First detection of resistance genes and virulence factors in Escherichia coli and Salmonella spp in Togo: the case of imported chicken and frozen by-products

Abstract

Background

The increasing importation of frozen poultry into Togo raises concerns about the microbiological safety and antimicrobial resistance of associated pathogens. Despite the public health risks posed by resistant foodborne bacteria, data on resistance profiles, resistance genes, and virulence factors in imported frozen chickens in Togo remain limited. This study aims to address this gap by characterizing these factors in pathogenic strains isolated from imported poultry.

Methods

A cross-sectional prospective study was undertaken to assess the microbiological quality and resistance profiles of imported poultry products. Samples were collected from seven cold storage facilities located within the Golfe prefecture of the Greater Lomé metropolitan area. In total, 285 poultry meat and cut samples were analyzed following standardized AFNOR microbiological protocols. Isolated Salmonella spp. and Escherichia coli strains underwent antibiotic susceptibility testing using the disk diffusion method, adhering to the guidelines established by the Comité de l’Antibiogramme de la Société Française de Microbiologie (CA-SFM). Furthermore, polymerase chain reaction (PCR) assays were employed to identify genetic determinants of antibiotic resistance and virulence factors in the bacterial isolates.

Results

Microbiological analysis revealed a prevalence of Escherichia coli of 32.98%, while Salmonella spp. were detected in 2.46% of the samples. Antibiotic susceptibility testing demonstrated resistance among isolates to several beta-lactams and quinolones. Specifically, resistance to cefoxitin was observed in 14.28% of strains, whereas resistance to cefalexin, cefuroxime, ceftazidime, ceftriaxone, and nalidixic acid was uniformly detected at a prevalence of 28.57%. Among the E. coli isolates, 9.44% exhibited multidrug resistance to both beta-lactams and quinolones. Molecular characterization identified class 1 integrons in 17.6% of isolates, with gene cassettes predominantly harboring aadA1 and dfr1, which encode resistance to streptomycin, spectinomycin, and trimethoprim. Notably, class 2 and class 3 integrons were absent. Additionally, the plasmid-mediated qnrB gene was detected in 5.9% of isolates. The study also documented the emergence of resistance to third-generation cephalosporins (C3G), primarily associated with extended-spectrum beta-lactamase (ESBL) production, as evidenced by the presence of blaCTX (35.3%) and blaTEM (58.8%) genes among ESBL-producing strains.

Conclusions

This study reveals a notable presence of antimicrobial-resistant Escherichia coli and Salmonella in imported frozen poultry in Togo, highlighting significant public health risks. The findings call for improved surveillance and stricter control measures to prevent the spread of resistant pathogens via the food supply.

Clinical trial number

Not applicable.

Introduction

Food quality and antimicrobial resistance have become significant concerns for consumers globally [1]. In particular, the microbiological safety of chicken meat warrants careful examination, as it represents a potential vector for transmitting resistant pathogenic bacteria to humans, thereby posing a public health risk [2]. Studies conducted in Africa have consistently reported a high prevalence of Escherichia coli and Salmonella species in poultry products [3]. This widespread presence of pathogenic bacteria is further complicated by elevated levels of resistance to commonly used antibiotics [4]. The mechanisms underlying antimicrobial resistance are diverse and include enzymatic drug inactivation, target site modification, active efflux, and reduced membrane permeability. These resistance traits may emerge either through spontaneous genetic mutations or via the horizontal acquisition of foreign genetic material [5, 6].

Bacteria possess various genetic elements, such as plasmids and transposons, which facilitate the horizontal transfer of antibiotic resistance genes across different bacterial genera. In the 1980s, a novel class of genetic elements termed integrons was identified; these structures are capable of acquiring and excising antibiotic resistance genes. Integrons carry resistance determinants embedded within mobile gene cassettes, enabling rapid dissemination of resistance traits [7, 8]. In Togo, a clinical study conducted in Lomé reported an exceptionally high prevalence (96%) of extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae [9]. However, comparable molecular data on resistance genes in foodborne pathogens, particularly Escherichia coli and Salmonella spp., remain scarce. The present study aims to address this gap by characterizing the antibiotic resistance profiles and associated genetic determinants in these bacteria isolated from food sources. Such data are crucial for optimizing therapeutic management of infections and informing strategies to control the spread of antimicrobial resistance.

Methods

Period, study area and sample collection

This cross-sectional prospective study was conducted from November 2020 to December 2023 in southern Togo, specifically within the Golfe prefecture of the Greater Lomé commune. Samples were collected using a probabilistic random sampling approach from seven major cold storage facilities in Lomé, which serve as primary distributors of frozen products. The geographic locations of the sampling sites are illustrated in Fig. 1.

Fig. 1.

Geographical map of Lomé (Golfe Prefecture) indicating the locations where samples were collected (sites A–G).

Source: QGIS Date: March 2023 Author: Kossi Touglo

To ensure adequate representativeness, the sampling interval (k) was calculated by dividing the total population size (N) by the predetermined sample size. An initial starting point was randomly selected within the range of 1 to k. Subsequently, every kth individual from this starting point was systematically included in the sample. Using this method, a total of 285 samples, comprising frozen chickens and their derived products, were collected.

Microbiological analyses

The preparation of the stock solution and successive dilutions was carried out in accordance with ISO 7218 (2007), which provides general requirements and guidelines for microbiological examinations of food and animal feeding stuffs. The specific standards followed include ISO 6579-1:2017 and its amendment Amd1:2020, which describe the horizontal method for the detection, enumeration, and serotyping of Salmonella species. Furthermore, the enumeration of β-glucuronidase-positive Escherichia coli was performed according to ISO 16649-2 (April 2001), which specifies the horizontal method using direct inoculation on Tryptone Bile X-glucuronide (TBX) agar, a selective medium that facilitates the identification and enumeration of E. coli through the detection of β-glucuronidase activity.

The selection of bacterial strains and dilution criteria adhered to standards established for meat products, as outlined in the Fédération du Commerce et de la Distribution (FCD, 2022) guidelines, which specify a maximum acceptable microbial limit (M) of 10 m [10]. Following verification of consistency between enumeration results obtained from duplicate plates, microbial counts were calculated using the following formula:

N represents the estimated number of microorganisms; V denotes the volume of the inoculum; a and b correspond to the colony counts obtained from the first and second dilutions, respectively; d indicates the dilution factor;0.1 refers to the specific dilution volume used; CFU stands for Colony-Forming Units.

Antibiogram and search for resistance genes

Antibiotic susceptibility testing of isolates

The antibiotic resistance profiles of the isolated strains were determined using the disk diffusion method on Mueller-Hinton agar. This standardized technique involves the placement of antibiotic-impregnated disks onto the surface of agar plates inoculated with bacterial isolates, followed by incubation to allow bacterial growth. The diameter of the resulting inhibition zones was measured to assess bacterial susceptibility to each antibiotic [11]. Antibiotic disks (Bio-Rad Laboratories, California, USA) were selected based on recommendations from the Antibiogram Committee of the Société Française de Microbiologie and the European Committee on Antimicrobial Susceptibility Testing, as well as antibiotics commonly employed in clinical practice in Togo. These disks were applied to test the susceptibility of Escherichia coli and Salmonella spp. isolates.

The antibiotic disks employed in this study encompassed a broad spectrum of agents, including: Amoxicillin (20 µg), Amoxicillin-clavulanic acid (20/10 µg), Ticarcillin (75 µg), Ticarcillin-clavulanic acid (75/10 µg), Piperacillin (30 µg), Piperacillin-tazobactam (30/6 µg), Cefalexin (30 µg), Cefoxitin (30 µg), Cefuroxime (30 µg), Ceftazidime (10 µg), Ceftriaxone (30 µg), Cefepime (30 µg), Aztreonam (30 µg), Imipenem (10 µg), Ertapenem (10 µg), Amikacin (30 µg), Gentamicin (10 µg), Tobramycin (10 µg), Chloramphenicol (30 µg), Trimethoprim-sulfamethoxazole (25 µg), Ciprofloxacin (5 µg), Ofloxacin (5 µg), Nalidixic acid (30 µg), Levofloxacin (5 µg), Fosfomycin (200 µg), and Nitrofurantoin (100 µg). Quality control was ensured using the reference strain Escherichia coli ATCC 25,922. Interpretation of susceptibility results followed the guidelines established by the Comité de l’Antibiogramme de la Société Française de Microbiologie and the European Committee on Antimicrobial Susceptibility Testing (CA-SFM/EUCAST, 2023).

Detection of resistance and virulence genes

The identification of antibiotic resistance genes and virulence factors was performed using both conventional and multiplex polymerase chain reaction (PCR) assays. Genomic DNA was extracted from multidrug-resistant Escherichia coli and Salmonella spp. isolates employing the Genolyse kit (Lifescience GmbH, Nehren, Germany) according to the manufacturer’s protocol. Briefly, bacterial colonies were suspended in the supplied lysis buffer, incubated at 95 °C for 10 min, and subsequently centrifuged to recover the supernatant containing purified DNA. The extracted DNA served as a template for PCR amplification targeting β-lactamase genes (blaTEM, blaCTX, blaSHV), quinolone resistance genes (qnrA, qnrB, qnrS), virulence-associated genes (spvR, invA, fimA), and integron integrase genes (Int1, Int2, and Int3). Primer sequences utilized for these amplifications are detailed in Tables 1 and 2. All primers were custom-synthesized by Invitrogen Custom Primers (Thermo Fisher Scientific, California, USA).

Table 1.

Primer sequences employed for the detection of antibiotic resistance genes

Genes	Primers	Primer sequences (5’ to 3’)	Size (bp)	Reference
BlaTEM	OT-F	5’-ATTGGGTGCACGAGTGGGTTAC-3’	465	Moussé et al. 2016
	OT-R	5’-ATAATTGTTGCCGGGAAGCTAG-3’
blaSHV	SHV-F	5’-CGCCGGGTTATTCTTATTTGTCGC-3’	1017	Moussé et al. 2016
	SHV-R	5’-TCTTTCCGATGCCGCCGCCAGTCA3’
blaCTX	CTX-F	5’-CGCTTTGCGATGTGCAG-3’	550	Moussé et al. 2016
	CTX-R	5’-ACCGCGATATCGTTGGTAAT-3’
qnrA	qnrA-F	F: 5’-ATTTCTCACGCCAGGATTTG-3’	516	Jacoby et al. 2014
	QnrA-R	R: 5’-GATCGGCAAAGGTTAGGTCA-3’
qnrB	QnrB-F	F :5’-GATCGTGAAAGCCAGAAAGG-3’	469	Buelow et al.2018
	QnrB-R	R :5’-ACGATGCCTGGTAGTTGTCC-3’
qnrS	QnrS-F	F :5’-ACGACATTCGTCAACTGCAA-3’	416	Buelow et al.2018
	QnrS-R	R :5’-TAAATTGGCACCCTGTAGGC-3’
IntI 1	IntI − 1-F	F: CCTCCCGCACGATGATC	280	Goldstein et al.2001
	IntI 1-R	R: TCCACGCATCGTCAGGC
IntI 2	IntI − 2-F	F: TTATTGCTGGGATTAGGC	233
	IntI-2-R	R: ACGGCTACCCTCTGTTATC
IntI 3	IntI-3-F	F: AGTGGGTGGCGAATGAGTG	600
	IntI-3-R	R: TGTTCTTGTATCGGCAGGTG

Table 2.

Primer sequences used for virulence factors

Genes	Primer sequences (5’ to 3’)	Size (bp)	Reference
invA	F–GTGAAATTATCGCCACGTTCGGGCAA	284	Nadi et al. 2020
	R–TCATCGCACCGTCAAAGGAACC
fimA	F–CCTTTCTCCATCGTCCTGAA	85	Fardsanei et al.2017
	R–TGGTGTTATCTGCCTGACCA
spvR	F–CAGGTTCCTTCAGTATCGCA	310	Pasmas et al. 2003
	R–TTTGGCCGGAAATGGTCAGT

Results

Distribution of samples in cold rooms

The distribution of 285 samples across seven cold rooms and eight sample types was analyzed. Cold room G accounted for the largest proportion of samples (24.56%), followed by cold rooms F (15.79%) and B (14.03%). The most frequent sample types overall were PIP (33.33%), CUP (19.30%), and API (10.53%) (Table 3).

Table 3.

Distribution of samples in cold rooms

Cold rooms	API	AIP	ARP	CUP	PIP	PEN	QAP	SAU	TOTAL (n)	TOTAL (%)
A	30	0	0	0	0	0	5	0	35	12.28
B	0	0	0	40	0	0	0	0	40	14.03
C	0	0	0	0	0	0	0	30	30	10.53
D	0	0	0	0	0	0	30	0	30	10.53
E	0	5	10	15	0	5	0	0	35	12.28
F	0	0	0	0	45	0	0	0	45	15.79
G	0	20	0	0	50	0	0	0	70	24.56
TOTAL(n)	30	25	10	55	95	5	35	30	285	100
TOTAL(%)	10.53	8.77	3.51	19.30	33.33	1.75	12.28	10.53	100	-

Legend: CUP: Chicken leg, ARP: Chicken breast, PEN: Whole chicken, SAU: Sausage, PIP: Chicken drumstick, AIP: Chicken wing, QAP: Chicken forequarter, API: Guinea fowl wing. A, B, C, D, E, F, G: names of cold rooms

The statistical analysis performed using XLSTAT software, employing the Chi-square test of independence revealed a significant association between the sample type and the cold room (p < 0.001). This indicates that the distribution of different poultry products varies according to the cold rooms. This variation may reflect organizational or logistical choices related to the specific characteristics of the cold rooms or the storage requirements of the different products. Specific cold rooms exhibited distinct sample type profiles, such as cold room A with a predominance of API samples, cold room G with elevated AIP and PIP samples, and cold room F mainly storing PIP samples.

Prevalence and assessment of the microbiological quality of imported chickens

Prevalence of Escherichia coli

Escherichia coli was detected in 94 of the 285 samples analyzed, yielding an overall prevalence of 32.98%. The carriage rates of Escherichia coli across different sample types and cold storage facilities are summarized in Fig. 2. Marked heterogeneity was observed among both sample types and cold room locations. The highest carriage rate was detected in API samples from cold room A (100%), followed by CUP samples from cold room E (86.67%) and QAP samples from cold room D (60.00%). In contrast, several combinations, such as SAU samples from cold room C and AIP samples from cold room E, exhibited no detectable E. coli contamination (0.00%).

Fig. 2.

Escherichia coli carriage rates. Legend: CUP: Chicken leg, ARP: Chicken breast, PEN: Whole chicken, SAU: Sausage, PIP: Chicken drumstick, AIP: Chicken wing, QAP: Chicken forequarter, API: Guinea fowl wing. A, B, C, D, E, F, G: names of cold rooms

Statistical analysis using the chi-square test for independence revealed a significant association between sample type, cold room location, and E. coli carriage rate (p < 0.05). This suggests that both the nature of the sample and the storage environment are important determinants of E. coli prevalence. Notably, the distribution pattern indicates that certain sample types (e.g., API and CUP) and specific cold rooms (e.g., A and E) are at higher risk for contamination, highlighting the need for targeted interventions in these areas (Fig. 2).

Prevalence of Salmonella

Salmonella spp. were detected in 7 out of 285 samples, corresponding to an overall prevalence of 2.45%. Analysis of Salmonella carriage rates across different sample types and cold room locations revealed a low overall prevalence, with positive detections confined to a limited number of sample/cold room combinations (Fig. 3). Notably, Salmonella was detected in QAP samples from cold room A (20.00%), CUP samples from cold room E (6.67%), and PEN samples from cold room E (100.00%). All other sample types and cold rooms exhibited a carriage rate of 0%.

Fig. 3.

Salmonella carriage rate. Legend: CUP: Chicken leg, ARP: Chicken breast, PEN: Whole chicken, SAU: Sausage, PIP: Chicken drumstick, AIP: Chicken wing, QAP: Chicken forequarter, API: Guinea fowl wing. A, B, C, D, E, F, G: names of cold rooms

To evaluate the association between Salmonella prevalence, sample type, and cold room location, a chi-square test of independence was performed. The analysis demonstrated a statistically significant association between these variables (χ² = 31.47, df = 14, p = 0.008), indicating that Salmonella contamination was not randomly distributed but was significantly associated with specific sample types and storage environments.

Antibiotic resistance profiles of the studied strains

Microbiological analysis of the 285 samples led to the isolation of 180 Escherichia coli strains and 7 Salmonella spp. strains. These strains were preserved in 20% glycerol in Heart Brain Broth (HBI) at -80 °C for subsequent antibiogram and molecular tests.

Antibiotic resistance profiles of Salmonella strains

Antibiogram testing of seven Salmonella isolates revealed variable resistance patterns across different antibiotic classes. Resistance to multiple β-lactam antibiotics-including amoxicillin, amoxicillin-clavulanic acid, cefalexin, cefuroxime, ceftazidime, ceftriaxone, and aztreonam-was observed in 2 out of 7 isolates (28.57%). Additionally, resistance to nalidixic acid was also detected in these same isolates. Cefoxitin resistance was noted in 1 isolate (14.29%). All isolates remained fully susceptible (100%) to carbapenems (imipenem, ertapenem), ticarcillin and its clavulanic acid combination, piperacillin, cefepime, aminoglycosides, chloramphenicol, trimethoprim-sulfamethoxazole, fluoroquinolones, fosfomycin, and nitrofurantoin.

Based on the criteria of resistance to at least one agent in three or more antimicrobial classes, 2 isolates (28.57%) were identified as multidrug-resistant (MDR) (Table 4). The MDR phenotype was primarily characterized by resistance to multiple β-lactams and nalidixic acid, suggesting potential extended-spectrum beta-lactamase (ESBL) activity. The absence of resistance to carbapenems and aminoglycosides indicates these antibiotics remain effective treatment options for the tested strains.

Table 4.

Summary of antibiotic resistance phenotypes in Salmonella isolates (n = 7)

Resistance Phenotype	Number of Isolates	Percentage (%)
Extended-spectrum beta-lactamase (ESBL)	1/7	14.29%
Cephalosporinase	1/7	14.29%
Carbapenemase	0/7	0.00%
Multidrug-resistant (≥ 3 classes)	2/7	28.57%

An exact binomial test was applied to each antibiotic to assess whether the proportion of resistance among the seven Salmonella strains differed significantly from random distribution (p = 0.5). Statistical analysis demonstrated significant susceptibility (p = 0.0156) for antibiotics with 100% efficacy, while resistance rates for other antibiotics did not reach statistical significance, likely due to the small sample size. Statistical analyses were performed using Python and the SciPy statistical library (v1.11). (Table 5).

Table 5.

Antibiotic susceptibility test results for isolated Salmonella strains (n = 7)

Antibiotic (dose)	Resistance (%)	Susceptibility (%)		p-value
Amoxicillin (AM, 10 µg)	28.57	71.43		0.4531
Amoxicillin + Clavulanic acid (AMC, 20 µg)	28.57	71.43		0.4531
Ticarcillin (TC, 75 µg)	0.00	100.00		0.0156
Ticarcillin + Clavulanic acid (TCC, 75/10 µg)	0.00	100.00		0.0156
Piperacillin (PRL, 30 µg)	0.00	100.00		0.0156
Cefalexin (CL, 30 µg)	28.57	71.43		0.4531
Cefoxitin (FOX, 30 µg)	14.29	85.71		0.1250
Cefuroxime (CXM, 30 µg)	28.57	71.43		0.4531
Ceftazidime (CAZ, 10 µg)	28.57	71.43		0.4531
Ceftriaxone (CTR, 30 µg)	28.57	71.43		0.4531
Cefepime (CEF, 30 µg)	0.00	100.00		0.0156
Aztreonam (ATM, 30 µg)	28.57	71.43		0.4531
Imipenem (IPM, 10 µg)	0.00	100.00		0.0156
Ertapenem (ETP, 10 µg)	0.00	100.00		0.0156
Amikacin (AK, 30 µg)	0.00	100.00		0.0156
Gentamicin (GEN, 10 µg)	0.00	100.00		0.0156
Tobramycin (TOB, 10 µg)	0.00	100.00		0.0156
Netilmicin (NET, 30 µg)	0.00	100.00		0.0156
Chloramphenicol (C, 30 µg)	0.00	100.00		0.0156
Trimethoprim-sulfamethoxazole (SXT, 1.25/23.75 µg)	0.00	100.00		0.0156
Nalidixic acid (NA, 30 µg)	28.57	71.43		0.4531
Ofloxacin (OFX, 30 µg)	0.00	100.00		0.0156
Norfloxacin (NOR, 30 µg)	0.00	100.00		0.0156
Ciprofloxacin (CIP, 5 µg)	0.00	100.00		0.0156
Fosfomycin (FF, 200 µg)	0.00	100.00		0.0156
Nitrofurantoin (N, 100 µg)	0.00	100.00		0.0156

Antibiotic resistance profiles of Escherichia coli isolates

The antimicrobial susceptibility testing of 180 Escherichia coli isolates demonstrated a high prevalence of resistance to several antibiotic classes. Resistance to amoxicillin was predominant, affecting 70.0% (126/180) of isolates, while resistance to β-lactam/β-lactamase inhibitor combinations ranged from 26.67 to 38.89%. Resistance to cefalexin was noted in 22.22% (45/180) of isolates. Notably, 9.44% (17/180) of isolates exhibited resistance to multiple β-lactams, including third-generation cephalosporins (ceftriaxone, ceftazidime, cefotaxime), cefepime, and aztreonam, consistent with extended-spectrum beta-lactamase (ESBL) production and meeting the criteria for multidrug resistance (MDR) (Table 6). Moderate resistance was observed for aminoglycosides (gentamicin 10.0%, tobramycin 10.56%) and fluoroquinolones (12.78–21.11%), while no resistance was detected against carbapenems, amikacin, fosfomycin, or nitrofurantoin (Table 7).

Table 6.

Summary of antibiotic resistance phenotypes among Escherichia coli isolates (n = 180)

Resistance Phenotype	Number of Isolates (n/N)	Percentage (%)
Extended-spectrum beta-lactamase (ESBL)	17/180	9.44%
Cephalosporinase	17/180	9.44%
Carbapenemase	0/180	0.00%
Multidrug-resistant (≥ 3 classes)	17/180	9.44%

Table 7.

Antibiotic susceptibility profiles of isolated Escherichia coli strains (n = 180)

Antibiotic (dose)	Resistance n (%)	Susceptibility n (%)	p-value
Amoxicillin (AM, 10 µg)	126 (70.00)	54 (30.00)	0.0000
Amoxicillin + Clavulanic acid (AMC, 20 µg)	57 (31.67)	123 (68.33)	0.0000
Ticarcillin (TC, 75 µg)	70 (38.89)	110 (61.11)	0.0035
Ticarcillin + Clavulanic acid (TCC, 75/10 µg)	54 (30.00)	126 (70.00)	0.0000
Piperacillin (PRL, 30 µg)	70 (38.89)	110 (61.11)	0.0035
Piperacillin + Tazobactam (PRL + TAZO, 30 µg)	48 (26.67)	132 (73.33)	0.0000
Cefalexin (CL, 30 µg)	45 (22.22)	135 (77.78)	0.0000
Cefoxitin (FOX, 30 µg)	2 (1.11)	178 (98.89)	0.0000
Cefuroxime (CXM, 30 µg)	17 (9.44)	163 (90.56)	0.0000
Ceftazidime (CAZ, 10 µg)	17 (9.44)	163 (90.56)	0.0000
Cefotaxime (CFT, 5 µg)	17 (9.44)	163 (90.56)	0.0000
Ceftriaxone (CTR, 30 µg)	17 (9.44)	163 (90.56)	0.0000
Cefepime (FEP, 30 µg)	17 (9.44)	163 (90.56)	0.0000
Aztreonam (ATM, 30 µg)	17 (9.44)	163 (90.56)	0.0000
Imipenem (IPM, 10 µg)	0 (0.00)	180 (100.00)	0.0000
Ertapenem (ETP, 10 µg)	0 (0.00)	180 (100.00)	0.0000
Amikacin (AK, 30 µg)	0 (0.00)	180 (100.00)	0.0000
Gentamicin (GEN, 10 µg)	18 (10.00)	162 (90.00)	0.0000
Tobramycin (TOB, 10 µg)	19 (10.56)	161 (89.44)	0.0000
Chloramphenicol (C, 30 µg)	7 (3.89)	173 (96.11)	0.0000
Trimethoprim-sulfamethoxazole (SXT, 1.25/23.75 µg)	24 (13.33)	156 (86.67)	0.0000
Ciprofloxacin (CIP, 5 µg)	23 (12.78)	157 (87.22)	0.0000
Ofloxacin (OFX, 30 µg)	24 (13.33)	156 (86.67)	0.0000
Pefloxacin (PEF, 5 µg)	38 (21.11)	142 (78.89)	0.0000
Levofloxacin (LEV, 5 µg)	24 (13.33)	156 (86.67)	0.0000
Fosfomycin (FF, 200 µg)	2 (1.11)	178 (98.89)	0.0000
Nitrofurantoin (NI, 100 µg)	0 (0.00)	180 (100.00)	0.0000

All antibiotics tested demonstrated statistically significant resistance or susceptibility patterns, with p-values ≤ 0.0035, and most at p = 0.0000, confirming the robustness and reliability of the observed resistance frequencies. The low p-values indicate that the resistance rates are unlikely due to random variation and reflect true resistance trends in the tested population (Fig. 4). An exact binomial test was applied to each antibiotic to assess whether the proportion of resistant strains differed significantly from a random distribution (p = 0.5). Statistical analyses were performed using Python and the SciPy statistical library (v1.11).

Fig. 4.

Resistance and susceptibility profile of E. coli isolates against various antibiotics

A thorough analysis of the antibiotic susceptibility testing results led to the selection of 17 multidrug-resistant Escherichia coli isolates (17/180) and 2 multidrug-resistant Salmonella isolates (2/7). These selected strains were subsequently subjected to molecular typing to investigate the presence of antimicrobial resistance genes and virulence factors.

Prevalence of virulence and resistance genes in isolates by PCR

Among the 187 strains subjected to antibiotic susceptibility testing, 19 multidrug-resistant isolates of Escherichia coli and Salmonella were identified, representing 10.16% of the total. These multidrug-resistant strains were further analyzed by PCR to detect genetic determinants associated with antibiotic resistance and virulence factors. The molecular screening targeted β-lactamase genes (blaTEM, blaCTX, blaSHV), quinolone resistance genes (qnrA, qnrB, qnrS), integron classes (Int1, Int2, Int3), and virulence genes (fimA, invA, spvR) (Table 8).

Table 8.

PCR results for resistance and virulence genes

Resistance and Virulence Genes researched
Isolate No.	Strain	blaTEM	blaCTX	blaSHV	fimA	invA	spvR	qnrA	qnrB	qnrS	Int 1	Int 2	Int 3
1	Escherichia coli	-	+	-	-	-	-	-	-	-	+	-	-
2	Escherichia coli	+	-	-	-	-	-	-	-	-	-	-	-
3	Escherichia coli	+	-	-	-	-	-	-	-	-	-	-	-
4	Escherichia coli	+	-	-	-	-	-	-	-	-	-	-	-
5	Escherichia coli	-	-	-	-	-	-	-	-	-	-	-	-
6	Escherichia coli	+	-	-	-	-	-	-	-	-	+	-	-
7	Escherichia coli	+	-	-	-	-	-	-	+	-	-	-	-
8	Escherichia coli	+	-	-	-	-	-	-	-	-	-	-	-
9	Escherichia coli	-	-	-	-	-	-	-	-		-	-	-
10	Escherichia coli	+	+	-	-	-	-	-	-	-	-	-	-
11	Escherichia coli	-	+	-	-	-	-	-	-	-	-	-	-
12	Escherichia coli	-	-	-	-	-	-	-	-	-	-	-	-
13	Escherichia coli	-	-	-	-	-	-	-	-	-	-	-	-
14	Escherichia coli	-	+	-	-	-	-	-	-	-	+	-	-
15	Escherichia coli	+	-	-	-	-	-	-	-	-	-	-	-
16	Escherichia coli	+	+	-	-	-	-	-	-	-	-	-	-
17	Escherichia coli	+	+	-	-	+	-	-	-	-	-	-	-
18	Salmonella spp	-	-	-	-	-	-	-	-	-	-	-	-
19	Salmonella spp	-	-	-	-	-	-	-	-	-	-	-	-

The PCR detection analysis of 19 bacterial isolates revealed that the blaTEM gene was the most prevalent resistance determinant, detected in 58.8% of isolates, followed by blaCTX in 35.3%, and class 1 integrons (Int1) in 17.6%. The qnrB and invA genes were each identified in 5.9% of isolates. Co-occurrence of two resistance or virulence genes was observed in 31.58% of strains. To assess potential associations among the most frequently detected genes (blaTEM, blaCTX, and qnrB), Fisher’s exact tests were performed, yielding odds ratios of 0.86 (blaTEM vs. blaCTX), infinity (blaTEM vs. qnrB), and zero (blaCTX vs. qnrB), with all corresponding p-values equal to 1.000 (Table 9; Fig. 5). These results indicate no statistically significant associations, suggesting an independent distribution of resistance and virulence genes within the examined isolate population.

Table 9.

Prevalence of resistance genes and virulence factors in isolates (n = 19)

Strain	Number of Isolates	blaTEM n (%)	blaCTX n (%)	Int 1 n (%)	qnrB n (%)	invA n (%)
Escherichia coli	17	10 (58.8%)	6 (35.3%)	3 (17.6%)	1 (5.9%)	1 (5.9%)
Salmonella spp	2	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)

Fig. 5.

PCR detection of resistance and virulence genes in individual isolates (n = 19)

Discussion

Surveillance studies on foodborne bacterial pathogens are routinely conducted in developed countries, facilitating effective monitoring and control of food safety risks. Poultry has been consistently identified as a major reservoir of microorganisms responsible for foodborne illnesses. In contrast, developing countries face significant financial and technological limitations that hinder the implementation of regular surveillance programs, resulting in a limited understanding of the epidemiology and etiology of foodborne infections, as emphasized by several authors [12, 13].

While poultry meat and eggs are important sources of high-quality protein and essential nutrients, contamination during processing, handling, marketing, and storage prior to cooking remains a critical public health concern, as it can lead to food poisoning outbreaks [14]. In Togo, despite governmental efforts, domestic production fails to meet the population’s demand for animal protein, necessitating substantial imports of meat products. This situation underscores the urgent need to assess the microbiological quality of broiler meat to evaluate the potential risks it poses to public health.

Regarding Escherichia coli, a well-established indicator of fecal contamination and inadequate hygiene during the production process, we observed a prevalence of 32.98%. This finding is comparable to the 26.2% prevalence reported by Yar et al. (2021) [15]. However, our prevalence is lower than those documented in previous studies on chicken thighs, including 48.4% by N. Cohen et al. (2007), 67% by Jansen et al. (2018), and 54% by Mahmoud et al. (2020) [16]; [17, 18].

Concerning bacterial load, Abraham Adu-Gyamfi et al. (2012) reported mean E. coli counts of 1.27, 2.59, and 2.74 log₁₀ CFU/g in poultry samples from supermarkets, local markets, and farms respectively in Accra, Ghana, which are lower than the levels observed in our study [19]. In contrast, Egervärn et al. (2014) reported a markedly higher prevalence of 95% in poultry meat imported predominantly from Brazil into Sweden, highlighting significant regional and supply chain differences in contamination levels [20].This difference in prevalence could be explained by the number of samples tested in these studies and in ours, as well as by the fact that Escherichia coli is a mesophilic germ whose growth is inhibited by negative cold [21]. Chicken meat is a valuable source of high-quality protein; however, its susceptibility to microbial contamination makes it a frequent vehicle for foodborne illnesses. Among the microbial contaminants, Escherichia coli serves as a primary indicator of fecal pollution and represents one of the most significant foodborne pathogens posing a threat to public health [22].

Salmonella is a major foodborne pathogen causing millions of gastroenteritis cases worldwide annually, primarily transmitted through contaminated poultry and meat products, posing a significant public health risk [23]. In our study, Salmonella spp. was detected in 7 out of 285 samples, corresponding to a prevalence of 2.45%. This finding aligns closely with the prevalences reported by N. Cohen et al. (2007) in Morocco (1.6%) and Jansen et al. (2018) (1.2%) [17, 24]. However, our prevalence is notably lower than those documented in similar studies on frozen chicken by L. Kozačinski et al. (2006), Adeyanju and Ishola (2014), Fernandes et al. (2016), and Mahmoud et al. (2020), who reported prevalences of 10.60%, 33%, 18%, and 9%, respectively indicating considerable variability in Salmonella contamination levels across different geographic regions and production systems [18, 25–27]. In contrast, significantly higher prevalences have been observed in other regions; for example, N. Alloui et al. (2013) reported 83.33%, and Elgroud et al. (2009) documented a 73.33% prevalence of Salmonella in poultry from the Constantine region of Algeria [28, 29]. Similarly, studies conducted in slaughterhouses in Thailand and Japan reported Salmonella prevalences of 41.2% and 40.7%, respectively [30]. Regardless of the prevalence rates, these findings are concerning given that Escherichia coli and Salmonella are among the most significant and frequent pathogens implicated in foodborne illnesses associated with chicken meat consumption [31, 32]. Notably, breaded frozen reformulated chicken products have been linked to human cases of salmonellosis; for instance, an outbreak involving multiple strains of Salmonella enterica serovar Enteritidis resulted in over 400 human cases in the United Kingdom in 2020, highlighting the significant public health risks associated with contaminated poultry products and the need for stringent food safety measures throughout the production and supply chain [33]. It is therefore imperative to recognize poultry as a critical reservoir of Salmonella, which poses a substantial risk for foodborne infections in humans [34].

In the present study, the hygienic quality assessment revealed contamination rates of Salmonella spp. at 1.82% in chicken thighs, 2.86% in chicken forequarters, and 100% in whole chickens. These findings underscore significant hygiene deficiencies during the slaughtering process. Critical steps such as slaughter conditions, plucking, evisceration, bleeding, washing, rinsing, preservation, and sale frequently fail to adhere to established good hygiene and manufacturing practices. Non-compliance with these hygiene protocols is widely recognized as the primary factor contributing to microbial contamination of poultry meat.

Beyond microbial contamination, antimicrobial resistance (AMR) has emerged as a critical issue of global concern, with serious implications for public health, food security, and sustainable development [35]. This phenomenon arises from the natural adaptive mechanisms of microorganisms exposed to antimicrobials, which are extensively used not only in human medicine but also in food-producing animals, agriculture, and for disinfection practices on farms and in households [36–38]. In recent years, the widespread use of antimicrobials in veterinary medicine and the food industry particularly as growth promoters and for prophylactic or therapeutic purposes has significantly contributed to the rise of antimicrobial resistance [39–41]. This alarming trend has raised concerns within the scientific community due to its dual impact: it complicates the treatment of bacterial infections and facilitates the dissemination of resistant bacteria through the food chain, especially via poultry and poultry products [39–41].

The transmission of antimicrobial resistant bacteria is further exacerbated by several factors, including the movement of carrier animals between herds and the role of vectors [42]. These bacteria are responsible for infections in both humans and poultry that are more difficult and costly to treat than infections caused by susceptible strains.

Moreover, resistant bacteria pose additional human health risks due to the potential presence of antimicrobial residues in animal derived products such as meat [43] and eggs [44]. Beyond public health implications, the economic consequences are substantial, stemming from increased treatment costs with ineffective antimicrobials and losses associated with unresolved poultry diseases [45].

Antibiogram analysis of Salmonella isolates revealed that 28.57% (2/7) exhibited resistance to multiple antibiotics, including amoxicillin, amoxicillin-clavulanic acid, cefalexime, cefuroxime, ceftazidime, aztreonam, nalidixic acid, and ceftriaxone, while 14.29% (1/7) demonstrated resistance to cefoxitin. No resistance was observed against the other tested antimicrobials. Notably, one isolate was identified as an extended-spectrum beta-lactamase (ESBL)-producing strain, characterized by resistance to amoxicillin, amoxicillin-clavulanic acid, cefalexime, and ceftriaxone, but susceptibility to cefepime. This ESBL phenotype was confirmed by the observation of a “champagne cork-like inhibition pattern” synergy between the amoxicillin-clavulanic acid disk and that of ceftriaxone or another third or fourth-generation cephalosporin. Additionally, a cephalosporinase hyperproducing strain was detected, exhibiting resistance to amoxicillin, amoxicillin-clavulanic acid, cefalexime, cefoxitin, and ceftriaxone, yet remaining sensitive to cefepime (Tables 4 and 5).

According to the 2023 CA-SFM veterinary guidelines [11], ESBL-producing E. coli must be considered resistant to all veterinary β-lactams, except amoxicillin-clavulanic acid. However, the in vivo efficacy of this combination remains undocumented in veterinary medicine. Similarly, hyperproduction of cephalosporinases indicates resistance to all β-lactams used in animals. In this study, both ESBL and cephalosporinase-hyperproducing strains exhibited full resistance to all tested β-lactams. These findings are consistent with current interpretive guidelines and highlight serious therapeutic limitations. They also reinforce the need for prudent antimicrobial use and further investigation into resistance mechanisms in foodborne pathogens of veterinary importance. These results confirm the extensive use of antibiotics in poultry production, both for therapeutic purposes and as over the counter growth promoters, practices largely driven by their low cost and widespread accessibility. This extensive antibiotic use has contributed to the emergence of antimicrobial resistance, exemplified by the observed resistance to cefotaxime. Comparable resistance patterns have been documented in local chicken products by Sasaki et al. (2021) [46], as well as in studies conducted in Japan and Brazil, highlighting the global scope of this issue [47, 48].Our findings show a 28.57% resistance rate to amoxicillin-clavulanic acid, which is significantly lower than the approximately 60% resistance reported by Caroline Bouda in poultry isolates from Burkina Faso [49]. This high resistance level in Burkina Faso reflects challenges related to antibiotic use and regulation in low-resource settings. In contrast, European data from the European Food Safety Authority (EFSA) and the European Centre for Disease Prevention and Control (ECDC) indicate generally lower resistance rates in foodborne bacteria, with average resistance rates around 22–30% depending on the bacterial species and antibiotic tested [50]. For example, resistance to critically important antibiotics in Salmonella and E. coli isolated from poultry meat remains below 30% in most European countries, thanks to stricter antibiotic stewardship and surveillance programs (EFSA & ECDC, 2022) [50]. Thus, our observed resistance rate positions the studied isolates in an intermediate range between the high resistance reported in Burkina Faso and the lower rates commonly seen in Europe, highlighting the need for ongoing monitoring and targeted interventions to control antimicrobial resistance in the food chain. Recent studies have demonstrated that the rising antimicrobial resistance among Salmonella strains is largely driven by selective pressure resulting from the inappropriate use of antibiotics in both veterinary and human medicine [51]. The resistance observed to nalidixic acid in our isolates suggests concomitant resistance to fluoroquinolones, as nalidixic acid serves as a reliable marker for the initial stages of quinolone resistance [11]. Consistent with our findings, high levels of nalidixic acid resistance have been reported in other regions, including 72.3% in China [52], and rates increasing from 57.5% in 2014 to 86.5% in 2017 in Brazil [53]. Additionally, Sasaki et al. (2021) documented elevated resistance rates to streptomycin (51.1%), tetracycline (33.1%), and kanamycin (18.4%) in Salmonella isolates from poultry products [46]. The high antibiotic susceptibility observed likely results from reduced or discontinued use in poultry farming, reflecting increased farmer awareness of antibiotic overuse risks. Nonetheless, 9.44% of Escherichia coli isolates exhibited multidrug resistance to key antibiotic classes-beta-lactams, cephalosporins, and quinolones-which is concerning due to their importance in treating enterobacterial infections.

The emergence of strains resistant to third-generation cephalosporins (C3G) is frequently associated with the production of extended-spectrum beta-lactamases (ESBLs), which are often co-expressed with resistance to fluoroquinolones [54, 55]. These resistance mechanisms may arise from chromosomal mutations, including target site modifications or active efflux systems; however, they are predominantly plasmid-mediated, facilitating horizontal gene transfer among bacteria of the same or even different species. Molecular characterization of resistance determinants in our study revealed the presence of class 1 integrons in 17.6% of isolates. The genetic cassettes within these integrons primarily harbored the aadA1 and dfr1 genes, which encode resistance to streptomycin, spectinomycin, and trimethoprim. Notably, class 2 and class 3 integrons were not detected in any of the isolates analyzed. The extrachromosomal resistance genes qnr, aac(6’)-Ib-cr, and qepA, first identified in 2002, are typically harbored on conjugative plasmids. Since their discovery, multiple qnr variants-including qnrA, qnrB, qnrS, qnrC, and qnrD have been characterized [56]. In the present study, 5.9% of isolates carried the qnrB gene. Previous investigations have demonstrated that these plasmid-mediated resistance determinants are predominantly associated with non-typhoidal Salmonella strains and are relatively rare or absent in typhoidal strains, consistent with our observations [57]. Moreover, plasmids encoding extended-spectrum beta-lactamases (ESBLs) frequently co-harbor genes conferring resistance to quinolones [58, 59]. This co-localization is reflected in our findings, which identified ESBL-producing strains harboring blaCTX (35.3%) and blaTEM (58.8%) genes.

In several exporting countries and across sub-Saharan Africa, the veterinary medicines market is hindered by significant organizational deficiencies [60]. These challenges include the absence of specific legislation tailored to the evolving context of veterinary pharmacy liberalization, inadequate enforcement of existing regulations, lack of routine inspections, and the absence of formal marketing authorization (MA) and registration procedures for veterinary pharmaceuticals. Furthermore, unofficial parallel distribution channels operate alongside official supply networks, complicating regulatory oversight [61].

Although data remain limited for low-income countries, the rising antimicrobial resistance in these regions, coupled with restricted access to effective and safe antibiotics, is likely contributing to substantial mortality. Routine surgical procedures and minor infections risk becoming fatal once again, threatening to reverse the substantial progress made against infectious diseases over the past five decades (WHO S.P.F. 2020). Without the implementation of proactive interventions to curb the spread of resistance, projections estimate a catastrophic increase in mortality potentially reaching 10 million deaths annually and an economic loss of approximately USD 100 trillion by 2050 [62].

The use of antimicrobials in animal husbandry significantly influences the prevalence of resistant bacterial strains, posing a critical public health challenge [63]. Consequently, monitoring and minimizing antimicrobial residues in livestock derived products have been emphasized as essential measures to safeguard consumer health [64, 65]. Strict adherence to the prescribed withdrawal periods following antimicrobial administration in poultry and other farm animals is vital to ensure the elimination or substantial reduction of these residues in animal products. Moreover, multiple studies have demonstrated that preventing horizontal transmission of resistant bacteria on farms necessitates rigorous cleaning and disinfection protocols, alongside the implementation of effective sanitary measures to control vectors responsible for pathogen dissemination [66–69].

Conclusion

Strict adherence to animal hygiene practices plays a crucial role in reducing the contamination of carcasses and limiting the spread of pathogenic bacteria. Conversely, the lack of rigorous hygiene throughout the production chain, combined with the absence of effective health monitoring programs and the inappropriate prophylactic use of antibiotics, likely contributes to the persistence of resistant bacteria in the samples analyzed.

The detection of multidrug-resistant (MDR) Escherichia coli strains in nearly 10% of isolates underscores the urgent need for continuous antimicrobial resistance surveillance and targeted stewardship interventions to preserve the efficacy of last resort antibiotics. The extensive use of antimicrobials in food producing animals has accelerated the emergence and dissemination of bacterial resistance.

Importantly, certain multidrug-resistant Salmonella and E. coli strains isolated from humans originate from animals, having acquired resistance genes prior to transmission through the food chain. Therefore, systematic monitoring of antimicrobial resistance in food animals is essential to inform effective control strategies and mitigate the public health risks associated with zoonotic transmission.

Abbreviations

CUP

Chicken Leg

ARP

Chicken Breast

PEN

Whole Chicken

SAU

Sausage

PIP

Chicken Drumstick

AIP

Chicken Wing

QAP

Chicken Forequarter

API

Guinea Fowl Wing

Author contributions

Kossi Touglo and Bouraïma Djeri designed the study. Kossi Touglo, Bawa Boya, and Haziz Sina conducted the study. Data collection was performed by Kossi Touglo and Bawa Boya. Data analysis was carried out by Kossi Touglo and Guy Géoffroy Ayadokoun. The data were interpreted by Kossi Touglo, Guy Géoffroy Ayadokoun, and Haziz Sina. Kossi Touglo wrote the manuscript, and Haziz Sina revised its content. Lamine Baba-Moussa and Yaovi Ameyapoh approved the final version of the manuscript. All authors read and approved the final manuscript.

Funding

The authors declare that they have received no financial support for the research, writing and/or publication of this article.

Data availability

The data used in this study are available from the corresponding author upon reasonable request (touglokossi30@gmail.com).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.