Phenotypic and Molecular Characterization of Carbapenem-Resistant Escherichia coli Isolated from Retail Meats in Hat Yai, Thailand

Abstract

(1) Background: Carbapenem-resistant Escherichia coli (CREC) is widespread and resistant to almost all available antimicrobial agents. In this study, we aimed to assess the phenotypic and molecular characteristics of CREC isolated from retail meats in Hat Yai, Songkhla, Thailand. (2) Methods: A total of 155 retail meat samples were randomly collected, and 412 presumptive carbapenem-non-susceptible isolates were screened via culturing on imipenem-containing eosin methylene blue (EMB) agar. Susceptibility to imipenem and meropenem was tested using the disk diffusion method, and carbapenemase and virulence genes in CREC isolates were detected using PCR. Phylogenetic groups and genetic relatedness of carbapenemase-positive CREC isolates were analyzed using gene markers and BOX-PCR, respectively. (3) Results: The results revealed a high prevalence of presumptive carbapenem-non-susceptible E. coli (CNSEC) isolates in beef samples. Over 89% of the CNSEC isolates from all meat types were identified as CREC. Of these, only 4.8% of the isolates from beef samples were positive for the bla_NDM gene, and one was also positive for the bla_VIM gene. These isolates carried only the fimH gene as a virulence factor. The bla_NDM-positive CREC isolates were classified in phylogenetic Group D. (4) Conclusions: Identifying antimicrobial-resistant pathogens, particularly CREC, in food-producing animals is critical due to potential risks to public health.

1. Introduction

Carbapenem resistance in Enterobacterales has become a global health problem; it is often associated with resistance to multiple antimicrobial agents, complicating treatment options and leading to high morbidity and mortality rates [1,2]. It is caused by various mechanisms, including the production of carbapenemases, which can degrade almost all β-lactams [3]. In Enterobacterales such as E. coli, the most predominant carbapenemase-encoding genes include bla_KPC, bla_VIM, bla_IMP, bla_NDM, and bla_OXA-48 [4,5,6,7,8,9,10], which represent different molecular classes that determine their unique characteristics. The bla_KPC gene is a class A serine carbapenemase, strongly inhibited by β-lactamase inhibitors, and hydrolyzes almost all β-lactams. The bla_VIM, bla_IMP, and bla_NDM genes are class B metallo-β-lactamases (MBLs), which rely on zinc for activity, hydrolyze all β-lactams except aztreonam, and are not inhibited by traditional β-lactamase inhibitors. Finally, the bla_OXA-48 gene is a class D oxacillinase that causes resistance by weakly hydrolyzing carbapenems, often sparing expanded-spectrum cephalosporins (ESBLs), which can lead to delayed detection in clinical settings. A previous study reported that the bla_KPC gene was the most frequent carbapenemase gene in carbapenem-non-susceptible Enterobacterales (CNSE) isolated from patients in US hospitals, followed by bla_NDM, bla_IMP, bla_VIM, and bla_OXA-48 [7]. Crucially, ceftazidime/avibactam was the most active drug overall, inhibiting 96.9% of CNSE isolates, including almost all KPC producers. However, the discovery of a new bla_KPC variant (bla_KPC-58) that causes ceftazidime/avibactam resistance signals the continued evolution of resistance, despite a slight overall decline in carbapenemase prevalence during the study period. Other mechanisms involve the overexpression of efflux pumps, the loss of porin activity, and the combination of these mechanisms with the expression of bla_ampC and/or the production of ESBLs [4,5,6,7,8,10]. Most genes encode these mechanisms through plasmids or other mobile genetic elements (MGEs) such as transposons and integrons, which can be transferred to other cells or bacterial species [10]. Horizontal gene transfer (HGT) is one of the important factors causing the spread of antimicrobial resistance (AMR) in bacteria, especially Gram-negative pathogens.

Besides the AMR, numerous virulence determinants of E. coli further complicate its management. Generally, E. coli can be categorized into diarrheagenic E. coli (DEC) and extraintestinal pathogenic E. coli (ExPEC), each group harboring distinct sets of virulence-associated genes [11,12,13]. In the six categories of DEC, the major virulence determinants, namely stx₁/stx₂, bfpA, eae, aggR, ipaH, elt/est, and daaE, encode Shiga toxin 1/Shiga toxin 2, bundle-forming pili, intimin, the transcriptional activator AAF/I, enteroinvasive protein, heat-labile/heat-stable enterotoxins, and F1845 fimbriae, respectively [14]. These genes play crucial roles in mediating bacterial adhesion, invasion, and subsequent destruction of host epithelial cells. Furthermore, the ExPEC-associated indicator genes, namely papA, papC, sfaDE, afa, kpsMTII, and iutA, encode the P fimbrial structural subunit, outer membrane usher protein, S fimbriae, Afa adhesin, capsular antigen, and the siderophore aerobactin, respectively; these genes also have important roles in the pathogenesis of extraintestinal infections [15]. Additionally, fimH, cnf1, lpf, and hlyA, which encode the type I fimbrial tip, cytotoxic necrotizing factor 1, long polar fimbriae, and α-hemolysin, respectively, can further exacerbate bacterial virulence, leading to more severe disease outcomes. The presence of these virulence genes collectively enhances the pathogenic potential of E. coli, emphasizing its remarkable clinical significance [13,16].

Carbapenem-resistant Enterobacterales (CRE) are considered an urgent threat by the Centers for Disease Control and Prevention (CDC) [17]. In 2019, the CDC reported an estimated 13,100 cases of CRE infections in hospitalized patients and 1100 related deaths in the US. A high estimated healthcare cost was also associated with these infections [17]. Carbapenem-resistant Klebsiella pneumoniae (CRKP) and carbapenem-resistant Escherichia coli (CREC) are the most frequent pathogens among CRE [17]. Several studies have examined the prevalence of CRE in Thailand. National antimicrobial resistance surveillance, Thailand (NARST), reported that the resistance rates of imipenem, meropenem, and ertapenem in CREC during 2000–2022 were 1% to 5.3%, 0.6% to 5.3%, and 0.6% to 6.3%, respectively. A study conducted at Siriraj Hospital, one of Thailand’s largest hospitals, reported a 2-year surveillance (2009–2011) of CRE clinical isolates. Of 12,741 non-duplicated Enterobacteriaceae isolates, 181 (1.4%) were classified as CRE, with 5 identified species, including Enterobacter cloacae, K. pneumoniae, E. coli, and Citrobacter freundii [4]. In addition to those in the clinical setting, CRE isolates identified in foods, animals, and environments are also of concern. Although carbapenems are not typically used in livestock, ESBL-producing CRE are common in veterinary settings [18]. Global antimicrobial use in food-producing animals is expected to reach 174,549 tons by 2030, with intensive use creating selective pressure that drives AMR [18]. This resistance can spread through the food chain and the environment, as livestock may acquire it from contaminated water or biological vectors [18]. Additionally, antibiotics can alter gut microbiota, increasing the diversity of AMR genes [18]. According to previous evidence, CRE, especially CREC, can transfer from animals to humans via food and the environment [18]. Several investigations have demonstrated a positive correlation between livestock production systems and the occurrence of CREC infections in humans. One study demonstrated the direct transmission of New Delhi metallo-β-lactamases-type carbapenem-resistant E. coli (NDM-EC) between humans and animals [19], while zooanthroponotic transmission of CREC from humans to companion animals has also been documented [20]. The presence of carbapenem-non-susceptible E. coli (CNSEC), which includes isolates exhibiting intermediate resistance or resistance to carbapenem, in animal-sourced foods is a significant concern for food safety and public health [21,22]. Thus, in this study, we aimed to investigate the phenotypic and molecular characteristics of CREC in retail meats from an open market in Hat Yai, Songkhla, Thailand.

2. Materials and Methods

2.1. Sample Collection and CNSEC Screening

Sampling was performed using a stratified random approach to ensure representative coverage of retail meat sources within the Hat Yai district, Songkhla, Thailand. Eight vendors located in major fresh markets were selected as independent sampling units to capture geographical and commercial variability. From each vendor, meat samples were randomly collected at regular weekly intervals over a 10-month period (August 2018 to May 2019). The sampling schedule and vendor selection were designed to minimize temporal and spatial bias, ensuring that collected specimens reflected routine market conditions across the district. A total of 155 raw meat samples (beef = 53; pork = 51; and chicken = 51) were collected and processed following a protocol described in a previous study [23]. Briefly, 10 g of meat was soaked in 90 mL of tryptic soy broth (TSB, BD Difco, Sparks, MD, USA). Afterwards, the solution was mixed and incubated at 37 °C for 6 h, then serially diluted 10-fold, and subsequently spread on eosin methylene blue (EMB, BD Difco, Sparks, MD, USA) agar supplemented with imipenem (final concentration of 2 µg/mL) to screen for CNSEC [24]. The plate was then incubated at 37 °C for 18 h. One to fifteen green metallic sheen colonies were selected to increase the probability of detecting presumptive CNSEC isolates and kept at −80 °C in 10% glycerol (final concentration) [25].

2.2. Antimicrobial Susceptibility Testing

Antimicrobial susceptibility in the presumptive CNSEC isolates was examined using a disk diffusion test (Kirby–Bauer Method). Briefly, a single colony was picked, inoculated in 3 mL of Mueller–Hinton broth (MHB, BD Difco, Sparks, MD, USA), and incubated at 37 °C for 3 h with shaking at 150 rpm. Subsequently, the bacterial culture was adjusted to 0.5 McFarland standards with sterile normal saline solution (approximately 1.5 × 10⁸ colony-forming units (CFU) per ml). The bacterial solution was swabbed onto the surface of Mueller–Hinton agar (MHA, BD Difco, Sparks, MD, USA). Antimicrobial disks (Oxoid, Basingstoke, Hampshire, UK), including ceftazidime (CAZ, 30 µg), cefotaxime (CTX, 30 µg), meropenem (MEM, 10 µg), and imipenem (IPM, 10 µg), were placed on the surface of MHA and incubated at 37 °C for 18 h. E. coli ATCC 25922 was used as a quality control. The diameters of inhibition zones were measured using a vernier caliper and interpreted according to the clinical breakpoints, according to CLSI guidelines [26].

The overall proportion of presumptive CNSEC isolates was determined by dividing the number of isolates that exhibited a non-susceptible phenotype (intermediately resistant or resistant) to at least one tested carbapenem (imipenem or meropenem) by the total number of E. coli colonies tested from each source. The proportion was calculated using the following formula:

$$\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{o}\mathrm{f}\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{C}\mathrm{N}\mathrm{S}\mathrm{E}\mathrm{C}\mathrm{i}\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{s}={\displaystyle \frac{\mathrm{A}}{\mathrm{B}}}\times 100\%$$

where

• A: The number of isolates confirmed as CNSEC (intermediately resistant or resistant to any carbapenem)
• B: The total number of E. coli isolates tested from the source.

2.3. DNA Extraction

The bacterial DNA of each presumptive CNSEC isolate was extracted using the boiling method [23]. Briefly, a single colony was inoculated into 1 mL of TSB and incubated at 37 °C for 3 h with shaking at 150 rpm. The bacterial culture was boiled at 100 °C for 10 min and immediately cooled on ice for 5 min. Then, it was centrifuged at 11,000× g for 5 min. A total of 10 µL of supernatant was collected and transferred into a sterile 1.5 mL microcentrifuge tube containing 90 µL of sterile deionized water. The DNA solution of each isolate was used as a DNA template for further experiments. The universal primers (27F, 5′-AGAGTTTGATCCTGGCTCAG-3′, and 1492R, 5′-CTACGGCTACCTTGTTACGA-3′) targeting the 16S rRNA gene were employed as an internal control to assess the integrity and adequacy of the extracted genomic DNA for subsequent PCR analyses [27]. A 25 µL reaction mixture contained 3.0 mM MgCl₂, 0.1 mM dNTPs, 0.4 µM of each primer, 0.5 U of GoTaq DNA polymerase (Promega, Madison, WI, USA), 1× GoTaq Flexi Green buffer, and 2 µL of DNA solution. Thermocycling was performed in a T100™ Thermal Cycler (Bio-Rad, Hercules, CA, USA) under the following conditions: an initial denaturation at 95 °C for 3 min; 35 cycles of 94 °C for 1 min, 50 °C for 1 min, and 72 °C for 1 min; and a final extension at 72 °C for 5 min.

2.4. Confirmation of E. coli Strain

The carriage of the uidA gene, specifically for the identification of E. coli, was investigated in all presumptive CNSEC isolates using the polymerase chain reaction (PCR) method [28]. The PCR mixture contained 1× of GoTaq Flexi green buffer, 0.1 mM of deoxynucleotide triphosphates (dNTPs), 3.0 mM of MgCl₂, 0.4 µM of forward and reverse primers, 0.5 units of GoTaq DNA polymerase (Promega, Madison, WI, USA), and 2 µL of DNA template. A pair of uidA primers and the PCR conditions are shown in Tables S1 and S2, respectively. The uidA gene was amplified using a T100^TM Thermal Cycler (Bio-Rad, Hercules, CA, USA). Afterwards, the amplicons were analyzed using 1% agarose gel electrophoresis at 100 V for 40 min. The gel was stained with ethidium bromide and visualized using the WSE-5200 Printpraph 2M gel imaging system (ATTO Corp., Tokyo, Japan).

2.5. Detection of Carbapenemase Genes in CNSEC Isolates

To understand the primary mechanisms of carbapenem resistance in CNSEC isolates, the most common carbapenemase genes (bla_KPC, bla_IMP, bla_VIM, bla_NDM, and bla_OXA-48) were investigated using the uniplex-PCR method [29]. The uniplex-PCR mixture was prepared, the genes were amplified, and the results were observed as described above. Tables S1 and S2 illustrate all pairs of primers and their PCR conditions. The representative PCR amplicons of the positive genes were further sequenced using Sanger sequencing. Afterwards, the sequences were searched against the National Center for Biotechnology Information (NCBI) database using BLASTn (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastn&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome, accessed on 11 November 2024).

2.6. Detection of Virulence Genes in CREC Isolates

The virulence factors contributing to the pathogenicity of E. coli were assessed. In the classification of diarrheagenic E. coli (DEC), the indicator genes were identified using the PCR method [30,31,32,33,34,35,36,37,38,39,40,41,42,43]. The genes were grouped into 6 pathogenic categories: category 1: stx and eae for enterohemorrhagic E. coli (EHEC); category 2: bfp and eae for typical enteropathogenic E. coli (tEPEC) or eae alone for atypical enteropathogenic E. coli (aEPEC); category 3: elt or est for enterotoxigenic E. coli (ETEC); category 4: aggR for enteroaggregative E. coli (EAEC); category 5: ipaH for enteroinvasive E. coli (EIEC); category 6: daaE for diffusely adherent E. coli (DAEC). Meanwhile, the extraintestinal pathogenic E. coli (ExPEC) indicator genes were also investigated. Six target genes were categorized into Group 1 (papA and/or papC), Group 2 (sfaD), Group 3 (afa), Group 4 (kpsMTII), and Group 5 (iutA). For interpretation, the CNSEC isolates with positive results for two or more groups were considered ExPEC [44]. Furthermore, the other 6 virulence genes, astA, agn43, cnf1, hly, fimH, and lpf, were also examined. The uniplex-PCR mixture was prepared, all 20 virulence genes were amplified, and the results were observed as described above. All pairs of primers and their PCR conditions are illustrated in Tables S1 and S2, respectively.

2.7. Determination of Phylogenetic Groups

The phylogenetic groups of carbapenemase-positive CREC isolates were determined using the PCR method targeting 3 genetic markers (chuA, yjaA, and TSPE4.C2). The specific interpretation scheme was followed rigorously according to Clermont et al. [45], as shown in Figure S1. The PCR mixture was prepared, the genes were amplified, and the results and PCR products were analyzed as described above. Tables S1 and S2 illustrate all pairs of primers and their PCR conditions.

2.8. Analysis of Genetic Relatedness

In addition to classifying phylogenetic groups, the DNA fingerprints of carbapenemase-positive CREC isolates were also analyzed using BOX-PCR [46]. The genomic DNA (gDNA) of all carbapenemase-positive CREC isolates was extracted using the miniprep spin column method (Geneaid, Taipei, Taiwan). The PCR mixture contained 0.2 µM of BOXA1R primer (5′-CTACGGCAAGGCGACGCTGACG-3′), 3.0 mM of MgCl₂, 1× GoTaq Flexi green buffer, 1.25 units of GoTaq DNA polymerase, 0.2 mM of dNTPs, and a DNA template [47]. Amplification conditions included an initial denaturation step at 95 °C for 3 min; this was followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 50 °C for 1 min, extension at 65 °C for 8 min, and final extension at 72 °C for 5 min. Afterwards, the amplicons were observed via 1.5% agarose gel electrophoresis at 80 V for 60 min. The gel was stained and visualized as described above. The resultant banding profiles (genomic fingerprints) were captured digitally. The digitalized gel images were normalized using a molecular weight marker ladder. The similarity between the banding profiles was determined using BioNumerics software version 7.0 (Applied Maths, St-Martens-Latem, Belgium).

3. Results

3.1. Prevalence of Presumptive CNSEC Isolates

The prevalence of suspected CNSEC is illustrated in

Table 1. Prevalence of presumptive carbapenem-non-susceptible Escherichia coli isolates in raw meats from August 2018 to May 2019.

Sample Source	No. of Positive Samples/No. of Total Samples (%)	No. of Isolates
Beef	30/53 (56.6)	185
Pork	30/51 (58.8)	139
Chicken	20/51 (39.2)	88
Total	80/155 (51.6)	412

. Of 155 meat samples, suspected CNSEC was found in 56.6% (30/53) of beef, 58.8% (30/51) of pork, and 39.2% (20/51) of chicken. Among these positive beef, pork, and chicken samples, 185, 139, and 88 isolates were obtained, respectively.

3.2. Antimicrobial Resistance Profiles

The antimicrobial susceptibility profiles of suspected CNSEC isolates are shown in

Table 2. Antimicrobial susceptibility in presumptive carbapenem-non-susceptible Escherichia coli isolates.

Antimicrobial Agents	Beef			Pork			Chicken
	S (%)	I (%)	R (%)	S (%)	I (%)	R (%)	S (%)	I (%)	R (%)
Cefotaxime	149 (80.5)	15 (8.1)	21 (11.4)	106 (76.3)	29 (20.9)	4 (2.9)	72 (81.8)	11 (12.5)	5 (5.7)
Ceftazidime	155 (83.8)	13 (7.0)	17 (9.2)	106 (76.3)	29 (20.9)	4 (2.9)	80 (90.9)	8 (9.1)	0 (0.0)
Imipenem	55 (29.7)	13 (7.0)	117 (63.2)	3 (2.2)	16 (11.5)	120 (86.3)	4 (4.5)	11 (12.5)	73 (83.0)
Meropenem	84 (45.4)	44 (23.8)	57 (30.8)	31 (22.3)	57 (41.0)	51 (36.7)	20 (22.7)	34 (38.6)	34 (38.6)

S, susceptible; I, intermediately resistant; R, resistant.

and Table S3. Regarding susceptibility to third-generation cephalosporins, it was found that 11.4%, 2.9%, and 5.7% of suspected CNSEC from beef, pork, and chicken, respectively, were resistant to cefotaxime, and 9.2% and 2.9% from beef and pork were resistant to ceftazidime. The majority of these isolates were susceptible to both cefotaxime and ceftazidime. Regarding carbapenem susceptibility, 7.0%, 11.5%, and 12.5% of suspected CNSEC isolates from beef, pork, and chicken were intermediately resistant to imipenem, and 23.8%, 41.0%, and 38.6% were intermediately resistant to meropenem. Moreover, 63.2%, 86.3%, and 83.0% of the strains from beef, pork, and chicken were resistant to imipenem, and 30.8%, 36.7%, and 38.6% were resistant to meropenem. The CNSEC isolates showing resistance to at least one of the third-generation cephalosporins (cefotaxime and ceftazidime) were considered to presumptively produce ESBLs [48]. This classification was based on the fact that this phenotypic resistance profile serves as a screening marker, although we acknowledge that non-susceptibility mediated by other mechanisms (e.g., AmpC overexpression, porin loss, and efflux pump activation) may lead to false-positive presumptions. Furthermore, the suspected CNSEC isolates showing intermediate resistance or resistance to at least one carbapenem were phenotypically confirmed as CNSEC strains, and those susceptible to both imipenem and meropenem were classified as carbapenem-susceptible E. coli (CSEC). The total numbers of isolates from beef, pork, and chicken were 185, 139, and 88, respectively. Among all sample sources, mostly CNSEC strains (defined as non-susceptible to at least one carbapenem) were found in pork (95.0%), followed by chicken (89.8%) and beef (59.5%), while 10.8%, 5.7%, and 4.3% of isolates from beef, chicken, and pork were classified as both CNSEC and ESBL-PEC (Figure 1). More importantly, 90.0%, 89.6%, and 89.1% of the suspected CNSEC isolates from beef, pork, and chicken were resistant to at least one carbapenem, subsequently identified as carbapenem-resistant E. coli (CREC) (Table S3). Nevertheless, although the suspected CNSEC isolates were screened on the EMB containing imipenem, some suspected CSEC and ESBL-PEC isolates were still observed in this study (Figure 1).

3.3. Carbapenemase Genes in CREC Isolates

Primary mechanisms of carbapenem resistance were investigated in all CREC isolates (n = 284). These were specifically defined as the subset of CNSEC isolates that showed a definitive resistant phenotype to at least one tested carbapenem (imipenem or meropenem) across all sample sources, ensuring that we focused only on those categorized as truly carbapenem-resistant for the mechanism analysis. The results demonstrated that, among five carbapenemase genes, bla_NDM was found in 15 (4.8%) CREC isolates from one beef sample (B14) (

Table 3. Carbapenemase genes in carbapenem-resistant Escherichia coli isolates.

Isolate Code	uidA	Carbapenemase Gene					Antimicrobial Resistance Pattern
		Class A	Class B			Class D
		blaKPC	blaNDM	blaVIM	blaIMP	blaOXA-48
B14.1	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.2	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.3	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.4	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.5	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.6	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.7	+	−	+	+	−	−	CTX, CAZ, IPM, MEM
B14.8	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.9	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.10	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.11	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.12	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.13	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.14	+	−	+	−	−	−	CTX, CAZ, IPM, MEM
B14.15	+	−	+	−	−	−	CTX, CAZ, IPM, MEM

+, positive; −, negative; CTX, cefotaxime; CAZ, ceftazidime; IPM, imipenem; MEM, meropenem.

). The bla_VIM gene was also detected in one bla_NDM-positive CREC isolate. These 15 CREC isolates were examined for further experiments.

3.4. Molecular Relationship Among bla_NDM-Harboring CREC Isolates

The molecular relationship among all bla_NDM-harboring CREC isolates was analyzed by detecting genetic markers. The classification of phylogenetic groups revealed that all isolates were positive for the chuA gene and the TSPE4.C2 fragment but negative for the yjaA gene (

Table 4. Virulence genes in 16 carbapenem-resistant Escherichia coli isolates, with at least one carbapenemase gene carriage.

Isolate Code	DEC Indicator Genes									ExPEC Indicator Genes						Other E. coli Virulence Genes						Phylogenetic Group
	EPEC and EHEC				ETEC		EAEC	EIEC	DAEC	G1		G2	G3	G4	G5
	stx1	stx2	eae	bfpA	elt	est	aggR	ipaH	daaE	papA	papC	sfaDE	afa	kpsMT II	iutA	agn43	astA	cnf1	fimH	hlyA	lpf	chuA	yjaA	TSPE4.C2	Group
B14.1	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.2	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.3	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.4	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.5	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.6	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.7	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.8	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.9	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.10	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.11	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.12	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.13	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.14	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D
B14.15	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+	−	−	+	−	+	D

+, positive; −, negative; DEC, diarrheagenic Escherichia coli; EPEC, enteropathogenic Escherichia coli; EHEC, enterohemorrhagic Escherichia coli; ETEC, enterotoxigenic Escherichia coli; EAEC, enteroaggregative Escherichia coli; EIEC, enteroinvasive Escherichia coli; DAEC, diffusely adherent Escherichia coli; ExPEC, extraintestinal pathogenic Escherichia coli; G1, Group 1; G2, Group 2; G3, Group 3; G4, Group 4; G5, Group 5.

). Based on these genetic markers, all isolates were classified as in phylogenetic Group D. Furthermore, the genetic relatedness among these isolates was also analyzed using the BOX-PCR method. The results showed that all bla_NDM-harboring CREC isolates exhibited indistinguishable DNA fingerprints (Figure S2).

4. Discussion

The contamination of pathogenic bacteria, especially antibiotic-resistant strains, in meat is a critical public health concern due to the potential for these bacteria to cause severe infections in humans [49]. Through HGT, bacteria isolated from food sources may share their AMR genes with antibiotic-resistant pathogens in humans [50]. This issue is exacerbated by the overuse of antibiotics in livestock for growth promotion and disease prevention, contributing to the development of resistance [51]. Furthermore, these pathogens in meat pose a food safety threat and have global implications, as these bacteria can spread across borders through trade and travel, worsening the global crisis of AMR [52]. Addressing this problem requires the responsible use of antibiotics in agriculture, improved surveillance of antibiotic-resistant bacteria in the food supply, and promotion of safe food handling practices. In this study, we report the prevalence of CREC in raw meats from an open market in Hat Yai, Songkhla, Thailand, and describe their phenotypic and molecular characteristics, including antimicrobial resistance profiles, carbapenemase genes, virulence genes, and phylogenetic relationships.

Our findings demonstrated that presumptive CNSEC isolates were mostly found in beef samples, followed by pork and chicken samples. The highest number of CNSEC isolates coming from beef might be explained by specific antimicrobial usage practices in local cattle production that co-select for carbapenem resistance genes (often carried on highly mobile plasmids) or by unique contamination risks during beef processing, especially for ground beef, through which intestinal flora is disseminated more thoroughly [53]. A very high proportion of presumptive CNSEC isolates were subclassified as CREC; this predominance highlights a significant public health concern. CNSEC refers to all isolates showing reduced carbapenem susceptibility, whereas CREC denotes those with confirmed resistance mechanisms and thus carries greater clinical and epidemiological relevance [54]. The high prevalence of CREC in beef and other meat types suggests the food chain as a potential reservoir for multidrug-resistant E. coli, consistent with global reports linking livestock production to the dissemination of carbapenem and broad-spectrum β-lactam resistance [55]. The antimicrobial susceptibility testing results, showing resistance to imipenem and/or meropenem, indicate a critical loss of efficacy in these last-line agents. The concurrent resistance to third-generation cephalosporins may support the likelihood of non-carbapenemase-mediated mechanisms [56]. While CNSEC may include isolates with borderline susceptibility or transient adaptive mechanisms, CREC represents those with stable, transferable resistance and higher public-health significance. The presence of CREC in retail meat poses a risk of HGT to human commensal E. coli, facilitating the emergence of difficult-to-treat infections (Feng et al., 2021) [9]. These findings support the need for integrated One-Health surveillance and stricter antimicrobial stewardship across the food production system.

In the identification of primary carbapenem resistance mechanisms, only 5.3% of the CREC isolates were positive for the bla_NDM gene, and one of them was also positive for bla_VIM. While the detection of these MBL genes confirmed the basis of resistance in the CREC from a single beef sample, the precise mechanisms driving the carbapenem resistance in the remaining isolates were not determined. The observed carbapenem resistance phenotype probably results from several factors, including, but not limited to, the co-existence of high-level AmpC and/or ESBL production combined with outer membrane porin loss and/or the enhanced activity of efflux pumps [57,58]. Future studies utilizing molecular sequencing or gene expression assays are required to definitively characterize the carbapenem resistance mechanisms in these isolates [57,58]. Several studies have reported the prevalence of this pathogen in food-producing animals. In 2018, Guo et al. screened CREC isolates from 125 duck meat samples obtained from five farms in China (25 meat samples per farm) [59]. They reported that 33.6% of the isolates carried the bla_NDM gene with high resistance to almost all tested antimicrobial agents. Notably, their findings revealed that the structures of bla_NDM-bearing plasmids were highly similar to those of plasmids from human-sourced isolates. Their phylogenetic tree also showed that duck-sourced isolates were clustered with human-sourced isolates from different areas in China. Zhai et al. (2020), from China, identified 14 sequence types (STs) in NDM-producing E. coli isolates from chicken and their environmental samples, with ST6751 being the most frequent [60]. Additionally, VIM-1-producing E. coli ST88 isolated from pigs was first reported in Germany in 2011 and was also previously identified in chickens, cattle, and humans [61]. Another study from Italy showed that E. coli isolates from pigs carried bla_OXA-181, bla_OXA-48, mcr-1 (mobilized colistin resistance gene), and qnrS1 (fluoroquinolone resistance gene) [62]. The findings also demonstrated that the bla_OXA-181 and qnrS1 genes were detected on the IncX3 plasmid, highly similar to the plasmid from human-sourced isolates. For the study of CREC in cattle, bla_OXA-48-, bla_OXA-181-, bla_NDM-, and bla_VIM-carrying E. coli isolates were reported in Egypt in 2014, South Africa in 2020, Italy in 2021, and Spain in 2022 [53,63,64,65]. Tello et al. (2022) reported the bla_NDM-1 gene being located on an IncC plasmid that harbored aminoglycoside, sulphonamide, and trimethoprim resistance genes [65]. This evidence probably confirms the dissemination of carbapenem resistance between human- and animal-sourced isolates. However, our findings showed the presence of carbapenemase genes, bla_NDM (n = 15) and bla_VIM (n = 1), in the isolates from only one beef sample. The reasons for the relatively low presence of carbapenemase genes in the CREC isolates from raw meats might include factors such as geographic variability, controlled antibiotic use in agriculture, and alternative resistance mechanisms. In regions where carbapenemase production is not predominant, porin loss and efflux pump overexpression, alongside the production of ESBLs and AmpC β-lactamases, might be frequently found mechanisms leading to carbapenem resistance without the need for carbapenemase production in CREC isolates [8]. We hypothesize that these mechanisms are common in hospital settings, where ESBL-producing E. coli is prevalent. Further investigation of other mechanisms would confirm this hypothesis, and the lack of this is acknowledged as a limitation of the present study.

In the detection of virulence genes, our findings revealed that all bla_NDM-harboring CREC isolates were positive for the fimH gene responsible for the production of type 1 fimbriae, which are adhesive structures important for attachment to host tissues, particularly in the urinary tract [66]. However, the presence of fimH alone does not classify a strain as particularly pathogenic or indicative of a specific pathotype; it is present in both pathogenic and commensal strains of E. coli [67]. In pathogenicstrains, such as uropathogenic E. coli (UPEC), fimH plays a role in urinary tract infections (UTIs) [68]. However, if fimH is the only virulence factor presented in studied isolates, it may not necessarily indicate high virulence.

The phylogenetic group was classified based on the presence or absence of 3 genetic markers [45]. First, chuA is part of the heme transport system (colicin-hydroxamate uptake) and is essential for iron acquisition from the host. Its presence is a strong indicator of isolates belonging to more virulent extraintestinal pathogenic groups, primarily Groups B2 and D. Second, yjaA encodes a hypothetical protein of unknown function in E. coli. It is specifically associated with isolates of the highly virulent Group B2. Third, TSPE4.C2 is an anonymous DNA fragment. Its profile helps to differentiate commensal strains (Groups A and B1) from the virulent strains, serving as a key marker for the less pathogenic Group A. The results exhibited that all bla_NDM-harboring CREC isolates were classified in phylogenetic Group D, which is often associated with ExPEC, including strains involved in urinary tract infections and other invasive infections [69]. However, these isolates were common E. coli since they were negative for ExPEC indicator genes. In addition, the investigation of genetic relatedness revealed that all bla_NDM-harboring CREC isolates belonged to a single, dominant pulsotype, indicating clonal dissemination. However, further high-resolution genetic testing, such as whole-genome sequencing, may be required for confirmation at a finer scale.

Zoonotic transmission of antibiotic-resistant bacteria, particularly CREC, from animal-derived foods poses a significant public health concern. Meats from animals treated with antibiotics or exposed to resistant bacteria can act as reservoirs, infecting humans via handling or consuming undercooked meat. The presence of such bacteria in the food chain exposes weaknesses in food safety, agricultural practices, and antimicrobial stewardship, underscoring the urgent need for stricter regulations on antibiotic use in food animals, robust surveillance systems, and improved control measures to combat the growing antimicrobial resistance crisis affecting both human and veterinary medicine.

5. Conclusions

This study reported the prevalence and characteristics of CREC isolates detected in raw meats from an open market in Hat Yai, Songkhla, Thailand. The number of CREC isolates was relatively high, especially in beef samples, and a minority of them carried carbapenemase genes. The presence of CREC, particularly carbapenemase-positive isolates, in meats poses a significant threat to public health, food safety, and efforts to combat antimicrobial resistance. This underscores the urgent need for effective antimicrobial stewardship in agriculture, stringent food safety protocols, and comprehensive surveillance systems to monitor and curb the spread of these harmful bacteria along the food supply chain, from farm to fork.