGCCCTCATTATT	498	Produces cytotoxic necrotizing factor, a toxin contributing to tissue damage and inflammation.	[28]
hlyA	AACAAGGATAAGCACTGTTCTGGCT ACCATATAAGCGGTCATTCCCGTCA	1177	Encodes hemolysin, a toxin responsible for disrupting host cell membranes.	[28]
fyuA	TGAGTGGGAAATACACCACC TTACCCGCATTGCTTAATGTC	715	Involved in iron acquisition, enabling bacterial growth in iron-limited environments like the urinary tract.	[30]
iroN	AAGTCAAAGCAGGGGTTGCCCG GACGCCGACATTAAGACGCAG	665	Encodes a siderophore receptor for iron uptake, essential for UPEC survival in iron-deprived conditions.	[29]
neuC	AGGTGAAAAGCCTGGTAGTGTG GGTGGTACATCCCGGGATGTC	676	Associated with sialic acid capsule synthesis, aiding immune evasion and persistence in the urinary tract.	[31]
sfa/focDE	CTCCGGAGAACTGGGTGCATCTTAC CGGAGGAGTAATTACAAACCTGGCA	410	Contributes to adhesion to host tissues and biofilm formation in the urinary tract.	[28]
focG	CAGCACAGGCAGTGGATACGA GAATGTCGCCTGCCCATTGCT	360	Plays a role in adhesion and invasion of urothelial cells, contributing to persistent infections.	[28]
ecpA	ACCTCGGGAAGAAAAGCAA CAATTCCGTCCAGGAATAAA	640	Important for biofilm formation, enhancing the ability of UPEC to persist in the urinary tract.	[32]
iutA	ATGAGCATATCTCCGGACG CAGGTCGAAGAACATCTGG	587	Aids in iron uptake, promoting bacterial survival in the nutrient-limited urinary environment.	[33]
ibeA	TGAACGTTTCGGTTGTTTTG TGTTCAAATCCTGGCTGGAA	814	Involved in invasion of epithelial cells, contributing to UPEC’s ability to establish persistent infections.	[33]
rfc	ATCCATCAGGAGGGGACTGGA AACCATACCAACCAATGCGAG	788	Plays a role in the synthesis of the capsular polysaccharide, contributing to immune evasion.	[28]
traT	GGTGTGGTGCGATGAGCACAG CACGGTTCAGCCATCCCTGAG	290	Protects bacteria from serum killing by preventing immune recognition.	[28]
papGII	TGGCATCCCCCAACATTATCG GGGATGAGCGGGCCTTTGAT	190	Involved in adhesion to uroepithelial cells, contributing to UPEC’s ability to persist in the urinary tract.	[34]
cdtB	AAATCACCAAGAATCATCCAGTTA AAATCTCCTGCAATCATCCAGTTTA	430	Encodes a toxin that disrupts host cell function, contributing to tissue damage and immune evasion.	[28]
cvaC	CACACACAAACGGGAGCTGTT CTTCCCGCAGCATAGTTCCAT	680	Involved in capsule synthesis, enhancing immune evasion and bacterial persistence.	[28]
chuA	GACGAACCAACGGTCAGGAT TGCCGCCAGTACCAAAGACA	279	Plays a role in iron acquisition, important for UPEC’s survival in the iron-limited environment of the urinary tract.	[33]
yjaA	TGAAGTGTCAGGAGACGCTG ATGGAGAATGCGTTCCTCAAC	211	A phylogeny marker gene, also involved in bacterial survival and persistence.	[33]
tspE4.C2	GAGTAATGTCGGGGCATTCA CGCGCCAACAAAGTATTACG	152	Another phylogeny marker gene, contributing to the genetic characterization of UPEC strains.	[33]

Determination of E. coli phylogenetic groups

Phylogenetic grouping of the E. coli isolates was conducted using a triplex PCR assay, as described by Clermont et al. [35]. This method amplified the chuA, yjaA, and tspE4.C2 genes to classify isolates into one of four main phylogenetic groups: A, B1, B2, or D. The primer sequences used for phylogenetic grouping are summarized in Table 1. DNA extraction for this assay was performed using the same method described in Sect. 2.5. Phylogenetic classification provides insight into the evolutionary background and pathogenic potential of UPEC isolates, with group B2 typically associated with increased virulence and MDR phenotypes. Reference strains ECOR62 (B1 group) and RS218 (B2 group) were obtained from the microbiology laboratory of Golestan University of Medical Sciences, Iran. Although the triplex PCR method remains widely used, it does not distinguish the more recently defined phylogroups C, E, and F described in the updated Clermont schemes (2013–2015). These groups may therefore be incorporated into A or D in our classification.

Biofilm formation assay

Biofilm formation was quantified using a microtiter plate assay. Isolates were cultured in 96-well plates containing tryptic soy broth with 1% glucose, incubated at 37^°C for 24 h, and stained with crystal violet. The optical density (OD) at 595 nm was measured, and values were normalized by subtracting the mean OD of sterile medium controls to correct for background absorbance. OD readings were obtained using a microplate reader (BioTek ELx800, BioTek Instruments, Winooski, Vermont, USA).

The optical density cut-offs were defined as follows: isolates with OD ≤ ODc were considered non-biofilm producers; ODc < OD ≤ 2×ODc were weak producers; 2×ODc < OD ≤ 4×ODc were moderate producers; and OD > 4×ODc were classified as strong biofilm formers, where ODc represents the mean OD of the negative control plus three standard deviations (SD). All assays were performed in triplicate, and the mean OD value from three independent experiments was used for analysis [36, 37]. A known biofilm-positive UPEC strain and a non-biofilm-forming E. coli strain were included as positive and negative controls, respectively. Reproducibility was monitored across the three independent assay runs.

Statistical analysis

Data were analyzed in SPSS version 18.0 (SPSS Inc., Chicago, IL, USA). Socio-demographic characteristics in Table 2 are presented descriptively as frequencies and percentages. Categorical variables were compared using Pearson’s Chi-square test; when any expected cell count was < 5, Fisher’s exact test was applied. For virulence gene distributions across phylogroups (Table 4), Pearson’s Chi-square test was used only when all expected cell counts were ≥ 5; for genes with sparse or zero counts in one or more phylogroups, global p-values were obtained from Fisher’s exact test. Normality of continuous variables was assessed by the Shapiro-Wilk test, and parametric or non-parametric tests were selected accordingly. Statistical significance was set at p < 0.05 (two-sided). Effect sizes for categorical associations are reported as odds ratios (OR) with 95% CIs where applicable.

Table 2.

Socio-demographic and clinical characteristics of pregnant participants (N = 913)

Characteristics	Value, n (%)
Maternal age (years, n = 913)
Mean (SD)	29 (4.0)
< 30	281 (30.7)
30–34	423 (46.3)
≥ 35	209 (22.8)
Occupation (n = 913)
Employee	451 (49.3)
Housewife	314 (34.3)
Student	148 (16.2)
Education (n = 913)
Non formal education	45 (4.9)
Primary school	108 (11.8)
Secondary school	248 (27.1)
University level education	512 (56.0)
Residence (n = 913)
Urban	617 (67.5)
Rural	296 (32.4)
Gestational age at recruitment (weeks, n = 913)
14–18	83 (9.1)
19–23	187 (20.4)
24–28	552 (60.4)
29–33	91 (9.9)
Diabetes (n = 913)
Gestational diabetes mellitus	169 (18.5)
No diabetes	744 (81.4)
BMI at booking (kg/m2, n = 913)
Mean (SD)	24.1 (3.5)
Obese (≥ 30)	45 (4.9)
Overweight (25.0–29.9 )	151 (16.5)
Normal (18.5–24.9)	643 (70.4)
Underweight (< 18.5)	74 (8.1)

Abbreviations: SD Standard deviation, BMI Body mass index

Table 4.

Prevalence of virulence genes among UPEC isolates (n = 360)

Gene	Phylogenetic group, n (%)				p-value*
	A (n = 111)	B1 (n = 34)	B2 (n = 198)	D (n = 17)
neuC	45 (40.5)	10 (29.4)	129 (65.1)	8 (47)	< 0.001
fimH	83 (74.7)	23 (67.6)	178 (89.8)	15 (88.2)	< 0.001
papC	69 (62.1)	20 (58.8)	135 (68.1)	9 (52.9)	0.406
sfaS	0 (0)	0 (0)	23 (11.6)	2 (11.7)	< 0.001
sfa/focDE	41 (36.9)	9 (26.4)	156 (78.7)	11 (64.7)	< 0.001
focG	0 (0)	0 (0)	18 (9)	0 (0)	0.001
kpsMTII	111 (100)	34 (100)	198 (100)	16 (94.1)	< 0.001
ecpA	83 (74.7)	28 (82.3)	196 (98.9)	14 (82.3)	< 0.001
iutA	41 (36.9)	19 (55.8)	132 (66.6)	10 (58.8)	< 0.001
ibeA	0 (0)	0 (0)	11 (5.5)	0 (0)	0.025
rfc	0 (0)	5 (14.7)	0 (0)	0 (0)	< 0.001
fyuA	50 (45)	18 (52.9)	166 (83.8)	8 (47)	< 0.001
traT	62 (55.8)	34 (100)	132 (66.6)	15 (88.2)	< 0.001
iroN	41 (36.9)	6 (17.6)	83 (41.9)	5 (29.4)	0.049
papGII	0 (0)	0 (0)	61 (30.8)	0 (0)	< 0.001
hlyA	0 (0)	0 (0)	32 (16.1)	0 (0)	< 0.001
cnf1	0 (0)	0 (0)	31 (15.6)	0 (0)	< 0.001
cdtB	0 (0)	0 (0)	15 (7.5)	0 (0)	0.005
cvaC	0 (0)	0 (0)	33 (16.6)	0 (0)	< 0.001

* P-values reflect global comparisons across phylogroups (A, B1, B2, D). Pearson’s Chi-square test was used for genes with adequate expected counts in all cells; for genes with sparse distributions or zero counts in one or more phylogroups, p-values were obtained from Fisher’s exact test

In sections reporting overall frequencies only (e.g., Table 2; Sect. 3.2), no inferential testing was performed. Inferential analyses were restricted to explicitly defined group comparisons. To maintain statistical rigor and avoid inflating Type I error, only a clinically pre-specified primary panel of virulence-resistance associations was subjected to multiple-testing correction. With a sample size of 360 isolates, the precision of key estimates was acceptable; for example, a 50% resistance proportion yields a 95% CI width of approximately ± 5%. P-values for the predefined primary analyses were corrected using the Benjamini-Hochberg false discovery rate (FDR) procedure (FDR threshold = 0.05). Analyses involving phylogroup distributions, biofilm formation, and co-harbored virulence profiles (Sect. 3.3–3.7) were considered exploratory and their p-values are presented descriptively without FDR adjustment. The predefined primary hypothesis panel included associations between seven virulence genes (fimH, papC, kpsMTII, iutA, fyuA, hlyA, neuC) and the following resistance outcomes: MDR status, resistance to ampicillin and ciprofloxacin, and ESBL production. Only this primary panel was subjected to Benjamini-Hochberg FDR correction; all other analyses were exploratory.

Results

Out of 913 urine samples collected from pregnant women, 360 non-duplicate isolates of E. coli were confirmed by conventional biochemical tests. This accounts for an overall prevalence of 39.4% for UPEC isolates among the study population. Each isolate corresponded to one patient, with no repeat samples.

Socio-demographic characteristics

The study included 913 pregnant women with a mean age of 29 years (SD = 4.0). Almost half of the participants (46.3%) were between 30 and 34 years old. Most women resided in urban areas (67.5%), and more than half had attained university-level education (56%). Regarding employment, nearly half of the participants were employed (49.3%), while 34.3% were housewives and 16.2% were students. The majority were enrolled during the second trimester (24–28 weeks), representing 60.4% of the sample. With respect to metabolic characteristics, 18.5% had a diagnosis of gestational diabetes mellitus. The mean BMI at the first antenatal visit was 24.1 kg/m² (SD = 3.5); most women were within the normal BMI range (70.4%), followed by overweight (16.5%) and obese categories (4.9%) (Table 2).

Antimicrobial susceptibility and ESBL production

Antibiotic susceptibility profiling of UPEC isolates (n = 360) showed substantial variability in resistance rates. The highest resistance frequencies were observed for nalidixic acid (78.3%), ampicillin (74.1%), and piperacillin (71.9%), whereas resistance to imipenem (3.6%) and nitrofurantoin (6.9%) was low. For each agent, the proportion of resistant isolates and its 95% confidence interval are presented in Table 3.

Table 3.

Antibiotic resistance frequencies and 95% confidence intervals for UPEC isolates (n = 360)

Antibiotic	Resistant n/N (%)	95% CI (resistance)	Susceptible n/N (%)
Ciprofloxacin	185 (51.3)	46.24–56.51	175 (48.6)
Ampicillin	267 (74.1)	69.41–78.42	93 (25.8)
Amoxicillin/clavulanic acid	185 (51.3)	46.24–56.51	175 (48.6)
Piperacillin	259 (71.9)	67.09–76.33	101 (28)
Piperacillin/tazobactam	244 (67.7)	62.78–72.40	116 (32.2)
Ceftriaxone	206 (57.2)	52.06–62.23	154 (42.7)
Cefotaxime	201 (55.8)	50.67–60.87	159 (44.1)
Gentamicin	74 (20.5)	16.70–25.03	286 (79.4)
Nalidixic acid	282 (78.3)	73.79–82.28	78 (21.6)
Trimethoprim/sulfamethoxazole	244 (67.7)	62.78–72.40	116 (32.2)
Aztreonam	47 (13)	9.96–16.93	313 (86.9)
Ceftazidime	167 (46.3)	41.30–51.55	193 (53.6)
Imipenem	13 (3.6)	2.12–6.08	347 (96.3)
Nitrofurantoin	25 (6.9)	4.75–10.05	335 (93.0)

Abbreviation: CI Confidence Interval

ESBL production was detected in 168 isolates (46.6%; 95% CI: 41.6–51.8). Multidrug resistance was observed in 51.4% of isolates (95% CI: 46.2–56.5), underlining the clinical relevance and growing public health concern of antimicrobial resistance in uropathogenic E. coli populations. A dedicated ESBL panel (Supplementary Table S1) summarizes this prevalence together with its association with MDR and resistance to key third-generation β-lactam antibiotics.

Distribution of virulence genes

Analysis of 360 UPEC isolates revealed a broad distribution of virulence-associated genes across different phylogenetic groups. Highly prevalent genes such as kpsMTII (100%), ecpA (98.9%), and fimH (89.8%) were particularly common among isolates from phylogroup B2 (p < 0.001) (Table 4). This group also exhibited the highest overall frequency of virulence-associated genes. Conversely, phylogroups A and B1 exhibited notably lower frequencies. For instance, neuC was present in only 40.5% of group A and 29.4% of group B1 isolates, compared to 65.1% in B2. Statistically significant differences were observed for most genes across phylogroups (p < 0.05), except for papC (p = 0.406), suggesting heterogeneous virulence gene distributions depending on the phylogenetic background. For virulence genes with sparse distributions or zero counts in some phylogroups (e.g., sfaS, focG, hlyA, cnf1), global p-values in Table 4 were derived from Fisher’s exact test rather than Pearson’s Chi-square test. These comparisons among phylogroups were prespecified as exploratory, and the corresponding p-values were interpreted descriptively without FDR adjustment.

Association between phylogenetic groups and MDR

The relationship between phylogenetic classification and MDR was assessed among the UPEC isolates. As shown in Table 5, group B2 had the highest MDR prevalence (58.3%), while group D had the lowest (2.7%). The overall association between phylogenetic group (A, B1, B2, D) and MDR status was not statistically significant (global Chi-square, p > 0.05); subgroup comparisons for B2 (p = 0.222) and D (p = 0.107) were likewise non-significant. This phylogroup-MDR analysis was exploratory in nature, and no FDR adjustment was applied to these p-values.

Table 5.

Multidrug resistance across phylogenetic group

Phylogenetic group	No-MDR (n = 175)	MDR (n = 185)	OR (95% CI)	p-value
A	51 (29.1%)	60 (32.4%)	1.17 (0.75–1.83)	0.574
B1	22 (12.5%)	12 (6.4%)	0.48 (0.23–1.01)	0.073
B2	90 (51.4%)	108 (58.3%)	1.32 (0.87–2.01)	0.222
D	12 (6.8%)	5 (2.7%)	0.38 (0.13–1.09)	0.107

Percentages in this table represent column percentages within MDR and non-MDR categories. They do not represent MDR prevalence within each phylogroup

Abbreviations: MDR Multidrug-resistant, OR Odds ratio

Correlation between virulence genes and clinical features

A very small subset of participants (≈ 1%) later reported mild urinary symptoms during follow-up. Isolates from these symptomatic cases showed higher frequencies of certain virulence genes (e.g., fimH, papC), whereas asymptomatic participants more commonly carried genes such as kpsMTII and neuC. Because this subgroup was extremely small, no inferential analyses were performed, and these observations should be interpreted as descriptive and hypothesis-generating only. A sensitivity analysis excluding these cases did not change the overall findings. A brief descriptive comparison is presented in Table 6.

Table 6.

Descriptive comparison between symptom-emergent and asymptomatic participants

Feature	Developed mild symptoms after enrollment (≈ 1% of isolates)*	Remained strictly asymptomatic (≈ 99% of isolates)
Clinical observation	Mild dysuria or flank discomfort reported during follow-up	No urinary symptoms during study period
Dominant virulence genes	Higher frequency of fimH, papC, iutA	Predominantly kpsMTII and neuC
Toxin genes	Occasional presence of hlyA and cnf1	Rare
Biofilm phenotype	More strong biofilm producers	Mostly moderate or weak producers
MDR patterns	Slightly higher MDR frequency	Comparable to overall cohort
Interpretation	Patterns suggest possible link between virulence load and symptom emergence	Findings remain descriptive; no inference made

* A very small proportion of participants (approximately 1% of the cohort) reported mild urinary symptoms after enrollment. Because of the small subgroup size, findings are descriptive only

Biofilm formation and antimicrobial resistance

Out of 360 isolates, 21.1% (n = 76) were strong biofilm producers, 49.1% (n = 177) were moderate, and 29.7% (n = 107) were weak or non-producers. Strong biofilm formation was significantly associated with MDR (Chi-square, p < 0.001). Associations with individual antibiotics were evaluated using Fisher’s exact test where needed due to sparse cells; resistance to ciprofloxacin (p = 0.003) and ampicillin (p = 0.012) was higher among strong biofilm producers. These biofilm-resistance associations were considered exploratory, and no FDR adjustment was applied.

Genotypic profiles of co-harbored virulence genes

Among the various co-harbored virulence gene patterns identified, the combinations fimH-kpsMTII-iutA (22.5%) and fimH-ecpA-iutA (18.3%) were the most frequently observed. Despite the high prevalence of certain gene clusters, no statistically significant correlations were found between genotypic profiles and either phylogenetic groups or antimicrobial resistance phenotypes (p > 0.05). This suggests that while co-occurrence of virulence factors is common, their distribution may be independent of both genetic lineage and resistance status. This assessment of co-harbored genotypic profiles was exploratory, and p-values were not corrected for multiple comparisons.

Association between virulence genes and antimicrobial resistance

The distribution of major virulence genes across phylogenetic groups is summarized in Table 4, showing that fimH and papC were most prevalent among group B2 isolates. We further examined the relationship between major virulence genes and resistance outcomes (Table 7). Adhesion genes, particularly fimH and papC, were significantly more common among ampicillin- and ciprofloxacin-resistant isolates than in susceptible counterparts. For example, fimH-positive isolates showed 74.2% resistance to ampicillin compared with 55.6% among fimH-negative isolates (p = 0.003). Similarly, papC-positive isolates were significantly associated with ciprofloxacin resistance (p = 0.004). Iron acquisition genes also showed strong associations. The prevalence of fyuA and iutA was markedly higher in ESBL-producing isolates (62.5% vs. 38.7%; p < 0.001 for fyuA; 58.3% vs. 39.5%; p = 0.004 for iutA). By contrast, capsule-related markers (kpsMTII and neuC) did not show statistically significant associations with resistance to any of the antibiotics tested (p > 0.05). Toxin genes such as hlyA showed a trend toward higher MDR prevalence (p = 0.025). Notably, the same adhesion genes (fimH and papC) that appeared more frequent among isolates from mildly symptomatic cases (Sect. 3.5) also showed significant associations with antibiotic resistance, indicating a possible overlap between virulence and resistance mechanisms in these UPEC isolates. These virulence-resistance associations formed part of the predefined primary analysis set, and the reported p-values were adjusted for multiplicity using the Benjamini-Hochberg FDR procedure.

Table 7.

Associations between selected virulence genes and antimicrobial resistance phenotypes in UPEC isolates (n = 360). Outcome: MDR, ampicillin resistance, Ciprofloxacin resistance, and ESBL production

Gene	Antibiotic resistance outcome	Resistant with gene, n/N (%)	Resistant without gene, n/N (%)	OR (95% CI)	BH-FDR adjusted p-value*
neuC	MDR	25 / 60 (41.7)	79 / 182 (43.4)	0.9 (0.5–1.7)	0.880
fimH	Ampicillin R	120 / 162 (74.2)	55 / 99 (55.6)	2.3 (1.3–4.0)	0.003
papC	Ciprofloxacin R	88 / 139 (63.3)	40 / 91 (44.0)	1.9 (1.1–3.3)	0.004
kpsMTII	Ampicillin R	39 / 78 (50.0)	136 / 222 (61.3)	0.6 (0.3–1.1)	0.085
iutA	ESBL+	70 / 120 (58.3)	47 / 119 (39.5)	2.1 (1.2–3.5)	0.004
fyuA	ESBL+	75 / 120 (62.5)	46 / 119 (38.7)	2.6 (1.5–4.5)	< 0.001
hlyA	MDR	42 / 78 (53.8)	62 / 164 (37.8)	1.9 (1.1–3.2)	0.025

“Resistant with gene” and “Resistant without gene” are expressed as n/N (%), where N represents the total number of isolates carrying or lacking the gene among all 360 UPEC isolates. OR estimates were calculated using these same denominators

Abbreviations: MDR Multidrug-resistant, ESBL Extended-spectrum β-lactamase, OR Odds ratio, CI Confidence interval, R Resistant, FDR False discovery rate

* Fisher’s exact test was applied for comparisons involving sparse or zero cell counts, and Pearson’s Chi-square test was used only when all expected frequencies met test assumptions. All p-values were adjusted for multiple testing using the Benjamini-Hochberg false discovery rate (FDR) procedure, applied exclusively to the predefined primary virulence-resistance panel (FDR threshold = 0.05)

Discussion

This discussion is organized around three central findings of the study: (i) the high burden of MDR and ESBL among UPEC isolates from pregnant women with asymptomatic bacteriuria, (ii) the predominance of the B2 phylogroup without a strong association with multidrug resistance, and (iii) the significant links between key virulence traits, particularly adhesins and iron-uptake systems, and antimicrobial resistance and strong biofilm formation. Together, these themes summarize how resistance, virulence, and phylogeny interact in shaping the pathogenic profile of UPEC in this population.

Except for the predefined primary virulence-resistance panel (for which FDR correction was applied), all other statistical associations reported in this study are exploratory and should be interpreted descriptively.

Our findings add to existing evidence on the challenges posed by UPEC infections in pregnant women with ASB. Antibiotic resistance remains a critical concern, as observed in our study, where high resistance rates to first-line antibiotics, such as nalidixic acid (78.3%), and ampicillin (74.1%), align with trends reported globally [12, 14]. Similar resistance patterns have also been described in Iran and neighboring regions, with elevated resistance to older β-lactams and fluoroquinolones, whereas nitrofurantoin has generally remained effective [38–40]. The observed resistance to ciprofloxacin (51.3%) is particularly concerning and reinforces the need for cautious prescribing of fluoroquinolones in maternal care [39]. Nitrofurantoin (6.9%) and imipenem (3.6%) showed the lowest resistance rates, supporting their continued usefulness in pregnancy when used judiciously. Taken together, these observations support our first key finding: a substantial burden of antimicrobial resistance among UPEC isolates in pregnant women, consistent with regional and global trends.

The detection of ESBL production in 46.6% of UPEC isolates underscores the escalating challenge of β-lactam resistance. ESBL-producing strains exhibit resistance to third-generation cephalosporins, complicating treatment options. This prevalence is consistent with previous Iranian reports, such as those by Kasraeian et al. and Azami et al., and aligns with global observations, highlighting the widespread dissemination of ESBL-positive UPEC strains [41–43]. These findings further reinforce the high burden of MDR and ESBL in this population, which represents the first major theme of our discussion. Co-resistance between ESBL production and other antibiotic classes further restricts therapeutic options and emphasizes the need for antimicrobial stewardship in antenatal care [43]. In our study, ESBL-producing isolates showed markedly higher odds of MDR and resistance to key third-generation cephalosporins compared with ESBL-non-producing isolates, as detailed in Supplementary Table S1; these ESBL-related associations were exploratory and should be interpreted with appropriate caution.

Our findings also underscore the diverse virulence arsenal of UPEC. Adhesion-related genes, such as fimH and papC, were predominant in isolates from symptomatic patients, highlighting their role in mediating bacterial colonization and persistence in host tissues [44]. These virulence patterns align with our third major theme, highlighting the interplay between adhesins, iron-uptake systems, and resistance phenotypes. Iron acquisition genes, such as iutA and fyuA, were also prevalent, emphasizing their importance in nutrient-limited environments like the urinary tract. These findings are consistent with previous studies demonstrating the critical role of adhesins and siderophores in UPEC pathogenesis [45, 46]. In line with our results, Eghbalpour et al. similarly reported high frequencies of fimH and kpsMTII, particularly in B2 isolates, as well as higher detection of hlyA and cnf1, reinforcing their association with more invasive infections [47].

Notably, the neuC gene was detected in a subset of isolates. Given its potential role in capsule synthesis and immune evasion, neuC may contribute to chronic persistence in the urinary tract. Its presence in UPEC warrants further investigation to explore whether it may represent a virulence factor of emerging relevance in pregnancy-related UTIs [31, 48]. Although neuC was less frequent, its potential role in immune evasion supports the broader theme of diverse virulence strategies contributing to persistence. Overall, this study provides one of the most comprehensive analyses of UPEC isolates from pregnant women with asymptomatic bacteriuria in Iran, integrating phenotypic resistance profiling with detailed virulence gene characterization and biofilm assessment, and extending previous regional work by including neuC, traditionally linked to neonatal meningitis E. coli.

The phylogenetic analysis showed that group B2 isolates were predominant, consistent with their recognized virulence potential. However, no significant association was observed between phylogroup and MDR, supporting our second central theme. This lack of association may reflect the uneven distribution of isolates across phylogenetic subgroups and the limitations of the triplex Clermont scheme, which does not distinguish the more recently defined C, E, and F groups and may lead to minor misclassification. In addition, widespread plasmid-mediated resistance and horizontal gene transfer can obscure lineage-specific patterns, allowing resistance traits to circulate independently of chromosomal background [49]. These methodological and ecological factors suggest that phylogroup-related comparisons should be interpreted cautiously. Overall, this supports our second central message: although the B2 phylogroup predominated, phylogeny alone did not strongly predict MDR patterns.

Biofilm formation was significantly associated with resistance, particularly to ampicillin and ciprofloxacin, and may serve as a reservoir enabling evasion from antibiotics and host immunity [15, 50]. This pattern aligns with our third central theme, indicating that specific virulence traits, such as adhesion and biofilm capacity, interact with resistance phenotypes and may contribute to persistence. However, the biofilm-resistance analyses were exploratory and not adjusted for multiple testing, and therefore should be interpreted with caution.

From a clinical perspective, strong biofilm-forming isolates may contribute to persistent or recurrent bacteriuria despite standard therapy, potentially increasing the risk of upper urinary tract infection or pregnancy-related complications [51]. However, these observations are exploratory and hypothesis-generating, and should not be interpreted as justification for modifying antenatal screening or management protocols. Any consideration of biofilm assessment would apply only to selected high-risk contexts and requires confirmation in multicenter longitudinal studies before broader clinical use.

The correlation between virulence genes and clinical features underscores the heterogeneity of UPEC pathogenesis. Symptomatic cases were more frequently associated with adhesin (fimH, papC) and toxin (hlyA, cnf1) genes, whereas asymptomatic cases more often carried immune evasion markers (kpsMTII, neuC). These contrasting patterns support the broader theme that distinct virulence strategies, ranging from adhesion and toxin production to capsule-mediated evasion, shape the clinical expression of UPEC infection, consistent with previous evidence of stratified pathogenic mechanisms [5, 52].

Our analysis showed that certain virulence determinants were significantly associated with resistance phenotypes in UPEC isolates. Adhesion-related genes (fimH, papC) were linked to higher resistance to ampicillin and ciprofloxacin, while iron-acquisition genes (fyuA, iutA) were strongly associated with ESBL production. In contrast, capsule-associated genes (kpsMTII, neuC) demonstrated no meaningful correlation with resistance, indicating that not all virulence traits influence antimicrobial susceptibility. These findings support the broader conclusion that resistance in UPEC is multifactorial, with adhesion and iron-uptake pathways contributing more prominently than capsule expression. Importantly, these associations derive from the predefined primary virulence-resistance panel and were evaluated with Benjamini-Hochberg FDR adjustment.

The socio-demographic characteristics of the participants (Table 2) provide contextual background for interpreting the microbiological findings. Because the study did not compare resistance across demographic subgroups, these variables were used only descriptively. Potential host or environmental contributors, such as prior antibiotic exposure or access to care, were not assessed analytically and warrant evaluation in larger future studies.

In high MDR settings, enhanced surveillance may hold conceptual value, but any potential application remains exploratory. Early identification of MDR or strong biofilm-forming isolates may help generate hypotheses about factors contributing to persistent bacteriuria, yet these observations should not be interpreted as grounds for modifying antenatal screening or management practices. Any consideration of targeted assessment would be limited to selected high-risk contexts and requires confirmation in multicenter longitudinal studies before clinical use.

This study had several limitations. Identification of E. coli relied on phenotypic methods rather than molecular confirmation, and a small degree of misclassification with other Enterobacterales cannot be excluded; the direction or magnitude of such bias is unknown and could affect prevalence and resistance estimates. Genotypic analysis of biofilm-related determinants was not performed, limiting interpretation of phenotype-genotype relationships. Phylogrouping was based on the triplex Clermont scheme, which does not differentiate the newer C, E, and F groups and may introduce minor classification imprecision. The single-center design and restriction to pregnant women limit generalizability. A small number of participants developed urinary symptoms after enrollment, but a sensitivity analysis excluding them did not alter overall trends. Despite these constraints, the inclusion of 360 non-duplicate isolates and extensive phenotypic profiling strengthens the robustness of the findings.

Overall, these findings highlight the value of continued surveillance in settings with a high MDR burden, while emphasizing that routine screening for virulence or biofilm traits remains premature. Conceptual strategies such as vaccines targeting conserved adhesins (e.g., fimH, papC) or biofilm-disrupting approaches may warrant future investigation but are not supported for clinical adoption at this stage. Given the single-center design and exploratory nature of several analyses, these considerations should be viewed as hypothesis-generating rather than prescriptive. By integrating clinical, molecular, and epidemiological perspectives, this study contributes to a broader understanding of UPEC pathogenicity in pregnancy and underscores the need for innovative approaches to address the challenges posed by MDR strains.

Conclusion

This study documents a high prevalence of antimicrobial resistance and ESBL production among UPEC isolates from pregnant women with asymptomatic bacteriuria. The coexistence of multiple virulence traits with MDR and strong biofilm formation suggests potential biological interactions that warrant further investigation, but causal relationships cannot be inferred from this cross-sectional design. While these findings may help generate hypotheses about factors relevant to risk stratification in pregnancy, they do not directly support changes to antenatal screening or management, and any potential application would require confirmation in multicenter longitudinal studies. Continued surveillance and molecular monitoring remain important for understanding local epidemiology and informing stewardship efforts.

Authors’ contributions

A.A. and L.Y. conceptualized and designed the study. A.A. was responsible for patient recruitment, sample collection, and laboratory diagnostics. S.F. assisted with sample collection and contributed to manuscript revision. F.A., A.H. and M.A. performed microbiological identification, antibiotic susceptibility testing, and molecular assays. F.A. and A.H. conducted the statistical analysis and data visualization. A.A. and L.Y. wrote the initial draft of the manuscript. A.A. and L.Y. provided critical revisions and contributed to the interpretation of the data. All authors reviewed and approved the final version of the manuscript.

Funding

Not applicable.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This research was approved in Ethics Committee of the Golestan University of Medical Sciences (Approval Code: IR.GOUMS.REC.1395.2). The authors confirm that all experiments were performed in accordance with relevant guidelines and regulations. Also, they confirming that informed consent was obtained from all subjects. All the methods/study was in accordance with the declaration of Helsinki, and all methods were conducted in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.