Skip to main content
NCBI home page
Genome Medicine logoLink to Genome Medicine
. 2026 Apr 1;18:61. doi: 10.1186/s13073-026-01635-9

Global survey of extraintestinal pathogenic Escherichia coli identifies stable serogroup-virulence-resistance linkages

Mengjie Li 1,2, Xiao Liu 2, Ruiting Lan 3, Yanwen Xiong 2, Yanqing Che 1,2, Yixin Mao 2,4, Hongzhou Li 1,2, Guodong Yan 2,5, He Gao 2, Xuemei Bai 2, Zhiwen Sun 2, Xiaoning Liu 1,✉, Duochun Wang 2,✉
PMCID: PMC13169692  PMID: 41918017

Abstract

Background

Extraintestinal pathogenic Escherichia coli (ExPEC) is a leading cause of urinary tract infections, sepsis, and neonatal meningitis globally. However, global patterns linking O-serogroups, sequence types (STs), virulence, and antimicrobial resistance (AMR) remain poorly resolved.

Methods

We analyzed 12,220 ExPEC genomes, including 12,160 publicly available genomes retrieved from the NCBI database and 60 representative isolates collected and sequenced in this study. Serogroups were assigned by in silico wzm/wzt/wzx/wzy and fliC gene analysis, and phylogenies constructed using maximum likelihood methods. Virulence and antibiotic resistance genes were annotated with the Virulence Factor Database and the Comprehensive Antibiotic Resistance Database. A set of ExPEC isolates collected from China was tested phenotypically for motility, iron acquisition and growth assays under standardized conditions.

Results

We identified 199 O-serogroups from the 12,220 genomes, with 17 dominant types accounting for 83% of isolates. O25 (30.75%) was globally prevalent, while O101 was more prevalent than O25 in China and India. Most of the isolates were from urine (44.49%) and blood (26.80%) as primary sources. Phylogenetic analysis revealed 52 clusters with stable serogroup-ST associations (e.g., O25-ST131, O2/O6-ST73). Virulence analysis uncovered 383 genes, with a conserved 24-gene core encoding adhesion and iron uptake genes. Resistance profiling revealed 308 genes, dominated by β-lactamases and efflux regulators. Integrated analysis identified iron acquisition (Vir_2) and efflux-mediated resistance (Res_5) as key genomic traits of high-risk serogroups, supported by phenotypic assays showing correlated motility and iron uptake patterns.

Conclusions

We propose a genomic survey that integrates serogroup profiling with virulence-resistance gene-content clusters to resolve ExPEC population structure and functional convergence. These insights may guide the rational development of next-generation ExPEC vaccines and precision surveillance strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13073-026-01635-9.

Keywords: Extraintestinal pathogenic Escherichia coli, O-serogroup; Phylogenomic analysis, Virulence, Resistance, Vaccine strategy

Background

Extraintestinal pathogenic Escherichia coli (ExPEC) is a significant group of opportunistic bacterial pathogens [1, 2], which are broadly classified into three pathotypes: uropathogenic E. coli (UPEC), sepsis-associated E. coli (SEPEC), and neonatal meningitis-associated E. coli (NMEC). They cause over 400 million urinary tract infections and approximately 230,000 related deaths annually, impose a substantial public health burden [3–6].

ExPEC pathogenicity is mediated by virulence factors such as adhesins, iron acquisition systems, and capsules, which enable the bacteria to colonize extraintestinal sites and evade host immunity [5, 7, 8]. Its clinical impact has been exacerbated by the rapid emergence of antimicrobial resistance (AMR). Extensive antibiotic use and horizontal gene transfer have driven resistance to key antimicrobial, including cephalosporins, fluoroquinolones, and carbapenems [9, 10]. Although high-risk clones such as ST131-O25b are well characterized, the co-evolution of virulence and resistance traits across diverse genetic backgrounds remains poorly understood, hindering predictions of emerging threats [11, 12].

The O-antigen, encoded by the O antigen biosynthesis gene cluster, is a critical component of the bacterial outer membrane and a primary target for host immune recognition, bacteriophage interaction, and vaccine development [13, 14]. While numerous studies have investigated the relationships between O-antigen variation, sequence types (STs), virulence factors, AMR, and vaccine escape, many have focused on a limited set of well-characterized serogroups such as O25, O2, and O6 [15, 16]. However, these studies have often been more targeted, addressing specific serogroups or lineages, rather than systematically integrating O-antigen profiles with a wide range of functional and phenotypic data across a global dataset.

In this study, we assembled a global dataset of 12,220 ExPEC genomes from 88 countries/regions spanning 27 years (1998–2025). By integrating O-serogroup profiling, phylogenetic reconstruction, virulence and resistance annotation, and phenotypic validation, we aimed to delineate both global distribution patterns and the descriptive gene clusters architectures driving ExPEC pathoadaptation. We explored whether conserved gene-content clusters could explain ExPEC’s ecological success across diverse serogroups. This work provides a comprehensive global survey for understanding ExPEC evolution and offers actionable insights for vaccine target updates and emerging serogroup monitoring.

Methods

Genome retrieval and inclusion criteria

A total of 12,220 ExPEC genomes were analyzed, including 12,160 publicly available genomes retrieved from the public database (Additional file 1: Table S1) and 60 representative Chinese isolates collected and sequenced in this study (Additional file 1: Table S2). As of March 10, 2025, we retrieved 12,160 genomes isolated from Homo sapiens from the NCBI database (https://www.ncbi.nlm.nih.gov/datasets/genome/), excluding samples labeled with intestinal (e.g., feces, gut, and intestine) or unknown source terms. Low-quality assemblies (contigs > 200) were filtered out. The 60 isolates collected in this study were sequenced on the Illumina HiSeq platform for downstream analysis. Raw sequencing data were evaluated using fastp v0.23.4 [17], and low-quality reads were filtered out. The filtered reads were de novo assembled using SPAdes v3.15.3 and the resulting assemblies were assessed using QUAST v5.2.0 to ensure assembly integrity and completeness [18, 19]. For comparative purposes, 200 intestinal E. coli isolates were randomly selected from the NCBI database (Additional file 1: Table S3).

Phylogenetic inference

Genome annotation was performed using Prokka v1.14.6 [20]. Persistent gene families were identified by PPanGGOLiN v1.2.10 [21], and multiple sequence alignments were generated for phylogenetic reconstruction. A maximum-likelihood tree was built using IQ-TREE v2.1.3 with ModelFinder for model selection and 1,000 ultrafast bootstrap and SH-aLRT replicates [22]. Phylogenetic clusters were defined based on patristic distance (cut-off = 0.05). Tree visualization and annotation were performed using iTOL v6 [23] (https://itol.embl.de). O-serogroup geographic prevalence was calculated based on geographic distribution of strains across continents. Mantel tests were conducted using Scikit-bio v0.5.8 [24], comparing phylogenetic and geographic distance matrices (999 permutations, Pearson correlation).

Serogroup, MLST, virulence, and AMR gene annotation

O- and H-serogroups were annotated using the Majorbio platform [25], based on wzm/wzt/wzx/wzy and fliC sequences (identity ≥ 95%, coverage ≥ 90%). Sequence types (STs) were assigned using the mlst v2.23.0 (identity ≥ 95%) software based on the PubMLST database [26, 27]. Virulence and AMR genes were annotated using the Virulence Factor Database with DIAMOND v2.1.8 and the Comprehensive Antibiotic Resistance Database with ABRicate v1.0.1, applying thresholds of identity ≥ 90% (virulence) or 95% (AMR) and coverage ≥ 90% [28–31]. Binary virulence and resistance gene matrices were visualized using hierarchical clustering (Ward.D2 method) via Hiplot Pro [32]. For descriptive gene-content clusters classification, K-means clustering (k = 5) was applied separately to virulence and resistance profiles using custom Python scripts. Clustering was performed on binary gene presence/absence matrices, with the number of clusters preset to five to balance biological interpretability and facilitate downstream analyses. The implementation used the KMeans algorithm in scikit-learn v1.7.0 [33]. For each cluster, the top five genes were identified according to their within-cluster frequencies. The resulting cluster labels were then assigned to each sample and used for subsequent gene-frequency analyses, which were summarized in the supplementary tables.

Phenotypic assays

Sixty representative strains were tested in triplicate for each assay. Swarming motility was assessed on 2% agar plates. 2 μL of bacterial culture was spotted and incubated at 37 °C. Colony diameters were measured at 8, 12, and 24 h using vernier calipers [34]. Swimming motility was evaluated on 0.35% semi-solid agar. 2 μL of bacterial culture was inoculated into the agar using a needle, and were incubated at 37 °C. Colony diameters were measured at 4, 8, and 12 h using vernier calipers [35]. Iron acquisition ability was assessed using the Chrome Azurol S (CAS) agar assay. 5 μL of bacterial culture was spotted onto CAS agar plates and incubated at 37 °C for 24 h. The ability of isolates to acquire iron was measured by the appearance of orange halos. Pseudomonas aeruginosa was used as a positive control to validate the assay’s specificity and reliability [36]. To assess bacterial growth, overnight bacterial cultures were inoculated (1:100) into fresh LB (200 μL/well in 96-well plates) and incubated at 37 °C with shaking. OD600 was measured every 30 min for 24 h [37]. Antimicrobial susceptibility testing (AST) and determination of minimum inhibitory concentrations (MICs) were performed using the VITEK 2 Compact automated system (bioMérieux, Marcy-l'Étoile, France), which is based on the broth microdilution method. A panel of 28 drugs across 14 categories (aminocoumarins, aminoglycosides, cephalosporins, diaminopyrimidines, fluoroquinolones, fosfomycins, macrolides, nitroimidazoles, nucleosides, peptides, phenicols, rifamycins, sulfonamides, and tetracyclines). MICs were interpreted according to Clinical and Laboratory Standards Institute (CLSI) guidelines, 2025, M100 [38]. E. coli ATCC 25922 was included as a quality control strain. All phenotypic assays were conducted in triplicate for each strain, and the average values were used for downstream analyses.

Statistical analysis

All statistical analyses were performed in R v4.2.3 and Python v3.10 [39, 40]. Phenotypic differences among O-serogroups were tested using one-way ANOVA. Phenotypic and genotypic resistance frequencies were compared using two-sided permutation tests with Benjamini–Hochberg correction for multiple testing. Pearson correlations assessed associations between gene frequencies and phenotypic traits.

Results

O-serogroup distribution and phylogenetic analysis

A total of 199 O-serogroups were identified among the 12,220 ExPEC strains. The 17 major serogroups accounted for 83.26% of all isolates (Table 1). The most prevalent O serogroup was O25 (n = 3,758; 30.75%), followed by O101 (n = 714; 5.84%), O6 (n = 663; 5.43%). Urine (n = 5,437; 44.49%) and blood (n = 3,275; 26.80%) were the most common sources, followed by respiratory (n = 412; 3.37%) and wound specimens (n = 341; 2.79%).

Table 1.

Global distribution of O-serogroups and isolation sources of extraintestinal pathogenic Escherichia coli$

Serogroup Isolation source Total (%)
urine blood respiratory wound others*
O25 1,643 1,225 181 88 621 3,758 (30.75)
O101 300 151 42 52 169 714 (5.84)
O6 303 178 6 5 170 662 (5.42)
O2 261 168 3 9 132 573 (4.69)
O16 264 141 6 14 127 552 (4.52)
O1 221 151 23 14 126 535 (4.38)
O8 218 134 13 12 145 522 (4.27)
O75 268 106 15 7 108 504 (4.12)
O102 172 100 19 15 59 365 (2.99)
O9 149 83 10 16 87 345 (2.82)
O77 98 91 4 10 90 293 (2.40)
O15 139 74 4 3 60 280 (2.29)
O153 119 59 8 17 57 260 (2.13)
O4 123 67 5 1 62 258 (2.11)
O18 119 66 0 2 62 249 (2.04)
O86 91 41 11 10 38 191 (1.56)
O45 52 40 3 4 31 130 (1.06)
O11 65 25 0 3 30 123 (1.01)
others 832 376 58 59 581 1,906 (15.60)
Total (%)

5,437

(44.49)

3,276

(26.81)

411

(3.37)

341

(2.79)

2,755

(22.55)

12,220

(100.00)
Open in a new tab

$based on 12,220 ExPEC genomes, including 12,160 from public databases and 60 representative isolates sequenced in this study

*Includes bodily fluids/exudates (e.g., ascites, bile, pleural fluid, cerebrospinal fluid), clinical or patient-related samples, urogenital tract specimens, skin/soft tissue, bone/joint, other tissues, and sources not otherwise classified

The temporal distribution of ExPEC genomes varied over time (Fig. 1A-B). From 1998 to 2006, annual isolate counts for all O-serogroups remained below 10 isolates. A sharp increase began in 2007, with peaks in 2018 (n = 4,312) and 2023 (n = 2,981). O25 emerged as the dominant serogroup since 2006, with a mean prevalence of 34.7% among sequenced isolates between 2006 and 2024. Data from 2025 (Jan-Mar only) were excluded from this average.

Fig. 1.

Fig. 1

Open in a new tab

Global distribution of ExPEC isolates. A Annual number of ExPEC isolates from (1998–2025). B Temporal trends in the relative abundance of major O-serogroups (1998–2025). C Geographic distribution of ExPEC and dominant O-serogroups across six continents

Geographically, isolates originated from 88 countries and regions across 6 continents (Fig. 1C), excluding 32 strains with unknown locations. The largest proportions of isolates were originated from the USA (n = 3,089; 25.28%), Canada (n = 1,473; 12.05%), the UK (n = 1,459; 11.94%). It is worth noting that in China and India, O101 (13.50% and 20.24%) was more prevalent than O25 (9.43% and 8.33%) as the dominant serogroup. By continent, North America (n = 4,592; 37.58%), Europe (n = 3,932; 32.18%), and Asia (n = 1,863; 15.25%) accounted for most isolates, while Oceania (8.11%), Africa (4.40%) and South America (2.23%) accounted for smaller proportions. O25 was dominant across all continents: North America (38.98%), Europe (29.17%), Asia (16.00%), Oceania (32.80%), Africa (21.56%), and South America (28.31%).

Phylogenetic structure and correlation between genetic and geographic distances

Phylogenetic tree constructed for 12,220 ExPEC genomes resulting in 52 phylogenetic clusters (Fig. 2A). Although isolates showed random distribution across sources, locations, and years, several clusters exhibited strong ST associations, indicating phylogenetic coherence. For example, cluster 36 (n = 4,217; 34.26%) was dominated by O25:H4-ST131 (n = 3,209; 76.09%); cluster 42 (n = 2,069; 16.81%) was enriched for O2/O6 (n = 842; 40.7%), with O2/O6:H1-ST73 (n = 484; 23.39%); cluster 28 (n = 1,411; 11.46%) included O8/O9:H9 (n = 161; 11.41%) and ST410 (n = 281; 19.91%). To compare source-specific clustering, we randomly selected 200 isolates each from urinary, blood, respiratory, wound, others and intestinal origins (total n = 1,200) (Fig. 2B). The resulting tree showed no distinct clustering by source, with intestinal and extraintestinal isolates randomly distributed across phylogenetic lineages. This topological pattern suggests that the potential to cause extraintestinal infections may be a polyphyletic trait distributed among multiple genetic backgrounds, rather than being strictly associated with specific phylogenetic lineages.

Fig. 2.

Fig. 2

Open in a new tab

Phylogenetic structure, source distribution, and spatiogenomic correlation of ExPEC. A Maximum-likelihood phylogenetic tree constructed from 12,220 ExPEC genomes (12,180 terminal branches; midpoint-rooted). B Distribution of 1,200 randomly selected E. coli isolates (from urinary, blood, respiratory, wound, and intestinal sources) mapped onto the phylogenetic tree. C Correlation between genomic and geographic distances across O-serogroups based on Mantel tests. South American data is not included due to insufficient variability in both Mantel r and sharing proportion values for the available samples, as indicated in the figure legend. Among 152 tested groups, only 14 showed significant correlation between genomic and geographic distances (p < 0.05). The median Mantel r was 0.0091, indicating overall weak associations. Each point represents an O-serogroup, color-coded by continent, with a fitted linear regression line. Key serogroups (e.g., O25, O101) are annotated

A Mantel test was performed to assess correlations between phylogenetic and geographic distances among isolates (Fig. 2C, Additional file 2: Table S4). Of 12,220 strains, 12,080 were retained after matrix harmonization. A total of 27 serogroups with > 50 isolates were included in the test. Most major serogroups (e.g., O25, O101, O6, O2) had Mantel r values between 0 and 0.2, indicating weak phylogeographic association. Nevertheless, 14 serogroups exhibited statistically significant geographic structure (p < 0.05), suggesting that the associations between genomic and geographic distances in these groups were non-random. Notably, serogroups (e.g., O64, O101;O8, O53) showed relatively higher Mantel r values, reflecting a more pronounced regional clustering. In contrast, some serogroups (e.g., O22, O11) had negative r values, possibly due to lineage admixture from cross-regional transmission. These findings are consistent with their scattered phylogenetic patterns.

Correlation of O-serogroup-virulence-resistance spectrum

A total of 383 virulence-associated genes were identified from 12,220 ExPEC genomes and classified into 13 functional groups (Fig. 3A, Additional file 2: Table S5). Core virulence genes were defined as those present in > 95% of isolates, with 24 genes meeting this criterion and being present in > 97.68% of isolates (Additional file 2: Table S6). The majority (n = 18; 85.71%) related to adhesion or nutrient acquisition. Hierarchical clustering revealed two major virulence profiles. Cluster I was dominated by O101 and O8 isolates, characterized by adhesins (fimH, yagX) and iron uptake effectors (entA, iucA). Cluster II exhibited a typical high-pathogenicity ExPEC profile, featuring yersiniabactin (irp1, ybtA) and heme uptake (chuA) systems. Subgroup IIa carrying colibactin biosynthetic genes (clbA-Q) which are located within a pathogenicity island involved in the production of the genotoxin colibactin, whereas subgroup IIb consisted mainly of O25-ST131 isolates.

Fig. 3.

Fig. 3

Open in a new tab

Genomic architecture and modular associations of virulence and resistance profiles in ExPEC. A Hierarchical clustering heatmap of virulence genes based on the ward.D2 method. B Hierarchical clustering heatmap of AMR genes. C Three-column Sankey diagram illustrating the relationships among O-serogroups, virulence gene-content clusters (Vir_1-5), and resistance gene-content clusters (Res_1-5), based on K-means clustering to simplify cluster classification and highlight modular cross-serogroup patterns. D PCA plot showing sample distribution in virulence-resistance feature space, colored by cluster category (defined by K-means). PERMANOVA analysis based on Euclidean distance revealed significant group-wise differences (pseudo-F = 4079.76, p = 0.01), indicating meaningful multidimensional structure across cluster combinations

AMR profiling revealed 308 resistance genes spanning 19 classes (Additional file 2: Table S7). Multiple drug resistance (MDR, n = 122), aminoglycoside resistance genes (n = 60), fluoroquinolone resistance (n = 46), and carbapenemases (n = 39) together accounted for 68.18% of all detections. Two major AMR clusters were identified (Fig. 3B): Class I was enriched for O25b-ST131, with high carriage of blaCTX-M-15 (n = 2,315; 56.65%) while Class II exhibited broader serogroup diversity and higher frequencies of chromosomally encoded efflux regulators (e.g., baeR, kdpE).

Integrated analysis using K-means clustering (k = 5) simplified the virulence-resistance spectrum into five gene-content clusters (Fig. 3C, Table 2). Iron acquisition (Vir_2, n = 4,506) was the most prevalent, enriched in serogroups O25, O101, O15, and O153. It comprised primarily iron uptake genes (fur, entE, fepB, fepD, entB), reflecting a conserved iron acquisition strategy. On the resistance side, efflux-mediated resistance (Res_5, n = 3,996) was the dominant AMR cluster, broadly distributed in O25, O2, O6, and O75. It included genes for multidrug efflux and global regulation (acrA, msbA, mdtH, hns, crp). Res_4 (n = 2,072), largely found in O25 (n = 1,566), was associated with cephalosporin resistance. Principal component analysis (PCA) further confirmed modular clustering, with distinct separation between Vir_2-Res_5 enriched isolates and other combinations (Fig. 3D).

Table 2.

Summary of virulence-resistance gene-content clusters identified by K-means clustering

Gene-content clusters Representative genes*
Vir_1 fur, ybtP, ybtA, ybtX, ybtQ
Vir_2 fur, ibeC, rpoS, entB, entE
Vir_3 fur, ybtS, fepB, entE, entC
Vir_4 fur, ibeC, fepD, fepA, entB
Vir_5 fur, entE, fepB, fepD, entB
Res_1 msbA, hns, crp, acrB, tolC
Res_2 msbA, acrA, hns, acrD, cpxA
Res_3 msbA, hns, crp, mdtA, ampH
Res_4 msbA, hns, crp, acrA, mdtA
Res_5 acrA, hns, msbA, mdtH, crp
Open in a new tab

*Representative genes were identified using K-means clustering (k = 5) according to their within-cluster frequency. Gene overlaps across clusters indicate shared genomic features

Experimental validation of virulence-resistance-associated phenotypes

To validate the phenotypic relevance of gene-content clusters, we tested 60 ExPEC strains collected from clinical samples. The 60 strains spanned 14 O-serogroups and covered multiple phylogenetic clusters defined in the global ExPEC tree (Fig. 4A, Additional file 1: Table S2), enabling exploratory comparisons between genotype and phenotype. No significant differences were observed in 24-h growth curves or swarming motility across serogroups (Fig. 4B-C). In contrast, swimming motility varied significantly among serogroups (Fig. 4D). O153 showed the highest ability, followed by O4, O15, O8, and O16. O1 and O6 displayed moderate motility, whereas O2, O162, O18, and O77 were weakest. Given the known association between motility and H (flagellar) antigens, H-antigen typing revealed a higher H5 prevalence (42.86%) in high-motility strains, though correlation was weak and nonsignificant (r = 0.18). Virulence gene analysis showed significant enrichment of fimbrial assembly genes, with the type 1 fimbrial genes fimI and fimG showed strong positive correlations with the average swimming diameter at 12 h (rfimI = 0.79; rfimG = 0.70), and higher carriage in high-motility serogroups (81% and 75%) compared to low-motility groups (20% and 25%). The iron-limitation assay revealed significant differences in growth performance after 24 h (Fig. 4E). O16, O75, and O153 strains exhibited the strongest growth, O1, O2, O15, and O162 showed intermediate performance, and O4, O25, O8, and O6 had the weakest growth. Although siderophore gene carriage was widespread, expression levels or functional compensation may vary, explaining the weak correlation (r = 0.12). Comparison of genotypic predictions and phenotypic results revealed that most serogroups resistant to cephalosporins, fluoroquinolones, and sulfonamides (Fig. 4F-G). Aminocoumarin and aminoglycoside resistance was largely confined to O4, O18, and O25, whereas tetracycline resistance predominated in O15, O16, and O153. No phenotypic resistance was observed for other classes. Genotypically, however, resistance genes were widely enriched across serogroups for all categories except nucleoside, phenicols, and rifamycins.

Fig. 4.

Fig. 4

Open in a new tab

Phenotypic validation of virulence-resistance gene-content clusters in 60 representative ExPEC strains. A Phylogenetic tree of the 60 tested strains. Phylo-cluster ID strips indicate phylogenetic cluster assignments based on the global ExPEC tree (Fig. 2A). B Growth performance after 24 h of incubation. Each strain was tested in triplicate (n = 3); error bars represent mean ± SD ANOVA revealed significant differences among serogroups (ANOVA, p = 0.02), although overall variation was minor and no clear group separation was observed. C Swarming motility assay. Data represent mean ± SD from triplicate experiments (n = 3). ANOVA revealed no significant differences among serogroups (ANOVA, p > 0.01). D Swimming motility assay. Data represent mean ± SD from triplicate experiments (n = 3). ANOVA identified significant differences (ANOVA, p < 0.01). E Iron-limitation assay. Iron diameter (mm) denotes the diameter of the halo on CAS agar in Iron acquisition assay. Data represent mean ± SD from triplicate experiments (n = 3). ANOVA indicated significant differences among serogroups (ANOVA, p < 0.01). F Distribution of predicted antimicrobial resistance genes across dominant O-serogroups. The presence of genes was inferred from whole-genome annotation against CARD databases. G Phenotypic AMR patterns from susceptibility testing. Triplicate experiments (n = 3). ANOVA revealed significant discrepancies between predicted and phenotypic resistance profiles across serogroups (Permutation test, with p and false discovery rate (FDR)-adjusted q). Full statistical details for phenotypic assays are provided in Additional file 3: Table S8-9

Discussion

This global genomic survey of 12,220 ExPEC genomes reveals significant associations between O-serogroups, STs, and virulence-resistance gene clusters, further supporting previously reported patterns in the literature [41–43]. These trends suggest potential adaptive routes contributing to the ecological success of distinct ExPEC groups. Integrating gene-content profiling with serogroup and ST monitoring may enhance genomic surveillance and inform risk-based interventions.

A critical implication of this work lies in the dynamic nature of O-antigen distribution among ExPEC populations. Previous global surveillance identified O25, O2, and O6 as dominant serogroups, which were later incorporated into candidate vaccines such as ExPEC10V [16, 44]. However, our temporal analysis highlights the recent rise of O101, O8 and O16, suggesting potential changes of O-serogroup composition. Notably, O101 rather than O25 was the most prevalent serogroup from China and India. This rise in O101 suggests its growing adaptation in these regions, potentially driven by a combination of factors. The high-frequency virulence and resistance genes found in O101 strains could contribute to their survival in iron-limited environments and resistance to multiple antibiotics, as reported in other high-prevalence serogroups [45]. Additionally, O101’s spread may be accelerated by high population density and frequent human-to-human contact, particularly in regions with high antibiotic usage. More importantly, by introducing a descriptive gene-content clusters, this study providing a complementary perspective to identify high-risk ExPEC strains, focusing on conserved virulence and resistance features that may persist independently of antigenic variation.

At the mechanistic level, the coupling of Vir_2 and Res_5 clusters supports a co-adaptive evolutionary model, particularly in high-risk lineages such as O25 and O101. Echoing findings by Behnoush et al., the widespread presence of iron uptake genes in Vir_2 suggests an evolutionarily conserved pathoadaptive mechanism, particularly effective in iron-limited niches like the urinary tract and bloodstream [46–48]. In parallel, Res_5 is enriched with efflux pump genes and global regulatory elements, constituting a typical multidrug resistance gene cluster [49–55]. This dual-functional arrangement may represent an evolutionarily stable strategy, balancing nutrient exploitation with antimicrobial defense, and shaping the persistence of high-risk ExPEC lineages.

Experimental assays further validated the functional significance of these clusters. Swimming motility assays revealed significant associations with fimI/fimG gene presence, particularly in O153 and O101 strains, suggesting enhanced early-stage colonization and surface migration capacities. Similarly, iron acquisition assays revealed that strains enriched for Vir_2 clusters, such as O16, O75, and O153, exhibited superior growth under iron-restricted conditions. While Vir_2 contains fewer iron uptake genes than other virulence gene-content clusters, its integration of stress response genes (rpoS) and host adaptation genes (ibeC) likely contributes to its ability to thrive in iron-limited environments, highlighting its broader ecological adaptability. This coordinated strategy likely facilitates ExPEC pathogenesis, where motility promotes initial mucosal colonization, followed by systemic dissemination supported by efficient iron acquisition. Together, these traits may model a two-phase infection process—initial migration followed by resource exploitation, potentially contributing to the global spread and clinical dominance of specific ExPEC clones, as observed in previous studies [12, 56, 57].

Although significant discrepancies between genotypic resistance predictions and phenotypic resistance profiles have been documented [58–60], our analysis shows that while genomic data predicted resistance to 14 antibiotic classes, phenotypic testing (Fig. 4G) revealed resistance primarily to cephalosporins, fluoroquinolones, and sulfonamides. Notably, strains from the O25 serogroup, often resistant to cephalosporins in clinical settings, did not exhibit resistance phenotypically. This highlights that the presence of resistance genes does not always correlate with phenotypic resistance. Gene expression, environmental factors, and unaccounted resistance mechanisms, such as efflux pumps and β-lactamases, may influence resistance expression. These discrepancies underscore the limitations of current genotypic prediction methods, particularly in predicting gene expression. Future studies should refine these models by incorporating phenotypic validation and gene expression analysis, as well as expanding sample sizes and conducting multi-center studies to improve prediction accuracy and clinical relevance.

Beyond these findings, our data also highlight the ecological adaptability and global dissemination of ExPEC. Phylogenetic analysis further revealed strong serogroup-ST associations (e.g., O25-ST131, O2/O6-ST73, O8-ST410), which are well-documented in the literature and reflect clonal stability [61–64]. However, isolates from different infection sites (e.g., urine, blood) were randomly distributed on the phylogenetic tree, consistent with the modular transfer of pathogenicity traits rather than strict tissue tropism [65, 66]. Similarly, the weak geographic structure observed in our study is consistent with prior research that suggests widespread lineage dissemination in ExPEC, likely driven by global mobility and healthcare-associated transmission [67, 68]. Temporal comparisons showed no major shifts of some ExPEC lineages across years, suggesting long-term stability in the human population. Likewise, intestinal and extraintestinal isolates showed no systematic divergence, underscoring the broad ecological adaptability and evolutionary stability of core ExPEC lineages. Moreover, certain serogroups such as O22 and O11 exhibited negative Mantel r values, suggesting globally dispersed distribution patterns with minimal geographic constraint, which may indicate broad ecological tolerance, although further evidence is needed to confirm generalist strategies. O101, although highly successful in regions such as China and India, did not show strong geographic correlation in the Mantel test. This suggests that even regionally dominant lineages like O101 undergo extensive inter-regional transmission, further reinforcing the idea that ExPEC’s global success is driven by high mobility and broad ecological adaptability rather than strict geographic confinement.

Although our dataset includes 12,220 genomes, public repositories are subject to sampling bias, with uneven geographic representation and over-representation of clinical isolates. Gene presence-absence profiling is informative but cannot capture expression dynamics, regulatory control, or post-transcriptional modifications. Future transcriptomic and proteomic studies are needed to clarify mechanisms of ExPEC adaptation and pathogenesis. Phenotypic assays were restricted to 60 Chinese strains, spanning diverse O-serogroups and virulence-AMR clusters, yet interpretations are exploratory and require validation in larger cohorts.

Together, our study defines a global ExPEC survey, linking O-serogroups to genomic lineages and highlighting the rise of emerging serogroups like O101. Conserved virulence and resistance gene-content clusters underscore complementary antigenic and functional dimensions, informing next-generation vaccines and precision surveillance strategies.

Conclusions

In the current study, we delineated the global population structure of ExPEC across 12,220 genomes. The O-serogroup distribution of ExPEC was broad yet highly structured, with a limited number of dominant serogroups accounting for most isolates worldwide, while clear geographic heterogeneity was also evident, particularly the prominence of O101 in China and India. Phylogenetic analysis revealed that major ExPEC lineages were characterized by stable serogroup-ST associations, and virulence and resistance gene-content profiling identified recurrent virulence and resistance genes, particularly those related to iron acquisition and multidrug resistance.

Supplementary Information

13073_2026_1635_MOESM1_ESM.xlsx (773.8KB, xlsx)

Additional file 1: Metadata used in this study. Table S1: Metadata of 12,160 ExPEC from NCBI. Table S2: Metadata of 60 ExPEC clinical isolates. Table S3: Metadata of 200 E. coli from NCBI.

13073_2026_1635_MOESM2_ESM.xlsx (23.1KB, xlsx)

Additional file 2: Genomic analysis results used in this study. Table S4: Mantel test results assessing correlations between phylogenetic and geographic distances across ExPEC O-serogroups. Table S5: Functional categories of virulence genes in ExPEC genomes. Table S6: Core virulence genes (cutoff of > 95%). Table S7: Functional categories of antimicrobial resistance genes in ExPEC genomes.

13073_2026_1635_MOESM3_ESM.xlsx (11.9KB, xlsx)

Additional file 3: Phenotypic analysis results used in this study. Table S8: ANOVA results for growth, motility, and iron-limitation phenotypic assays across ExPEC serogroups. Table S9: Comparison of phenotypic and genotypic resistance frequencies across 14 antibiotic classes.

13073_2026_1635_MOESM4_ESM.zip (17.3MB, zip)

Additional file 4: Source data for main figures.

Acknowledgements

We are grateful to Chenhong Guan for his contributions to the preparation of the figures in this manuscript.

Abbreviations

ExPEC

Extraintestinal pathogenic Escherichia coli

UPEC

Uropathogenic Escherichia coli

SEPEC

Sepsis-associated Escherichia coli

NMEC

Neonatal meningitis-associated Escherichia coli

STs

Sequence types

AMR

Antimicrobial resistance

Authors’ contributions

M.L., X.L., D.W. designed the research project. M.L. performed the experiments and wrote original draft. M.L., Y.C., Y.M., H.L., G.Y. analyzed and discussed the data. M.L., X.L., R.L., Y.X., D.W., H.G., X.B., Z.S. reviewed and discussed the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (32570220), the Prevention and Control of Emerging and Major Infectious Diseases-National Science and Technology Major Project (2025ZD01900100), the National Pathogen Resource Center (NPRC-32), and the screening of Streptococcus agalactiae capsular polysaccharide vaccine production strain (HX2025-001) was funded by the National Institute for Communicable Disease Control and Prevention, China CDC.

Data availability

The 60 genome assemblies generated in this study are publicly available at NCBI GenBank under BioProject accession number PRJNA1297337, with GenBank accession numbers JBPZJR000000000 to JBPZLY000000000 [69]. All relevant metadata, analysis scripts, and summary tables have been made publicly available on GitHub (https://github.com/mjlee-micro/expec-global-genomic-survey) [70].

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the Chinese Center for Disease Control and Prevention (approval number ICDC-LSM-2025009). Informed consent was obtained from participants, where applicable. This study conformed to the principles of the Helsinki Declaration.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Xiaoning Liu, Email: liuxn@lzu.edu.cn.

Duochun Wang, Email: wangduochun@icdc.cn.

References

  • 1.Qiu L, Chirman D, Clark JR, Xing Y, Hernandez Santos H, Vaughan EE, et al. Vaccines against extraintestinal pathogenic Escherichia coli (ExPEC): progress and challenges. Gut microbes. 2024;16(1):2359691. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Denamur E, Clermont O, Bonacorsi S, Gordon D. The population genetics of pathogenic Escherichia coli. Nat Rev Microbiol. 2021;19(1):37–54. [DOI] [PubMed] [Google Scholar]
  • 3.Manges AR, Geum HM, Guo A, Edens TJ, Fibke CD, Pitout JDD. Global Extraintestinal Pathogenic Escherichia coli (ExPEC) Lineages. Clin Microbiol Rev. 2019;32(3):e00135-18. [DOI] [PMC free article] [PubMed]
  • 4.Foxman B, Bangura M, Kamdar N, Morgan DM. Epidemiology of urinary tract infection among community-living seniors aged 50 plus: population estimates and risk factors. Ann Epidemiol. 2025;104:21–7. [DOI] [PubMed] [Google Scholar]
  • 5.Wijetunge DS, Gongati S, DebRoy C, Kim KS, Couraud PO, Romero IA, et al. Characterizing the pathotype of neonatal meningitis causing Escherichia coli (NMEC). BMC Microbiol. 2015;15:211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Rudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR, et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. Lancet. 2020;395(10219):200–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sarowska J, Futoma-Koloch B, Jama-Kmiecik A, Frej-Madrzak M, Ksiazczyk M, Bugla-Ploskonska G, et al. Virulence factors, prevalence and potential transmission of extraintestinal pathogenic Escherichia coli isolated from different sources: recent reports. Gut Pathog. 2019;11:10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Terlizzi ME, Gribaudo G, Maffei ME. Uropathogenic Escherichia coli (UPEC) infections: virulence factors, bladder responses, antibiotic, and non-antibiotic antimicrobial strategies. Front Microbiol. 2017;8:1566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Darby EM, Trampari E, Siasat P, Gaya MS, Alav I, Webber MA, et al. Author correction: molecular mechanisms of antibiotic resistance revisited. Nat Rev Microbiol. 2024;22(4):255. [DOI] [PubMed] [Google Scholar]
  • 10.Pitout JD. Extraintestinal pathogenic Escherichia coli: an update on antimicrobial resistance, laboratory diagnosis and treatment. Expert Rev Anti-Infect Ther. 2012;10(10):1165–76. [DOI] [PubMed] [Google Scholar]
  • 11.Chorro L, Li Z, Chu L, Singh S, Gu J, Kim JH, et al. Preclinical Immunogenicity and efficacy of optimized O25b O-Antigen glycoconjugates to prevent MDR ST131 E. coli infections. Infection and immunity. 2022;90(4):e0002222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Riley LW. Pandemic lineages of extraintestinal pathogenic Escherichia coli. Clin Microbiol Infect. 2014;20(5):380–90. [DOI] [PubMed] [Google Scholar]
  • 13.Liu B, Furevi A, Perepelov AV, Guo X, Cao H, Wang Q, et al. Structure and genetics of Escherichia coli O antigens. FEMS Microbiol Rev. 2020;44(6):655–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Royer G, Clermont O, Condamine B, Mercier-Darty M, Laouenan C, Lefort A, et al. O-antigen targeted vaccines against escherichia coli may be useful in reducing morbidity, mortality, and antimicrobial resistance. Clin Infect Dis. 2022;74(2):364–6. [DOI] [PubMed] [Google Scholar]
  • 15.Poolman JT, Wacker M. Extraintestinal pathogenic Escherichia coli, a common human pathogen: challenges for vaccine development and progress in the field. J Infect Dis. 2016;213(1):6–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Weerdenburg E, Davies T, Morrow B, Zomer AL, Hermans P, Go O, et al. Global distribution of o serotypes and antibiotic resistance in extraintestinal pathogenic escherichia coli collected from the blood of patients with bacteremia across multiple surveillance studies. Clin infect Dis. 2023;76(3):e1236–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Chen S. Ultrafast one-pass FASTQ data preprocessing, quality control, and deduplication using fastp. Imeta. 2023;2(2):e107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Prjibelski A, Antipov D, Meleshko D, Lapidus A, Korobeynikov A. Using SPAdes de novo assembler. Curr Protoc Bioinform. 2020;70(1):e102. [DOI] [PubMed] [Google Scholar]
  • 19.Gurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics (Oxford, England). 2013;29(8):1072–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics (Oxford, England). 2014;30(14):2068–9. [DOI] [PubMed] [Google Scholar]
  • 21.Gautreau G, Bazin A, Gachet M, Planel R, Burlot L, Dubois M, et al. PPanGGOLiN: depicting microbial diversity via a partitioned pangenome graph. PLoS Comput Biol. 2020;16(3):e1007732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic Era. Mol Biol Evol. 2020;37(5):1530–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Letunic I, Bork P. Interactive tree of life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res. 2024;52(W1):W78–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Borry M. Sourcepredict: prediction of metagenomic sample sources using dimension reduction followed by machine learning classification. J Open Source Software. 2019;4(41):1540.
  • 25.Han C, Shi C, Liu L, Han J, Yang Q, Wang Y, et al. Majorbio Cloud 2024: update single-cell and multiomics workflows. iMeta. 2024;3(4):e217. [DOI] [PMC free article] [PubMed]
  • 26.Seemann T. mlst. GitHub. Available from: https://github.com/tseemann/mlst.
  • 27.Jolley KA, Bray JE, Maiden MCJ. Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications. Wellcome open research. 2018;3:124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Buchfink B, Reuter K, Drost H-G. Sensitive protein alignments at tree-of-life scale using DIAMOND. Nat Methods. 2021;18(4):366–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Zhou S, Liu B, Zheng D, Chen L, Yang J. VFDB 2025: an integrated resource for exploring anti-virulence compounds. Nucleic Acids Res. 2025;53(D1):D871–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Alcock BP, Raphenya AR, Lau TTY, Tsang KK, Bouchard M, Edalatmand A, et al. CARD 2020: antibiotic resistome surveillance with the comprehensive antibiotic resistance database. Nucleic Acids Res. 2020;48(D1):D517–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Seemann T. Abricate. GithubAvailable from: https://github.com/tseemann/abricate.
  • 32.Li J, Miao B, Wang S, Dong W, Xu H, Si C, et al. Hiplot: a comprehensive and easy-to-use web service for boosting publication-ready biomedical data visualization. Brief Bioinform. 2022;23(4):bbac261. 10.1093/bib/bbac261. [DOI] [PubMed] [Google Scholar]
  • 33.Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: machine learning in Python. JMLR 12, pp. 2011;12(J. Mach. Learn. Res.):2825–30.
  • 34.Partridge JD, Harshey RM. Swarming motility assays in Salmonella. Methods Mol Biol. 2023;2646:147–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ha DG, Kuchma SL, O’Toole GA. Plate-based assay for swimming motility in Pseudomonas aeruginosa. Methods Mol Biol. 2014;1149:59–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Louden BC, Haarmann D, Lynne AM. Use of blue agar CAS assay for siderophore detection. J Microbiol Biol Educ. 2011;12(1):51-3. [DOI] [PMC free article] [PubMed]
  • 37.Rogers AT, Bullard KR, Dod AC, Wang Y. Bacterial growth curve measurements with a multimode microplate reader. Bio-Protoc. 2022;12(9):e4410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.CLSI. Performance Standards for Antimicrobial Susceptibility Testing. ed. t, editor. Wayne, PA: Clinical and Laboratory Standards Institute; 2025.
  • 39.Team RC. R: A language and environment for statistical computing. R Foundation for Statistical Computing. 2022. Available from: https://www.R-project.org/.
  • 40.Foundation PS. Python Language Reference, version 3.10.2021. Available from: https://docs.python.org/3.10/reference/.
  • 41.Mamani R, Flament-Simon SC, García V, Mora A, Alonso MP, López C, et al. Sequence types, clonotypes, serotypes, and virotypes of extended-spectrum β-Lactamase-Producing Escherichia coli Causing Bacteraemia in a Spanish Hospital Over a 12-Year Period (2000 to 2011). Front Microbiol. 2019;10:1530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Blanco J, Mora A, Mamani R, López C, Blanco M, Dahbi G, et al. Four main virotypes among extended-spectrum-β-lactamase-producing isolates of Escherichia coli O25b:H4-B2-ST131: bacterial, epidemiological, and clinical characteristics. J Clin Microbiol. 2013;51(10):3358–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Fierz L, Cernela N, Hauser E, Nüesch-Inderbinen M, Stephan R. Characteristics of Shigatoxin-producing Escherichia coli strains isolated during 2010-2014 from human infections in Switzerland. Front Microbiol. 2017;8:1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Saade E, Gravenstein S, Donskey CJ, Wilson B, Spiessens B, Abbanat D, et al. Characterization of Escherichia coli isolates potentially covered by ExPEC4V and ExPEC10V, that were collected from post-transrectal ultrasound-guided prostate needle biopsy invasive urinary tract and bloodstream infections. Vaccine. 2020;38(33):5100–4. [DOI] [PubMed] [Google Scholar]
  • 45.Bunduki GK, Heinz E, Phiri VS, Noah P, Feasey N, Musaya J. Virulence factors and antimicrobial resistance of uropathogenic Escherichia coli (UPEC) isolated from urinary tract infections: a systematic review and meta-analysis. BMC Infect Dis. 2021;21(1):753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Khasheii B, Mahmoodi P, Mohammadzadeh A. Siderophores: importance in bacterial pathogenesis and applications in medicine and industry. Microbiol Res. 2021;250:126790. [DOI] [PubMed] [Google Scholar]
  • 47.Garcia EC, Brumbaugh AR, Mobley HL. Redundancy and specificity of Escherichia coli iron acquisition systems during urinary tract infection. Infect Immun. 2011;79(3):1225–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Hashimoto M, Ma YF, Wang ST, Chen CS, Teng CH. Iron acquisition of urinary tract infection Escherichia coli Involves pathogenicity in caenorhabditis elegans. Microorganisms. 2021;9(2):310. [DOI] [PMC free article] [PubMed]
  • 49.Atac N, Kurt-Azap O, Dolapci I, Yesilkaya A, Ergonul O, Gonen M, et al. The role of AcrAB-TolC efflux pumps on quinolone resistance of E. coli ST131. Curr Microbiol. 2018;75(12):1661–6. [DOI] [PubMed] [Google Scholar]
  • 50.Xia Z, Zhou J, Gao N, Li G, Liu R, Lu G, et al. AcrAB-TolC efflux pump overexpression and tet(A) gene mutation increase tigecycline resistance in Klebsiella pneumoniae. World J Microbiol Biotechnol. 2024;40(8):233. [DOI] [PubMed] [Google Scholar]
  • 51.Alav I, Sutton JM, Rahman KM. Role of bacterial efflux pumps in biofilm formation. J Antimicrob Chemother. 2018;73(8):2003–20. [DOI] [PubMed] [Google Scholar]
  • 52.Alenazy R. Drug efflux pump inhibitors: a promising approach to counter multidrug resistance in gram-negative pathogens by targeting acrb protein from acrab-tolc multidrug efflux pump from Escherichia coli. Biology (Basel). 2022;11(9):1328. [DOI] [PMC free article] [PubMed]
  • 53.Weston N, Sharma P, Ricci V, Piddock LJV. Regulation of the AcrAB-TolC efflux pump in Enterobacteriaceae. Res Microbiol. 2018;169(7–8):425–31. [DOI] [PubMed] [Google Scholar]
  • 54.Reyes CL, Ward A, Yu J, Chang G. The structures of MsbA: insight into ABC transporter-mediated multidrug efflux. FEBS Lett. 2006;580(4):1042–8. [DOI] [PubMed] [Google Scholar]
  • 55.Mi W, Li Y, Yoon SH, Ernst RK, Walz T, Liao M. Structural basis of MsbA-mediated lipopolysaccharide transport. Nature. 2017;549(7671):233–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Xia F, Cheng J, Jiang M, Wang Z, Wen Z, Wang M, et al. Genomics analysis to identify multiple genetic determinants that drive the global transmission of the pandemic st95 lineage of extraintestinal pathogenic Escherichia coli (ExPEC). Pathogens (Basel, Switzerland). 2022;11(12):1489. [DOI] [PMC free article] [PubMed]
  • 57.Pitout JD, Peirano G, DeVinney R. The contributions of multidrug resistant clones to the success of pandemic extra-intestinal pathogenic Escherichia coli. Expert Rev Anti Infect Ther. 2023;21(4):343–53. [DOI] [PubMed] [Google Scholar]
  • 58.Eladawy M, Heslop N, Negus D, Thomas JC, Hoyles L. Phenotype–genotype discordance in antimicrobial resistance profiles of Gram-negative uropathogens recovered from catheter-associated urinary tract infections in Egypt. J Antimicrob Chemother. 2025;80(11):3123–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Gomes CN, Campioni F, Barker DOR, Che EV, Duque SDS, Taboada EN, et al. Antimicrobial resistance genotypes and phenotypes of Campylobacter coli isolated from different sources over a 16-year period in Brazil. J Glob Antimicrob Resist. 2023;33:109–13. [DOI] [PubMed] [Google Scholar]
  • 60.Manikandan R, Rajagunalan S, Malmarugan S, Gupta C. First report on whole genome sequencing and comparative genomics of Salmonella enterica serovar Abortusequi isolated from Donkey in India. Sci Rep. 2024;14(1):23455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Iguchi A, Iyoda S, Kikuchi T, Ogura Y, Katsura K, Ohnishi M, et al. A complete view of the genetic diversity of the Escherichia coli O-antigen biosynthesis gene cluster. DNA Res. 2015;22(1):101–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Kyung SM, Lee J, Lee ES, Hwang CY, Yoo HS. Genomic molecular epidemiology of carbapenemase-producing Escherichia coli ST410 isolates by complete genome analysis. Vet Res. 2023;54(1):72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Overdevest I, Haverkate M, Veenemans J, Hendriks Y, Verhulst C, Mulders A, et al. Prolonged colonisation with Escherichia coli O25:ST131 versus other extended-spectrum beta-lactamase-producing E. coli in a long-term care facility with high endemic level of rectal colonisation, the Netherlands, 2013 to 2014. Euro Surveill. 2016;21(42):30376. 10.2807/1560-7917.ES.2016.21.42.30376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.ArconadaNuin E, Vilken T, Xavier BB, Doua J, Morrow B, Geurtsen J, et al. A microbiological and genomic perspective of globally collected Escherichia coli from adults hospitalized with invasive E. coli disease. J Antimicrob Chemother. 2024;79(9):2142–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Gogarten JP, Doolittle WF, Lawrence JG. Prokaryotic evolution in light of gene transfer. Mol Biol Evol. 2002;19(12):2226–38. [DOI] [PubMed] [Google Scholar]
  • 66.Dagan T. Phylogenomic networks. Trends Microbiol. 2011;19(10):483–91. [DOI] [PubMed] [Google Scholar]
  • 67.Quilodrán CS, Currat M, Montoya-Burgos JI. Benchmarking the Mantel test and derived methods for testing association between distance matrices. Mol Ecol Resour. 2025;25(2):e13898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Zeller KA, Creech TG, Millette KL, Crowhurst RS, Long RA, Wagner HH, et al. Using simulations to evaluate Mantel-based methods for assessing landscape resistance to gene flow. Ecol Evol. 2016;6(12):4115–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Wang D. Genome assemblies of 60 extraintestinal pathogenic Escherichia coli isolates generated in this study. NCBI BioProject.2025. https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1297337
  • 70.Li M. ExPEC global genomic survey: analysis scripts, metadata, and summary tables. GitHub.2026. Available from: https://github.com/mjlee-micro/expec-global-genomic-survey.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

13073_2026_1635_MOESM1_ESM.xlsx (773.8KB, xlsx)

Additional file 1: Metadata used in this study. Table S1: Metadata of 12,160 ExPEC from NCBI. Table S2: Metadata of 60 ExPEC clinical isolates. Table S3: Metadata of 200 E. coli from NCBI.

13073_2026_1635_MOESM2_ESM.xlsx (23.1KB, xlsx)

Additional file 2: Genomic analysis results used in this study. Table S4: Mantel test results assessing correlations between phylogenetic and geographic distances across ExPEC O-serogroups. Table S5: Functional categories of virulence genes in ExPEC genomes. Table S6: Core virulence genes (cutoff of > 95%). Table S7: Functional categories of antimicrobial resistance genes in ExPEC genomes.

13073_2026_1635_MOESM3_ESM.xlsx (11.9KB, xlsx)

Additional file 3: Phenotypic analysis results used in this study. Table S8: ANOVA results for growth, motility, and iron-limitation phenotypic assays across ExPEC serogroups. Table S9: Comparison of phenotypic and genotypic resistance frequencies across 14 antibiotic classes.

13073_2026_1635_MOESM4_ESM.zip (17.3MB, zip)

Additional file 4: Source data for main figures.

Data Availability Statement

The 60 genome assemblies generated in this study are publicly available at NCBI GenBank under BioProject accession number PRJNA1297337, with GenBank accession numbers JBPZJR000000000 to JBPZLY000000000 [69]. All relevant metadata, analysis scripts, and summary tables have been made publicly available on GitHub (https://github.com/mjlee-micro/expec-global-genomic-survey) [70].

Articles from Genome Medicine are provided here courtesy of BMC

RESOURCES