Emergence of virulent ESBL-producing Escherichia coli in meat with human health implications and control using liposomal cinnamon, oregano, and clove within a One Health framework

Abstract

Transmission of extended-spectrum β-lactamases-producing (ESBL) Escherichia coli from meat and meat by-products to humans has emerged as a major public health issue, requiring a One Health framework to address this menace. Hence, we investigated the phylotypes, antimicrobial resistance profiles, and virulotypes of E. coli isolates from chicken meat, beef burger, and human stool samples, besides investigating the in vitro antimicrobial and antivirulence efficacies of liposomal cinnamon, oregano, and clove essential oils (LCOC). A total of 90 isolates (28.1%) were phenotypically and molecularly identified as E. coli, and they were classified into four phylogenetic groups: B1, B2, A, and D (40, 35.6, 14.4, and 10%, respectively). The majority of the isolates were multidrug-resistant (MDR) (97.8%), with remarkable resistance against ampicillin, tetracycline, and cefotaxime (95.6, 84.4, and 81.1%, respectively). The blaTEM, tetA, aadA1, and blaCTX-M were the most prevalent resistance genes (96.7, 85.6, 81.1, and 81.1%, respectively). Virulotyping revealed that 70% of the isolates were multi-virulent, with iroN being the most prevalent one (92.2%). LCOC demonstrated in vitro antibacterial and antivirulence properties via inhibiting the growth of MDR, multi-virulent, and ESBL meat E. coli isolates and downregulating the expression of their investigated virulence genes. Concisely, the observed 28.1% prevalence of ESBL-producing E. coli indicates a notable public health concern associated with meat contamination. Our findings demonstrate that LCOC exhibited antimicrobial activity against the tested E. coli isolates and may help reduce their antibiotic resistance and virulence potential.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-48140-y.

Introduction

Foods of animal origin, including meat and meat by-products, are high in proteins that are vital for bodily development and growth¹. Unsanitary slaughter and processing procedures can cause meat contamination with enteric bacteria such as Escherichia coli, which can lead to worldwide human foodborne illnesses². Intestinal and extraintestinal pathogenic E. coli strains (InPEC and ExPEC, respectively) are capable of causing different intestinal and extraintestinal diseases³, and they can be readily spread through the food chain to several ecosystems⁴ through ingestion of contaminated undercooked or raw meat products or via direct contact with the bacteria during food preparation⁵.

E. coli is classified into four primary phylotypes (A, B1, B2, D) depending on the existence or lack of the TspE4.C2, yjaA, and chuA genes⁶. Most virulent, highly pathogenic strains are found in phylotypes B2 and D⁷. ExPEC strains, distinguished by their array of virulence factors, predominantly belong to phylotypes B2 and D, while InPEC may belong to any of the phylotypes, with phylotype B2 strains exhibiting superior abilities to endure within the gut microbiota compared to strains from other phylotypes⁸. On the other hand, commensal strains often fall into phylotypes A and B1, where phylotype A is non-pathogenic, and phylotype B1 is less pathogenic (mostly environmental strains)⁹. E. coli pathogenicity is associated with several virulence attributes, such as shiga toxins (stx1 and stx2), outer membrane protein (ompA), intimin (eaeA), siderophores (iroN), and temperature-sensitive hemagglutinin (tsh) encoding genes¹⁰. Several E. coli virulence factors collaborate during host invasion to allow the bacteria to escape, destroy the host’s immune defenses, and provoke inflammatory reactions, thus leading to clinical illnesses¹¹.

Of note, E. coli strains obtained from tainted meat and its byproducts exhibit resistance to widely utilized antimicrobials, and contaminated meats and their byproducts are considered reservoirs for multidrug-resistant (MDR) E. coli strains². β-lactam resistance in E. coli is evolving because of the heightened development of extended-spectrum β-lactamases (ESBL), and broad-spectrum cephalosporins are hydrolyzed by these ESBL enzymes. The majority of ESBLs in the Enterobacteriaceae family are Ambler class A β-lactamases (KPC, CTX-M, TEM), class B metallo-β-lactamases (MBL; IMP, VIM, and NDM), class C (AmpC) (DHA, and CMY), and class D (OXAs) enzymes, and they are encoded on mobile genetic elements (integrons or plasmids)^2,12,13.

The global emergence of MDR bacterial infections highlights the critical need to decrease the utilization of antimicrobial agents by looking into alternative strategies^14,15. Phytochemicals have emerged as viable substitutes for antimicrobials in managing infectious pathogens^16,17. In contrast to chemical antibiotics, essential oils (EOs) have strong antimicrobial and antivirulence characteristics against resistant bacteria and enhance the currently utilized antimicrobials¹⁸. Carvacrol, trans-cinnamaldehyde, and eugenol are the primary phenolic constituents of oregano (Origanum glandulosum), cinnamon (Cinnamomum zeylandicum), and clove (Syzygium aromaticum) EOs^19,20, and their subinhibitory concentrations (SICs) reduced the virulence genes’ transcriptional levels and prevented the formation of toxins in different bacterial species^19,21–23. These compounds are known to target multiple bacterial pathways, including cell membrane integrity, quorum sensing, and virulence factor expression. Their combination is therefore expected to exert synergistic effects, enhancing antibacterial efficacy while potentially reducing the required effective dose, as recent data showed that combinations of EOs yield superior results compared to the use of a single addition²⁴. Nonetheless, the antimicrobial effectiveness of EOs may be diminished by issues of low stability, volatile nature, and inadequate water solubility, as well as their unpleasant flavor^1,15. To address these challenges, liposomes, colloidal structures characterized by a spherical shape composed of phospholipid bilayer membranes and an internal aqueous compartment, may be employed to encapsulate and regulate the release of EO components, thereby enhancing their stability, functioning, and biological effects while mitigating their drawbacks^1,25,26. The influence of EOs on the transcription of pathogenic bacteria’s virulence genes was studied in broth or food-like broth^27,28. Nonetheless, external factors like the kind of food or different sample matrices could significantly impact the transcription of virulence genes. Comprehending the bacterial survival and pathogenicity patterns in the food matrices in relation to EOs is essential for the implementation of efficient decontamination and prevention techniques²⁷.

To address the difficulties of antimicrobial resistance (AMR) phenomena, a multidisciplinary One Health perspective has been advocated to prevent and battle the emergence and dissemination of MDR strains across humans, animals, and their surroundings²⁹. Recent findings indicate that when a blend of EOs, such as oregano, cinnamon, and clove, is employed, the phenolic bioactive constituents may exert synergistic effects on their mechanisms of action^14,20,24. A few pieces of research have discussed the in vitro antibacterial efficacies of oregano, clove, and cinnamon EOs against ESBL-E. coli, but to our knowledge, there has been no investigation into the antibacterial and anti-virulence impact of liposomal cinnamon, oregano, and clove essential oils (LCOC) against carbapenem-resistant (CR) and ESBL-producing E. coli strains in meat. Hence, the present research was undertaken to detect the prevalence, phenotypic and genotypic patterns of antimicrobial resistance, virulotypes, and phylotypes of E. coli strains in chicken meat, beef burger, and human stool samples in Egypt under the One-Health framework. Additionally, we aimed to assess, for the first time, the in vitro antimicrobial and antivirulence effectiveness of LCOC against multi-virulent, MDR, CR, and ESBL-E. coli strains in chicken meat and beef burger samples.

Materials and methods

Ethics considerations

All methods were done following the relevant guidelines and regulations. For human and animal samples, the collection protocols were undertaken and approved in compliance with the recommendations of the Suez Canal University Ethics Review Committee (AERC-SCU), Egypt, under reference No. AERC-SCU-VET 2,024,012, and informed consent was obtained from all participants prior to their inclusion in the study. In our research, sample collection was solely used for patient care and antimicrobial susceptibility testing to ensure proper diagnosis and treatment. All procedures were carried out following ARRIVE recommendations for animal samples and local and institutional regulations for both animal and human samples.

Sampling

A total of 320 samples were randomly and aseptically retrieved from chicken meat (n = 150), beef meat by-products (n = 100), and diarrheic humans (n = 70) during the period from April 2024 to August 2024. Each sample represented a single independent source, corresponding to one patient, one animal, or one individual retail beef burger package. The chicken meat and beef burger samples were randomly and aseptically obtained, commercially, in sterile plastic bags from major slaughterhouses and retail poultry shops in Ismailia and Sharkia Governorates, Egypt. Additionally, 70 human stool samples were obtained from diarrheic patients attending private hospitals and laboratories in Ismailia and Sharkia Governorates, Egypt. All participants granted their informed permission for inclusion before their participation in the research. Meat specimens and human stool swabs were aseptically transported in an icebox to the bacteriology laboratory for additional examination as promptly as possible.

Isolation and identification of E. coli

For E. coli isolation, human stool swabs and 10 g of each meat sample were aseptically homogenized in 90 mL of buffered peptone water and incubated at 37°C/24 h. Following incubation, 10 µL of the prepared samples were inoculated onto the surface of MacConkey and eosin methylene blue agar (Oxoid, UK) plates, and incubated for 24 h at 37°C. After purification on tryptone soy agar (Oxoid, UK), suspected colonies were identified by Gram’s staining, cultural features on blood agar, and biochemical tests comprising sugar fermentation, nitrate reduction, hydrogen sulfide production, catalase, oxidase, indole, citrate, urease, Voges-Proskauer, and methyl red³⁰.

Molecular confirmation and phylotyping of recovered E. coli isolates

The QIAamp DNA Mini kit (Qiagen, Germany) was utilized for the extraction of DNA from preliminarily identified E. coli colonies, complying with the instruction manual. One pair of oligonucleotide primers (Thermo Fisher Scientific, Massachusetts, USA) targeting the phoA gene was utilized in conventional uniplex PCR procedures to verify E. coli. The confirmed E. coli strains were phylotyped via amplification of tspE4.C2, yjaA, and chuA genes using the phylotype categorization scheme⁶. All PCR reactions were performed in triplicate utilizing the Emerald Amp GT PCR master mix (Takara, CA, USA) as per the manufacturer’s regulations. The primer sequences used in all PCR protocols are depicted in Table 1, and the amplification procedures were executed as previously delineated^6,31. DNA of Staphylococcus aureus ATCC 6538 and E. coli ATCC 25,922 were utilized as negative and positive controls, respectively, in all PCR procedures.

Table 1.

Primer sequences and amplified PCR products of target genes utilized in PCR assays.

Specificity/ target gene	Primer sequence (5`–3`)	Amplified product (bp)	References
E. coli
phoA	F: CGATTCTGGAAATGGCAAAAG	720	31
	R: CGTGATCAGCGGTGACTATGAC
Phylotyping
tspE4.C2	F: GAGTAATGTCGGGGCATTCA	152	6
	F: CGCGYCAACAAAGTATTRCG
yjaA	F: TGAAGTGTCAGGAGAYGCTG	211
	R: ATGRAGAATGCGTTCCTCAAC
chuA	F: GACGAACCAACGGTCAGGAT	279
	R: TGCCGCCAGTACCAAAGACA
Integron
Int1	F: CCTCCCGCACGATGATC	280	32
	R: TCCACGCATCGTCAGGC
Antimicrobial resistance genes
sul1	F: CGGCGTGGGCTACCTGAACG	433	33
	R: GCCGATCGCGTGAAGTTCCG
aadA	F: TATCAGAGGTAGTTGGCGTCAT	484	34
	R: GTTCCATAGCGTTAAGGTTTCATT
tetA	F: GGTTCACTCGAACGACGTCA	576
	R: CTGTCCGACAAGTTGCATGA
qnrA	F: ATTTCTCACGCCAGGATTTG	516	35
	R: GATCGGCAAAGGTTAGGTCA
qnrS	F: ACGACATTCGTCAACTGCAA	417
	R: TAAATTGGCACCCTGTAGGC
blaDHA	F: ATGTCACTGTATCGCCGTCT	405	36
	R: TTACTGCCCGTTGACGCCC
blaCMY	F: GCTGCTCAAGGAGCACAGGAT	520
	R: CACATTGACATAGGTGTGGTG
blaCTX-M	F: ATGTGCAGYACCAGTAARGTKATGGC	593	37
	R: TGGGTRAARTARGTSACCAGAAYCAGCGG
blaTEM	F: ATCAGCAATAAACCAGC	800	38
	R: CCCCGAAGAACGTTTTC
blaOXA	F: ATATCTCTACTGTTGCATCTCC	564
	R: AAACCCTTCAAACCATCC
blaVIM1	F: AGTGGTGAGTATCCGACAG	261	39
	R: ATGAAAGTGCGTGGAGAC
blaIMP1	F: CCGCAGCAGAGTCTTTGCC	587
	R: ACAACCAGTTTTGCCTTACC
	F: CCGCAGCAGAGTCTTTGCC
blaNDM1	F: GGCGGAATGGCTCATCACGA	287	40
	R: CGCAACACAGCCTGACTTTC
Virulence genes
tsh	F: GGTGGTGCACTGGAGTGG	620	41
	R: AGTCCAGCGTGATAGTGG
ompA	F: AGCTATCGCGATTGCAGTG	919	42
	R: GGTGTTGCCAGTAACCGG
iroN	F: ATCCTCTGGTCGCTAACTG	847
	R: CTGCACTGGAAGAACTGTTCT
eaeA	F: GTGGCGAATACTGGCGAGACT	890	43
	R: CCCCATTCTTTTTCACCGTCG
stx1	F: ACACTGGATGATCTCAGTGG	614	44
	R: CTGAATCCCCCTCCATTATG
stx2	F: CCATGACAACGGACAGCAGTT	799
	R: CCTGTCAACTGAGCAGCACTTTG
Housekeeping gene
16S rRNA	F: CGATGCAACGCGAAGAACCT	178	45
	R: CCGGACCGCTGGCAACAAA

Antimicrobial susceptibility testing

The antimicrobial susceptibility profiles of the molecularly confirmed E. coli isolates were ascertained through the application of the standardized Kirby-Bauer disc diffusion technique⁴⁶, utilizing Muller-Hinton agar (Oxoid, Hampshire, UK) and eleven antimicrobial agents from nine different antimicrobial classes as per the recommendations of the Clinical and Laboratory Standards Institute (CLSI)⁴⁷. Eleven antimicrobial discs (Oxoid, Hampshire, UK) were utilized; ampicillin (AMP, 10 µg), tetracycline (TE, 30 µg), ciprofloxacin (CIP, 5 µg), trimethoprim/sulfamethoxazole (SXT, 1.25 + 23.75 µg), amoxicillin/clavulanic acid (AMC, 20 + 10 µg), gentamicin (CN, 10 µg), meropenem (MEM, 10 µg), tigecycline (TGC, 15 µg), ceftazidime (CAZ, 30 µg), cefotaxime (CTX, 30 µg), and levofloxacin (LEV, 5 µg). Quality control was performed using Escherichia coli ATCC 25,922, and the obtained inhibition zone diameters for all tested antibiotics were within the acceptable ranges specified by CLSI guidelines. The multiple antibiotic resistance index (MARI) for each isolate was estimated by dividing the count of antimicrobials to which the investigated isolate was resistant by the total count of antimicrobial agents evaluated²⁰. For the purpose of MARI calculation, isolates exhibiting intermediate susceptibility were considered susceptible. The MDR phenomenon was described as an isolate’s resistance to at least one agent in three or more unrelated antimicrobial classes⁴⁸.

Antimicrobial and antivirulence impact of liposomal cinnamon, oregano, and clove essential oils on multi-virulent, MDR, CR, and ESBL-producing meat E. coli isolates

The liposomal encapsulated cinnamon, oregano, and clove EOs (LCOC) employed in this investigation were acquired from Hopkinton Drug Inc. (Hopkinton, MA, USA), which included 200 mg of cinnamon (Cinnamomum verum) extract, 150 mg of oregano (Origanum minutiflorum) extract, and 50 mg of clove (Syzygium aromaticum) extract. LCOC was utilized to evaluate their antimicrobial and antivirulence efficacies against multi-virulent, MDR, CR, and ESBL-producing meat E. coli isolates carrying all the examined virulence genes: tsh, iroN, ompA, eaeA, stx1, and stx2.

Evaluation of antimicrobial effects of liposomal cinnamon, oregano, and clove essential oils by agar well diffusion and broth microdilution techniques

The antimicrobial effects of LCOC against multi-virulent, MDR, CR, and ESBL-E. coli isolates were assessed by the broth microdilution technique as described previously⁴⁷. Every experiment was conducted in triplicate. The bacterial inoculum was standardized to approximately 1 × 10⁶ CFU/mL using freshly prepared cultures. Microdilution plates were incubated at 37°C for 18–24 h under aerobic conditions. LCOC was prepared using an appropriate solvent (dimethyl sulfoxide, DMSO), ensuring that the final solvent concentration had no inhibitory effect on bacterial growth. Appropriate solvent controls were included.

The minimal inhibitory concentration (MIC) of LCOC was the lowest concentration that entirely inhibited the growth of E. coli⁴⁹, and the sub-MIC of LCOC was identified as its respective SIC²³. Bacterial viability at the MIC endpoint was confirmed by subculturing aliquots from wells showing no visible growth onto fresh agar plates.

Assessment of the antivirulence impact of liposomal cinnamon, oregano, and clove essential oils on virulence gene expression of E. coli isolates in chicken and beef meat via reverse transcription quantitative PCR technique

The reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis was employed to assess the impacts of SIC of LCOC on the transcription levels of E. coli virulence-related genes in multi-virulent, MDR, CR, and ESBL-E. coli isolates in chicken meat and beef burger, as earlier pronounced²⁷. In brief, the examined LCOC at its SIC was combined with 5 g of meat samples in stomacher bags. To disperse the LCOC and E. coli inoculum, meat specimens were manually massaged for roughly two minutes after being inoculated with ∼10⁹ CFU of E. coli inoculum/g. Infected meat specimens were stabilized with 10 mL of RNA stabilization reagent (RNAlater; Qiagen, Germany), and the meat was macerated with 40 mL of sterile peptone water (0.1%) in the stomacher for two min. Filtrated meat juice aliquots were centrifuged at 7000 rpm for five min at 4 °C^27,28, and then the pellets were collected for RNA extraction using the QIAamp RNease Mini Kit (Qiagen, USA) as per the manufacturer’s guidelines. The RNA’s purity and concentration were evaluated utilizing a NanoDrop 2000 spectrophotometer. Using the appropriate primers (Table 1) and QuantiTect SYBR Green PCR Master Mix (Qiagen, USA), the expression levels of the virulence genes of E. coli isolates that had been pretreated or not with LCOC were ascertained using the SYBR Green RT-qPCR technique, in triplicate, in a Stratagene MX3005P real-time PCR machine (Thermo Fisher, USA), and the amplified DNAs’ melting curves were produced to evaluate the specificity of the reaction. DNA of S. aureus ATCC 6538 and E. coli ATCC 25,922 were utilized as negative and positive controls, respectively, in each RT-qPCR run. The 2^−ΔΔCt approach was employed to determine the mRNA transcription levels of virulence genes in E. coli isolates treated with LCOC in comparison to the control non-treated isolates⁵⁰. The 16S rRNA⁴⁵ was employed as a housekeeping gene and endogenous control for the transcript levels of the studied genes. The end outcomes were displayed as fold changes of the investigated genes` transcription levels in the analyzed isolates compared to the control isolates.

Molecular characterization of genes related to integron 1, antimicrobial resistance, and virulence in retrieved E. coli isolates

Conventional PCR amplification techniques were conducted on the identified E. coli isolates for determining the presence of genes conferring resistance to aminoglycosides: aadA, sulfonamides: sul1, tetracycline: tetA, quinolones: qnrA, and qnrS, β-lactams: blaCTX-M, blaCMY, blaTEM, and blaDHA, carbapenems: blaOXA, blaNDM, blaVIM1, and blaIMP1, and integron: int1. Virulotyping was also conducted by detecting six important virulence genes, including tsh, iroN, ompA, eaeA, stx1, and stx2, via PCR assays. The primer sequences employed in all PCR procedures are depicted in Table 1, and the amplification procedures were executed as previously stated^32–44.

Evaluation of sensory properties of meat

For sensory analysis, the SIC of LCOC was applied to the chicken meat and beef burger samples. Chicken meat and beef burger samples were cooked for 15 min at 205 °C, and when the internal temperature reached 80°C, they were cooked for an additional 30 s. The samples were presented in individual cups with sealed lids and were designated by three random digits⁵¹. Ten trained panelists evaluated the appearance, juiciness, taste, tenderness, odor, and overall acceptability of grilled meat samples, assigning scores from 1 to 5; 1 = Very unfavorable, 2 = Moderately unfavorable, 3 = Neutral, 4 = Moderately favorable, 5 = Very favorable⁵². The sensory assessment was conducted following a standardized sensory descriptive methodology by ISO (the International Organization for Standardization)⁵³.

Statistical analysis

Differences in the prevalence of antimicrobial resistance (AMR) phenotypes, resistance genotypes, integron 1, and virulence genes in human and animal samples were assessed using the Chi-square test. All analyses were conducted at a significance threshold of p < 0.05 using SPSS Inc. 26 software (IBM Corp., NY, USA). The frequencies of resistance phenotypes, genotypes, and virulence genes were visualized as stacked bar graphs and boxplots using the ggplot package⁵⁴ in R software version 4.3.3 (https://www.r-project.org/).

Heatmaps, hierarchical clustering dendrograms, and principal component analysis (PCA) biplot were used to investigate the overall distribution of variables across isolates from human and animal sources and their interrelationships, employing the pheatmap⁵⁵ and factoextra⁵⁶ packages in R software version 4.3.3.

Associations between variables were evaluated using Pearson’s correlation analysis to explore genotype–phenotype relationships. Prior to correlation analyses, normality of variables was assessed using Q–Q plots. Pearson correlation coefficients (r) were calculated and visualized using the Hmisc package⁵⁷ in R software version 4.3.3, with significant correlations detected at p < 0.05.

Pair scatter plot matrices were used to estimate the correlation between isolates from different sources based on the prevalence of resistance phenotypes, genotypes, and virulence genes using the ggplot package in R software version 4.3.3. These integrative analyses were performed to correlate data across sample sources, thereby reinforcing the One Health perspective of the study.

The independent sample T-test was applied to assess the effects of LCOC on the expression of virulence genes in multi-virulent, MDR, CR, and ESBL-producing meat E. coli isolates. Graphs were generated using the pheatmap package⁵⁵ in R software version 4.3.3 and GraphPad Prism 8 (San Diego, USA).

Results

Prevalence and phylotyping of E. coli isolates

Out of 320 samples obtained from chicken meat, beef meat by-products, and diarrhetic humans, 90 isolates (28.1%) were phenotypically identified as E. coli, with one representative isolate selected per positive sample for subsequent analyses (Supplementary Fig. S1). The highest incidence of E. coli isolates was detected among human samples (32/70, 45.7%), followed by chicken meat (40/150, 26.7%) and beef burger (18/100, 18%) samples. Statistical analysis exhibited a significant difference in the prevalence of E. coli isolates from chicken meat, beef burger, and human stool samples (p = 0.0004).

All 90 retrieved E. coli isolates possessed the phoA species-specific gene and were molecularly verified as E. coli. The phylogenetic grouping of the 90 molecularly confirmed E. coli isolates revealed that 36 (40%), 32 (35.6%), 13 (14.4%), and 9 (10%) isolates were categorized in phylogenetic groups B1, B2, A, and D, respectively. The majority of E. coli isolates from beef burger and chicken meat samples belonged to phylotypes B1 (66.7% and 35%, respectively) and A (33.3% and 15%, respectively). Most human E. coli isolates (n = 32) belonged to phylotype B2 (65.6%), followed by phylotypes B1 (31.3%) and A (3.1%) (Supplementary Fig. S2). Statistical analysis exhibited substantial differences in the prevalence of phylotypes A, B1, B2, and D among E. coli isolates from various sources (p < 0.05).

Antimicrobial susceptibility profiles of E. coli isolates

As indicated in Table 2 and Supplementary Fig. S3, all 90 recovered E. coli isolates were assessed for their susceptibility to eleven antimicrobial agents from nine relevant antimicrobial classes. Regardless of the sample source, it was noticed that most E. coli isolates revealed remarkable resistance to ampicillin (95.6%), followed by tetracycline and cefotaxime (84.4 and 81.1%, respectively). On the other hand, relatively low resistance rates were observed against meropenem (21.1%) and ciprofloxacin (44.4%). Interestingly, all the examined E. coli isolates exhibited complete susceptibility to tigecycline (100%), establishing it as the preferred treatment option for E. coli infections.

Table 2.

Antimicrobial resistance patterns of the retrieved E. coli isolates from different sources.

Antimicrobial class	Antimicrobial agent	No. of resistant E. coli isolates (%)			p-value	Total no. of E. coli isolates (%) (n = 90)
		Chicken (n = 40)	Beef burger (n = 18)	Human (n = 32)
Penicillins	AMP	37 (92.5)	18 (100)	31 (96.9)	0.533	86 (95.6)
β-lactams combination agents	AMC	26 (65)	12 (66.7)	27 (84.4)	0.172	65 (72.2)
Cephalosporins ӀӀӀ	CTX	27 (67.5)	14 (77.8)	32 (100)	0.002**	73 (81.1)
	CAZ	26 (65)	14 (77.8)	32 (100)	0.001**	72 (80)
Carbapenems	MEM	14 (35)	0 (0)	5 (15.6)	0.006**	19 (21.1)
Fluoroquinolones	CIP	19 (47.5)	0 (0)	21 (65.6)	< 0.0001***	40 (44.4)
	LEV	21 (52.5)	0 (0)	21 (65.6)	< 0.0001***	42 (46.7)
Tetracyclines	TE	31 (77.5)	18 (100)	27 (84.4)	0.098	76 (84.4)
Glycylcyclines	TGC	0 (0)	0 (0)	0 (0)	NA	0 (0)
Folate pathway antagonists	SXT	28 (70)	16 (88.9)	26 (81.3)	0.244	70 (77.8)
Aminoglycosides	CN	31 (77.5)	16 (88.9)	24 (75)	0.56	71 (78.9)

AMP, Ampicillin; AMC, Amoxicillin/clavulanic acid; CIP, Ciprofloxacin; CTX, Cefotaxime; CAZ, Ceftazidime; MEM, Meropenem; SXT, Trimethoprim/sulfamethoxazole, TGC, Tigecycline; TE, Tetracycline; LEV, Levofloxacin; CN, Gentamicin; n, Number; NA, Non-applicable.

*p > 0.05, **p > 0.01, ***p > 0.001.

Regarding the isolation source, elevated resistance rates were noted in human E. coli isolates compared to those from chickens and beef cattle for the examined antimicrobials, except for ampicillin, meropenem, trimethoprim-sulfamethoxazole, and gentamicin (Table 2 and Supplementary Fig. S3). The majority of chicken E. coli isolates were resistant to ampicillin (92.5%), followed by tetracycline and gentamicin antimicrobials (77.5% each). Among beef burger isolates, all E. coli isolates were resistant to ampicillin and tetracycline, while all isolates were sensitive to meropenem, ciprofloxacin, levofloxacin, and tigecycline antimicrobials. All human E. coli isolates were resistant to cefotaxime and ceftazidime antimicrobials. Statistical analysis indicated significant differences in the antimicrobial resistance profiles of E. coli isolates from different sources to meropenem, cefotaxime, ceftazidime, ciprofloxacin, and levofloxacin antimicrobials (p < 0.05) (Table 2). Our findings exhibited significantly (p < 0.05) higher resistance rates of phylotype B2 E. coli isolates for ceftazidime, ciprofloxacin, and levofloxacin antimicrobials. Significantly (p = 0.0002) higher resistance rates of phylotype B1 E. coli isolates for cefotaxime antimicrobial (Table 2 and Supplementary Fig. S3).

Most E. coli isolates were resistant to 6 antimicrobial classes (36.7%) and 7 antimicrobial agents (28.9%) with a MARI of 0.64. The majority of chicken E. coli isolates were resistant to 7 antimicrobial agents (30%) and 6 antimicrobial classes (35%). Most of the beef E. coli isolates were resistant to 7 antimicrobial agents, and 5 and 6 antimicrobial classes (44.4% each). Most of the human E. coli isolates were resistant to 9 antimicrobial agents (34.4%), and 7 antimicrobial classes (37.5%) (Fig. 1 and Supplementary Table S1). There were statistically significant variations (p < 0.05) in the prevalence of resistance to 4, 6, and 9 antimicrobials, and 4, 5, and 7 antimicrobial classes among E. coli isolates from different sources (Supplementary Table S1). The majority of E. coli isolates belonging to phylotype A were significantly (p = 0.02 and 0.009, respectively) resistant to 6 antimicrobial agents (38.5%) and 5 antimicrobial classes (53.8%), while most phylotype D E. coli isolates (44.4%) were significantly (p = 0.019 and 0.003, respectively) resistant to 4 antimicrobial agents and classes. In total, resistance to 10 antimicrobial agents (MARI = 0.91) and 8 antimicrobial classes was detected among four E. coli isolates (4.4%) obtained from three chicken meat and one human stool samples. Overall, 97.8% (88/90) of the investigated E. coli isolates were MDR, having MARI ≥ 0.27, with 15.6 and 16.7% of E. coli isolates being resistant to 8 and 9 antimicrobials (MARI of 0.73 and 0.82, respectively) (Fig. 1).

Fig. 1.

Multiple antibiotic resistance (MAR) indices (A) and prevalence of resistance to various antimicrobial agents (B) and classes (C) of E. coli isolates obtained from chicken meat, beef burger, and human stool samples. MAR index: multiple antibiotic resistance index, No: number. In stacked bar (B, C), the frequency was calculated concerning the total number of the examined isolates (n = 90) for each parameter, and sub-columns are calculated as a part of the total column.

Virulotyping of E. coli isolates from meat and human samples

Among the 90 tested E. coli isolates, the iroN gene was the most prevalent one (92.2%), followed by ompA (86.7%), eaeA (55.6%), tsh (40%), stx1 (33.3%), and stx2 (14.4%) ones (Figs. 2 and 3). All the examined E. coli isolates (100%) carried at least one virulence gene, and only one chicken E. coli isolate (1.1%) belonging to phylotype B1 possessed all six investigated virulence genes (Figs. 2 and 4A). Additionally, multi-virulent criteria (carrying 3 or more genes) were present among 63 E. coli isolates (70%), and most isolates possessed three examined virulence genes (34.4%). Fourteen (15.6%) and 15 (16.7%) of the investigated E. coli isolates possessed four and five virulence genes, respectively (Fig. 4A and Supplementary Table S2).

Fig. 2.

Heat map illustrating the antimicrobial resistance patterns and the occurrence of virulence, integron 1, and antimicrobial resistance-associated genes among various E. coli isolates from chicken meat, beef burger, and human stool samples. Various categories of phylotypes, sample source, antimicrobial classes, antimicrobial resistance genes, integron 1, and virulence genes are color-coded on the right of the heatmap. The isolate code numbers on the right of the heatmap represent the chicken meat (CM), beef burger (BB), and human stool (H) samples.

Fig. 3.

Heatmap exhibiting the prevalence of virulence, integron 1, and antimicrobial resistance-associated genes among various E. coli isolates from chicken meat, beef burger, and human stool samples.

Fig. 4.

Distribution of the investigated virulence (A) and antimicrobial resistance (B) genes among E. coli isolates from chicken meat, beef burger, and human stool samples. In the stacked bar plots, the frequency was calculated concerning the total number of the examined isolates (n = 90) for each parameter, and sub-columns are calculated as a part of the total column.

Notably, 20 virulence gene profiles were determined across the investigated E. coli isolates. Sixteen E. coli isolates (17.8%) showed the most common virulence gene profile (19: ompA, iroN) (Table 3). There were statistically significant differences in the distribution of seven virulence genes’ profiles (5: eaeA, ompA, tsh, iroN; 7: eaeA, ompA, iroN; 9: eaeA, stx1, iroN; 14: stx1, stx2, iroN; 16: eaeA, ompA; 18: tsh, iroN; 19: ompA, iroN) among E. coli isolates from chicken, beef, and human origins (p = 0.006, 0.013, 0.038, 0.038, 0.038, 0.038, and 0.002, respectively). Moreover, there were significantly higher differences in the prevalence of two virulence genes’ profiles (6: eaeA, stx1, ompA, iroN, and 12: ompA, tsh, iroN) among E. coli isolates from chicken, beef, and human origins (p > 0.0001 each) (Table 3). Two virulence gene profiles (5: eaeA, ompA, tsh, iroN and 12: ompA, tsh, iroN) were more prevalent among chicken meat isolates than human ones, and four virulence gene profiles (14: stx1, stx2, iroN; 16: eaeA, ompA; 18: tsh, iroN; and 19: ompA, iroN) were more prevalent among beef burger isolates than human ones, which suggests a possible public health hazard due to the risk of dissemination along the food chain (Table 3).

Table 3.

Distribution of virulence gene profiles among E. coli isolates from chicken meat, beef burger, and human stool samples.

Virulence gene profile	No. of E. coli isolates harboring virulence gene profile (%)			p-value	Total no. of E. coli isolates (%) (n = 90)
	Chicken meat (n = 40)	Beef burger (n = 18)	Human stool (n = 32)
1: eaeA, stx1, stx2, ompA, tsh, iroN	1 (2.5)	0	0	1	1 (1.1)
2: eaeA, stx1, ompA, tsh, iroN	5 (12.5)	0	3 (9.4)	0.319	8 (8.9)
3: eaeA, stx2, ompA, tsh, iroN	2 (5)	0	0	0.357	2 (2.2)
4: stx1, stx2, ompA, tsh, iroN	3 (7.5)	2 (11.1)	0	0.209	5 (5.6)
5: eaeA, ompA, tsh, iroN	7 (17.5)	0	0	0.006**	7 (7.8)
6: eaeA, stx1, ompA, iroN	0	0	9 (28.1)	< 0.0001***	9 (10)
7: eaeA, ompA, iroN	1 (2.5)	2 (11.1)	8 (25)	0.013*	11 (12.2)
8: eaeA, ompA, tsh	1 (2.5)	0	0	1	1 (1.1)
9: eaeA, stx1, iroN	0	2 (11.1)	0	0.038*	2 (2.2)
10: eaeA, stx1, ompA	0	0	1 (3.1)	0.556	1 (1.1)
11: eaeA, tsh, iroN	1 (2.5)	0	0	1	1 (1.1)
12: ompA, tsh, iroN	12 (30)	0	0	< 0.0001***	12 (13.3)
13: stx1, ompA, iroN	0	0	1 (3.1)	0.556	1 (1.1)
14: stx1, stx2, iroN	0	2 (11.1)	0	0.038*	2 (2.2)
15: eaeA, iroN	4 (10)	0	0	0.074	4 (4.4)
16: eaeA, ompA	0	2 (11.1)	0	0.038*	2 (2.2)
17: eaeA, stx1	1 (2.5)	0	0	1	1 (1.1)
18: tsh, iroN	0	2 (11.1)	0	0.038*	2 (2.2)
19: ompA, iroN	1 (2.5)	6 (33.3)	9 (28.1)	0.002**	16 (17.8)
20: ompA	1 (2.5)	0	1 (3.1)	1	2 (2.2)

*p > 0.05, **p > 0.01, ***p > 0.001.

The eaeA, stx1, ompA, and iroN virulence genes were more abundant among human E. coli isolates than those obtained from meat samples. While stx2 and tsh genes were more frequent among E. coli isolates obtained from beef burger and chicken meat samples, respectively. All human E. coli isolates possessed the ompA gene, while none of the isolates carried the tsh gene. Among chicken and beef E. coli isolates, iroN was the most predominant gene (92.5% and 88.9%, respectively) (Fig. 3). There were statistically significant differences in the prevalence of ompA (p = 0.003) and tsh (p < 0.0001) virulence genes across E. coli isolates retrieved from different sources. Among the 90 examined E. coli isolates, 15 isolates (16.7%) possessed five investigated virulence genes (Fig. 4A and Supplementary Table S2). The presence of two and four virulence genes was more significantly (p = 0.006 and 0.046, respectively) prevalent between isolates obtained from beef burger (55.6%) and human stool (28.1%) samples, respectively.

Screening of genes encoding integron 1 and antimicrobial resistance in the retrieved E. coli isolates from meat and human samples

Among the analyzed 90 E. coli isolates, 78 (86.7%) and 77 (85.6%) were found to carry int1 and tetA genes, respectively. Additionally, 73 (81.1%), 71 (78.9%), and 44 (48.9%) isolates were positive for aadA, sul1 gene, and qnrA genes, respectively. Meanwhile, the qnrS gene was absent among all tested E. coli isolates (Figs. 2 and 3). All the investigated E. coli isolates possessed at least two β-lactams resistance genes, and only two chicken E. coli isolates (2.2%) belonging to phylotypes A and B1 harbored all eight investigated β-lactams resistance genes. Concerning the cephalosporins and penicillin resistance genes in ESBL-E. coli, the recovered isolates possessed blaTEM, blaCTX-M, blaCMY, and blaDHA resistance genes with prevalence rates of 96.7, 81.1, 72.2, and 42.2%, respectively. Concerning carbapenemase genes in CR strains of E. coli, blaOXA, blaIMP-1, blaVIM-1, and blaNDM-1 genes were found among 68.9, 47.8, 27.8, and 23.3% of the examined E. coli isolates, respectively (Fig. 3), and 85.6% (77/90) of the investigated isolates possessed a minimum of one carbapenemase gene. Notably, all the examined E. coli isolates (100%) harbored at least three resistance genes. Only two chicken E. coli isolates (2.2%) belonging to phylotypes A and B1 possessed 12 investigated resistance genes, and only two human E. coli isolates (2.2%) belonging to phylotypes B1 and B2 carried 11 investigated resistance genes (Fig. 4B and Supplementary Table S2).

Concerning various isolation sources, the examined tetA, sul1, aadA, blaCMY, blaOXA, blaDHA, and genes were more abundant among isolates obtained from beef burger samples than those obtained from chicken meat and human stool samples. The blaNDM gene was more significantly (p = 0.002) abundant among chicken isolates (40%) than human and cattle ones (15.6 and 0%, respectively), which suggests a possible public health hazard due to the risk of dissemination along the food chain. The tested blaCTX-M and qnrA genes were more significantly (p < 0.05) predominant among human isolates (100% and 65.6%, respectively) than chicken and cattle isolates. The int1 gene was more frequent among cattle isolates (88.9%) than chicken and human ones (87.5 and 84.4%, respectively) (Fig. 3). The majority of chicken and beef isolates carried six and seven investigated resistance genes (22.5% and 33.3%, respectively), while the majority of human isolates harbored seven and eight resistance genes (25% each). The presence of four resistance genes was significantly (p = 0.038) predominant among beef burger isolates (11.1%), indicating a potential public health threat due to the danger of transmission through the food chain (Fig. 4B and Supplementary Table S2).

The blaTEM resistance gene was more significantly (p = 0.017) abundant among isolates belonging to phylotypes A and B1 (100% each) than those belonging to phylotypes B2 and D (96.7% and 77.8%, respectively). The blaCTX-M gene was more significantly (p = 0.0004) abundant among isolates of phylotype B1 (91.7%), followed by those belonging to phylotypes B2, A, and D. The blaDHA, and blaVIM1 genes were more significantly (p = 0.032, and 0.02, respectively) predominant among isolates of phylotype A (76.9, and 46.2%, respectively) than those belonging to phylotypes B2, D, and B1.

Phenotypic and genotypic diversity of the investigated E. coli isolates

Figures 5 and 6 depict the clustering patterns of the investigated variables, along with E. coli isolates from different sources. Integrative analyses across human and animal isolates revealed significant patterns of antimicrobial resistance and virulence. The five variables (Phylotypes, antimicrobial resistance phenotypes and genotypes, integron, and virulence genes) produced six main branches and two clusters. The qnrS gene and tigecycline resistance were gathered in the same cluster, and the blaCTX-M gene and cefotaxime resistance were identical (Fig. 5A). The examined E. coli isolates displayed polyclonality and high diversity as determined by the investigated variables (Figs. 5 and 6). Hierarchical clustering of the 90 investigated isolates produced 12 main branches and one cluster, with 88 isolates belonging to various lineages, and only two isolates obtained from beef burger samples were gathered in the same cluster (Fig. 5B). As depicted in Fig. 6, the PCA (principal component analysis) biplot of the investigated E. coli isolates showed a specific clustering pattern of isolates from different sources.

Fig. 5.

Hierarchical clustering dendrograms viewing the interrelatedness of the examined variables: phylotypes, antimicrobial resistance phenotypes and genotypes, integron 1, and virulence genes (A), and E. coli isolates retrieved from chicken meat (CM), beef burger (BB), and human stool (H) samples on the basis of the investigated variables (B).

Fig. 6.

Principal component analysis (PCA) biplot illustrating the comprehensive dispersion of E. coli isolates from different sources based on the frequency distribution of antimicrobial resistance phenotypes, and genotypes, phylotypes, integron 1, and virulence genes. Each dot represents an isolate, whereas the arrows indicate the correlation of each variable with either dimension 1 or 2.

Correlation analysis between various E. coli phylotypes, antimicrobial resistance phenotypes, integron 1, antimicrobial resistance genes, and virulence gene profiles.

Figure 7 exhibited the correlation analysis among the investigated antimicrobial resistance phenotypes, phylotypes, integron 1 gene, antimicrobial resistance, and virulence genes. The resistance to tigecycline, prevalence of qnrS, and phoA genes were excluded from the correlation analysis as they were identical in all isolates. There were significant weak positive associations between eaeA, stx1 gene (r = 0.25, p = 0.016), and the resistance to levofloxacin (r = 0.21, p-value = 0.047). The stx1 gene was significantly and positively related to the stx2 gene (r = 0.45, p = 0.00001), phylotype B1 (r = 0.34, p = 0.001), and the resistance to amoxicillin/clavulanic acid (r = 0.28, p = 0.007), but it was significantly (p = 0.007) and negatively correlated with phylotype B2 (r = -0.28). The ompA gene was significantly (p < 0.05) and positively related to the blaCTX-M gene, phylotype B2, and the resistance to cefotaxime and ceftazidime (r = 0.23, 0.22, 0.23, and 0.21, respectively), but it was significantly (p = 0.003) and negatively correlated with phylotype A (r = -0.3). Positive significant (p > 0.05) relationships between tsh, blaNDM1 gene, phylotype D, and the resistance to meropenem (r = 0.25, 0.26, and 0.24, respectively), and significant (p > 0.05) negative associations between tsh, blaTEM, blaCTX-M genes and the resistance to cefotaxime, and ceftazidime (r = -0.23, 0.24, -0.24, and − 0.27, respectively) were also recorded. In addition, the int1 gene was significantly (p = 0.012) and positively correlated (r = 0.26) with the resistance to trimethoprim/sulfamethoxazole. Statistically significant and positive associations were recorded among the blaOXA gene and the resistance to amoxicillin/clavulanic acid (r = 0.39, p < 0.001), and the iroN gene r = 0.25, p < 0.05) (Fig. 7).

Fig. 7.

Correlation (r) among antimicrobial resistance phenotypes, and genotypes, integron 1 gene, and virulence genes of E. coli isolates from chicken meat, beef burger, and human stool samples.

There were significant (p > 0.05) positive correlations between trimethoprim/sulfamethoxazole phenotype in animal isolates and ampicillin, amoxicillin/clavulanic acid, levofloxacin, and ciprofloxacin phenotypes, eaeA, stx1, qnrA genes, and phylotype B1 in human isolates (r = 0.41, 0.01, 0.09, 0.22, 0.27, 0.33, and 0.29, respectively). The sul1 gene in animal isolates was significantly (p < 0.05) and positively related to the ampicillin, amoxicillin/clavulanic acid phenotypes, stx1 gene, and phylotype B1 in human isolates (r = 0.35, 0.3, 0.38, and 0.38, respectively), but it was significantly (p > 0.05) and negatively correlated with blaCMY, blaVIM, blaIMP genes and phylotype B2 in human isolates (r =  − 0.3, − 0.42, − 0.29, and − 0.39, respectively). The sul1 gene in human isolates was significantly (p < 0.05) and positively related to the cefotaxime, levofloxacin phenotypes, and blaCTX gene in animal isolates (r = 0.27 each) (Supplementary Fig. S5).

Figure 8 depicts the correlation between E. coli isolates from human stool, chicken meat, and beef burger samples based on the prevalence of antimicrobial resistance phenotypes, resistance, and virulence genes. The pairwise scatter plot matrix showed significant (p < 0.001) strong positive correlations between human E. coli isolates and isolates retrieved from chicken meat and beef burger samples (r = 0.888 and 0.835, respectively), suggesting a possible public health hazard due to the danger of transmitting MDR and virulent strains through the food chain (Fig. 8).

Fig. 8.

Pairs plot showing the correlation between E. coli isolates from human stool, chicken meat, and beef burger samples based on the prevalence of antimicrobial resistance phenotypes, resistance, and virulence genes. ***p < 0.001.

The color key denotes the correlation coefficient (r). The deeper red and blue colors signify more robust positive and negative correlations (details of the r value are displayed in Fig. S4). Stars refers to significant correlations; *p < 0.05, **p < 0.01, ***p < 0.001.

Impact of liposomal cinnamon, oregano, and clove essential oils on multi-virulent, MDR, and ESBL-producing E. coli isolates from meat samples

The antimicrobial and antivirulence efficacies of LCOC at its SIC were determined on 13 multi-virulent, MDR, CR, and ESBL-producing meat E. coli isolates carrying six and five investigated virulence genes. These isolates were obtained from 11 chicken meat and 2 beef burger samples, and they belonged to phylotypes B1 (n = 7), B2 (n = 3), D (n = 2), and A (n = 1). LCOC significantly inhibited the examined E. coli isolates (MIC values ranging from 0.03125–0.25 µg/mL) using the broth microdilution assay (Supplementary Fig. S6).

The antivirulence impact of LCOC (at its SIC) was ascertained via significant (p < 0.0001) downregulation of the investigated virulence-related genes (up to 0.16-fold) of the 13 multi-virulent, MDR, CR, and ESBL-producing meat E. coli isolates (Fig. 9 and Supplementary Fig. S7) via the RT-qPCR technique. Interestingly, LCOC showed the highest significant (p < 0.0001) repression levels for tsh (up to 0.16-fold), followed by ompA (up to 0.18-fold), eaeA (up to 0.27-fold), iroN (up to 0.26-fold), stx2 (up to 0.37-fold), and stx1 (up to 0.39-fold) virulence genes (Figs. 9 and S8). It was evident that treatment of the investigated meat E. coli isolates with the tested LCOC could significantly (p < 0.0001) attenuate the examined virulence genes compared with the untreated isolates (Supplementary Fig. S8).

Fig. 9.

Heatmap of the thirteen investigated multi-virulent, multidrug-resistant, carbapenem-resistant, and extended-spectrum β-lactamases producing Escherichia coli isolates from chicken meat and beef burger samples based on the fold change in eaeA, stx1, stx2, ompA, tsh, and iroN virulence genes expression in response to the SICs of liposomal cinnamon, oregano, and clove EOs (LCOC), demonstrating their antivirulence characteristics. The code numbers for chicken meat (CM) and beef burger (BB) isolates are on the right of the heat map.

For sensory analysis of meat, the average panel scores for the flavor of chicken meat and beef burger samples are presented in Table 4. No statistically significant difference in sensory qualities was seen between LCOC-treated and control untreated groups of meat (p > 0.05), which suggests that LCOC did not alter the sensory properties of meat and that meat had acceptable sensory characteristics.

Table 4.

Mean panel scores for the flavor of meat treated with liposomal cinnamon, oregano, and clove essential oils (LCOC).

Sensory properties	Control	LCOC	p-value
Appearance	4 ± 0.35	3.6 ± 0.35	0.633
Odor	3.5 ± 0.28	3.6 ± 0.26	0.397
Juiciness	3.4 ± 0.34	3.6 ± 0.29	0.811
Tenderness	4.2 ± 0.23	3.8 ± 0.17	0.438
Taste	3.4 ± 0.23	3.6 ± 0.23	0.319
Overall acceptability	3.8 ± 0.17	3.9 ± 0.17	0.681

Results were expressed as mean ± SEM (standard error of the mean).

Sensory properties order; 5 = Best, 1 = Worst. LCOC: liposomal cinnamon, oregano, and clove essential oils.

Discussion

Handling and manufacturing of meat and its by-products in unsanitary conditions can spread MDR pathogens, such as ESBL-producing E. coli, to humans. Therefore, enforcing stringent monitoring protocols and encouraging hygienic practices in processing and distributing meat is imperative to stop the spread of MDR microorganisms⁵⁸. In the current work, we characterized, at both phenotypic and genetic levels, 90 E. coli isolates with an occurrence rate of 28.1% from human stool (45.7%), chicken meat (26.7%), and beef burger (18%) samples. Accordingly, a recent work carried out in Egypt stated that E. coli isolates were more frequent among diarrheic human samples (36.7%), followed by chicken meat and beef burger samples (8 and 2.7%, respectively)⁵⁹. Likewise, previous research in Malawi⁶⁰ and Nigeria⁶¹ had determined E. coli isolates in human stool samples from diarrheic patients (41.8% and 50%, respectively). Similar to this, prior research in Brazil had detected E. coli isolates among 58.7% of the examined chicken meat samples⁶². In the same line, prior work in Bangladesh⁶³ and Egypt⁶⁴ had identified E. coli isolates among beef burger samples with percentages of 20% and 33.3%, respectively. Our data showed that the majority of diarrheic human E. coli isolates belonged to phylotype B2 (65.6%), followed by phylotypes B1 and A. Similarly, a prior investigation in Pakistan indicated that most diarrheic human E. coli isolates belonged to phylotype B2, followed by phylotype A⁶⁵. In general, factors such as conventional and molecular identification procedures, geographic location, climate, and contamination status may account for the variability in the prevalence of E. coli in various studies⁶⁶.

Notably, there are differences in AMR between and within nations, which are closely linked to the medications administered and the differences in their recommendations and administration practices^49,67. In this regard, the high rates of ampicillin and tetracycline resistances that we recorded in our examined isolates (95.6 and 84.4%, respectively) were less than those found in prior investigations conducted in China (98.9 and 97.6%, respectively)⁶⁸ and Bangladesh (98.9 and 85.3%, respectively)². All our examined E. coli isolates showed 100% susceptibility to tigecycline, a last-resort antimicrobial used for treating severe multidrug-resistant infections, highlighting its continued clinical effectiveness; this finding aligns perfectly with a prior report from China⁶⁹. Unsettlingly, a global alert has been issued on the emergence of MDR strains⁴⁸, which was evidenced in the present research, where 97.8% of the tested E. coli isolates were MDR. This finding was partially similar to those of earlier reports carried out in the UK⁷⁰ (100%). The unchecked use of antimicrobials as growth enhancers in livestock and as non-prescription drugs for animals and humans may cause the high resistance rates of E. coli isolates in developing countries²³. Additionally, the high levels of antimicrobial resistance observed in this study are concerning because these drugs are employed in the treatment of human E. coli infections, creating a significant challenge to effective antimicrobial therapy. Hence, it is essential to limit the utilization of antimicrobials in both animals and humans⁷¹, along with the widespread application of substitute medications derived from medicinal plants¹⁶.

It has been shown that the presence of virulence-related genes, like eaeA, stx1, and stx2, is essential for E. coli pathogenicity⁷². Virulotyping of our investigated E. coli isolates stated that all the isolates harbored a minimum of one virulence gene, with iroN being the most prevalent gene (92.2%). Similarly, prior reports in Algeria⁷³ and China⁷⁴ reported that 98% and 81.8% of the tested E. coli isolates harbored at least one virulence gene, and the majority of the isolates were positive for the iroN gene (86.9% and 63%, respectively). Shiga toxin-producing E. coli isolates carrying the stx2 gene have been correlated with more severe illness than those having the stx1 gene, and the eae and stx genes combination has been linked to higher virulence⁷⁵. Herein, 55.6, 33.3, and 14.4% of our investigated E. coli isolates carried eaeA, stx1, and stx2 genes, respectively, which is higher than the outcomes of a prior work in Egypt, where 13.8, 36.9, and 0% of the examined E. coli isolates possessed the respective genes, respectively⁷⁶. The multi-virulent criterion was present among 70% of our examined E. coli isolates, surpassing the outcomes of a prior research in the Czech Republic (32%)⁷⁷. Various study sample sources and geographic locations may have contributed to the differences in the prevalence of target virulence genes¹⁹.

Interestingly, our results revealed high prevalence rates of non-β-lactam resistance genes, including tetA, aadA, and sul1 genes (85.6, 81.1, and 78.9%, respectively) between the analyzed E. coli isolates, surpassing the outcomes of a prior study in Bangladesh (77.2, 34.7, and 45.9%, respectively) ². Of note, none of our investigated E. coli isolates carried the qnrS gene, aligning with findings from prior research in Mexico and Egypt^78,79. All our investigated E. coli isolates carried a minimum of one β-lactams resistance gene, surpassing the prevalence stated in other research in Malaysia (37.5 and 24.4%)^80,81. Concerning the β-lactam resistance genes in the tested E. coli isolates, the most prevalent ESBL-encoding genes were blaTEM, blaCTX-M, blaCMY, and blaOXA genes (96.7, 72.2, 81.1, and 68.9%, respectively). These findings surpass those reported in a previous study in Egypt (52.4, 42.9, 0, and 14.3%, respectively)⁷⁸. In Brazil, Crecencio et al.⁶² stated that blaTEM-1 (73.33%) was the most frequent among E. coli isolates, followed by blaCMY-2 gene (6.7%); while none of the isolates carried blaCTX-M2 or blaOXA-1 genes.

Antimicrobial resistance and unfavorable patient outcomes may result from the large dosages of antimicrobials used to treat ESBL-producing E. coli infections⁷⁶. Previous studies have stated the safe impact of LCOC⁸², liposomal cinnamon EO⁸³, and clove EO nanoemulsion⁸⁴ in an in vivo chicken model. Our outcomes revealed that the investigated LCOC exhibited no significant difference in the sensory properties of meat, including juiciness, tenderness, taste, and overall acceptability, in comparison to control untreated samples, which suggests that LCOC did not negatively impact the sensory characteristics of meat. Accordingly, previous reports depict that oregano, clove, and cinnamon EOs^52,85,86 had acceptable sensory characteristics of meat, along with enhancing the physical and chemical properties of meat when compared with the control group. In the current work, LCOC demonstrated potential in vitro antibacterial efficacy against the analyzed MDR, multi-virulent, and ESBL-producing E. coli isolates. This may be associated with the efficacy of liposomal encapsulation in enhancing the bioactivity and bioavailability of EOs, which promotes cellular absorption and deeper tissue penetration, thereby augmenting the immune response. Accordingly, previous work noted that liposomal clove EO^87,88, cinnamon EO-loaded nanoliposomes^89,90, and liposomal oregano^91,92 had antibacterial properties against E. coli. However, to the best of our knowledge, no research has been conducted on the antibacterial effects of LCOC against ESBL-E. coli isolates. Oregano, clove, and cinnamon EOs have antimicrobial properties because of their lipophilic characteristics, which allow them to accumulate in biological membranes and alter the structure of membrane fatty acids in microbial cells^19,93.

Several EOs at their SICs may lessen the virulence and pathogenicity of microorganisms by changing their virulence gene expressions^19,23,94. Our results showed that LCOC, at its SIC, had a diminishing effect on the pathogenicity of multi-virulent and ESBL-E. coli isolates from chicken meat and beef burger samples, by downregulating the expression levels of their important virulence genes, including eaeA, stx1, stx2, ompA, tsh, and iroN. In accordance with this vital finding, a previous study conducted in Belgium showed that exposure to oregano EO reduced stx2 gene expression in E. coli isolates from human origin⁹⁵. Likewise, an earlier report in the USA indicated that carvacrol downregulated the transcription levels of eaeA, stx1, and stx2 genes in E. coli isolates from beef meat and salami samples⁹⁶. In the same line, previous research showed that exposure to cinnamon EO reduced the expression levels of stx2 and ompA genes in E. coli isolates of human origin⁹⁷. Nonetheless, as far as we know, no studies have been done on the impacts of LCOC on the transcriptional levels of E. coli virulence genes among ESBL-E. coli isolates in chicken meat and beef burger samples. Nevertheless, confirmation through in vivo studies, toxicity and safety assessments, and larger-scale field investigations is required before practical application can be recommended for improving food safety and meat quality.

Conclusions

Our interesting findings demonstrated an alarming prevalence of MDR, multi-virulent, CR, and ESBL-E. coli isolates in chicken meat, beef burger, and human stool samples in Egypt. All the examined isolates possessed a minimum of one virulence gene, with the iroN and ompA genes being the most prevalent ones. All the investigated E. coli isolates possessed a minimum of two β-lactams resistance genes, with blaTEM, blaCTX-M, and blaCMY being the most prevalent ones. Thus, food handlers must regularly examine chicken meat and implement sanitary food safety standards to safeguard consumers. Notably, LCOC exhibited potent antimicrobial and antivirulence activities against multi-virulent, MDR, CR, and ESBL-producing E. coli isolates in chicken meat and beef burger with acceptable sensory characteristics of meat. The present study suggests that LCOC could serve as a potential alternative approach for controlling MDR, virulent, and ESBL-E. coli in meat systems, thus ensuring food safety for the public and providing high-quality meat and meat by-products. Future research should focus on in vivo validation, safety and toxicity assessments, and large-scale field studies to confirm its practical applicability for improving food safety and meat quality.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Marwa I. Abd El-Hamid: Conceptualization, methodology, validation, formal analysis, investigation, resources, data curation, writing – review and editing, visualization, supervision, project administration, funding acquisition. Reham M. ELTarabili: Conceptualization, methodology, validation, formal analysis, investigation, resources, data curation, writing – review and editing, visualization, supervision, project administration, funding acquisition. Ghada A. Ibrahim: methodology, validation, resources. Eman M. Elmehrath: methodology, data curation, visualization. Marwa E. E. Mansour: methodology, data curation, visualization. Elsayyad M. Ahmed: resources, data curation, visualization. Saqer S. Alotaibi: project administration, funding acquisition. Amirah Albaqami: project administration, funding acquisition. Wedad Mawkili: project administration, funding acquisition. Musaad B. Alsahly: project administration, funding acquisition. Thamer A. Alsufayan: project administration, funding acquisition. Sherief M. Abdel-Raheem: project administration, funding acquisition. Waleed R. El-Ghareeb: project administration, funding acquisition. Rania M. S. El-Malt: Conceptualization, methodology, software, validation, formal analysis, investigation, resources, data curation, writing – original draft preparation, writing – review and editing, visualization, supervision, project administration, funding acquisition.

Funding

This work was supported by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia [Grant No: KFU261697].

Data availability

The datasets generated and/or analyzed during the current study are included in the submitted manuscript or supplementary material and are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

All methods were done following the relevant guidelines and regulations. For human and animal samples, the collection protocols were undertaken in compliance with the ARRIVE recommendations and approved by the Suez Canal University Ethics Review Committee (AERC-SCU), Egypt, under reference No. AERC-SCU-VET 2024012, and oral informed consent was obtained from all participants prior to their inclusion in the study.

Consent for publication

It was applicable and available upon demand.