Silent Reservoirs: Antibiotic-Resistant Escherichia coli in Autochtonous Portuguese Laying Hens

Abstract

Antimicrobial Resistance (AMR) is a critical public health challenge requiring a coordinated One Health approach. Escherichia coli is a key indicator of AMR and fecal contamination, as well as a zoonotic pathogen transmissible from animals to humans, often through contaminated products like meat and eggs. This study assessed the presence of antibiotic-resistant E. coli and associated resistance genes in 248 cloacal/eggshell samples collected from four autochthonous Portuguese laying hen breeds (Preta Lusitânica, Amarela, Branca, and Pedrês Portuguesa) raised under low antibiotic exposure. A total of 81 E. coli isolates were analyzed for phenotypic antibiotic susceptibility (EUCAST/CLSI) and genotypic resistance, using PCR. Resistance to at least one antibiotic was observed in 98.0% of the isolates. Gentamicin resistance was particularly high (97.1% cloacal; 95.7% eggshell isolates), followed by tetracycline (31.0% cloacal; 41.0% eggshell) and ampicillin (14.0% cloacal; 24.0% eggshell). Multidrug resistance (MDR) was observed in 14.3% of cloacal and 17.4% of eggshell isolates. Notably, no resistance was found against critically important antibiotics. The most prevalent resistance genes were sul2 (45.0% cloacal; 48.0% eggshell) and bla_TEM (45.0% cloacal; 36.0% eggshell). Detection of resistant and MDR E. coli in low input systems suggests environmental acquisition, with chickens as reservoirs, highlighting the need for One Health surveillance.

1. Introduction

Antimicrobial Resistance (AMR) has become one of the most critical and complex public health challenges of our time, representing a transboundary threat that undermines decades of progress in both human and veterinary medicine [1,2]. Tackling AMR effectively requires coordinated, multidisciplinary efforts grounded in the One Health approach, which recognizes the intrinsic interconnectedness of human, animal, plant, and environmental health. This integrated framework is formally embodied in the quadripartite collaboration between the Food and Agriculture Organization of the United Nations (FAO), the World Health Organization (WHO), the World Organization for Animal Health (WOAH), and the United Nations Environment Programme (UNEP) [3,4].

This crisis is primarily driven by the intensive and often inappropriate use of antibiotics, which accelerates natural selection and promotes the emergence of resistant bacteria [1,3,5]. It is estimated that, within the next three decades, AMR could cause up to 10 million deaths annually, potentially becoming more lethal than cancer [1,2]. Beyond its profound health implications, AMR also imposes a considerable economic burden, estimated at approximately €11.7 billion annually in Europe [6].

The widespread use of antibiotics in intensive animal farming is widely recognized as a major driver of AMR. It is estimated that between 50% and 80% of antibiotics used globally are administered in this sector [7]. Furthermore, antibiotic consumption in animal agriculture has historically exceeded that in human medicine [8,9]. However, recent European surveillance data show a positive trend, with antibiotic consumption in food-producing animals declining by 44% between 2014 and 2021 and now falling below levels reported for humans across EU/EEA countries [10]. Despite these reductions, food-producing animals can still act as important reservoirs of antibiotic-resistant bacteria. Environmental contamination resulting from the application of organic fertilizers derived from animal manure may introduce antibiotic residues, antibiotic-resistant bacteria, and resistance genes into agricultural soils, water systems, and the food chain [11].

Among the microbial agents of concern, Escherichia coli is a Gram-negative bacterium naturally present in the gastrointestinal tract of animals and humans [12], widely used as an indicator organism to monitor resistance trends. Escherichia coli is a major pathogen in poultry production and a significant cause of foodborne and other types of infections in humans [13,14]. This species is frequently used as an indicator of antimicrobial resistance due to its wide distribution and its ability to harbor multiple resistance genes located on mobile genetic elements, such as plasmids, transposons, and integrons, thereby acting as a reservoir of antimicrobial resistance genes for pathogenic strains [10,15]. In particular, β-lactamases, enzymes capable of hydrolyzing β-lactam antibiotics, are among the most important mechanisms contributing to resistance in E. coli. Moreover, E. coli may acquire resistance through the development of mutations when exposed to antimicrobial selective pressures exerted on their host [15,16]. Several studies [15,17] have evaluated the resistance profiles of E. coli isolates obtained from chickens and chicken meat, reporting resistance to aminoglycosides, β-lactam antibiotics, including penicillin and cephalosporins, and fluoroquinolones in all tested isolates. The excessive and often inappropriate use of antibiotics has favored the emergence of resistant and multidrug-resistant (MDR) strains [18], defined as those resistant to three or more classes of antimicrobials [19]. MDR E. coli strains pose a public health threat because they can be transmitted from animals to humans through the consumption of contaminated animal products (e.g., meat, milk, and eggs), direct contact with animals, or exposure to contaminated environments [20].

Studies describing AMR profiles of bacteria isolated from commercial poultry and intensive production systems are relatively limited in Portugal [21,22,23], and the role of autochthonous Portuguese laying hens as potential reservoirs of antibiotic-resistant E. coli remains largely unexplored. Portugal is home to four autochthonous breeds, namely Branca, Pedrês Portuguesa, Amarela and Preta Lusitânica (Figure 1). The minimal or absent use of antibiotics observed in these breeds is primarily associated with the production systems in which they are raised. These breeds are typically raised in extensive farming systems, including in low-input production or backyard systems, where chickens are allowed free-range access during the day and are housed at night, with feed and water provided indoors to supplement scavenging. Production usually occurs on a small scale, within household poultry keeping that relies on family labour and, whenever possible, locally available feed resources, and are commonly referred to as family poultry systems [24,25,26]. The relevance of these native Portuguese laying hens is multifaceted. The production systems used for their breeding make them a valuable model for studying the natural ecology of AMR in the absence of strong selective pressure. Knowledge generated from these dual-purpose breeds (meat and eggs) is essential for establishing baseline AMR data and supporting the implementation of more sustainable livestock practices [27]. Beyond their biological relevance, with unique physical and genetic characteristics, Portuguese native chicken breeds hold significant historical, cultural, and socio-economic value, being deeply rooted in the traditions of small family farms. However, the intensification of industrial poultry farming throughout the 20th century led to the marginalization of these breeds, resulting in severe population declines and placing several of them at risk of extinction, particularly the Branca and Preta Lusitânica breeds [24].

In line with the One Health approach, this study aimed to assess the occurrence of E. coli in cloacal and eggshell samples from these four autochthonous Portuguese laying hen breeds with minimal antibiotic exposure, followed by the characterization of their phenotypic antimicrobial resistance profiles and the detection of associated resistance genes. Sampling both cloaca and eggshells allows assessment of gut colonization and potential contamination of food products, providing an integrated view of the ecological dynamics of AMR in these native breeds and their potential implications for animal, human, and environmental health.

2. Materials and Methods

2.1. Sample Origin, Collection, and Characterization

A total of 248 swab samples (124 cloacal and 124 eggshells) from the four autochthonous Portuguese laying hen breeds, including Branca (n = 64), Pedrês Portuguesa (n = 64), Amarela (n = 64), and Preta Lusitânica (n = 56), collected from 31 flocks across 12 farms were analyzed in this study.

The characteristics of farms and flocks, as well as the original sample collection procedures, were previously described by Miranda et al. [28]. Briefly, all farms operated under extensive rearing systems characterized by partial outdoor access and small flock sizes (≤50 birds per farm), with a male-to-female ratio of 1:10–15. No antibiotic administration was reported during the three months preceding the sampling period. From each farm, which included separate flocks of one to four breeds, four cloacal and four eggshell samples were randomly collected using sterile swabs [28].

2.2. Isolation and Identification of Escherichia coli

The stored cloacal and eggshell samples were inoculated into Brain Heart Infusion (BHI) broth at a 1:10 (v/v) ratio and incubated at 37 °C for 18 to 24 h. Following pre-enrichment, 20 μL (corresponding to two loops) of each BHI culture was streaked onto MacConkey agar plates using the streak plate technique to allow colony isolation. MacConckey agar is both selective and differential: it contains crystal violet and bile salts to inhibit the growth of Gram-positive bacteria, while lactose and neutral red enable differentiation based on lactose fermentation. Plates were incubated at 37 °C for 24 h.

From each positive MacConkey plate, one to three representative pink colonies, indicativee of lactose fermenters, such as E. coli, were selected for further confirmation on Chromogenic Coliform Agar (CCA). This medium is selective and differential for coliform bacteria, and E. coli typically forms distinctive blue colonies, with or without a pink halo. These plates were incubated at 37 °C for 24 h. Presumptive E. coli isolates were then plated onto BHI agar for further analyses. DNA was extracted using the boiling method, and multiplex PCR [29] was performed for both species confirmation and the detection of antibiotic resistance genes. PCR was carried out using the Xpert Fast Hotstart Mastermix (GRiSP Research Solutions, Porto, Portugal), according to the manufacturer’s instructions. Species identification was confirmed using the uidA and uspA primers [29], while additional primers targeting ESBL genes were used as described below (Table 1).

2.3. Antimicrobial Susceptibility Testing and ESBL Detection

The antimicrobial susceptibility of E. coli isolates was evaluated using the disk diffusion method on Mueller–Hinton II Agar (Alliance Bio Expertise, Guipry Messac, France), following the guidelines of the European Committee on Antimicrobial Susceptibility Testing (EUCAST, 2025) and the Clinical and Laboratory Standards Institute (CLSI, 2020) [30,31]. Escherichia coli ATCC 25922 was used as the quality-control strain in all assays. Bacterial suspensions were prepared in 0.9% NaCl and adjusted to a turbidity equivalent to the 0.5 McFarland standard using a densitometer. These suspensions were inoculated onto Mueller–Hinton II agar plates using sterile swabs and the lawn culture technique, and antibiotic-impregnated disks were subsequently placed on the surface using an automatic dispenser.

Table 1. Nucleotide sequences with annealing temperature (AT) and expected amplicon size, used in this study.

Primer	Nucleotide Sequence (5′ to 3′)	Size (bp)	AT (°C)	Reference
blaOXA-1	ACACAATACATATCAACTTCGC AGTGTGTTTAGAATGGTGATC	813	61	Steward et al. [32]
blaCTX-M	ATGTGCAGYACCAGTAARGT TGGGTRAARTARGTSACCAGA	593	50	Lim et al. [33]
blaSHV	GGTTATGCGTTATATTCGCC TTAGCGTTGCCAGTGCTC	867	60	Lim et al. [33]
blaTEM	GAGTATTCAACATTTTCGT ACCAATGCTTAATCAGTGA	857	50	Maynard et al. [34]
tet(B)	CTCAGTATTCCAAGCCTTTG CTAGCACTTGTCTCCTGTT	416	57	Guardabassi et al. [35]
sul2	CGGCATCGTCAACATAACCT TGTGCGGATGAAGTCAGCTC	721	55	Fazel et al. [36]
aac(3′)-IV	CTTCAGGATGGCAAGTTGGT TCATCTCGTTCTCCGCTCAT	286	60	Sáenz et al. [37]
aac(6′)-Ib-cr	CGGCATCGTCAACATAACCT TGTGCGGATGAAGTCAGCTC	482	55	Park et al. [38]

Table 1. Nucleotide sequences with annealing temperature (AT) and expected amplicon size, used in this study.

Primer	Nucleotide Sequence (5′ to 3′)	Size (bp)	AT (°C)	Reference
blaOXA-1	ACACAATACATATCAACTTCGC AGTGTGTTTAGAATGGTGATC	813	61	Steward et al. [32]
blaCTX-M	ATGTGCAGYACCAGTAARGT TGGGTRAARTARGTSACCAGA	593	50	Lim et al. [33]
blaSHV	GGTTATGCGTTATATTCGCC TTAGCGTTGCCAGTGCTC	867	60	Lim et al. [33]
blaTEM	GAGTATTCAACATTTTCGT ACCAATGCTTAATCAGTGA	857	50	Maynard et al. [34]
tet(B)	CTCAGTATTCCAAGCCTTTG CTAGCACTTGTCTCCTGTT	416	57	Guardabassi et al. [35]
sul2	CGGCATCGTCAACATAACCT TGTGCGGATGAAGTCAGCTC	721	55	Fazel et al. [36]
aac(3′)-IV	CTTCAGGATGGCAAGTTGGT TCATCTCGTTCTCCGCTCAT	286	60	Sáenz et al. [37]
aac(6′)-Ib-cr	CGGCATCGTCAACATAACCT TGTGCGGATGAAGTCAGCTC	482	55	Park et al. [38]

The antibiotic panel included representatives of five major antimicrobial classes: β-lactams (ampicillin 10 µg, amoxicillin + clavulanic acid 20 + 10 µg, cefotaxime 30 µg, aztreonam 30 µg, imipenem 5 µg), folate pathway antagonists (trimethoprim + sulfamethoxazole 1.25 + 23.75 µg), quinolones (ciprofloxacin 5 µg), aminoglycosides (gentamicin 10 µg, amikacin 30 µg), and tetracyclines (tetracycline 30 µg).

Cultures were incubated at 37 °C for 18 to 24 h. After incubation, the diameters of the inhibition zones surrounding each antibiotic disk were measured in millimeters to determine the susceptibility profile of each isolate. Screening for extended-spectrum beta-lactamase (ESBL) production was performed using the double-disk synergy test. In this assay, an amoxicillin-clavulanic acid disk was placed at the center of the plate, with cefotaxime and aztreonam disks positioned 25 mm apart. The presence of an enhanced inhibition zone, known as keyhole effect, between the central and surrounding disks was interpreted as indicative of ESBL production.

2.4. Detection of Antibiotic Resistance Genes

Detection of antibiotic resistance genes was performed in E. coli isolates selected based on their phenotypic antibiotic resistance profiles and breed/farm’s distribution.

Genomic DNA was extracted from E. coli isolates using the boiling method. The detection of antimicrobial resistance genes was performed by PCR using the Xpert Fast Hotstart Mastermix (GRiSP Research Solutions, Porto, Portugal) according to the manufacturer’s instructions. Amplification was conducted in an thermocycler (MJ Mini, Bio-Rad, Hercules, CA, USA) under the following conditions: an initial denaturation at 95 °C for 3 min, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing temperature ranging from 50 °C to 60 °C (depending on the specific primer pair used) for 15 s, and extension at 72 °C for 2 s. A final extension step was performed at 72 °C for 5 min. The target genes included bla_OXA-1, bla_CTX-M, bla_SHV, and bla_TEM (associated with β-lactam resistance), tet(B) (tetracyclines), sul2 (sulfamethoxazole), aac(3′)-IV (aminoglycosides), and aac(6′)-Ib-cr (fluoroquinolones). Primer sequences and annealing temperatures were based on previously published studies [32,33,34,35,36,37,38] are presented in Table 1.

PCR products were analyzed by electrophoresis on a 1.2% agarose gel containing 3 µL of GelRed^® fluorescent dye (Biotium, Fremont, CA, USA). A molecular weight marker (NzyDNA Ladder V, Nzytech, Lisbon, Portugal) was used to estimate the size of the amplified fragments. Electrophoresis results were visualized using a ChemiDoc transilluminator (Bio-Rad, Hercules, CA, USA). Positive controls, kindly provided by the CCP, Culture Collection of Porto, were included in all PCR assays.

3. Results

3.1. Breed Distribution of Escherichia coli Isolates

A total of 81 E. coli isolates were successfully obtained from 53 out of 248 (21.4%) samples, including 28/124 (22.6%) cloacal samples and 25/124 (20.2%) eggshell samples collected from laying hens of the four autochthonous Portuguese breeds. Most samples (78.6%, 195/248) showed no growth compatible with E. coli. Among cloacal samples, 35 E. coli isolates were identified: Preta Lusitânica (n = 7), Amarela (n = 11), Pedrês Portuguesa (n = 5), and Branca (n = 12). From eggshell samples, 46 E. coli isolates were obtained: Preta Lusitânica (n = 15), Amarela (n = 9), Pedrês Portuguesa (n = 10), and Branca (n = 12).

3.2. Phenotypic Antimicrobial Resistance

All obtained E. coli isolates, originating from different breeds and farms, were subjected to antimicrobial susceptibility testing. Among the 35 cloacal isolates, most (97.1%, 34/35) were resistant to gentamicin. All isolates (100.0%, 28/28) from the Amarela, Pedrês Portuguesa, and Branca breeds exhibited resistance, while 85.7% (6/7) of isolates from Preta Lusitânica hens were resistant. Resistance to tetracycline was observed in 31.4% (11/35) of isolates, distributed as follows: Amarela, 8.3% (1/12); Pedrês Portuguesa, 60.0% (3/5); and Branca, 63.6% (7/11). Resistance to ampicillin was detected in 14.3% (5/35) of isolates: Preta Lusitânica, 14.3% (1/7); Amarela, 16.7% (2/12); and Branca, 18.2% (2/11). Isolates from Amarela, Pedrês Portuguesa, and Branca breeds also showed resistance to trimethoprim/sulfamethoxazole (11.4%, 4/35).

Resistance to amoxicillin/clavulanic acid (5.7%, 2/35) and amikacin (8.6%, 3/35) was only observed in isolates from the Amarela and Branca breeds. It is important to note that no resistance was observed to cefotaxime, imipenem, aztreonam, or ciprofloxacin (Figure 2). Overall, the results indicate that phenotypic resistance varied among breeds, with gentamicin resistance being the most prevalent across all cloacal isolates.

Among the 46 eggshell isolates, 95.7% (44/46) were resistant to gentamicin. All isolates from Preta Lusitânica and Branca breeds exhibited resistance, whereas resistance in Pedrês Portuguesa and Amarela isolates was 90.0% (9/10) and 88.9% (8/9), respectively.

Tetracycline resistance was detected in 41.3% (19/46) of isolates across all four breeds. Resistance to amikacin (4.3%, 2/46) was observed only in isolates from Preta Lusitânica, and resistance to amoxicillin/clavulanic acid (4.3%, 2/46) was detected in Amarela isolates. Trimethoprim/sulfamethoxazole resistance (2.2%, 1/46) occurred exclusively in an isolate from the Branca breed Ampicillin resistance was observed in 23.9% (11/46) of isolates: Preta Lusitânica, 13.3% (2/15); Amarela, 55.6% (5/9); and Pedrês Portuguesa, 40.0% (4/10). No resistance was detected to cefotaxime, imipenem, aztreonam, or ciprofloxacin (Figure 3). Overall, these results indicate that gentamicin resistance was highly prevalent across eggshell isolates, while resistance to other antimicrobials varied among breeds.

Resistance to at least one antibiotic and an MDR phenotype were observed in 98.0% (79/81) and 16.0% (13/81), respectevely, among all recovered E. coli isolates, whereas only two isolates exhibited pan-susceptibility.

Among cloacal E. coli isolates, only one isolate (2.9%, 1/35) from the Preta Lusitânica breed was susceptible to all antibiotics tested. In contrast, all isolates from the remaining breeds exhibited resistance to at least one antibiotic. The MDR phenotype was observed in 14.3% (5/35) of isolates: Branca, 27.3% (3/11) and Amarela, 16.7% (2/12) (Figure 4a). Among eggshell E. coli isolates, only one isolate (2.2%, 1/46) obtained from the Amarela breed was susceptible to all antibiotics, while isolates from the other breeds showed resistance to at least one antibiotic. The MDR phenotype was observed in 17.4% (8/46) of eggshell isolates: Amarela, 55.6% (5/9); Preta Lusitânica, 13.3% (2/15); and Pedrês Portuguesa 10.0% (1/10) (Figure 4b).

No phenotypic resistance to third-generation cephalosporins, such as cefotaxime, was observed in any isolates. Consistently, no ESBL-producing E. coli were detected by the double-disk synergy test. Overall, these results indicate that MDR E. coli isolates were relatively uncommon among both cloacal and eggshell samples, with breed-specific differences in resistance patterns.

3.3. Antibiotic Resistance Genes

Molecular screening for antibiotic resistance genes was performed in selected E. coli isolates and compared with their phenotypic profiles. Although no ESBL phenotype was detected using the double-disk synergy testing, screening for ESBL-related genes (bla_TEM, bla_CTX-M, bla_SHV) was conducted. Representative isolates from different breeds and farms were analyzed, including 20/35 cloacal isolates for the dectection of resistance genes. Screening for bla_TEM, bla_CTX-M, bla_OXA-1, and bla_SHV genes was performed in 44/46 cloacal isolates, and screening of tet(B), sul2, aac(6′)-Ib-cr, and aac(3′)-IV genes was conducted in 21/46 eggshell isolates.

Among cloacal E. coli isolates, not all targeted resistance genes were detected. No isolates harbored the bla_OXA-1, bla_SHV, tet(B), or aac(6′)-Ib-cr genes. Overall, 13/20 (65.0%) cloacal isolates carried one to three of the analyzed resistance genes. The most frequently detected genes were bla_TEM (45.0%, 9/20) and sul2 (45.0%, 9/20). The bla_CTX-M gene was identified in two isolates (10.0%), obtained from Preta Lusitânica and Pedrês Portuguesa breeds. The aac(3′)-IV gene was detected in two isolates (10.0%) from Preta Lusitânica and Branca breeds (Figure 5). These findings indicate that despite the absence of ESBL phenotypes, cloacal E. coli isoaltes may act as a reservoir of hidden or low-level β-lactam resistance.

Overall, the genotypic results were consistent with the phenotypic resistance patterns. Several cloacal E. coli isolates exhibiting phenotypic antibiotic resistance also carried one or more corresponding antibiotic resistance genes (

Table 2. Phenotypic and genotypic characteristics of E. coli isolates obtained from cloacal samples (n = 20).

Isolate	Breed	Antibiotic Resistance Phenotype *	Antibiotic Resistance Genotype
S37.F1	Preta Lusitânica	GN	blaTEM, sul2
S41.F1	Preta Lusitânica	AMP, GN	Negative
S65.F1	Preta Lusitânica	GN	Negative
S69.F1	Preta Lusitânica	GN	sul2, aac(3′)-IV
S72.F1	Preta Lusitânica	GN	blaTEM, blaCTX-M, sul2
S25.F1	Amarela	AMP, GN, TE, SXT	sul2
S45.F1	Amarela	GN, TE	Negative
S120.F2	Amarela	AMC, GN	blaTEM, sul2
S122.F1	Amarela	AMP, GN, SXT	blaTEM
S124.F2	Amarela	GN, AK	Negative
S9.F1	Pedrês Portuguesa	GN, TE	blaTEM
S104.F1	Pedrês Portuguesa	GN	Negative
S114.F1	Pedrês Portuguesa	GN, SXT	blaTEM
S114.F2	Pedrês Portuguesa	GN, TE	blaTEM, blaCTX-M, sul2
S116.F1	Pedrês Portuguesa	GN, TE	blaTEM, sul2
S53.F3	Branca	GN, TE, SXT	sul2
S60.F1	Branca	GN, AK	Negative
S61.F1	Branca	GN, TE	sul2, aac(3′)-IV
S106.F1	Branca	AMP, AMC, GN, TE	blaTEM
S112.F1	Branca	AMP, GN, TE	Negative

* AMP ampicillin; AMC, amoxicillin/clavulanic acid; GN, gentamicin; AK, amikacin; TE, tetracycline; SXT, trimethoprim/sulfamethoxazole.

).

As observed for cloacal isolates, none of the isolates carried the bla_OXA-1 or bla_SHV genes. The most frequently detected genes were bla_TEM (36.4%, 16/44) and sul2 (47.6%, 10/21). The bla_CTX-M gene was found in 11.4% (5/44) of isolates obtained from the Preta Lusitânica and Branca breeds (Figure 6a). The Pedrês Portuguesa breed was the only one with isolates carrying the tet(B) (20.0%, 1/5) and aac(6′)-Ib-cr (20.0%, 1/5) genes. The aac(3′)-IV gene was detected in one isolate from the Branca breed (20.0%, 1/5) (Figure 6b). Overall, 16/21 (76.2%) eggshell isolates harbored one to three of the analyzed resistance genes.

Genotypic profiles were generally consistent with phenotypic resistance, and several antibiotic-resistant E. coli strains from eggshell samples harbored one or more antibiotic resistance genes (

Table 3. Phenotypic and genotypic characteristics of E. coli isolates obtained from eggshell samples (n = 21).

Isolate	Breed	Antibiotic Resistance Phenotype *	Antibiotic Resistance Genotype
S66.02	Preta Lusitânica	GN, TE	Negative
S68.01	Preta Lusitânica	GN	blaTEM, blaCTX-M, sul2
S69.03	Preta Lusitânica	AK, GN	blaTEM, blaCTX-M, sul2
S96.01	Preta Lusitânica	AMP, GN, TE	blaCTX-M
S96.02	Preta Lusitânica	AMP, AK, GN, TE	Negative
S117.01	Preta Lusitânica	GN	Negative
S75.01	Amarela	AMP, GN, TE	Negative
S119.01	Amarela	GN, TE	sul2
S120.02	Amarela	AMP, GN, TE	Negative
S121.01	Amarela	AMP, AMC, GN, TE	Negative
S124.01	Amarela	AMP, AMC, GN, TE	Negative
S79.01	Pedrês Portuguesa	AMP, GN	sul2
S113.01	Pedrês Portuguesa	GN	blaTEM, sul2
S114.01	Pedrês Portuguesa	AMP, GN, TE	Negative
S115.01	Pedrês Portuguesa	GN, TE	blaTEM, sul2, aac(6′)-Ib-cr
S116.04	Pedrês Portuguesa	AMP, TE	tet(B)
S105.01	Branca	GN, TE	blaCTX-M
S107.01	Branca	GN	blaTEM, sul2
S109.01	Branca	GN, TE	blaTEM, sul2, aac(3′)-IV
S110.01	Branca	GN, TE, STX	sul2
S111.01	Branca	GN, TE	blaTEM, sul2

* AMP, ampicillin; AMC, amoxicillin/clavulanic acid; GN, gentamicin; AK, amikacin; TE, tetracycline; SXT, trimethoprim/sulfamethoxazole.

).

4. Discussion

This study is pioneering in evaluating the AMR profile of E. coli isolates obtained from autochthonous Portuguese laying hens. It fills a critical knowledge gap, as no previous studies have specifically targeted these native breeds. These autochthonous chickens (Branca, Pedrês Portuguesa Amarela, and Preta Lusitânica) are typically raised in extensive or semi-extensive production systems characterized by low inputs and minimal or no antibiotic administration. Nevertheless, the detection of antibiotic-resistant and MDR E. coli in all breeds studied emphasizes that AMR is not limited to intensive production systems and can persist even in ecological niches with low direct antibiotic selective pressure.

Poultry production is widely recognized as an important reservoir of resistant-bacteria, including E. coli, which poses a growing threat to global public health. These bacteria may be transmitted to humans through the food chain, particularly via contaminated eggs, or through direct contact with animals, contaminated fomites, and the surrounding environment [1,20]. In Portugal, studies investigating AMR in native poultry raised under extensive systems remain scarce, with most research focusing predominantly on broilers from intensive production systems, which limits direct comparisons [39]. Despite this, one Portuguese study conducted in breeding and commercial laying hens reared under intensive production reported that 81.0% of E. coli isolates obtained from fecal samples were resistant to at least one antimicrobial [40], a finding that corroborates the results of the present study.

In the present work, resistance to at least one antibiotic was observed in 98.0% of the isolates, including both cloacal and eggshell samples. The prevalence of MDR phenotypes was 17.4% among eggshell isolates from the Preta Lusitânica, Branca, and Pedrês Portuguesa breeds, and 14.3% among cloacal isolates from the Amarela and Branca breeds. These MDR values are lower than those reported in some international studies, such as the 34.9% MDR observed in E. coli isolated from cloacal swabs and internal organs of backyard chickens in small scale poultry farms [41], and the 66.9% reported for Enterobacteriaceae isolated from liver samples of commercial broilers and backyard chickens [42]. This difference may reflect the minimal antibiotic exposure characteristic of Portuguese extensive production systems.

Worldwide, tetracyclines, fluoroquinolones, sulfonamides/trimethoprim, macrolides, aminoglycosides (e.g., neomycin and gentamicin), polymyxins (colistin), and β-lactams (amoxicillin/penicillin) are among the most frequently used antimicrobial classes in chickens. This pattern is consistent with studies reporting resistance of E. coli to aminoglycosides, β-lactams (including penicillin and cephalosporins), and fluoroquinolones in poultry isolates [15,17]. Gentamicin is reported as one of the most widely used antimicrobials in broiler production, where it is often administered prophylactically to day-old chicks to prevent bacterial infections [43]. Gentamicin belongs to the aminoglycoside class and is classified by the WHO as a critically important antimicrobial, defined as a drug essential for the treatment of serious human infections with limited or no effective alternatives. In the present study, resistance levels to gentamicin were substantially higher than those previously reported in Portuguese poultry (17.0%) [44], as well as, in international studies, such as 23.4% [42], or 39.5% for gentamicin and 29.1% for tetracycline [41]. In addition, resistance to gentamicin is common among enteric bacteria associated with poultry. Avian pathogenic E. coli isolates obtained from avian colibacillosis cases in North Georgia between 1996 and 2000, showed that 69% of isolates were resistant to gentamicin [45]. In Nigeria, a recent study reported that isolates from local chickens in smallholder poultry systems showed higher resistance to aminoglycosides (22.6%) compared with tropically adapted chickens [46]. A German study conducted in commercial broiler flocks between 2013 and 2018 reported aminoglycosides (25.6%) and lincosamides (25.6%) as the most frequently used antimicrobial classes, followed by polypeptides (21.4%) and β-lactams (16.2%). During this period, the use of aminoglycosides and lincosamides tended to increase, whereas fluoroquinolones, macrolides and polypeptides have decreased [47]. Importantly, Price et al. [43] demonstrated that poultry workers were 32 times more likely to carry gentamicin-resistant E. coli than individuals in the general community, highlighting the occupational exposure associated with the broiler production environment. Although aminoglycosides are poorly absorbed in the gastrointestinal tract and gentamicin is not administered orally, limiting community-level selection pressure, this markedly increased carriage among poultry workers supports the hypothesis that direct contact with poultry environments plays a key role in the dissemination of gentamicin-resistant E. coli [43].

Despite minimal antibiotic use in these low-input production systems, the persistent presence of resistant E. coli, particularly to gentamicin, suggests that resistance acquisition of E. coli in Portuguese native breeds may originate from environmental sources or external contamination. Environmental factors such as soil, water, feed, and human contact may act both as sources of contamination for animals and eggshells and as recipients of resistant bacteria shed by the colonized hens [20]. Scavenging behavior, an adaptive trait of local chickens shaped by multiple genetic factors, contributes to their survivability and resilience compared with exotic and improved breeds. However, this characteristic may increase exposure to environments contaminated with antibiotics through direct sources, such as feed and water, or indirect sources, including human sewage and wastewater [46,48].

In autochthonous chickens, antibiotic resistance is likely driven by antibiotic selection pressure from direct or environmental sources, together with a dense and diverse gut microbiome that acts as a reservoir of antibiotic resistance genes, and efficient horizontal gene transfer mediated by mobile genetic elements. These processes are further modulated by the production system ecotype and scavenging behavior, which increase contact with contaminated environments and resistance determinants. Overall, these findings highlight the complex ecological interactions underlying antimicrobial resistance dissemination. Consequently, native laying hens may serve as reservoirs of resistant E. coli and resistance genes, with potential implications for the human food chain through contaminated eggs.

A particularly encouraging finding was the absence of resistance to imipenem, cefotaxime, aztreonam, and ciprofloxacin in all isolates. The full susceptibility to imipenem and aztreonam, antibiotics generally restricted to human medicine, is especially noteworthy.

Differences were observed between sample types. Eggshell isolates showed a slightly higher prevalence of resistance to ampicillin (23.9%) and tetracycline (41.3%), compared with cloacal isolates (14.3% and 31.4%, respectively). This discrepancy suggests that environmental contamination of eggs during or after laying may contribute to the observed resistance patterns. Molecular screening for ESBL-related genes (bla_TEM, bla_CTX-M, bla_SHV) was conducted despite the absence of an ESBL phenotype detected by the Double-Disk Synergy Test. The presence of these genes without phenotypic expression suggests silent gene carriage or the production of narrow-spectrum β-lactamases, a phenomenon also reported in another Portuguese study [40]. Some β-lactamase genes may be expressed at low levels, be non-functional, or fail to produce a detectable phenotype under the tested conditions. Moreover, certain genes, such as bla_TEM, encode narrow-spectrum β-lactamases that do not confer an ESBL phenotype but may still contribute to antibiotic resistance. Thus, molecular screening provides complementary information and improves the detection of hidden resistance potential.

Molecular analysis identified sul2 and bla_TEM as the most frequently detected genes in isolates from both cloacal and eggshell samples. Although detected at low frequency, the bla_CTX-M gene followed a similar trend. Notably, genes such as tet(B) and aac(6′)-Ib-cr were detected exclusively in eggshell isolates from the Pedrês Portuguesa breed, suggesting a contamination source specifically associated with the eggshell environment.

This study provides novel insights into antimicrobial resistance in autochthonous chicken breeds, which are under-studied and poorly characterized in the literature. Some limitations should still be acknowledged. First, the relatively small number of isolates and farms included may limit the generalizability of the findings to broader poultry populations. Second, a selected set of antimicrobial resistance genes was screened, and other resistance determinants may exist that were not investigated, limiting the completeness of the molecular characterization. Third, sampling was conducted at a single time point, which may not capture temporal or seasonal variations in antimicrobial resistance. Fourth, only lactose-fermenting colonies were included, so atypical E. coli isolates with delayed or absent lactose fermentation may have been missed, potentially affecting the detection of some resistance phenotypes. Approximately 10% of E. coli strains may exhibit delayed or absent lactose fermentation, which could affect colony color interpretation on selective media [49,50]. Despite these limitations, the study fills an important gap in knowledge, providing baseline data on antimicrobial resistance in autochthonous breeds and supporting future research in other breeds and production systems.

5. Conclusions

This study represents the first assessment of AMR in E. coli isolates from autochthonous Portuguese laying hen breeds, addressing a critical knowledge gap, as previous investigations have predominantly focused on intensive production systems. The detection of resistant E. coli strains, even under low-input rearing conditions, indicates that AMR can persist in traditional systems, although the absence of resistance to critically important antimicrobials and the low prevalence of MDR are encouraging. These findings highlight the value of native breeds as models for understanding baseline AMR dynamics under reduced selective pressure and emphasize the importance of continued surveillance.

Beyond their microbiological relevance, Portuguese autochthonous chicken breeds represent an important genetic, cultural, and socio-economic heritage, deeply embedded in traditional farming systems and rural livelihoods. Many of these breeds are currently at risk of extinction due to population decline and genetic erosion, making their conservation a priority not only for biodiversity preservation but also for the sustainability of traditional agricultural systems. The present results demonstrate that conservation efforts should be aligned with AMR monitoring, as these populations provide unique insights into resistance ecology in low-intensity agroecosystems.

Overall, the study reinforces the importance of integrated strategies grounded in the One Health approach, recognizing the interconnectedness of human, animal, and environmental health. Understanding AMR dynamics in autochthonous poultry contributes to the development of informed mitigation strategies that support sustainable livestock production, protect public health, and promote the preservation of culturally significant and endangered poultry breeds.