Background

For several decades, β-lactam antibiotics have been considered the antimicrobial agents of choice in humans and veterinary medicine. However, the effectiveness of these drugs has been diminished due to the emergence of β-lactamase-producing bacteria, particularly within the Enterobacterales [1]. Recently, as a result of the increased production of extended-spectrum β-lactamases (ESBL) a rise in β-lactam resistance in E. coli has increased [2]. ESBL-producing E. coli or their resistance genes can potentially be transmitted through direct contact, the food chain, or environmental sources [3]. The emergence of ESBL-producing bacteria linked to cattle, poultry, and pigs may be linked to the gradual increase in the usage of third-generation cephalosporins in food animal production [4, 6].

The genome of E. coli consists of a mobile gene core that determines the strain pathotypes. Various virulence factors determinants have been attributed to E. coli pathogenicity, including Shiga toxin-associated genes (stx1 and stx2), toxin production genes such as hemolysin (hly) and the astA gene encoding enteroaggregative E. coli heat-stable enterotoxin, intimin encoding gene (eae), and fimbrial H gene (fimH). E. coli are classified into eight phylotypes (A, B1, B2, C, D, E, F, clade 1) based on the presence or absence of ChuA, yjaA, TspE4.C2, and arpA genes [5, 6]. Phylotypes B2 and D are frequently found in humans, with phylotype B2 being associated with extraintestinal disease in both animals and humans [7]. Different properties of isolates are indicated by their typing, which is helpful for identifying the circulation pattern across various sources and hosts [8]. The presence of similarities between clinical and foodborne ESBL isolates suggests that food products can serve as a reservoir for ESBL-producing bacteria and their genes [9, 10, 13].

ESBL enzymes hydrolyze broad-spectrum cephalosporins, including ceftazidime, cefotaxime, cefuroxime, ceftriaxone and cefepime. Most ESBLs are classified as Ambler class A enzymes, which includes the sulfhydryl variable (SHV), Temoneria (TEM) and cefotaxime (CTX-M) types [11]. Unlike other bacterial families, Enterobacterales encode SHV, TEM and CTX-M genes on plasmids rather than the chromosome [12], ESBL-producing E. coli is becoming more resistant to fluoroquinolones and aminoglycosides, leading to the evaluation of colistin as an important alternative antimicrobial for human therapy [13, 17].However, colistin has been extensively used in veterinary medicine for years to prevent and treat gastrointestinal infections caused by Enterobacterales in food producing animals [14]. The mcr-1 gene, responsible for mobile colistin resistance, has been detected in Enterobacterales isolates from humans, food, companion animals, meat, and the environment in various studies [15].

Cephalosporins are widely used to treat various infections caused by both Gram-negative and Gram-positive bacteria [16]. They possess a broader activity range and are less susceptible to inactivation by β-lactamase enzymes compared to other β-lactam antibiotics, with successive generations of cephalosporins having a wider spectrum of activity [17]. The World Health Organization has prioritized third-generation cephalosporins (such as cefotaxime and ceftriaxone) in monitoring and stewardship programs for antibiotic resistance due to their significance to human health [18, 23].ESBLs, which hydrolyze third-generation cephalosporins, can transmit acquired resistance to other bacterial populations, leading to the spread of antibiotic resistance [12, 19].

In the past, the most common ESBL genes found in K. pneumoniae were the allelic variants of TEM-1, TEM-2, and SHV-1 that are the traditional TEM- and SHV-types, exhibit enhanced activity against extended-spectrum cephalosporines, and are primarily found in the hospital environment. However, in recent times, ESBL genes have been more frequently identified in E. coli, with the majority belonging to the CTX-M- group [20, 21].

The increasing prevalence of infections caused by MDR bacteria, coupled with the rising problem of acquired resistance resulting from improper antibiotic use, highlights the urgent need to reduce antibiotic consumption by exploring alternative approaches [22]. Among the potential alternatives, nanoparticles have emerged as a promising alternative to antibiotics for controlling infectious agents. Silver nanoparticles (AgNPs) exhibit biocidal effects against various foodborne bacteria [23]. They can interact with the cell surface of Gram-negative bacteria, causing damage and structural changes that enhance bacterial permeability [24]. Therefore, this study aims to (i) determine the occurrence, pathotypes, virulotypes, genotypes, and antimicrobial resistance patterns of ESBL-producing E. coli in retail meat samples and workers in contact in Egypt, and (ii) evaluate the bactericidal efficacy of silver nanoparticles (AgNPs-H2O2) against MDR ESBL-producing E. coli.

Materials and methods

Sample collection

The study received approval from the Zagazig University Institutional Animal Care and Use Committee (ZU-IACUC) under approval number ZU-IACUC/2/F/214/2022. Animal procedures were conducted following the ARRIVE guidelines. A total of 250 meat samples (chicken breast, duck breast, turkey breast, beef meat, and camel meat) were collected from retail meat shops in Zagazig City, Sharkia Governorate, Egypt. Samples were collected aseptically in sterile plastic bags between January to March 2022. Additionally, a total of 100 hand and stool swabs (50, each) were collected from workers at the retail shops where meat samples were collected. Hand swabs were collected by rolling a sterile swab moistened in buffered peptone water over the palmar surface of the workers’ hand. The swab was then placed in the collection tube containing buffered peptone water. Stool samples were collected in clean sterile cups, and a swab was taken from each sample and inserted in a collection tube containing buffered peptone water. All samples and swabs were transported in an icebox at 4 °C to the laboratory and processed within 12–24 h.

E. coli isolation and identification

From each meat sample, 25 g were aseptically homogenized in 1:10 buffered peptone water and incubated for 24 h at 37 °C. Hand and stool swabs were also incubated in buffered peptone water. A loopful of incubated samples was then streaked onto Chromocult® Tryptone Bile X-glucuronide agar (Sigma Aldrich, Millipore, ISO 16,649) supplemented with 2 μg/mL of cefotaxime (CTX) and incubated for 24 h at 37 °C. Green blue colonies were picked and purified by streaking into Tryptone Soy Agar (TSA) (Merck, Darmstadt, Germany). The purified colonies were then identified by biochemical tests for indole, oxidase, catalase, Methlye-Red (MR), Voges-Proskauer (VP), citrate, urease, nitrate reduction and sugar fermentation.

To confirm the presence of E. coli, bacterial DNA from presumptive E. coli colonies was extracted using the QIAamp DNA Mini kit according to the manufacturer’s instructions (Qiagen GmbH, Hilden, Germany, Catalogue no. 51,304). Specific primers targeting the phoA gene (Table 1) were employed for amplification to confirm E. coli isolates [25].

Table 1 Primers, product size and annealing temperatures used for identification, virulence and antimicrobial resistance genes reported in the present study
Full size table

Characterization of ESBL-producing E. coli

The Double Disc Synergy Test (DDST) was used to verify the confirmed E. coli strains for phenotypic ESBL expression [26].

ESBL-producing E. coli phylotyping, virulotyping and genotyping

The confirmed strains were phylotyped using primers for the amplification of chuA, yjaA, and, tspE4C2 genes (Table 1), according to the phylotype classification scheme previously described [27].

Virulotyping was also performed using specific primers for amplification of the virulence associated genes (Table 1), including hly [28],, astA [29], stx1 and stx2 [30], stx2f [31],, fimH [32], eaeA [33].

Genotyping was performed on extracted DNA through fingerprinting PCR using REP-primers synthesized by Metabion (Germany) with the following sequences: Rep1R-I 5’- III ICG ICG ICA TCI GGC-3’ and Rep2-I 5’- ICG ICT TAT CIG GCC TAC-3’ [34]. The primers were included in a 25- μL reaction mixture containing 12.5 μL of EmeraldAmp Max PCR Master Mix (Takara, Japan), 1 μL of each primer ( 20 pmol), 4.5 μL of water, and 6 μL of DNA template. The PCR procedure was carried out using an Applied Biosystem 2720 thermal cycler.

To assess the discriminatory power of the REP-PCR fingerprinting data, the Simpson’s index of diversity (D) was utilized. The fingerprinting data was converted into a binary code, indicating the presence or absence of each band. A D value greater than 0.9 indicated good differentiation [35].

Antimicrobial susceptibility testing

The Kirby-Bauer disc diffusion method was used to evaluate the isolates’ antibiotic susceptibility in accordance with the standards established by the National Committee for Clinical Laboratory Standards (NCCLS). Nineteen different antimicrobial agents were tested, and the zones of inhibition were measured and interpreted based on the guidelines provided by the Clinical and Laboratory Standards Institute (CLSI) [26]. In accordance with CLSI recommendations, the double fold dilution procedure (0.125-256 g/mL) was used to establish the minimum inhibitory concentration (MIC) for colistin [36]. The antimicrobial agents used included penicillin (PEN), ampicillin (AMP), amoxicillin (AMX), streptomycin (STR), erythromycin (ERY), nalidixic acid (NAL), amikacin (AMK), trimethoprim-sulfamethoxazole (SXT), kanamycin (KAN), neomycin (NEO), gentamicin (GEN), ciprofloxacin (CIP), tetracycline (TET), colistin (CST), imipenem (IPM), chloramphenicol (CHL), cefotaxime (CTX), ceftriaxone (CRO), ceftazidime (CAZ). E. coli ATCC 25,922 and Staphylococcus aureus ATCC 25,923 served as the microorganisms for quality control.

Multiple antibiotic resistance (MAR) index was calculated by dividing the number of antibiotics to which E. coli isolates showed resistance by the total number of drugs tested [37]. Multidrug resistance (MDR) was defined as the resistance of an isolate to at least one agent in three or more antibiotic classes [38].

Antimicrobial resistance genes

Bacterial DNA from the E. coli confirmed isolates was also screened for ESBL encoding genes (Table 1); blaIMP, blaVIM, and blaNDM [39], blaTEM, blaOXA-1, and blaSHV [40], blaCMY [41], blaCTX-M-1 [42], tetracycline; tetA(A) and tetA(B) [43], sulfonamides; sul [44], chloramphenicol; cmlA [45], and florquinolones; floR [46]. The presence of colistin resistance genes mcr-1 to mcr-5 was also examined [47].

Antimicrobial effect of silver nanoparticles on MDR ESBL-producing E. coli

AgNPs-H2O2 (Top Superpower-vision) was obtained as a commercial product from El-Delta Center for Nanosilver Technology Company, Mansoura, Egypt. The stock solution of the product contained 45-nm silver nanoparticles (0.00004467 mL/liter), hydrogen peroxide (50% per liter) and natural herb mint (1 mL/liter) at a concentration of 5 mL/liter of water. The particles size was previously determined to be 30.17–67.92 nm with a zeta potential estimation of − 0.192 mV [48]. The AgNPs-H2O2 mixture was prepared by diluting the stock solution in sterile distilled water to achieve the desired commercial concentration. The minimum inhibitory concentrations (MIC50 and MIC90) of AgNPs-H2O2 were determined against MDR ESBL-E. coli by the broth microdilution method [36]. Briefly, microtiter plate wells were supplemented with various concentrations of AgNPs-H2O2 ranging from 100, 50, 25, 10, 5, 2.5 1.25, 0.625, 0.312, 0.156 and 0.078 μg/mL. MDR- ESBL-producing colonies were added in Muller Hinton broth and adjusted to the density of a 0.5 McFarland standard (1 × 108 cfu/ml). Each well received a final inoculum of 5 × 105 cfu/mL, and the plates were incubated for 24 h at 37 °C. A well with MHB alone and another well included MHB with AgNPs-H2O2 were used as reference control. The lowest agent concentration that entirely prevents an organism’s observable growth is known as the MIC endpoint. The MIC50 and MIC90 were calculated using an orderly array method [49], where the middle value was selected as MIC50. The MIC90 was determined in the same way by selecting the appropriate value from the orderly array.

Data analysis

For statistical analysis and data visualization, R software was used (R Core Team, 2022; version 4.2.0). The E. coli isolation rate was calculated by dividing the number of E. coli positive samples by the total number of samples tested for each source. Pearson’s chi-square test was employed to assess any variations in the E. coli isolation rates among different sample types. The heatmap was created using the “Complex heatmap” package [50], and the dendrogram was created using the “hclust” function of the stats package. Furthermore, one-way analysis of variance was conducted to compare the MAR index of isolates from different sources. P-values < 0.05 were considered statistically significant.

Results

E. coli isolation and identification

Out of the 350 samples tested, 112 (32%) were E. coli positive with 68.75% (77/112) were isolated from retail meat samples and the remaining 31.25% (35/112) were recovered from retail market workers. E. coli was identified in various types of retail meat and market workers samples. The E. coli isolation rates exhibited significant variation (P = 0.0003) among the different sample sources. The highest isolation rate (54%) was observed in beef meat samples followed by workers stool samples (48%) (Table 2).

Table 2 Proportion of E. coli and ESBL-producing E. coli isolated from retail meat samples and retail shop workers
Full size table

ESBL-producing E. coli

Overall, 58% (65/112) of E. coli isolates were ESBL producers, with 45 (69.2%) and 20 (30.8%) isolates recovered from retail meat and market workers samples, respectively (Table 2). The highest proportion of ESBL-producing E. coli was observed in duck meat and worker hand swabs (81.8%), followed by camel meat samples (61.5%).

Phylotyping, virulotyping and genotyping

The phylogenetic grouping of the 65 ESBL-producing E. coli showed that 4 (6.2%) isolates belonged to group A and 61 (93.8%) to group B1 (Table 2). All group A isolates were recovered from workers stool samples (Fig. 1). None of the isolates belonged to group B2 or D. All the ESBL-producing E. coli isolates (100%) harbored at least one virulence gene, but only 25 (38.5%) of the isolates harbored two or more virulence genes (Fig. 2). The stx2 gene was amplified in 24 (36.9%) of the isolates, followed by eaeA and hlyA genes in 9 (13.8%) and 4 (6.2%) isolates, respectively. The stx2f gene was identified in only four isolates that tested positive for stx2 gene (one each from chicken, turkey, duck, and beef meat samples). None of the isolates carried the astA or stx1genes, however, all of them tested positive for the fimH gene (Fig. 2).

Fig. 1
figure 1

Heatmap representation of ESBL-producing E. coli phylotypes, virulotypes, genotypes, antimicrobial resistance patterns and genes

Full size image
Fig. 2
figure 2

Frequency of virulence genes of ESBL-producing E. coli isolated from retail meat samples and retail shop workers

Full size image

A single amplification profile was used to analyze the REP-PCR patterns of the 65 ESBL-producing E. coli isolates. The profiles were distinguished based on the number and position of the amplified fragments, which ranged in size from 290 to 1600 bp. Visual examination of the banding patterns revealed that five profiles (E1 to E5) were generated (Fig. 3). Simpson’s index of diversity was used to evaluate the discriminatory power of the REP-PCR, and the results showed that it had relatively low discriminatory power with a D value of 0.42. There were three main clusters revealed by the dendrogram analysis of the 65 analyzed isolates. (Figures 1 and 3). Notably, isolates from workers and meat within the same cluster exhibited 100% similarity.

Fig. 3
figure 3

REP-PCR based dendrogram for ESBL-producing E. coli isolated from retail meat samples and retail shop workers

Full size image

Antimicrobial susceptibility testing

Table 3 presents the antimicrobial susceptibility profiles of the 65 ESBL-producing E. coli isolates. The isolates showed high frequencies of resistance to PEN, AMP, AMX, CTX, CAZ and CRO (100%, each), while the lowest resistance rate (21.5%) was observed for GEN. All isolates were susceptible to IMP. Figure 4 displays the frequency of antimicrobial resistance by source of isolation. However, no significant difference (P > 0.05) was found between resistance rates of ESBL-producing E. coli isolated from retail meat and market workers. All isolates were classified as MDR with MAR index ranging from 0.32 to 0.95 and an average of 0.70 (Fig. 5). The average MAR index of isolates recovered from various sources differed significantly, with workers’ stool having the highest MAR index (0.86) followed by camel meat (0.85).

Table 3 Frequency of antimicrobial resistance profiles of ESBL-producing E. coli isolated from retail meat samples and retail shop workers
Full size table
Fig. 4
figure 4

Frequency of antimicrobial resistance of ESBL-producing E. coli isolated from retail meat samples and retail shop workers

Full size image
Fig. 5
figure 5

Multiple antibiotic resistance (MAR) index box plot of ESBL-producing E. coli isolated from retail meat samples and retail shop workers

Full size image

Antimicrobial resistance genes

The frequency of antimicrobial resistance genes detected in ESBL-producing E. coli isolates obtained from retail meat and market workers is shown in Fig. 6. The ESBL encoding genes (blaIMP, blaTEM, and blaCTX-M-1) were found in 100% of the isolates, while blaVIM, and blaNDM were detected in 93.8% of the isolates. However, none of the isolates were positive for blaOXA-1, blaSHV, or blaCMY. Tetracycline resistance genes (tetA and tetB) were detected in 66.2% and 27.7% of the tested isolates, respectively The resistance genes for sulfonamides (sul), and florquinolones (floR) were identified in 86.2% and 69.2% of the isolates, respectively. Only mcr-1 was detected among the tested colistin resistance genes (mcr-1 and mcr-5), and it was found in 60% of the isolates.

Fig. 6
figure 6

Frequency of antimicrobial resistance genes of ESBL-producing E. coli isolated from retail meat samples and retail shop workers

Full size image

Antimicrobial effect of silver nanoparticles

The antimicrobial activity of AgNPs-H2O2 against ESBL-producing E. coli was evaluated using the broth microdilution method. The MIC values of different concentrations of AgNPs-H2O2 against ESBL-producing E. coli isolated from retail meat and market workers are illustrated in Table 4. AgNPs-H2O2 concentrations of 0.625, 1.25, 2.5 and 5 μg/mL showed complete bacterial growth inhibition (no turbidity). The MIC50 and MIC90 were 0.625 and 2.5 μg/mL, respectively.

Table 4 The distribution of minimum inhibition concentration (MIC) values of AgNPs concentrations against ESBL-producing E. coli isolated from retail meat samples and retail shop workers
Full size table

Discussion

Unhygienic food handling and processing procedures facilitate the dissemination of MDR bacteria, including ESBL-E. coli to human consumers. Thus, enforcing strict monitoring measures and promoting sanitary procedures for meat distribution are necessary to prevent the proliferation of antibiotic-resistant bacteria [4].

In this study, ESBL-E. coli prevalence in duck meat was higher compared to the other sources (81.8%), followed by camel meat (61.5%). Similar results were reported in other studies in China [51] and Thailand [52], where MDR E. coli isolates were more common in ducks than in chicken. The higher isolation rate of ESBL-E. coli from ducks could be attributed to their nature as waterfowl, which excretes feces in water, thus enhancing the spread of the pathogens within the duck population [52].

A high ESBL-producing E. coli isolation rate of 86.7 from chicken meat in Turkey was reported by Kürekci, et al. [53]. The authors attributed the high isolation rate from chicken meat due to the extensive use of fluoroquinolones, cephalosporines and aminoglycosides in poultry industry which resulted in selection pressure for the high carriage rate of ESBL-producing E. coli in chickens [54].

The prevalence of ESBL-producing E. coli in beef was lower than that in chicken, which is consistent with the findings of Randall, et al. [55] who found rates of 20% in beef samples compared to 63% in chicken samples. Rao, et al. [56], also reported that only 1.9% of beef samples were positive for ESBL-producing E. coli, whereas 65.4% of chicken samples tested positive. Meanwhile, the prevalence of ESBL-producing E. coli in beef samples in the current study (48.1%) was much higher than the average (7.1%) of ESBL/AmpC-producing E. coli found in beef samples purchased at retail in the EU [57].

Meat is highly susceptible to microbial contamination due to multiple contacts from the slaughterhouse until consumption as food [4]. The prevalence of ESBL-E. coli was found to be high in camel (61.5%) and beef meat (48.1%) samples, this is in line with another study [4]. Similarly, El-Ghareeb, et al. [58] reported the isolation of ESBL-producing E. coli from 11.3% camel minced meat samples in Saudi Arabia. The high water and protein content of red meat provide a favorable environment for bacterial growth and may contribute to the high contamination rates [59].

In the present study, ESBL-E. coli was found to be prevalent in workers hand swabs (81.8%) and stool samples (45.8%). Similar findings were reported in studies conducted in Thailand [52, 60]. In Ethiopia, E. coli was isolated from 20% of hand swabs, with ESBL-E. coli accounting for 25% of the isolates [4]. Variations in the hygienic procedures of meat handlers and the sanitary standards of meat retail shops may be the cause of the observed differences in E. coli prevalence. Inadequate hygienic procedures of meat handlers and insufficient sanitation standards in meat retailer stores may lead to cross-contamination of meat with E. coli [4].

E. coli can be classified into four main phylogenetic groups named A1, B1, B2, and D, which can be identified by PCR of four genes [6]. Generally, commensal strains belong to groups A1 and B1, while most virulent strains belong to groups B2 and D [6, 61]. In the present study, all the isolates were of the commensal groups A (6.2%) and B1 (93.8). Previous studies have also reported that the majority of the E. coli phylotypes from different sources were of the commensal groups A1 and B1 [61,62,63]. In Egypt, Abdallah, et al. [64] reported that 80% of ESBL-producing E. coli isolated from chicken meat in the same study area were also of the commensal groups. The high proportion of commensal strains highlight their critical silent role in the spread and dissemination of ESBL-resistance genes [61, 64].

Virulotyping of the isolates revealed that all had at least one virulence gene, with stx2 (36.9%) being the most frequently identified gene, and four isolates being positive for stx2f gene. The existence of genes related to virulence such as stx1, stx2, and eae have been found to be crucial factors for the pathogenicity of E. coli strains [65]. Stx2-encoding strains of Shiga toxin producing E. coli (STEC) have been linked to more severe infections than those only possessing stx1 [66]. In our study, stx2 was the most common virulence gene (36.9%), consistent with reports from Germany, Argentina, and China [67,68,69]. In line with our findings, a study conducted in Egypt reported that 36% of MDR E. coli isolates obtained from chicken meat carried the stx2 gene [70].

In China, 28.57% and 51.02% of E. coli isolates had stx1 and stx2 genes, respectively [69]. Although Shiga toxin is necessary for STEC pathogenicity, it is not sufficient. Therefore, we examined four additional virulence-associated genes: hly, ast, eae, and fimH, that are linked to bacterial virulence. The combination of the eae and stx genes has been associated with increased virulence [71]. None of the isolates in our investigation carried the astA or stx1genes, while Nong, et al. [69] found astA in 20.41% and Ali, et al. [70] detected stx1 gene in 4% of the E.coli strains from chicken meat. All tested isolates in our study were positive for fimH gene, which is responsible for adhesion (the first step in the colonization process). Similar results were reported in poultry meat from Brazil [53, 72]. This is expected because this gene has been reported from both clinical and commensal E. coli isolates from different sources [53, 57]. The eaeA and hlyA genes were found in 13.8% and 6.2% of our isolates, respectively. A lower proportion of eaeA and hlyA encoding genes (5 and 4%, respectively) among E. coli isolates was reported previously in Egypt by Ali, et al. [70]. However, a higher percentage of eae -encoding genes (34.69%) among the E. coli isolates was observed in China [69]. In contrast, none of ESBL-producing E. coli recovered from retail meat in Mexico were positive of stx1, stx2, hlyA, or eae, virulence genes [61].

Recent studies highlight the significance of comparing isolates from various sources to evaluate the relevance of the foodborne pathway in human infection [73]. This study utilized REP-PCR to investigate the genetic relatedness between the isolates from different sources. The isolates were grouped into five profiles in three clusters. The 100% similarity between human and meat isolates within the same cluster indicates genetic relatedness and the possibility of transmission of strains from meat to humans. This is supported by the isolation of the same ESBL-associated genes from human and meat isolates. A similar study conducted in the Netherlands also reported a partially close genetic relationship between strains obtained from human carriers and chicken meat samples [74]. A Japanese study also reported 80% similarity index between E. coli isolates harboring β-lactamase obtained from domestic and imported chicken meat samples [75]. Similarly, a study in Sweden revealed that less than 0.1% of the population carried ESBL-producing isolates associated with poultry, but 5% of individuals were colonized with ESBL-encoding plasmids that were identical to those found in chicken meat and poultry isolates [76]. Additionally, a study conducted in rural Ghana demonstrated genetic links between ESBL-producing E. coli, suggesting possible transmission between poultry and human populations [77].

However, Belmar Campos, et al. [21] reported that ESBL genes produced by E. coli from chicken meat are different from those found in human stool samples and their data do not support the notion that ESBL strains from chicken meat significantly contribute to human colonization. The authors attributed this discrepancy to the fact that the chicken meat samples were collected more than six months after the human feces samples were obtained. Similarly, other studies have reported no genetic relatedness between ESBL-producing isolates from food animals and humans in close contact [73, 76]. In a study conducted in the Netherlands, no evidence was found to support the hypothesis of clonal transmission of ESBL-producing E. coli isolates between humans and poultry [78].

All isolates tested in our study were found to be MDR, with an average MAR index ranging from 0.32 to 0.95. The isolates showed 100% resistance to penicillins, beta-lactams and cephems. The MDR isolates are considered reservoirs for both resistance and virulence genes and can be transferred to other strains of the same species and other species, thereby increasing the source of antibiotic resistance [61]. Similar results were reported by other studies [52, 55]. Abayneh, et al. [4] reported that the majority of E. coli isolates (74.3%) exhibited resistance to three or more classes of antibiotics, including TET, ERY and cotrimoxazole. Additionally, 85.7% of ESBL-producers were resistant to CTX and CRO, while 71.4% were resistant to CAZ. Another study reported high resistance of ESBL-E. coli to AMP (69.4%), SXT (66.7%), TET (88.9%) and Sulfonamide (75%) [79]. In China, ESBL-producing E. coli were found to be resistant to different antibiotics such as AMP (98.9%) and TET (97.6%) [80]. Another study in Egypt TET (80.9%), STR (61.9%) and SXT (61.9%) [81]. All our isolates were susceptible to imipenem, this is consistent with another study conducted in Turkey on chicken meat samples [53]. The difference between the resistance rates of ESBL-producing E. coli to antimicrobials could be attributed to antibiotic administration practices in the veterinary field as growth promoters or for therapy, also due to geographic distribution [3].

Colistin resistance has been observed in E. coli isolates from food-producing animals, especially poultry [82]. In this study, 60% of the isolates were resistant to colistin. The resistance to colistin is mediated by chromosomal mutations in pmrA/B, phoP/Q, and mgrB genes. However, an acquired colistin-resistance gene mcr-1 has also been identified in E. coli [83, 84]. To date, nine variants of mcr have been identified in humans and different animals [85]. The existence of mcr genes in mobile genetic elements raises concerns about their potential horizontal transfer in the food chain, posing a risk to public health [14]. Since 2006, the coexistence of mcr genes with other resistance determinants, such as ESBL and/or carbapenemase genes, has been reported in Enterobacterales. Notably, an increase in the prevalence of mcr-1 genes has been observed among ESBL-producing E. coli strains in animals, while their occurrence remains low in non-ESBL-producing E. coli strains. This suggests that the use of extended-spectrum cephalosporins may have contributed to the dissemination of mcr-1 [86, 87]. A study conducted in Turkey revealed a close genetic relationship between mcr-1 genes found in chicken meat and isolates of human origin, indicating the emergence and spread of mcr-mediated colistin resistance in E. coli across various sources with zoonotic potential in the food chain [88].

In this study, none β-lactamase resistance genes including tetA, tetB, sul and flor were identified with high isolation rates ranging from 27.7% to 86.2%. In the same line, a previous study has reported high prevalence of tetA (72.58%), and sul1 (44.67%) [89]. The investigation of resistance determinants in our study indicated that ESBL-encoding genes were highly prevalent in the isolates. TEM, SHV, OXA, CMY, and CTX-M beta-lactamases are the most prevalent beta-lactamases in Gram-negative bacteria. In accordance, a study in Egypt reported that 57.55%, 46.23%, and 23.58% of the isolates had TEM, CTX-M, and SHV genes, respectively [64], In Bangladesh, Rahman, et al. [89] reported only blaSHV gene from ESBL-E. coli isolates, while in the Netherlands, blaCTX-M-1(58.1%) is the most prevalent gene found in chicken meat, followed by blaTEM-52 (14%) and blaSHV-12 (14%) [90]. In Singapore, Guo, et al. [91] found that out of 225 ESBL-producing E. coli isolates, 76.4% carried blaCTX-M genes, 45.3% carried blaTEM genes and 23.1% carried blaSHV genes. Additionally, Lim, et al. [92] emphasized the prevalence of CTX-M genes as ubiquitous ESBL genes in ESBL-producing E. coli.

Although our isolates were sensitive to imipenem by phenotypic test, carbapenem resistance genes were detected in the isolates by PCR. This indicates that not all carbapenemase-producing isolates exhibit phenotypic resistance to carbapenems due to either lack of expression or the level of gene expression is less than the required to exhibit phenotypic resistance [93, 94].

Treatment of infections caused by ESBL-producing E. coli requires high doses of antibiotics, which can lead to antibiotic resistance and adverse effects on patients [95]. Nanoparticles, such as silver nanoparticles, are considered alternatives to antibiotics for treating various infections. Silver nanoparticles have large surface area to volume ratio, allowing for increased contact with bacteria and resulting in direct interaction with the bacterial cell wall to produce antibacterial activity [96]. Our results showed that the MIC50 of AgNPs-H2O2 was 0.625 μg/mL against ESBL-producing E. coli. Another study in Egypt reported that the average MIC value of AgNPs against ESBL-producing E. coli was 27 μg/ml [95]. In India, an average MIC values of 11.25-45 μg/mL was reported [97],[103] while in Mexico, an MIC of 10 μg/mL was demonstrated for AgNPs [98]. The discrepancies in MIC values could be attributed to differences in the particle size of AgNPs [95]. Shafreen, et al. [99] argued that silver nanoparticles suspensions prepared by biological methods and with concentrations higher than 100 μg/mL may lose their antibacterial effect on microorganisms.

Conclusions

Results of present study showed high prevalence of virulent MDR ESBL-producing E. coli in retail meat products and workers in retail meat shops in Egypt. Therefore, regular monitoring of retail meat and application of hygienic food safety practices by food handlers are required for protecting consumers. Silver nanoparticles are considered a promising alternative for treating MDR ESBL-producing E. coli infections and reducing the risk of antimicrobial resistance.