Prevalence and molecular characterization of ESBL-producing Escherichia coli isolated from broiler chicken and their respective farms environment in Malaysia

Abstract

Background

Extended spectrum beta-lactamase-producing Escherichia coli (ESBL-EC) is an increasing public health threat. This study aimed to determine the prevalence and characterization of ESBL-producing Escherichia coli (E. coli) isolated from broiler chicken and their farm environment, in Kelantan Malaysia.

Methods

Escherichia coli was isolated from 453 collected samples, including 210 cloacal swabs and 243 environmental samples. The antimicrobial susceptibility profile of the E. coli isolates was assessed for sixteen antibiotics using the disc diffusion method. The E. coli isolates were evaluated for phenotypic ESBL production using modified double disc synergy. After extraction of genomic DNA, ESBL resistance genes, phylogenetic group, and virulence genes were detected by PCR using appropriate primers. ESBL genes were further confirmed by sequencing. The molecular typing of E. coli strains was determined by Multilocus Sequence Typing (MLST).

Results

A total of 93.8% (425/453) E. coli were isolated from the collected samples. Out of 334 E. coli isolates screened, 14.7% (49/334) were phenotypically ESBL producers. All the ESBL-EC were resistant to tetracycline, ciprofloxacin, and ampicillin. Thus, 100% of the ESBL-EC were multidrug resistant. Of the ESBL-EC 81.6% were positive for at least one ESBL encoding gene. The most prevalent ESBL gene detected was bla_TEM (77.6%; 38/49) followed by bla_CTX−M (32.7%; 16/49) and bla_SHV (18.4%; 9/49). The majority of ESBL-EC belonged to phylogenic groups A followed by B1 accounting for 44.9% and 12.2%, respectively. The most frequently identified sequence types were ST10 (n = 3) and ST206 (n = 3). The most detected virulence genes in the E. coli isolates were astA (33.3%; 22/66) followed by iss (15.2%; 10/66).

Conclusions

Our results show both broiler chicken and their respective farms environment were reservoirs of multi-drug resistant ESBL-producing E. coli and ESBL resistance genes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-024-03653-2.

Introduction

Antimicrobial resistance (AMR) is an increasing public health threat globally. The emergence and spread of antibiotic-resistant bacteria in food-producing animals are of particular concern. AMR worryingly spreads due to the misuse and excessive use of antimicrobials in humans and food animals [1]. The increasing number of animals bred for food production leads to more use of antimicrobial substances [2]. This overuse of antimicrobials may induce the intestinal microorganisms to acquire resistance and transmit to humans and animals.

Some evidence showed that animal gut microbiota and animal farms are the reservoirs of resistant bacteria and antimicrobial resistance genes (ARGs) because of antibiotic utilization [3–6]. The use of antibiotics for disease prevention and growth promotion in animals has been shown to cause antibiotic-resistant infections in humans [7]. Resistant gut bacteria are shed through feces and urine, causing environmental contaminations [8, 9]. In addition, several reports indicate that animal wastes such as manure play vital role in the spread of antimicrobial-resistant bacteria and ARGs to the environment [10–12].

Multidrug resistance has been increasing all over the world and is considered a public health threat. Several recent investigations reported the emergence of multidrug-resistant bacterial pathogens from different origins that increase the necessity of the proper use of antibiotics [13–19]. Besides, the routine application of antimicrobial susceptibility testing to detect the antibiotic of choice as well as the screening of the emerging MDR strains is crucial.

Extended-spectrum beta-lactamase (ESBL) is one of the most potent enzymes produced by certain resistant bacteria inducing antibiotic resistances that poses the most public health concerns. ESBL is a highly potent bacterial enzyme that confers resistance to β-lactam antibiotics including cephalosporines which are classified as the highest priority as well as critical important antimicrobials for human medicine by WHO [20]. It is mainly produced by Enterobacterales, specifically by E. coli and Klebsiella pneumoniae [21]. According to the CDC 2019 report, ESBL-producing Enterobacterales are among the most serious threats to human health [22]. ESBL-producing E. coli is one such group of bacteria that is increasingly detected in food-producing animals, including broiler chickens.

Escherichia coli is one of the most common normal inhabitants of human and animal intestines [23]. However, E. coli can also be a potentially pathogenic bacterium that colonizes the gastrointestinal tract and may disseminate to extra-intestinal organs of humans and various animals. Escherichia coli causes various diseases such as urinary tract infection (UTI), neonatal meningitis, septicemia in humans, and avian colibacillosis in poultry [24, 25]. Avian Pathogenic E. coli (APEC) is part of Extraintestinal pathogenic E. coli (ExPEC). These bacteria are the main cause of poultry extraintestinal bacterial infection [26]. Virulence factors contribute in the pathogenesis of E.coli, can be grouped into adhesion, protection, iron acquisition factors, and toxins production [27]. The iron acquisition systems, colicins (CvaC), increased serum resistance proteins, capsule as well as lipopolysaccharide complexes, and temperature-sensitive hemagglutinin (Tsh) are virulence related to APEC [28, 29]. The P-fimbriae (PAP) helps the bacteria for adhesions to the host cell to initiate infection [30]. Capsule and lipopolysaccharide protect the bacteria from the host immune system [29]. Antibiotic-resistant E. coli can cause disease in humans through direct contact or via food chain. Escherichia coli can mutate and horizontal transfer antibiotic resistance genes to inter- and intra-bacterial species [31]. Such characteristics of this bacteria facilitate the rapid spread of antibiotic resistance. Poultry is the most common source of ESBL-EC for humans among food-producing animals [32]. ESBL producing E. coli is increasingly reported in food-producing animals, including chickens. Furthermore, it is the main cause of infections and death in poultry farms and may transmit to humans and cause life threatening disease [33]. In poultry farms, ESBL-EC can contaminate the farm environment through fecal shedding of these resistant bacteria. It has been shown that ESBL-EC is highly prevalent in chicken meat compared to other meat-producing animals [34].

ESBL-producing strains of bacteria are the most common ESBL-encoding genes of bla_CTX−M, bla_TEM, and bla_SHV. Among the ESBL encoding genes, bla_CTX−M has been widely reported and is responsible for the increasing ESBL related resistance in pathogens infecting humans and animals throughout the world [35, 36]. The ESBL genes are encoded by plasmids which can spread easily between commensal and pathogenic bacteria in poultry farms and the environment [37].

According to Clermont O classification, E. coli can be classified into eight phylogenetic groups A, B1, B2, C, D, E, F, and clade I [38]. Commensal E. coli colonizing the gut lining are usually grouped as phylogroup A or B1. Meanwhile, pathogenic E. coli that causes gut infections is usually assigned to groups A, B1, or D. Escherichia coli that causes ExPEC-related infections are usually grouped as phylogroup B2 and D. Group E is closely related to group D, while group F is closer to group B2 [39–41]. Multi-locus sequence typing (MLST) is important to determine the phylogenetic relationship and evolution of bacterial lineages.

In Malaysia, ESBL colonization was reported in clinical patients, communities, farm animals, and from food [36, 42–45]. In this study, we investigated the prevalence and molecular characterization of ESBL-EC in broiler chickens and their respective farms environment. The phylogenetic group distribution and virulence gene of E. coli isolates were determined. The presence of bla_CTX−M, bla_TEM and bla_SHV ESBL genes in the E. coli isolates were investigated.

Method and materials

Sample collection

A total of four hundred fifty-three samples were collected from six different broilers chicken farms and farm environments in Kota Bharu and nearby areas (Fig. 1). Samples were collected from February to November 2021 using a simple random sampling technique. Two hundred ten cloacal swabs, 95 environmental swabs from feeding and drinking troughs, 27 drinking water, 32 feed, 20 litter, 55 freshly passed fecal samples, and 14 sewage glasses of water were collected. Cloacal, fecal, and environmental swabs samples were collected using a sterile cotton swab with an Amies transport medium (Oxoid, Manchester, UK). Feed and litter were collected in sterile zip lock bags. All samples were collected aseptically using sterile and appropriate containers. Collected samples were labeled individually with identification number and date of collection and were transported to the laboratory using a cold box within 6 h. In the case of delay, samples were stored at 4 °C until processed.

Fig. 1.

A map showing location of sampling area for this study and the locality is boxed in yellow, which includes the Kota Bharu city and its surrounding

Isolation and identification of E. Coli

Collected samples were enriched in Buffered Peptone Water (BPW) (Oxoid, Manchester, UK) by incubating at 37 °C for 24 h as described previously [46]. The isolation and identification of E. coli were performed as described previously [36]. The identified presumptive E. coli was further tested for nine biochemical characteristics including triple sugar iron agar (TSI) for glucose fermentation, citrate metabolism, methyl red, Voges-Proskauer, motility, urease production, indole production, gas, and H₂S production as described previously [47]. The media used for the biochemical test were from Oxoid, Manchester, UK. E. coli ATCC^® 25,922 was used as a positive control strain.

PCR confirmation of isolated E. Coli

The genomic DNA of isolated E. coli was extracted by boiling method as described previously by Mahmud et al. [48]. Extracted DNA was amplified with species-specific primer for the Pho gene encoding for E. coli as described previously [44, 49].

Antimicrobial susceptibility profile

E. coli isolates were tested for their susceptibility to 16 antibiotics belonging to 11 different antibiotic classes (Table 1). Antimicrobial susceptibility test was determined using the Kirby-Bauer disk diffusion method following the recommendation of the Clinical Laboratory Standards Institute (CLSI) [50]. E. coli isolate was grown on nutrient agar (NA) (Oxoid, Manchester, UK) at 37 °C for 18 h. Pure colonies from NA were suspended in 2 ml of sterile normal saline (0.85%) with turbidity equivalent to 0.5 MacFarland Standard. The bacterial suspension was then spread evenly on Mueller-Hinton agar (MHA) (Oxoid, Manchester, UK) using a sterile cotton swab. Then antibiotic disks were dispensed on the surface of the agar plate and incubated at 37 °C for 18 h. The susceptibility profile of the E. coli isolates was done by measuring the zone of inhibition and interpreted by comparing with CLSI standard breakpoint to determine as sensitive, intermediate, and resistant as per CLSI guidelines [50]. All isolates with intermediate resistance were considered as resistance in calculating Multiple antibiotic resistance (MAR) index as well as in defining multi-drug resistance (MDR), extensively drug resistance (XDR), and pan drug resistance (PDR). The MDR is defined as “non-susceptibility” to at least one agent in three or more antibiotic classes and XDR is defined as “non-susceptibility” to at least one antibiotic in all the antimicrobial classes but isolates remain susceptible to only one or two antibiotic classes. PDR is “non-susceptibility” to all antibiotics in all the antibiotic classes based on Magiorakos et al. definition described previously [51]. The MAR index calculation was done as described in Tambekar et al. [52]. All the antibiotic discs used were from Oxoid, Manchester, UK. Standard E. coli strain ATCC^® 25,922 was used as a control strain.

Table 1.

List of antibiotics and antibiotic classes used for antibiotic susceptibility test

Antibiotic (µg)	Antibiotic class	Antibiotic (µg)	Antibiotic class
Ampicillin (AMP10)	Penicillin	ceftriaxone (CRO30)	Cephalosporine
Amoxicillin/clavulanic acid (AMC30)	Beta-lactam inhibitor	meropenem (MEM10)	Carbapenem
Gentamicin (GEN10)	Aminoglycoside	cefotaxime (CTX30)	Cephalosporine
Imipenem (IPM10)	Carbapenem	tetracycline (TET30),	Tetracycline
Ciprofloxacin (CIP5)	Fluoroquinolones	ceftazidime (CAZ30)	Cephalosporine
Trimethoprim/sulfamethoxazole (SXT25)	Sulphonamides	cefuroxime (CXM30)	Cephalosporine
Chloramphenicol (CHL30)	Chloramphenicol	Aztreonam (ATM30)	Beta-lactam
Streptomycin (STR10)	Aminoglycoside	nalidixic acid (NAL)	quinolones

ESBL production

E. coli isolates with zone of inhibition of cefpodoxime ≤ 17 mm, ceftazidime ≤ 22 mm, aztreonam ≤ 27 mm, ceftriaxone ≤ 25 mm, and cefotaxime ≤ 27 mm considered as a potential ESBL producer according to CLSI, 2021 guideline and in addition, the isolates encoding ESBL genes and/or, virulence genes were selected for ESBL production test using Modified Double Disc Synergy (MDDS). ESBL production of E. coli isolates were tested using MDDST using amoxicillin-clavulanate (30 µg) along with four cephalosporins; 3GC-cefotaxime, ceftriaxone, aztreonam, cefpodoxime and 4GC-cefepime as previously described [53]. Briefly, 3–5 freshly grown colonies were suspended into 0.85% sterile normal saline. The turbidity of bacterial suspension was equivalent to 0.5 McFarland standard, the bacterial inoculum was lawn on the surface of Mueller Hinton agar. Amoxicillin-clavulanate (AMC30µg) was placed in the center of the plate with 3GC and 4GC discs placed 15 mm apart around AMC and the 3G and 4G cephalosporin were 20 mm apart from each other. Isolates were determined as ESBL producers when the zone of inhibition increased towards the AMC disc [53]. Escherichia coli 25,922 was used as a negative control for ESBL production.

Detection of ESBL encoding genes

All the confirmed E. coli isolates were screened for ESBL-encoding genes bla_CTX, bla_TEM, and bla_SHV. The PCR amplification of these ESBL genes were performed following the previous protocol [54]. The PCR amplification protocol used for bla_TEM, and bla_CTX−M genes was as described previously [36]. The PCR protocol for bla_SHV (768 bp) amplification was an initial denaturation step at 95 °C for 15 min, followed by 30 times of denaturation at 94 °C for 1 min, annealing at 58 °C for 30 s, and extension at 72 °C for 1 min and the final extension 72 °C for 10 min. Primers used in this study are summarized in Table 2.

Table 2.

List of primers and genes used in this study

Gene Name	Sequence	Amplicon size(bp)	Annealing T(°C)	Reference
pho	F: 5′- GTGACAAAAGCCACACCATAAATGCCT-3′	903	56	[49]
	R: 3′-TACACTGTCATTACGTTGCGGATTTGGCGT-5′
E. coli	F: 5’-TGACGTTACCCGCAGAAGAA-3’	832	55	[44]
	R: 3’-CTCCAATCCGGACTACGACG-5’
bla CTX-M	F: 5’-ATGTGCAGYACCAGTAARGTKATGGC-3’	592	65	[54]
	R: 3’-TGGGTRAARTARGTSACCAGAAYSAGCGG-5’
bla TEM	F: 5’-GCGGAACCCCTATTTG-3’	964	55
	R: 3’-ACCAATGCTTAATCAGTGAG-5’
bla SHV	F: 5’-TCGCCTGTGTATTATCTCCC-3’	768	58
	R: 3’-CGCAGATAAATCACCACAATG-5’
chu A	F:5’-TGCCATCAACACAGTATATCC-3’	288	59	[38]
	R:3’-TCAGGTCGCGAGTGACGGC-5’
yja A	F:5’-ATCACATAGGATTCTGCCG-3’	211
	R:3’-CAGCGGAGTATAGATGCCA-5’
TspE4.C2	F:5’-AAGGATTCGCTGTTACCGGAC-3’	152
	R:3’-AACTCCTGATACAGGTGGC-5’
arpA	F:5’-TGATATCACGCAGTCAGTAGC-3’	400
	R:3’-CCGGCCATATTCACATAA-5’
trpAgpC	F:5’-ACAAAAAGTTCTATCGCTTCC-3’	219	62
	R:3’-CCTGATCCAGATGATGCTC-5’
ArpAgpE	F:5’-ACTATTCTCTGCAGGAAGTC − 3’	301	59
	R:3’-CTTCCGATGTTCTGAACGT-5’
trpBA	F:5’-TCCTGGGACATAATGGTCAG-3’	489
	R:3’-GTGTCAGAACGGAATTGT-5’

E. coli isolates that are positive for the ESBL gene (bla_CTX−M, bla_TEM, and/or bla_SHV) were further confirmed by DNA sequencing and the sequenced DNA was analyzed using BLAST analysis in a website: http://www.ncbi.nlm.nih.gov/. The nucleotide sequence of the ESBL genes is available in GenBank accession no. PQ522631-PQ522632.

Phylogenetic typing

Following Clermont O et al. classification, E. coli was classified into eight phylogenetic groups A, B1, B2, C, D, E, F, and clade I. Quadruplex PCR was used to assign the isolates into phylogroups A, B1, B2, and D which target chuA, yjaA, TspE4.C2, and arpA genes as described previously [38]. The isolates of phylogroup A were separated from phylogroup C by trpAgpC primer which is a C-specific primer. Similarly, phylogroup D isolates were distinguished from E using ArpAgpE primer, and phylogroup F was differentiated from phylogroup D in the quadruplex PCR as F does not contain the ArpA gene. The PCR protocol used was based on Clermont et al. [38].

Virulence gene

The ESBL genes-positive E. coli isolates were assessed for APEC-associated virulence genes. The APEC virulence factors include genes related to adhesion (papC and tsh), toxin production (astA, vat and cvaC/cvi), iron uptake (iucD, irp2), and invasion (iss). The papC gene encodes for P fimbrial adhesion, iucD encodes the aerobactin operon, irp2 encodes an iron-repressible protein, tsh encodes for temperature-sensitive hemagglutinin, vat encodes for vacuolating autotransporter toxin, astA for enteroaggregative heat-stable toxin, iss increases serum survival and cva/cvi encodes for colicin V plasmid operon [55]. Multiplex PCR was performed to test papC, iucD, irp2, tsh, vat, astA, iss, and cva/cvi virulence genes associated with virulence factors. The PCR protocol used was as described previously [55].

Multi-locus sequence typing (MLST) of ESBL producing E. Coli

The E. coli isolates positive for the ESBL gene and/or carry virulence genes were selected for MLST analysis. The detailed MLST analysis method was done as described previously [36]. The PCR conditions used were as follows: initial denaturation at 95 °C for 2 min; 30 cycles of initial denaturation at 95 °C for 1 min, annealing at 57 °C (adk) or 64 °C (fumC, and purA) or 68 °C (recA), 72 °C (gyrB), 69 °C(icd) and 71 °C (mdh) for 1 min and elongation 72 °C for 2 min; followed by a final extension step at 72 °C for 5 min. The amplified PCR products were sent to Apical Scientific SDN. BHD, Malaysia, for a sequence analysis. The alleles and Sequence types (ST) were assigned based on the E. coli database at the MLST website, http://enterobase.warwick.ac.uk/.

Statistical analysis

The presented data set is categorical data type since it has a well-known non-numerical set of values. Hence tools applicable to this type of data were utilized. Correlations between the phenotypic antibiotic resistance, and antibiotic resistance genes as well as correlation among the antibiotics were determined using Python version 3.11. Python-Pandas, One-hot-encoding approach use applied to generate the correlation matrix between the phenotypic antibiotic resistance, antibiotic resistance genes as well as the correlation among the antibiotics as explained previously [56]. The correlation ratio was calculated based on the Pearson correlation coefficient (r) using the same matrix table generated using the one-hot-encoding of the Pandas library. The resulting matrix was plotted using the Python seaborn library heat map visualization tool [57].

The Venn diagram showing the overlap of the ESBL genes such as bla_CTX−M, bla_TEM, and bla_SHV genes was plotted using R software version 4.3.3, the package of ggVennDiagram [58].

Results

Phenotypic characteristics of the recovered isolates

Out of the 453 samples tested, 431 (95%) were positive for E. coli using routine microbiological methods. Of the isolated E. coli, 98.6% (425/431) were confirmed as E. coli by PCR (Table 3). Majority of the confirmed E. coli were assigned to phylogenetic group A accounting for 52.2% (222/425) followed by B1 (11%, 48/425).

Table 3.

Isolation and PCR confirmed E. Coli strains in broiler chickens from Kelantan Malaysia

Sample type	Isolated E.coli	PCR-confirmed E. coli	Percentage%
Cloacal Swab	205	203	99.0
Drinking water	24	23	95.8
Environmental	85	83	97.6
Fecal	55	55	100.0
Food	31	31	100.0
Litter	20	20	100.0
Sewage	11	10	90.9
Grand Total	431	425	98.6

Of the E. coli confirmed by PCR, 334 of them were evaluated for phenotypic ESBL production. Out of the 334 E. coli isolates, 14.7% (49/334) were phenotypically ESBL producers. Among these, 57.1% (28/49) of the ESBL-EC were detected from cloacal samples. Meanwhile, 42.9% of the ESBL-EC were identified from various farm environment samples. These include environmental swabs (14.3%, 7/49), litter samples (6.1%, 3/49), drinking water (6.1%, 3/49), freshly passed fecal samples (10.2%, 5/49) and food (6.1%, 3/49) were found positive for ESBL-EC.

Antimicrobial susceptibility profile

The 334 E. coli isolates were subjected to antimicrobial susceptibility profiles of sixteen antibiotics belonging to eleven different classes. E. coli isolates were highly resistant to tetracycline, streptomycin, chloramphenicol, ampicillin, ciprofloxacin, and trimethoprim/sulfamethoxazole (Fig. 2). Meanwhile, the isolates showed least resistance towards meropenem and imipenem in both ESBL-EC and non-ESBL-EC. ESBL-EC were shown to be highly resistant towards ciprofloxacin (100%), nalidixic acid (88%), cefotaxime (57%), cefuroxime (47%), ceftriaxone (43%), aztreonam (31%), amoxicillin/clavulanic acid (24%), and ceftazidime (22%) compared to non-ESBL producing E. coli (Fig. 3). AST profile of the antibiotic ciprofloxacin, tetracycline, ampicillin, chloramphenicol streptomycin, and trimethoprim/sulfamethoxazole showed similarly high resistance among all sample types. In addition, isolates were highly susceptible to meropenem in all sample types. However, amoxicillin-clavulanic acid showed higher resistance and intermediate (non-susceptible) rate in sewage (71%), cloacal swabs (66%), Environmental swab (62%), litter (55%), food (53%), and fecal (51%) compared to drinking water (33%). The detail antibiotic susceptibility profile of each sample type is described in supplementary material (Supplementary Fig. 1).

Fig. 2.

Antimicrobial resistance profiles of non-ESBL-EC isolates from broiler chicken and farm environment in Kelantan, Malaysia (n = 285)

Fig. 3.

Antimicrobial resistance profiles of ESBL-EC isolates from broiler chicken and farm environment in Kelantan, Malaysia (n = 49)

All the AST-tested E. coli isolates (334) were MDR. Of these 15% (50/334) of them were XDR and one E. coli isolate was classified as PDR. In addition, 100% of the ESBL-EC were MDR of which 28.6% (14/49) of them were classified as XDR. However, none of the ESBL-EC were classified as PDR.

All (100%) of the E. coli isolates were with MAR index value > 0.2. The majority of the isolates (64.4%) had MAR index values in the range 0.3–0.6, while the remaining isolates (35.7%) had MAR index values in the range 0.7-1. Among the ESBL-EC isolates, 46.9% had MAR index values ranging from 0.4 to 0.6, while the majority, 53.1%, had MAR index values ranging from 0.7 to 0.9. None of the isolates showed a MAR index value ≤ 2. The phenotypic multi-drug resistance type, resistance antibiotics MAR index values and the antibiotic resistance genes in the ESBL-EC isolates is summarized in supplementary material (Supplementary Table 1).

Extended spectrum beta lactamase (ESBL) encoding genes

All the isolated E. coli were screened for bla_CTX-M, bla_TEM and bla_SHV ESBL encoding genes. 79.5% (338/425) of the isolated E. coli were positive for the bla_TEM gene. In addition, 9.9% (42/425) of the isolates were positive for the bla_SHV gene, and 7.3% (10/425) were positive for bla_CTX−M genes. Out of the ESBL-EC, 81.6% were positive for at least one ESBL-encoding gene. In this study, bla_TEM (77.6%, 38/49), bla_CTX−M (32.7%; 16/49) and bla_SHV (18.4%; 9/49) were found harboring ESBL genes in the ESBL-EC isolates (Table 4). The bla_TEM was the predominant ESBL gene detected. Moreover, 20.41% (10/49) of the ESBL-EC isolates were positive for both bla_CTXM and bla_TEM, 8.16% (4/49) carried both bla_SHV and bla_TEM, and another 8.16% (4/49) carried all the ESBL bla_CTX−M, bla_TEM and bla_SHV genes (Fig. 4). However, 18.4% (9/49) of the ESBL-EC were negative for the ESBL genes. The co-occurrence of both mcr and ESBL encoding genes was observed in eight isolates.

Table 4.

ESBL encoding genes in ESBL-EC from broiler chicken and farm environment in Kelantan, Malaysia

ESBL genes	Cloacal	EVT	Food	Litter	Drinking water	Fecal	Grand total	Percentage (%)
bla TEM	11	4	2	2	-	1	20	40.82
blaCTX−M, blaTEM	8	-	1	-	-	1	10	20.41
blaSHV, blaTEM	-	-	-	1	3	-	4	8.16
blaCTX−M, blaSHV	1	-	-	-	-	-	1	2.04
blaCTX−M, blaSHV, blaTEM	1	3	-	-	-	-	4	8.16
bla CTX−M	1	-	-	-	-	-	1	2.04
Negative	6	1	1	-	-	1	9	18.37
Grand Total	28	8	4	3	3	3	49	100.00

Fig. 4.

A Venn diagram showing overlap ESBL genes such as bla_CTX-M, bla_TEM, and bla_SHV in ESBL-EC. Each circle is labeled with the respective gene name, the numbers in the non- overlapping region show the count of isolates carrying respective single ESBL gene. Numbers in the overlapping region indicate the number of isolates carrying the respective genes commonly

The correlation coefficient of the phenotypic (the resistant antibiotics) and the genotypic (positive resistance genes) resistance of the ESBL-EC showed that CRO as well as ATM have a strong correlation with an ESBL gene bla_CTX−M positive. In addition, CTX30 and CAZ30, CRO30 and CAZ30, ATM30 and CAZ30, NAL30 and SXT25, ATM30 and CTX30, CRO30 and CTX30, CRO30 and ATM30 resistant antibiotics showed strong correlation pattern (Fig. 5).

Fig. 5.

Correlation heatmap of resistant antibiotic with resistant genes

Multilocus sequence typing (MLST)

The MLST result of the isolated E. coli in this study is highly diverse. The most dominantly identified sequence types were ST10 (n = 3) and ST206 (n = 3). Two isolates with ST10 were detected from cloacal and one was from fecal samples. Isolates assigned to ST48, ST10, ST648, ST469, and ST165 were positive for more than three virulence genes. ST10 Cplx comprises ST10, ST1638 and ST48 sequence types (Table 5).

Table 5.

Phenotypic and genotypic characteristics of selected isolated E.coli positive for ESBL genes

Sample ID	Source	ESBL- EC	ESBL gene	ST	ST Complex	Phylogenetic group	Virulence genes	MAR I
1b	Cloacal	Positive	blaCTX−M,blaTEM, blaSHV	ST117	-	F	ast	0.8
B6H2	Cloacal	Positive	blaCTX−M, blaTEM	ST48	ST10 Cplx	B2	ast, papC, iuc D	0.9
C34H4	Cloacal	Positive	blaCTX−M, blaSHV	ST5229	ST101 Cplx	B1		0.8
C39H4	Cloacal	Negative	blaCTX−M, blaTEM	ST1771		A	ast	0.5
E24	EVT swab	Positive	blaCTX−M,blaTEM, blaSHV	ST1638	ST10 Cplx	A		0.9
E43	EVT swab	Negative	blaCTX−M,	ST398	ST398 Cplx	F	ast, iss	0.6
F3	Fecal	Negative	blaCTX−M,blaTEM, blaSHV	ST10	ST10 Cplx	A	ast	0.8
Food9	Feed	Positive	blaCTX−M,blaTEM, blaSHV	ST3896	-	Unknown	papc, ast	0.8
KB25	Cloacal	Positive	blaCTX−M, blaTEM	ST10	ST10 Cplx	C	ast, irp2,papC, iucD	0.8
KB29	Cloacal	Positive	blaCTX−M, blaTEM	ST648	ST648 Cplx	D	ast, iss, papC, iucD, tsh	0.8
KB4	Cloacal	Positive	blaCTX−M,	ST162	ST469 Cplx	B1	ast, iss, iuc D	0.7
KBF1	Fecal	Positive	blaCTX−M, blaTEM	ST165	ST165 Cplx	A	ast, iss, irp2	0.8
L25	Litter	Positive	blaCTX−M,blaTEM, blaSHV	ST106	ST69 Cplx	D	iss, irp2	0.9
Drw3	Drinking water	Positive	blaCTX−M, blaTEM	ST206	ST206 Cplx	A	ast	0.8
Drw7	Drinking water	Negative	blaTEM, blaSHV	ST1285	-	B1	-	0.8
E26	EVT swab	Negative	blaTEM, blaSHV	ST206	ST206 Cplx	A	iss, irp2	0.7
Food1(2)	Feed	Negative	bla SHV	ST155	ST155 Cplx	B1		0.5
KB10	Cloacal	Negative	blaTEM, blaSHV	ST10	ST10 Cplx	A		0.8
ZC4H1	Cloacal	Negative	blaTEM, blaSHV	ST206	ST206 Cplx	C	ast	0.5

Phylogenetic group and the virulence genes distribution

The PCR identification of phylogenetic typing of the ESBL-EC showed that phylogroup A was the most frequent group followed by B1 in all cloacal and farm environment samples, accounting for 44.9% (n = 22) and 12.2% (n = 6) of the ESBL-EC (Table 6). Moreover, B2 (6.1%; n = 3), D (8.2%; n = 4), and F (4.1%; n = 2) phylogroups of ESBL-EC isolates were detected. The virulence genes distribution in the E. coli isolates were astA (33.3%; 22/66), iss (15.2%; 10/66), irp2 (15.2%, 10/66), Papc (10.6%; 7/66), iucD (10.6%; 7/66) and tsh (1.5%; 1/66).

Table 6.

Phylogenetic typing of ESBL-EC from broiler chickens and farm environments in Kelantan, Malaysia n = 49

Row labels	A	B1	B2	C	Clad I or II	D	E	F	Unknown	Grand total
Cloacal	11	4	3	3	1	1	1	2	2	28
Drinking Water	1	1	-	-	-	1	-	-	-	3
EVT	4	-	-	1	-	1	-	-	1	7
Fecal	3	1	-	-	1	-	-	-	-	3
Food	2	-	-	-	1	-	-	-	-	5
Litter	1	-	-	-	-	1	1	-	-	3
Grand Total	22	6	3	4	3	4	2	2	3	49

Discussion

Escherichia coli is a common inhabitant of warm-blooded animal guts and has the potential to cause serious infections in humans and animals. The ability of E. coli to mutate, and acquire mobile genetic elements carrying resistance genes through horizontal gene transmission is a major contributing factor in the emergence and spread of antibiotic resistance [31].

In our study, among the presumptive E. coli isolates 14.7% (49/334) were phenotypically ESBL-EC. This is in line with a previous study from South Korea, where ESBL-EC was 14% [59]. But, this prevalence is lower than those reported in previous studies from Malaysia (48.8%), Indonesia (28.75%), India (25%) and Zambia (20.1%) [44, 60, 61]; however the current prevalence is higher than what has been reported from Tanzania (4.7%) and Mexico (5%) [62, 63]. These differences in the prevalence of ESBL-EC might be due to the difference in geographic location, time study difference, and different methods used. In this study, we determined the phenotypic ESBL-EC by MDDS, while most of the above studies used double disk synergy (DDS). The majority of the ESBL-EC were recovered from cloacal swab samples, 53.1% followed by environmental swab samples (14.3%). In addition to broiler chickens, feces, litter, water, and food were positive for ESBL-EC. Similarly, feces and litter were reported as a source of ESBL-EC in Germany [64]. It was also reported that healthy dairy cattle, farm environment, and milk were reservoirs of ESBL-EC [65]. This highlights that healthy livestock animals could be the reservoirs and sources of ESBL-EC to humans. A study from Singapore showed that ESBL-EC prevalence was high in chicken meat compared to pork and beef [66].

In the current study, we observed that the resistance patterns of E. coli isolates against tetracycline, ampicillin, meropenem, streptomycin, trimethoprim-sulfamethoxazole, and chloramphenicol antibiotics are similar in all sample types. The ESBL-EC strains were 100% resistant to ciprofloxacin, ampicillin, and tetracycline as well as 98% resistant to chloramphenicol. This 100% correlation between ESBL-EC and resistance towards these antibiotics might be due to the cross-resistance of different antibiotic classes in the plasmids and could result in multi-drug resistance [67, 68]. Higher resistance to streptomycin, ampicillin, chloramphenicol, tetracycline, and trimethoprim-sulfamethoxazole is repeatedly reported in several studies from poultry origin E.coli [36, 69, 70]. This might be caused by the common use of these antibiotics in food animal production. Isolates also showed resistance to trimethoprim-sulfamethoxazole (92%), and nalidixic acid (88%). Furthermore, 100% of the ESBL-EC were multi-drug resistant and 14% of them were XDR based on the classification of Magiorakos et al. [51]. In agreement with our findings, 96.2% of ESBL-EC were reported to be MDR from a study on broiler chickens from Thailand [71]. This could be due to the excessive use of antibiotics in animal production which may result in limited options for the treatment of multi-drug-resistant infections. ESBL-EC showed higher resistance to ciprofloxacin, cefotaxime, cefuroxime, ceftazidime, amoxicillin/clavulanic acid, aztreonam, ceftriaxone and nalidixic acid compared to none ESBL-EC isolates. In the current study, the MAR index value of all the E. coli isolates was greater than 0.2, indicating that the E. coli strains including the ESBL-EC might have originated from highly antibiotic resistant bacteria contaminated environments or overuse of antibiotics.

The isolates from the current study were found to contain the most common ESBL genes, bla_CTX−M, bla_TEM, and bla_SHV. Majority of the isolated E. coli (79.5%) were positive for the bla_TEM gene followed by bla_SHV (9.9%) and then bla_CTX−M (7.3%). Meanwhile, 77.6% (38/49) of the ESBL-EC isolates harbored the bla_TEM gene, 32.7% (16/49) contained the bla_CTX−M gene, and 18.4% (9/49) contained the bla_SHV genes. The prevalence bla_CTX−M is higher in ESBL-EC compared to the prevalence of bla_CTX−M in the non-ESBL-EC isolates. This could be due to the fact that bla_CTX−M is related to cefotaxime resistance, cefotaxime resistance is higher in ESBL-EC than non ESBL-EC. Moreover, 20.41% (10/49) of the ESBL-EC isolates were positive for both bla_CTX−M and bla_TEM, 8.16% (4/49) carried both bla_SHV and bla_TEM, and another 2.04% (1/49) had bla_CTX−M, and bla_SHV genes. In addition, another 8.16% (4/49) carry all the three ESBL genes, bla_CTX−M, bla_TEM and bla_SHV genes. A study from Spain reported 20.7% ESBL producing E. coli harboring bla_SHV−12 gene from broiler chicken [72]. In a similar study, isolates were typed ST68 with phylogroup-E and ST117 with phylogroup-B2 from chicken manure. The results also showed that 18.4% of the ESBL-EC did not harbor any of the tested ESBL genes, this discrepancy of the phenotypic ESBL production and genotypic result might be due to other mechanisms of resistance or mutations that increase ESBL activity including other ESBL types such as AmpC β-lactamase or mutation in the outer membrane porin [73]. AmpC β-lactamase is a chromosomal mediated resistance that enhance the ESBL production but does not encode these specific ESBL genes [74].

In our study, bla_TEM was the predominant ESBL gene followed by bla_CTX−M and bla_SHV. Similarly, a study from China reported bla_TEM gene (82.3%) as the predominant ESBL type, followed by bla_CTX−M (43.5%) and bla_SHV (19.4%) [75]. In agreement with our findings, recent studies from Ecuador and Nigeria show that bla_TEM was the most prevalent ESBL gene for E.coli [67, 76]. A study conducted from hospital patients in Malaysia reported that 87.5% of ESBL-EC were harboring bla_TEM genes [77]. Moreover, we found high prevalence of bla_TEM genes in 100% ampicillin resistance ESBL-EC, in agreement with our finding, literatures reported in 98–100% ampicillin-resistant E. coli with high bla_TEM prevalence from chickens and humans [76, 78]. However, most of the broiler chicken origin studies around the globe including our previous study reported that bla_CTX−M was the dominant ESBL gene [36, 60, 61]. It has been reported that bla_CTX−M is an increasingly reported type of ESBL gene in food-producing animals. In general, the spread of these ESBL genes among the chickens and farm environment could be due to the fact these genes being plasmid mediated so that they can spread easily.

Four ESBL-EC isolates in the current study were positive for all bla_CTX−M, bla_TEM, and bla_SHV ESBL genes. These isolates were assigned to sequence types and phylogroup, ST117-F, ST1638-A, ST106-D, and ST3896-unknown, respectively. In addition, these isolates were sourced from cloacal swabs, environmental swabs, and litter. This indicates that these ESBL genes could spread easily from the chickens to the environment and vice versa. The ESBL-EC isolate ST117 was positive for astA, and ST3896 was found positive for papC and astA virulence genes. This indicates that these virulent ESBL-EC strains have an elevated risk of causing serious infection to the birds and possibility of zoonotic transmission to humans through the food chain.

In our study, Phylogenetic group A (44.9%) was the most frequent group followed by B1 (12.2%) among the ESBL-EC isolates. On top of that, we found 6.1% (n = 3) of B2, 8.2% (n = 4) of D, and 4.1% (n = 2) of F phylogroups. In agreement with these findings, another similar studies from Malaysia reported phylogroup A was the dominant phylogroup accounting for 45.1% [79–81]. It is widely reported that phylogenetic groups B2, F, and D are potential pathogens that can cause ExPEC infections in humans. Meanwhile, phylogenetic groups A and B1 are associated with commensal bacteria [82]. Even though phylogroups A and B1 are commensals these isolates were positive for virulence genes and ESBL genes. This could be due to the character of E. coli serving as a transporter of resistance genes to other or the same species of bacteria. This indicates that commensal E.coli isolates can be the reservoir of ESBL determinants that can spread to other pathogenic bacteria and disseminate antibiotic resistance.

The sequence typing of the E.coli isolates in this study showed high diversity. The most common STs we found were ST206 (n = 3) and ST10 (n = 3). Among the identified STs in this study, ST10, ST48, and ST117 were the most frequently reported STs in E.coli isolates from poultry [83]. Of these ST48 and ST10 were found in both poultry and human ExPEC isolates [84]. Moreover, ST10 and ST117 E. coli isolates harbored bla_CTX−M, bla_TEM, and bla_SHV, meanwhile, ST48 was positive for bla_CTX−M, and bla_TEM. ST117 has been reported in Avian pathogenic E. coli (APEC) associated in study from Italy [85].  Moreover, ST117 was reported as an emerging ExPEC lineage [86].

In the present study, an ESBL-EC with ST648 was found harboring both bla_CTX−M, and bla_TEM, ESBL genes and it was positive for astA, iss, papC, iucD, and tsh APEC related virulence genes with phylogenetic group D. This lineage was reported previously in highly virulent multidrug resistant pathogenic ESBL-EC with multiple resistant genes, and it is an emerging high-risk virulent strain with similar features to the ST131 [87]. Similarly, this lineage was reported to carry antibiotic resistance genes and has been reported from human in China [14]. The most dominant APEC related virulence genes detected were in the E. coli isolates were ast A (33.3%; 22/66), followed by iss (15.2%; 10/66), and irp2 (15.2%, 10/66).

In this study, we initiated a detailed investigation into the prevalence and molecular characterization of ESBL-EC in broiler chickens and their environment. Our main objective was to determine the existence of ESBL-EC and the molecular characteristics of these strains. We postulated that both the broiler chickens and their environment might serve as significant sources of ESBL-EC, thereby facilitating the spread of antibiotic resistance. To test our hypothesis, we conducted a comprehensive field study involving the collection of samples from broiler chickens and their surrounding farm environment. From these samples E. coli was isolated, and were tested for antimicrobial resistance profile, phenotypic ESBL production, ESBL gene, phylogenetic group, virulence genes, and sequence typing to identify and characterize the ESBL-EC strains.

Conclusion and recommendations

This study showed both the broiler chickens and their contaminated farm environments could be significant sources of ESBL-producing E. coli in Kelantan, Malaysia. Our results show that feces, litter, drinking water, as well as chicken’s food were reservoirs for ESBL-EC and ESBL genes. The bla_TEM was the most common ESBL gene detected. All the ESBL-EC were multi-drug resistant and more than half of them were XDR. Moreover, the E. coli isolates were highly resistant to tetracycline, ampicillin, streptomycin, and chloramphenicol. Multi-drug resistant ESBL-EC have been found positive for virulence genes that can cause disease in birds, and humans. ESBL-EC from broiler chickens and their contaminated farm environment may pose a potential risk of transmission to humans through food chain or direct /indirect contact. Based on our findings, we recommend the implementation of a strict regulation of antibiotic use in broiler chicken production focusing on both the chickens and their contaminated environments.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

ML conducted the sample collection and laboratory analysis, wrote the main manuscript text, and prepared figures and tables. EA supervised the laboratory and provided funding for the laboratory consumables. EA, MM and NFK, supervise and edit the manuscript. SSD and HL assist in laboratory analysis and sample collection. AAK= assists in sample collection. All authors have reviewed the manuscript. ML=Mulu Lemlem, EA= Erkihun Aklilu, MM=Maizan Mohamed, NFK=Nor Fadhilah Kamaruzzaman, SSD=Susmita Seenu Devan, HL= Habiba Lawal, AAK= Abubakar Abdulkarim Kanamma.

Funding

The Laboratory reagents and consumables of this research was funded by Ministry of Higher Education Malaysia (MOHE) under the fundamental research grant scheme (FRGS), grant no. FRGS/1/2019/WAB01/UMK/02/6.

Data availability

Data is provided within the manuscript as well as attached in the supplementary section. The nucleotide sequence of the ESBL genes is available in GenBank accession no. PQ522631-PQ522632.

Declarations

Ethics approval and consent to participate

The Institutional Animal Care and Use Committee of Universiti Malaysia Kelantan approved this study (UMK/FPV/ACUE/PG/2/2019). The animal subjects (chicken from commercial poultry farms) were only used to collect cloacal swabs, and no invasive or harmful procedures were used to handle the birds. Informed consent was obtained from the farm owners for the use of their animals in the study prior to study commencement. All methods conducted in this manuscript are based on standard guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.