Phylogenetic and pathogenic description of Enterococcus faecalis at a Northern Chinese hospital Authors: Jingliang Qin, Jiayi Zhang, Jing Wang, Yongming Shen, Nan Chen, Boyang Cao, Xi Guo Published: 2025 ABSTRACT BackgroundEnterococcus faecalis is one of the leading causes of bacterial infectious diseases. Describing this infection-inducing pathogen can facilitate evidence-based infection prevention and control. Yet clinical isolates of E. faecalis are difficult to structurally distinguish from non-clinical isolates, so a more precise and comprehensive typing assay is needed to equip us for this challenge. With its high accuracy and completeness, Whole-genome sequencing (WGS) has initiated a technological revolution in exploring the genetic variation of microbes. We aimed to investigate the use of WGS in characterizing the phylogenetic and pathogenic features of E. faecalis clinical isolates in the northern Chinese city of Tianjin. We analyzed these isolates’ sequence types (STs), their phylogenetic relationship to their global counterparts, their possible domestic spread events, their genetic linkage and the possible within-hospital transmission. We also screened their profiles of antibiotic resistance genes (ARGs), virulence genes, pathogenicity islands (PAIs), and chromosome-encoded type IV secretion system (CE-T4SS). Furthermore, we verified the ARG profiles with the antimicrobial susceptibility results.ResultsAmong the 53 clinical E. faecalis isolates, we identified 15 known STs and reported 2 novel STs (ST1970 and ST1971). Three dominant STs exhibited different clustering patterns: ST16 isolates grouped with other Chinese isolates, ST745 isolates formed a separate cluster, and ST179 isolates overlapped with other global isolates. Isolates within some STs exhibited closer pairwise single nucleotide polymorphism (SNP) distances, but direct transmission was unlikely due to the lack of coincidence between the phylogenetically close isolates. Although local isolates generally carried a small number of ARGs and multiple virulence genes, isolates assigned to distinct STs varied in the distributions of ARGs, virulence genes, PAIs, and CE-T4SS. Also, the ARG profiles correlated well with their antibiotic-resistant phenotypes.We demonstrated for the first time the phylogenetic and pathogenic characterization of E. faecalis clinical isolates in the Tianjin area. Additionally, we revealed a robust correlation between these features and the STs of the isolates, indicating clinical isolates belonging to different STs might vary in pathogenic ability.ConclusionsWGS exhibited excellent capabilities in characterizing local clinical E. faecalis, making it a potential tool for epidemiological studies and infection management. In the meantime, fully exploiting the use of genomic data from local clinical strains can enhance antimicrobial resistance surveillance and provide insights into potential pathogenicity, aiding the development of more effective treatment strategies.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12866-025-04487-2. Background Enterococcus faecalis (E. faecalis) is an opportunistic pathogen causing community-acquired and hospital-acquired infections, such as bloodstream infections, urinary tract infections, and endocarditis [1, 2]. It is the most frequently isolated species of enterococci and the second most commonly isolated Gram-positive bacteria in major Chinese hospitals [3]. However, we still struggle to characterize local E. faecalis clinical strains to improve regional infection management. On the one hand, E. faecalis genotypes correlate limitedly with hosts or regions: some reports suggested some epidemic genotypes dispersed globally, while other studies showed that the enriched sequence types (STs) of E. faecalis clinical isolates varied by region [4–7]. On the other hand, unlike another clinically important enterococci, Enterococcus faecium (E. faecium), which has a significantly larger genome in clinical isolates than in non-clinical isolates, E. faecalis appears to have no such structural distinction [8]. Recently, whole-genome sequencing (WGS) enabled an efficient tool for genome-based microbial characterization with higher resolution than the classical seven-gene multi-locus sequence type (MLST) scheme [5, 9]. This high resolution means that WGS can provide explicit sequence-based epidemiological surveillance, depict detailed phylogenetic relationships, and trace possible transmission events. For example, a WGS-based study in the UK and Ireland revealed three hospital-associated epidemic lineages of E. faecalis and traced the evolution of vancomycin resistance [10]. WGS has also been applied in many countries to optimize prevention strategies for nosocomial E. faecalis infection [9, 11, 12]. Besides phylogenetic divergence, WGS can simultaneously provide the opportunity to define the pathogenic features of pathogens, like antibiotic resistance and virulence factor genes. These genes, closely associated with the larger genomic size of E. faecium clinical isolates and their hospital environmental adaptation, are similarly present in E. faecalis clinical isolates. Harboring both inherent and acquired antibiotic resistance genes, E. faecalis is resistant to many classes of antibiotics and thus limits the options of antibiotic drugs capable of treating E. faecalis infections [13]. Although vancomycin-resistant genes are less prevalent in E. faecalis than in E. faecium, they complicate the treatment and deteriorate the economic burden of E. faecalis infection [14]. As a reservoir of these antibiotic determinants, E. faecalis might disseminate antibiotic resistance across other Gram-positive bacteria with the aid of mobile element genes, such as transposons and plasmids [15, 16]. Furthermore, the over-representation of virulence factor genes emphasizes the primary role of E. faecalis in nosocomial infections [10]. In this study, we performed WGS in 57 clinical E. faecalis isolates from a local hospital, defining their STs and phylogenetic relationships, describing their critical pathogenic features, and verifying the phenotypes of ARG profiles. The detailed genomic analysis of our study can reveal epidemic populations of local E. faecalis strains and deeply understand their antibiotic-resistant/pathogenic features, thus supporting epidemiological studies, enhancing infection management, and improving treatment strategies. Methods and materials Isolates and antimicrobial susceptibility tests Fifty-seven E. faecalis isolates were collected from pediatric inpatients at Tianjin Children’s Hospital (TCH) from July 2020 to March 2021 (Supplementary Table 1). The median age of the patients was 7 months (range: 15 days to 12 years), with a slight predominance of males (53%). Clinical specimens were cultured on blood agar plates (Autobio, Zhengzhou, China) overnight at 37℃, and isolates were identified by the MALDI-TOF technique using a VITEK MS system (BioMérieux, France). Antimicrobial-susceptibility was tested using the VITEK 2 AST-GP67 Test Kit and a VITEK 2 Compact System (BioMérieux) according to the manufacturer’s protocol. Genomic analyses Genomic DNA of each isolate was extracted from 5 ml of overnight culture with the Wizard Genomic DNA Purification Kit (A1120, Promega) as per the manufacturer’s instructions, and DNA libraries were prepared using the Nextera XT DNA preparation kit (Illumina). Next, 2 × 150 bp paired-end genome sequencing was carried out using the NextSeq platform (Illumina). Raw data were trimmed using Trimmomatic [17]. After filtering, cleaned data of each sample was used for species-level analysis based on the standard database of the Kraken2 software (v2.1.2) [18]. All genomes obtained in this study were de novo assembled using SPAdes (v3.15.5) [19]. Protein-encoding genes were predicted and annotated with the Prokka annotation pipeline (v1.14.6) [20]. To place the TCH collection into a global context, we downloaded 993 E. faecalis genomes from a broad range of sources, including clinical, animal, environmental, and food sources. After quality control, we excluded 46 genomes that formed exceptionally long branches in a preliminary phylogenetic tree, as well as those with anomalous genome lengths or GC content compared to the typical range for E. faecalis, to minimize the impact of sequencing errors or contamination in subsequent analyses. Among the filtered 947 isolates, 632 were collected from other regions of China between 2002 and 2024, and the remaining 315 being recovered between 1996 and 2021 were mainly from Europe, America, Africa, Asia, and New Zealand (Supplementary Table 1). All assemblies were aligned against the chromosome of reference genome E. faecalis OG1RF (GCF_000172575.2) using snippy (v4.6.0) (https://github.com/tseemann/snippy). Then, the recombination regions were removed from the core alignment file using Gubbins (v2.3.4) [21]. Finally, IQ-TREE (v1.6.12) was used to construct the Maximum-Likelihood (ML) tree under the Generalized Time Reversible (GTR) model, with a bootstrap value of 1,000 [22]. To explore the population structure, all isolates were clustered using the Bayesian Analysis of Population Structure (BAPS) software [23] using three strains of Enterococcus faecium (GCF_009734005.1, GCF_016415425.1, GCF_016406585.1) as outgroups. To obtain a high-resolution phylogeny for TCH strains, we recalled the single nucleotide polymorphism (SNP) for the TCH assemblies using the same pipeline. The pairwise SNP distances from the core-gene alignment were identified using snp-dists (v0.8.4) (https://github.com/tseemann/snp-dists). Spread pattern inference To investigate the spread of E. faecalis isolates within China, we analyzed a dataset comprising 685 Chinese isolates, which includes the 53 TCH isolates collected in this study and the public sequences of Chinese isolates from NCBI. Analysis was conducted utilizing the R packages “evobiR” (v1.1) and “phytools” (v2.0-3) [24, 25]. Specifically, the tips of the Maximum Likelihood (ML) tree were aligned with the source records (provinces and cities) through the ReorderData function within EvobiR. Subsequently, stochastic source mapping was executed under the All-Rates-Different (ARD) model for ancestral state reconstruction, with 100 iterations, by leveraging the “make.simmap” function available in Phytools. Multi-locus sequence type (MLST) The STs of TCH and publicly available isolates were identified by comparing the assemblies to the MLST database (https://pubmlst.org/efaecalis). A neighbor-joining tree was constructed to visualize the genetic relationship among strains using FastME (v2.0) [26]. Antibiotic resistance genes (ARGs), virulence genes, pathogenicity islands (PAIs), and chromosome-encoded type IV secretion system (CE-T4SS) profiles Presence of ARGs and virulence genes was screened from assembled sequences against ResFinder (v4.4.0) and VirulenceFinder (v2.0) databases via BLASTN (v2.13.0+), both using a cutoff of 100% coverage and 90% identity, respectively. To complement the result generated from ResFinder, we also used BLASTN to investigate a linezolid resistance-conferring gene cfrD (accession number PHLC01000011), and a reference vanA transposon Tn1546 (accession number M97297.1). Variability of the PAIs among the TCH isolates was investigated using the prototype PAI of strain MMH594 as a reference (accession number AF454824). To determine the presence of CE-T4SS genes, a BLASTN search was performed using assembled sequences against the CE-T4SS reference genes (virB4/6/8/11, EF1877 to EF1880) from E. faecalis strain V583 (accession number NC_004668), using a cutoff of 75% coverage and 75% identity, respectively, as indicated by Hua et al. [27]. Results General features, MLST distribution, and phylogenetic analyses The genomic features of all 57 TCH isolates were summarized (Supplementary Table 2). Among them, four isolates (7008, 3791, 7799, and 7941) with abnormally large genome sizes and gene numbers, were identified as species contamination using Kranken2 and then excluded from further study (Supplementary Fig. 1). For the 53 isolates used in the subsequent analyses, the numbers of contigs are < 150, and the contig N50 is > 0.15 Mb, indicating overall genomic completeness. The genome sizes range from 2.7 Mb to 3.3 Mb, and the numbers of coding genes vary from 2,497 to 3,252. Based on the MLST database, 48/53 isolates were assigned to 15 STs, while 5/53 isolates (7431, 7801, 7853, 8166, and M4355) were unable to match the current database. The most common STs were ST16 (n = 17), ST745 (n = 9), and ST179 (n = 8), a human-associated single-locus variant of ST16, while the other 12 STs were underrepresented STs, each represented by only 1 to 3 isolates. 2/5 unmatched isolates (7853 and M4355) had an allele not scanned due to a gap between sequencing contigs. Therefore, according to the phylogenetic clustering data, we assumed that 7853 belonged to ST179 and M4355 belonged to ST295. The other three unmatched isolates, which held novel allele profiles not reported previously, were submitted to the pubMLST platform to obtain new sequence types: 7431 and 7801 (allele profile 5-–21-6-1-11-1-4) were allocated to ST1970, and 8166 (allele profile 2-–4-3-1-1-1-6) to ST1971. We next put our isolates into a global context. Within the mapping-based core genome, a total of 216,246 SNPs were identified, and an ML phylogeny was constructed. The phylogeny comprised of 1,000 isolates: 53 isolates from our study and 947 isolates from online sources. 869 isolates were grouped into 209 STs, while the remainder could not be assigned. Two of the dominant STs in our study, ST16 and ST179, were also among the top four STs which accounted for a quarter of the assigned isolates. TCH isolates and the global E. faecalis collection generally overlapped across different sources, host types, and regions, but TCH isolates assigned to different STs showed distinct clustering patterns (Fig. 1). ST16 isolates clustered with strains from other parts of China, while ST179 isolates overlapped with a worldwide collection of strains from the UK, France, and the USA. In contrast, ten ST745 isolates (nine from TCH and one from a single publicly available strain isolated from food in Zhejiang Province, China) formed a unique cluster, distinctive from other globally retrieved strains. The 12 underrepresented STs were expanded from 14 TCH isolates to 84 isolates after we incorporated publicly available sequences alongside TCH strains. Among these STs, ST480 (44%, 37/84) and ST81 (13%, 11/84) were the most prevalent, suggesting that these STs may have epidemiological significance in other regions. Moreover, BAPS delineated all E. faecalis into 14 clusters, among which TCH isolates mainly belonged to group 13 (seventeen ST16 isolates, 17/53, 32%), group 9 (eight ST179 and six other ST isolates, 14/53, 26%), and group 6 (nine ST745 and one ST30 isolates, 10/53, 19%). Fig. 1Global population structure of E. faecalis. Phylogeny of 53 TCH isolates combined with global isolates from geographically diverse locations obtained from the NCBI database. Isolates from this work were indicated as red dots at the end of each branch. Colored rings from the inner to the outer indicate the top 15 STs, the country of isolation, the host of isolates, the isolation source, and clusters generated by BAPS, respectively. Three outgroup strains of Enterococcus faecium were deleted Global population structure of E. faecalis. Phylogeny of 53 TCH isolates combined with global isolates from geographically diverse locations obtained from the NCBI database. Isolates from this work were indicated as red dots at the end of each branch. Colored rings from the inner to the outer indicate the top 15 STs, the country of isolation, the host of isolates, the isolation source, and clusters generated by BAPS, respectively. Three outgroup strains of Enterococcus faecium were deleted We next performed a phylogenetic reconstruction analysis of E. faecalis isolates from China to investigate the genetic relationships between strains from Tianjin and those from other Chinese regions. By tracing the ancestral states of 685 Chinese E. faecalis isolates, comprising 53 TCH strains and 632 Chinese isolates from NCBI, we found that the majority of ancestral nodes, representing inferred evolutionary ancestors, were associated with strains from Guangdong Province, whereas Tianjin strains predominantly occupied terminal branches in the phylogenetic tree (Fig. 2A). Between 2020 and 2021, multiple clonal groups of E. faecalis occurred in Tianjin, and twelve interregional spread events were identified: three from Tianjin to Beijing, Zhejiang, and Guangdong; one from Beijing to Tianjin, and eight from Guangdong to Tianjin (Fig. 2B). Fig. 2Possible spread events of E. faecalis within China from 2002 to 2024. (A) Phylogeny based on 192,341 SNPs extracted from 685 Chinese isolates, including 53 TCH isolates. Pie chart represents the ancestral states of the isolates. Isolates related to spread events in Tianjin are highlighted with dashed lines and enlarged. (B) Sampling distribution of Chinese isolates. The spread routes involving Tianjin are zoomed in, with the possible number of spread events being indicated Possible spread events of E. faecalis within China from 2002 to 2024. (A) Phylogeny based on 192,341 SNPs extracted from 685 Chinese isolates, including 53 TCH isolates. Pie chart represents the ancestral states of the isolates. Isolates related to spread events in Tianjin are highlighted with dashed lines and enlarged. (B) Sampling distribution of Chinese isolates. The spread routes involving Tianjin are zoomed in, with the possible number of spread events being indicated Genomic content comparison and potential transmission To explore the interlinkages within the TCH isolates, we compared their core alignments based on pairwise SNP distances, and the genomic linkage analysis categorized them into three main clusters (Cluster I, II, and III), similar to ST grouping (Fig. 3, left). Isolates within Cluster I, II, and III differed by 3 to 52 sites, 0 to 2 sites, and 5 to 93 sites, respectively. The prominent STs composing these clusters are ST16/ST179, ST745, and ST776 and ST1970 (three alleles distinct from ST776). Fig. 3Pairwise SNP distances between the core alignments of TCH isolates. The heatmap color indicates the number of pairwise SNPs. Vertical bars (middle) indicate the ST of each isolate. Horizontal stripe indicates the time scale of each hospitalized patient, with the corresponding department being depicted in the right column Pairwise SNP distances between the core alignments of TCH isolates. The heatmap color indicates the number of pairwise SNPs. Vertical bars (middle) indicate the ST of each isolate. Horizontal stripe indicates the time scale of each hospitalized patient, with the corresponding department being depicted in the right column To investigate the potential for within-hospital transmission, we incorporated the clustering results with patients’ wards and hospitalization periods of each isolate (Fig. 3, right), yet we found a general lack of overlapping in both location and time period within each phylogenetically close cluster, suggesting direct transmission was unlikely. ARG profiles are well consistent with the results of antimicrobial susceptibility tests The ARG profiles suggested that TCH isolates were susceptible to many commonly used antibiotics, and none of them appeared to be vancomycin-resistant enterococci (VRE). These genetic results were confirmed by the antimicrobial susceptibility tests. In addition, the prevalence of ARGs in different ST groups also differed considerably: high in ST16/ST179 but low in ST745 isolates (Fig. 4). Fig. 4Antibiotic susceptibility and ARGs patterns of TCH isolates, aligned with the ML phylogeny on SNP-based core-genome alignment. Top left, ML phylogeny of 53 TCH isolates; top middle, text means the presence of ARGs, and different color blocks display relevant phenotypical antibiotic resistance; top right, ST, and source of each isolate. Bottom left, Summary of antimicrobial resistance of each ST group; bottom right, ARG profiles of each ST group. P values are given to access statistical differences among ST groups using Fisher’s exact test. * Point mutated gene Antibiotic susceptibility and ARGs patterns of TCH isolates, aligned with the ML phylogeny on SNP-based core-genome alignment. Top left, ML phylogeny of 53 TCH isolates; top middle, text means the presence of ARGs, and different color blocks display relevant phenotypical antibiotic resistance; top right, ST, and source of each isolate. Bottom left, Summary of antimicrobial resistance of each ST group; bottom right, ARG profiles of each ST group. P values are given to access statistical differences among ST groups using Fisher’s exact test. * Point mutated gene None of the TCH isolates had resistance genes related to ampicillin, penicillin, nitrofurantoin, tigecycline, and vancomycin, but all had lsa(A), conferring intrinsic resistance to quinupristin/dalfopristin [28]. Accordingly, all isolates were susceptible to ampicillin, penicillin, nitrofurantoin, tigecycline, and vancomycin but resistant to quinupristin/dalfopristin. To investigate the potential occurrence of vancomycin-susceptible vanA-positive enterococci, which might revert to vancomycin resistance strains, we also screened for the presence and variation of the Tn1546 vanA operon, but no vancomycin resistance genes were identified, totally consistent with the phenotype of vancomycin susceptibility. The distributions of resistant genes tet(M), erm(B), and aac(6’)-aph(2’’) in TCH isolates were substantially consistent with the corresponding antibiotic resistance against tetracycline, erythromycin, and high-level streptomycin. Other rarely detected resistant genes (str, lsa(E), erm(A), tet(L)) might only play a minor role in local healthcare settings. It has been reported that mutations in gyrA and parC genes are closely related to high-level quinolone resistance [6, 29]. All TCH isolates possess gyrA and/or parC genes; however, only five of them, which were ciprofloxacin/levofloxacin resistant, carry quinolone resistance-associated mutations within gyrA and gyrA/parC. Therefore, the occurrence of mutated gyrA/parC genes is almost totally consistent with the profile of ciprofloxacin/levofloxacin susceptibility tests. An exception was M7418, which had neither mutations in gyrA nor parC, but exhibited ciprofloxacin/levofloxacin resistant in susceptibility tests. It was reported that the development of quinolone antibiotic resistance in E. faecalis was also related to the expression of efflux pump genes [30]. However, examining the expression of efflux pump genes through reverse transcription and real-time quantitative PCR is beyond the scope of this study. Although all TCH isolates were susceptible to commonly used antibiotics, our results revealed different ARGs prevalence rates across diverse STs (Fig. 4). ST16/ST179 isolates carried more ARGs, while most ST745 isolates had almost no ARGs other than genes related to intrinsic resistance. The rest presented an intermediate ARGs prevalence rate, which might partly reflect the wide variety of local STs. We also compared the presence of ARGs among different age groups of patients, but found no statistically significant difference (Supplementary Table 3). Furthermore, we investigated the ARGs of isolates from other regions within and outside of China (Supplementary Table 4). The result of the t test indicated that ARGs of isolates from other Chinese regions were more abundant than those from TCH and other global regions (Supplementary Fig. 2), suggesting that the risk of antibiotic resistance spread due to the local prevalence of E. faecalis is relatively low. Meanwhile, the distributions of quinolone resistance associated mutations within gyrA and parC are very distinct among different regions: 9.4% in TCH strains, 37.8% in other Chinese regions, and 23% in other global regions (Supplementary Table 5). This difference suggests that although the proportion of TCH and local E. faecalis resistant to quinolones remains relatively low, there is still a significant risk of the spread of quinolone-resistant strains in China. Prevalence of virulence genes, PAIs regions, and CE-T4SS Our findings revealed that TCH isolates had an overall abundance of virulence genes: At least nine virulence genes (cCF10, cOB1, cad, camE, ebpA, efaAfs, elaA, srtA, and tpx) were prevalent in almost all TCH isolates, and others were less frequent but distributed significantly different among ST groups (Fig. 5). While virulence genes ace and ebpC, coding cell surface proteins associated with adherence and biofilm formation, were only absent in a few Others/No match isolates, the other two genes with similar functions were exclusively found in a particular ST group: agg in ST16/ST179, and ebpB in Others/No match subgroup. Virulence genes coding secreted proteins, however, were even more distinctive across ST groups: ST16/ST179 isolates were enriched for cylA, cylL, and cylM, coding cytolysins with a bactericidal and cytolytic activity; ST745 isolates were abundant in gelatinase coding genes fsrB and gelE, as well as hyaluronidase coding genes hylA and hylB; another cytolysin coding gene cylB was only present in one isolate M7270 belonging to ST295. Similar to ARGs, the presence of virulence genes had no statistically significant difference among different age groups of patients (Supplementary Table 6). We also screened the virulence patterns of isolates from other regions within and outside of China (Supplementary Table 7). Compared to those from other regions, the TCH strains possessed less virulence genes (Supplementary Fig. 3). Fig. 5Virulence profiles and PAI patterns of TCH isolates. Top, distribution of virulence genes and distinct PAI regions aligned with the ML phylogeny on SNP-based core-genome alignment with ST and source of each isolate shown next to the phylogeny; Bottom, graphs showing the percentage of isolates assigned to certain STs in which virulence genes or PAI regions were detected. Genes that were ubiquitous or nearly absent in the isolates are not shown. P values are given when a significant difference was gained among ST groups using Fisher’s exact test (the P value of each PAI region among ST groups is < 0.001) Virulence profiles and PAI patterns of TCH isolates. Top, distribution of virulence genes and distinct PAI regions aligned with the ML phylogeny on SNP-based core-genome alignment with ST and source of each isolate shown next to the phylogeny; Bottom, graphs showing the percentage of isolates assigned to certain STs in which virulence genes or PAI regions were detected. Genes that were ubiquitous or nearly absent in the isolates are not shown. P values are given when a significant difference was gained among ST groups using Fisher’s exact test (the P value of each PAI region among ST groups is < 0.001) PAIs, which not only encode virulence factors but also act as mobility elements in E. faecalis [31, 32] were only abundant in TCH isolates assigned to ST16/ST179 (Fig. 5). Yet unlike ST16 isolates, which generally harbored all regions (A to F) of PAIs, ST179 isolates are totally absent of regions D and F. By contrast, none of PAIs presented in any ST745 isolates, and they only sporadically existed in isolates assigned to other STs. The CE-T4SS, which is widely distributed among 40% of clinical E. faecalis isolates, and in some cases flanked by multiple-drug-resistant (MDR) genes, shows the potential high risk of triggering MDR gene transmission [27]. We found a similar proportion (32%, 17/53) of TCH isolates carried CE-T4SS. Further analyses support a strong association of distribution between CE-T4SS and STs: all isolates assigned to ST179, ST776, ST30, and ST295 carried the CE-T4SS, whereas the remainder were all CE-T4SS negative (Fig. 6). The presence of CE-T4SS differed, even between closely related strains, for instance, those of ST16 and ST179. Although the occurrence of CE-T4SS in strains of ST776 and ST295 has not been reported, the overall results herein were consistent with the previous study [27]. However, we observed no MDR genes flanking CE-T4SS within the TCH isolates. Fig. 6CE-T4SS patterns among the TCH isolates. The distribution of CE-T4SS-positive and -negative strains assigned by MLST analysis. STs are shown in each circle and larger circles indicate greater numbers of strains; ND, STs not matched. Circle color indicates the presence (pink) or absence (green) of CE-T4SS CE-T4SS patterns among the TCH isolates. The distribution of CE-T4SS-positive and -negative strains assigned by MLST analysis. STs are shown in each circle and larger circles indicate greater numbers of strains; ND, STs not matched. Circle color indicates the presence (pink) or absence (green) of CE-T4SS Discussion This study presented phylogenetic features (STs and phylogenetic relationships) and pathogenic features (ARGs, virulence genes, PAIs, and CE-T4SS) of E. faecalis clinical strains in the Northern Chinese city of Tianjin. WGS could provide us with comprehensive genomic information, yet due to the differences in genetic traits across local dominant STs, MLST might still be practical in clinical E. faecalis characterization. Characterizing local clinical E. faecalis is critical to regional infection control [10]. Although a few previous studies about linazolid and vancomycin resistant E. faecalis have documented the existence of ST16 in different regions of China [33–35], the ST179 lineage has been reported only recently. A study in Shanghai characterizing randomly selected clinical strains has identified that ST16 and ST179 were the main clinical lineages in a more susceptible collection of clinical isolates [36], a similar result to another East Asian study performed in Japan [6] and our study here. Our phylogenetic analysis revealed that local ST16 isolates clustered with some Chinese strains, while local ST179 isolates were closely related to other global strains. This result suggests that, unlike the ubiquitously identified ST16 isolates within China, the domestic ST179 isolates became prevalent in very limited areas only recently. To our knowledge, the other dominant ST presented in our study, ST745, has never been reported, and its origin remains unclear. Since the ST745 isolates in our study were collected only from TCH, the possibility of a localized outbreak cannot be excluded without comparative data from other healthcare facilities and communities in the same region. In our single-center study, only two strains belong to ST480 and ST81, but they seemed more prevalent in other publicly available assemblies. E. faecalis strains of ST480 have been previously reported linked to linezolid resistance, both within and outside of China [37, 38]. Especially, a nine-year epidemiological study in Belgium reported that approximately two-thirds of linezolid-resistant E. faecalis strains were classified as ST480 [39]. Although the ST480 strain isolated in our study was sensitive to linezolid, its relatively high prevalence in other parts of China raises alarms about the potential spread of linezolid resistance. In contrast, the eleven ST81 strains were distributed sporadically across six countries (China, the UK, South Africa, Denmark, Thailand, and Malaysia). This scattered distribution and the lack of detailed epidemiological data make it difficult to evaluate the potential impact of this genotype. Another finding of our analysis was that during recent years, the majority of the Chinese isolates were spread from Guangdong (Fig. 2, Supplementary Table 8), a southern coastal province with the highest population mobility in China. In contrast, the vast majority of strains isolated from Tianjin were located at the terminal branches of the phylogenetic tree, and their evolutionary history suggests a pattern of local transmission following the initial introduction, with limited spread to other provinces (Beijing, Zhejiang, and Guangdong). Despite conducting research on the potential dissemination of the TCH strain within the context of Chinese indigenous isolates, we acknowledge that the advanced facilities, well-established monitoring systems, comprehensive staffing, and heightened scientific research awareness of the disease control departments in given regions, such as Beijing and Guangdong, have facilitated the collection of a relatively large number of samples from these regions. However, the relationships of isolates between Tianjin and other regions remain inconclusive, and delineating the contributions of strains from other regions to Tianjin has yet to be determined. Although the evolution of hospital-associated lineages was connected closely to environmental selection [5], we excluded the possibility of clonal spread and direct transmission based on the isolation time and wards. Our collection included a diverse range of dominant and sporadic STs and thus may represent regional clinical genotypes. Our collection also showed a general lack of ARGs against commonly used antibiotics and the pattern of virulence genes, both aligned with the recent report in the Japanese population [6]. Though the strains included in this study were isolated from a children population, mainly infants, the distribution of ARGs and virulence genes had no difference between age groups. Since the first publication of the whole genome sequence of E. faecalis in 2003, WGS has exhibited excellent performance in systematically describing this important pathogen [40]. Several studies managed to assign STs to E. faecalis isolates using WGS, which could effectively compensate for the discriminative power deficiency of traditional assays, construct more detailed genetic associations, and trace possible nosocomial transmission [5, 6]. WGS could also make accurate predictions about drug resistance, supported by the consistency of antibiotic resistance genotype and phenotype in both literature and our study. However, since our results suggested that local ST groups had predominantly diverse pathogenic features, like the over-representation of resistance and virulence genes in ST16/ST179 compared to ST745, MLST might still be considered as an accessible alternative to predict pathogenic traits of E. faecalis in clinical practice. The clinical detection rates of E. faecalis in China were high, but their antimicrobial resistance rates remained low, with vancomycin resistance reported at only 0.1% in 2022 [3]. The insufficient transmission evidence observed in our isolates indicates that infection prevention should focus more on personal hygiene, hospital disinfection, and reducing excessive use of broad-spectrum antibiotics. Our lower-than-average resistance rate and the absence of VRE highlighted the importance of antibiotic stewardship. Besides choosing the first-line antibiotics, policymakers should also consider the high prevalence of virulence genes when choosing infection control measures, especially those against nosocomial infections. To our knowledge, this study is the only comprehensive description of the gene content of clinical E. faecalis isolates in the northern Chinese city of Tianjin. The results provide a solid basis for local infection prevention and treatment. Our study also unveils the diverse pathogenic traits in different STs and promotes more precise infection management. Our study also has several limitations. Since TCH is a children’s hospital with 1,721 beds, the features of clinical isolates from this single hospital may be distinct from other epidemic populations, thus weakening our conjecture on the local genotype patterns of E. faecalis. In addition, the lack of within-hospital environmental isolates may restrict the identification of indirect transmission. Conclusion We applied WGS to characterize the antimicrobial susceptibility, virulence genes prevalence, and evolutionary potentials in local clinical E. faecalis isolates. The comprehensive genome-level data provides valuable insights for epidemiological studies, infection management, and the development of advanced treatment strategies. Supplementary Information Supplementary Material 1. Supplementary Material 1.