Enterotoxigenic Escherichia coli (ETEC) is a significant cause of diarrhea in both humans and animals (Francis, 2002; Esfandiari et al., 2017). Post-weaning diarrhea (PWD) typically manifests within 2 weeks following weaning, and is characterized by clinical signs including diarrhea, dehydration, and mortality. ETEC is the most frequently isolated pathogen from PWD outbreaks (Fairbrother et al., 2005), and is distinguished by its production of fimbriae and enterotoxins. It can be spread through feces, water, and food. Due to its high morbidity and mortality rates, it significantly affects piglet performance and causes significant economic losses in pig industry globally (Luppi, 2017; García-Meniño et al., 2021).

Although antimicrobial agents may provide temporary relief and help control the disease burden in piglets, their prophylactic and therapeutic use may lead to the establishment of antimicrobial-resistant bacteria in the gut microbiota, which poses a threat to both environment and public health. PWD is one of the most common reasons for the use of antimicrobial agents in the pig industry worldwide and contributes significantly to the problem of antimicrobial resistance (AMR) development (Luppi, 2017; Rhouma et al., 2017; Luise et al., 2019; Laird et al., 2021). Chromosomes and plasmids play crucial roles in the field of public health, particularly when it comes to antibiotic resistance and the presence of virulence genes. Genes responsible for virulence in E. coli strains can frequently be transmitted between them, residing either on chromosomes, plasmids, or phages. The emergence of antibiotic-resistant bacteria can be attributed to genetic mutations within chromosomes or the acquisition of genetic material, such as resistance plasmids (Bakkeren et al., 2019). AMR genes such as mcr, tet(X4), ESBL genes like bla_CTX-M and bla_TEM, and carbapenem-resistant Enterobacteriaceae (CRE) associated genes like bla_KPC and bla_NDM present substantial threats to public health (Faccone et al., 2019; Nguyet et al., 2022; Wang et al., 2022; Zhang et al., 2022).

ETEC from pigs produce two enterotoxins, namely, heat-labile enterotoxin (LT) and heat-stable enterotoxin (STa and STb), both of which are encoded by plasmids (Johnson and Nolan, 2009). There are distinct disparities in the LT and ST enterotoxins produced by porcine and human ETEC strains. Despite a high degree of nucleotide sequence similarity between the eltAB genes (pLT and hLT) and the estA genes (pSTa and hSTa), the absence of hLT and hSTa toxins in porcine ETEC strains causing diarrhea is noteworthy. Additionally, there are variations in epitopes and antigenicities between the toxins of porcine and human origin (Zhang et al., 2008). ETEC pathotypes in pigs are characterized by specific fimbrial adhesins that allow bacterial colonization of the gut mucosal surface. ETEC uses fimbriae to adhere to specific receptors on the target host cells and subsequently colonizes piglet small intestinal epithelial cells (Smith and Linggood, 1971). Once established in the intestine, ETEC secretes enterotoxins that disrupt fluid homeostasis, resulting in watery diarrhea by increasing secretion and inhibiting absorption in the intestine (Dubreuil et al., 2016; Sun and Kim, 2017). ETEC strains causing porcine diarrhea include five different fimbrial subtypes, namely K88 (F4), K99 (F5), F41, F18, and 987P (F6) (Isaacson, 1977; Isaacson and Richter, 1981; De Graaf and Roorda, 1982; Bakker et al., 1991; Nagy et al., 1997). Compared with ETEC strains that infect humans, the predominant adhesins found in ETEC strains isolated from pigs with PWD are fimbriae F4 and F18 (García et al., 2020). In contrast, fimbriae F5, F6 and F41 are typically associated with ETEC-induced neonatal diarrhea in piglets, and have little correlation with PWD (Frydendahl, 2002; Zhang et al., 2007). The fimbriae of ETEC have both immunogenic and antigenic properties and are the primary target for developing vaccines against this pathogen (Ntakiyisumba et al., 2022). However, due to the prevalence of mixed infections and the diversity of ETEC pilus types, it remains challenging to determine the specific types of pili carried by infected individuals, which presents a significant obstacle to effective disease prevention and control.

Shiga toxin-producing Escherichia coli (STEC) strains are globally recognized as significant causes of diarrhea in both humans and animals (Lee et al., 2023). These strains are known for their ability to produce one or two distinct types of Shiga toxin (Stx), which serve as the primary virulence factor. Shiga toxins, classified as Stx1 and Stx2, play a crucial role in the pathogenicity of STEC (Melton-Celsa, 2014). Previous studies have shown that ETEC strains, capable of producing Stx toxins, are often classified as ETEC/STEC hybrid strains (Fairbrother et al., 2005).

Whole-genome sequencing (WGS) has become increasingly accessible to clinical and microbiological laboratories in recent decades due to a decrease in cost. This technology is a powerful tool for analyzing bacterial genomes rapidly (Kwong et al., 2015). In this study, WGS technology was employed in conjunction with sophisticated bioinformatics analysis to comprehensively investigate a multitude of pivotal aspects pertaining to ETEC isolates. The investigative scope encompassed an in-depth exploration of serotype prevalence, antimicrobial resistance phenotypes, antimicrobial resistance genes (ARGs), virulence factors (VFs), and plasmid incompatibility types. Furthermore, a comparative analysis was conducted to elucidate the phylogenetic relationship between ETEC strains under investigation and ETEC strains originating from diverse geographic regions worldwide.

To enhance understanding of porcine ETEC, it is crucial to monitor its entire genome. However, the current knowledge is hindered by some gaps and limitations. The gaps include limited knowledge of genomic diversity, detection of different serotypes in different regions, geographical representation in public databases, and comparative analysis with human ETEC strains. This hampers the ability to accurately assess global genetic diversity, represent various lineages and regional variations, and may result in a geographic bias in available genomes. The primary objective of this study was to achieve a thorough genomic characterization of ETEC strains isolated in China, thus contributing reference data that can significantly inform the development and implementation of targeted disease control strategies.

Diarrhea resulting from ETEC infection is prevalent among pigs and imposes significant economic burdens to the pig industry globally. The ensuing diarrhea in piglets is largely attributed to ETEC and can have detrimental effects on growth and production performance, presenting a significant technological hurdle for contemporary swine production (Gresse et al., 2017). However, despite its importance, the genomic characteristics and variation of porcine-origin ETEC have been poorly investigated. To address this gap, WGS was applied to identify ETEC isolates from pig feces samples obtained from large-scale farms in Sichuan, China. Our study aimed to clarify the serotypes, AMR, ARGs, VFs, and plasmid incompatibility types of ETEC associated with diarrhea in pigs.

ETEC strains have significant implications for both animal and human health. In pigs, ETEC colonization can lead to severe enteric diseases, resulting in adverse impacts on growth rates, feed conversion, and economic losses in the livestock industry. Similarly, in humans, ETEC is known to cause acute diarrhea, particularly affecting vulnerable populations like infants, young children and individuals in developing regions with limited access to clean water and sanitation facilities. Notably, ETEC is a prominent causative agent of traveler’s diarrhea, posing a substantial threat to human health (Guiral et al., 2019). Effective measures to mitigate ETEC transmission encompass improved sanitation practices, access to clean water, and adherence to proper hygiene protocols for both animal and human populations. Furthermore, implementing targeted vaccination programs that specifically address ETEC strains can significantly contribute to infection prevention and overall reduction in disease burden. Hence, it is of paramount importance to gain insights into the AMR profiles and complete genomic characteristics of ETEC.

In the present study, the predicted dominant serotypes in large-scale pig farms in Sichuan Province of China were O105, O141, H8, H10, H16 and H29, which differed from previous studies conducted in other regions of China. For example, in large-scale pig farms in northern Jiangsu, the dominant serotypes of ETEC were O8, O101, and O128 (Yang et al., 2020). Similarly, in certain areas of Anhui Province, pig farms showed a prevalence of O8, O20, and O128 as the predominant serotypes of ETEC, with generally high levels of virulence (Ren et al., 2020). These results indicated that ETEC with a diverse group of serotypes were circulating in pigs, with significant diversity in serotypes observed both within and across regions. Furthermore, our WGS analysis revealed the distribution of ST100 among MDR ETEC isolates from pig farms, which was considered to be a classic strain of porcine ETEC (non-zoonotic animal pathogen) (Abraham et al., 2015). Existing studies have demonstrated that the O149 serogroup (ST100) is the predominant cause of PWD in Australian pigs (Abraham et al., 2014). The variety of serotypes in ETEC is significant for disease severity, surveillance and vaccine development. It helps track strain spread, identify outbreaks, and select relevant antigens for vaccines targeting specific serotypes.

The analysis encompassed a comprehensive evaluation of 121 ETEC isolates worldwide utilizing cgMLST analysis, a highly precise method for subtype characterization based on genomic variations. The minimum spanning tree, derived from this analysis, provided insights into the genetic relatedness among ETEC isolates. However, the analysis failed to reveal distinct patterns of relatedness, as different STs were dispersed across various clusters. Furthermore, the same cluster accommodated isolates originating from diverse sources, including animals, environment, food and humans. These results indicated a wide distribution of ETEC across multiple reservoirs and transmission routes, with limited association between the genetic diversity of ETEC isolates and their geographical or temporal origins.

The absence of conspicuous patterns in the cgMLST data may be attributed to a variety of underlying factors, collectively contributing to the elusive nature of the observed relationships. Primarily, the substantial genetic diversity prevalent among ETEC isolates unquestionably played a pivotal role in the intricate network of associations within their genomic profiles. ETEC displays notable genetic diversity that emerges from mechanisms like mutations and genetic rearrangements. Consequently, distinct sets of genes, including those encoding VFs or AMR, can be found among ETEC strains. The extensive range of genetic variations posed significant challenges in deciphering the definitive relationships among these isolates. Moreover, the recurrent occurrence of horizontal transfer events involving mobile genetic elements further complicates the genomic makeup of ETEC isolates. These mobile genetic elements have the capacity to move across bacterial strains, leading to a mosaic-like configuration of genomes. This facilitates the exchange of genetic material, including VFs and ARGs, thereby shaping the genetic composition of ETEC populations. These genetic exchange events tended to blur the demarcations between different lineages, thereby obscuring any anticipated straightforward associations. Furthermore, the complex evolutionary history of this pathogenic microorganism, sculpted by diverse selective pressures, potentially exerts a profound impact on the observed genetic patterns. Adaptation plays a key role, driving the development of traits that enhance survival and reproduction in specific environments. In the case of ETEC, adaptation enables colonization of host environments, evasion of immune responses, and acquisition of nutrients. Genetic drift, a stochastic process, introduces random fluctuations in allele frequencies over time, leading to the divergence of lineages and contributing to genetic diversity. The combined effects of adaptation, genetic drift and other forces ultimately contribute to the observed genetic patterns, influencing the evolution and diversity of ETEC populations. Consequently, the resulting intricate genetic architecture posed a significant obstacle in identifying unambiguous associations within the cgMLST data.

In vitro antimicrobial susceptibility testing is a crucial tool providing guidance for veterinary clinical practice. The detection of MDR ETEC isolates from pigs with diarrhea on farms raised a concerning situation of AMR. The study demonstrated that all ETEC isolates were MDR. In comparison, the resistance rates of ETEC strains collected in Bangladesh to ampicillin (66%), cotrimoxazole (46%), doxycycline (44%), erythromycin (96%), nalidixic acid (83%), streptomycin (48%) and tetracycline (42%) differed to some extent from our findings (Begum et al., 2016). Geographic variation in antimicrobial usage and resistance patterns is a likely explanation, as different regions may have different levels of antibiotic use in agriculture, leading to variations in the prevalence and types of resistant bacteria. Local antibiotic prescription practices and veterinary antibiotic use also contribute to these differences. Additionally, factors such as the specific strains of ETEC circulating in different populations, the genetic mechanisms of resistance acquired by the bacteria, and variations in surveillance methods used in different investigations can contribute to the disparities in resistance rates observed. Despite the ban on the usage of CHL in animal production in China since 2002 (Peng et al., 2022), CHL-resistant isolates were found on pig farms in Sichuan in this study, potentially attributed to continued use of FFC, a member of the chloramphenicol class (Yang et al., 2019). In contrast, IPM-resistant isolates were not found in the study, most likely due to the official policy prohibiting the use of carbapenem in livestock in China. Although the ban on AMK usage in agriculture in 2017 resulted in a low percentage (10.53%, 2/19) of AMK-resistant isolates on pig farms, the high PMB resistance rate (78.95%, 15/19) of ETEC isolates in this study indicated that drug misuse in pig industry is still of a significant concern (Shen et al., 2020; Wang et al., 2020).

Fortunately, despite the majority of strains showing resistance to PMB, mcr-1 gene was not observed in the present study, which was first identified in 2016 in bacterial isolates from animal foods (chickens and pigs) in China (Liu et al., 2016). This plasmid-mediated resistance gene encodes an enzyme that belongs to the phosphoethanolamine transferase family and is responsible for synthesizing and conjugating pEtN to lipid A (Ayoub Moubareck, 2020). Other mcr genes, including mcr-2 and mcr-3, have also been found in plasmids among enterobacteria (Zhang et al., 2018). These resistance genes suggested the presence of various pathways for the horizontal transmission of colistin resistance, which could lead to its high potential for dissemination. Based on our results, it was suspected that the PMB-resistance observed in the ETEC isolates may not be attributed to horizontal transfer. Longitudinal studies in China found that reduced colistin sales were associated with decreased occurrence of mcr-1-producing E. coli in pig feces and human intestines. Human infections with colistin-resistant E. coli also declined after 2017, possibly due to the Chinese government’s ban on colistin as an animal growth promoter (Ewers et al., 2022).

It is worth noting that we detected the presence of tet(X4) in 2 (10.53%) isolates of ETEC. According to previous research, E. coli strains carrying tet(X4) may pose a risk of transmission to humans, as they exhibited a significant genetic resemblance to human commensal E. coli. Although tet(X4) was not extensively distributed across the country, it was highly endemic in northwestern China (Sun et al., 2020). Among 19 ETEC isolates, tet(A) was the most frequently detected ARGs, accounting for 68.42%. One potential explanation for the relatively low prevalence of tigecycline-resistance gene was the presence of TET-related ARGs, such as tet(A), tet(B), tet(C), and/or tet(M), which are often carried by E. coli strains from animals (Foong et al., 2020). The global spread of colistin-resistant E. coli and CRE poses major public health challenges. Consequently, tigecycline has become a crucial last-resort antibiotic for treating certain bacterial infections (Zhang et al., 2022). Given the emergence of tigecycline resistance mediated by tet(X4), our findings underscored the urgent need for heightened surveillance and intervention measures to prevent a potential dissemination of this ARGs.

Identifying the VFs of ETEC in pigs is crucial for understanding its pathogenicity and transmission, as well as developing effective prevention and treatment strategies. This study identified several essential genes, including cgsG, elfG, hcpB, hcpC, papC, fimD, ibeB, ibeC, espL1, espL4, espR1, espX1, and espX5, that play a significant role in pathogenesis and virulence. The presence of these genes in all ETEC isolates demonstrated their critical role in ETEC virulence. Furthermore, the investigation unveiled the capacity of the ETEC isolates to potentially produce either ST alone or both LT and ST toxins concurrently. Additionally, this observation underscored the significance of multiple virulence genes, such as eltA, eltB, eltIAB-4, estap-STa1, estap-STa6, estb-STb1, and astA, in potential toxin synthesis. STa has been implicated in human disease, while STb is predominantly associated with ETEC infection in pigs, although the presence of the stb gene has been observed in human ETEC isolates. Previous studies have shown that STa4 is frequently linked to diarrhea (Leonard et al., 2016). Plasmids are the predominant site for the presence of STa (von Mentzer et al., 2014), and the transfer of plasmids carrying ST toxin genes sta between different strains of E. coli is possible (Turner et al., 2006). Two distinct genetic variants of STa, namely STp and STh, are identified in human ETEC isolates. The findings indicate that STa3/4 (STh)-producing strains are correlated with disease in children, while STa5 (STp) strains are associated with disease in adults (Joffré et al., 2016). Extensive investigations have established the presence of astA gene, which encodes the enteroaggregative heat-stable enterotoxin EAST1 in ETEC linked to the occurrence of diarrheal disease in swine (Renzhammer et al., 2020). However, the precise role of this toxin in the disease remains a subject of ongoing debate, as neither EAST1-positive strains nor culture supernatants consistently induce explicit diarrhea in animal models (Dubreuil, 2019). Notably, a significant proportion of the strains examined in this study, specifically 57.89% (11/19) of the analyzed isolates, exhibited the presence of astA gene. This finding highlighted the widespread occurrence of this critical genetic determinant among ETEC isolates. Adhesin-related VFs, such as ECP, ELF, and HCP, also contributed to ETEC pathogenicity by interacting with enterotoxin, while Shiga-like toxin-related virulence factors, including stx2a, stx2b, and stx2e-ONT-2771, were identified with enterotoxic, cytotoxic and neurotoxic effects (Ahsan et al., 2020). In addition, the emergence of four hybrid isolates were confirmed in this study, which may be attributed to the frequent incidence of horizontal gene transfer (HGT) events. During HGT, specific genetic backgrounds acquire pathotype-defining virulence genes that are frequently harbored on mobile genetic elements (Leonard et al., 2016). Stx2e, in particular, is a crucial toxin responsible for edema disease in piglets, which often accompanies piglet diarrhea (Tseng et al., 2014). Hence, analyzing piglet symptoms alone is inadequate for diagnosing the disease accurately. A comprehensive investigation of ETEC VFs can provide an in-depth examination for the development of effective diagnostic methods and multivalent vaccines with extensive protection against this pathogen and other similar bacterial infections.

The acquisition of ARGs is facilitated by a diverse array of mobile genetic elements. Among these elements, plasmids are particularly significant in promoting ARGs dissemination (Bennett, 2008). The assessment of plasmid diversity in diverse cohorts can be achieved by utilizing plasmid incompatibility groups. The present study revealed significant differences in incompatibility groups prevalent in large-scale pig farms in southwestern China. The dissemination of epidemic plasmids has been linked to specific genes, and in this study, common replicators such as IncFIB, IncFII, and IncX1 were identified. Furthermore, certain groups of these plasmids, such as IncC, IncFIB, IncX1, IncFIIA, and IncY, have the ability to carry transfer, multidrug resistance and virulence functions across a broad range of hosts (Johnson and Nolan, 2009). IncFII has been recognized as a plasmid that disseminates multiple drug resistance (Wagner et al., 2014), and IncF is implicated in the transmission of drug resistance and virulence genes of Enterobacteriaceae (Hopkins et al., 2006). The possibility of transmission of MDR E. coli strains, especially those that produce extended-spectrum β-lactamases (ESBLs), from animals to humans is a growing concern (Jackson et al., 2015). IncC and IncX plasmid types have been recently reported as the most common among carbapenem-resistant E. coli (Kopotsa et al., 2019). This study found that the presence of the IncC plasmid was exclusive to hybrid strains, which was in line with the existence of tet(X4) gene. It was previously reported that the existence of a 119-kb multi-virulence IncFII/IncX1 hybrid ETEC/STEC plasmid (p15ODTXV) that carried virulence genes of both ETEC and STEC pathotypes played a pivotal role in the emergence of hybrid pathotypes, emphasizing the significance of plasmids in this mechanism (Brilhante et al., 2019). In this study, a comparable multi-virulence IncFII/IncX1 hybrid ETEC/STEC plasmid in the hybrid strain was clarified, providing additional evidence for the significant role of plasmids in the emergence of hybrid pathotypes. The transfer and transmission of plasmids played a critical role in the dissemination of virulence and drug resistance genes in pathogenic E coli. Plasmid transmission from animals to humans can lead to the spread of AMR, zoonotic infections, increased virulence and challenges in treating ETEC infections. Thus, understanding the incompatibility groups of porcine ETEC plasmids can inform the development of strategies to control and prevent ETEC infections in pigs.

In this study, an investigation was employed to analyze ETEC isolates from pigs with diarrhea in China. ETEC isolates exhibited notable levels of multidrug resistance, with the emergence of mobile tet(X4) genes in specific hybrid isolates, posed a challenge to swine industry and public health. Moreover, the study identified multiple plasmids capable of facilitating the horizontal transfer of multidrug resistance and virulence genes across diverse host organisms, which has the potential to exacerbate the dissemination of resistance traits, thereby posing a substantial threat to the swine industry. In addition, the strains under investigation were found to harbor various VFs associated with the pathogenesis and severity of animal diseases. The absence of conspicuous patterns in the cgMLST data among the ETEC isolates, suggested their wide distribution across different reservoirs and transmission routes. These results demonstrated the urgent need for intensified surveillance measures to effectively monitor the emergence, resistance and potential transmission of ETEC among porcine herds in China.

The datasets generated for this study can be found in the NCBI Bioproject with the accession number PRJNA983238.

JH: investigation, software, data curation, writing-original draft, and funding acquisition. JL: formal analysis. XHu and QZ: methodology. JX, MC, YX, and SC: writing-review and editing. YH and YW: visualization. LZ: methodology, supervision, review and editing, and funding acquisition. XHa: conceptualization, project administration, resources, and funding acquisition. All authors contributed to the article and approved the submitted version.

This work was supported by the grant from Natural Science Foundation of Sichuan, China (2023NSFSC0175) and Sichuan Science and Technology Program (2022ZDZX0017).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.