Prevalence of Escherichia coli in Marked Poultry Carcasses in Egypt

| This study evaluated the incidence of E. coli in retail chicken meat in El-SharkiaEgypt, and the effectiveness of zinc peroxide nanoparticles (ZnO2-NPs) against pan drug-resistant bacteria (PDR) isolated from chicken meat. A total of 403 chicken meat cuts-up samples were collected. The presence of E. coli was tested by the growth on selective medium and biochemical identification followed by antibiogram assay. Various virulence factors were phenogenotypically analyzed. The data revealed a 66.3% incidence of E. coli. The incidence in the fresh cuts-up sample was significantly higher than other samples (chilled and frozen). The incidence of multidrug-resistant (MDR) strains was found to be 81%. Genotypic characterization revealed the absence of stx1 from all strains, while the stx2 gene was found in 29% of PDR isolates. Also, ZnO2-NPs showed an efficiency against all PDR E. coli strains.


CollECTion of samplEs
A total of 403 marketed chicken meat samples, including breast, thigh, lung, liver, and gizzard cuts-up, were purchased from different local markets in El-Sharkia, Egypt. The samples were transported to the laboratory in a sterilized bag to avoid cross-contamination. A 25 gram of different parts of each sample was separately grounded in a sterile mechanical blender, mixed with 225 ml of sterile buffered 0.1 % peptone water, and mixed in a stomacher for 2 min (Seward, BA6021, UK) (ICMSF, 1986). One ml of the homogenate was used to prepare tenfold serial dilution (10 -4 -10 -8 ).

baCTEriologiCal ExaminaTion
One ml from each dilution was inoculated in MacConkey broth and incubated at 44°C for 24-48 h. After that, the inoculum was streaked onto eosin methylene blue (EMB) agar plates and incubated for 24 h at 37°C. The selected isolates were subculture on blood agar plates (BAP) to detect the positive isolates for β-hemolysis and these positive isolates were subjected to l-pyrrolidonylβ-naphthylamide (PYR) test based on the identification scheme of York et al. (2000). The selected isolates were subjected to bacterial cell motility, and the biochemical analyses, including Methyl Red/Voges-Proskauer (MR/ VP), hydrogen sulfide production, indole production, oxidase, catalase, urease, citrate, and gelatinase reaction as per the method described previously (Holt et al., 2000). The presumptive E. coli colonies were selected based on phenotypic characterization and subcultured on tryptic soy agar (TSA) medium to obtain pure bacterial cultures.

anTimiCrobial sEnsiTiviTy TEsTing
Drug sensitivity was evaluated by the disc diffusion method. The bacterial culture was inoculated onto Muller-Hinton agar along with 11 antibiotic discs containing; gentamicin (CN), imipenem (IPM), cephalothin (KF), ciprofloxacin (CIP), vancomycin (VA), colistin sulfate (CT), ampicillin (AMP), chloramphenicol (C), trimethoprim/sulfamethoxazole (SXT), tetracycline (TE) and amoxicillin/clavulanic acid (AMC). The results were interpreted as resistant, sensitive, or intermediate depending on the Clinical and Laboratory Standards Institute (CLSI) guidelines (CLSI, 2017) as shown in Table 1. MDR was defined as resistance to at least three or more classes of antimicrobial agents. However, pan drug-resistant (PDR) was defined as non-susceptibility to all agents in all antimicrobial categories El-Shouny et al., 2016). The multiple antibiotic resistance index (MAR) was also calculated as previously reported by . In order to perform the haemolysin test, the presumptive E. coli isolates were inoculated onto sheep blood agar (5%) and incubated overnight at 37 ºC. Haemolysin production was detected by the presence of lysis, a zone of complete lysis around the colony.
The gelatinase production was evaluated using gelatine agar. The plate was inoculated with the bacterial suspension and incubated at 37 ºC for 24 h. The plate was then flooded with a mercuric chloride solution. The halo zone around the colonies was considered to be positive for gelatinase.
For mannose resistant fimbrial haemagglutination (MRHA) test, E. coli growing colonies were inoculated into 5 ml of Muller Hinton broth (MHB) to give a turbid suspension of the standard McFarland's solution (2.4 x 10 9 CFU/ml) then incubated for 5 days at 37ºC. The pellicle on the surface was noted and it was subcultured onto Colonization Factor Antigen agar (CFA) and incubated overnight at 37 °C. The group A positive venous blood was added to equal amounts of the Alsever's solution, this was washed three times and an erythrocyte suspension (3%) was made in phosphate-buffered saline (PBS; pH 7.4). A volume of 40 μl of this erythrocyte suspension was added to 40 μl of PBS and 40μl of D-mannose (3%) in a different well of the same VDRL slide and mixed with the colonies from the CFA agar plates in both the wells. The slides were placed on a VDRL rotator for 4 min. and the haemagglutination reactions were recorded after 2, 5, 10, and 15 min. The haemagglutination was considered to be mannose resistant when it occurred in the presence of D-mannose and it was considered to be mannose sensitive when it was inhibited by D-mannose (Shruthi et al., 2012).
Cell surface hydrophobicity (CSH) was detected using salt aggregation tests. E.coli grown isolates were inoculated into 1 ml of phosphate buffer (pH 6.8) and the turbidity was adjusted to Mcfarland's standard 6 to get a colony count of 5×10 9 CFU/ml. Different molar concentrations of ammonium sulfate solutions ((NH 4 ) 2 SO 4 ; 1M, 1.4 M and 2 M) were prepared. A volume of 40 μl of the bacterial suspension was mixed with an equal volume of the (NH 4 ) 2 SO 4 of different molarities on a VDRL slide. The slides were placed on a VDRL rotator for 4 minutes and the clumps which were formed in different concentrations of the (NH 4 ) 2 SO 4 were observed. The E. coli strains were considered as hydrophobic if they aggregated in the (NH 4 ) 2 SO 4 solution of concentration, <1.4M.

gEnoTypiC CHaraCTErizaTion
Following (Pincus, 2006), serotyping was determined for PDR E. coli isolates using the VITEK ® 2 system (bioMérieux, Marcy l'Etoile, France; AST-GN65 card) at the Laboratory of Mabaret El-Asafra, Alexandria, Egypt based on the manufacturer's guidelines. The virulence factors-encoding genes presence; namely Shiga toxin 1 and 2 (stx1 and stx2) were detected by PCR assay as previously described by El-Shouny et al. (2018) using specific oligonucleotide primers (Table 2). PCR was performed in a reaction mixture (25 μl) containing primers (10 mM), buffer (1X), MgCl 2 (25 mM), dNTPs (10 mM), Taq polymerase (5U) and sterile distilled water in a final volume of 25 μl. PCR amplification conditions were denaturation step at 95ºC for 10 min, 35 amplification cycles at 95 ºC for 45 s, annealing temperature depending on each primer pair of resistance-encoding genes and virulence factors-encoding genes for 30 s and 72 ºC for 45 s with a final elongation step at 72 ºC for 10 min. Amplicons were revealed by gel electrophoresis on a 1.2% agarose stained with ethidium bromide at 100 V for 30 min.

synTHEsis of zno 2 -nps
With a transition temperature (211°C) and size (15-25 nm), a pure phase of ZnO 2 -NPs was obtained as a generous gift from Prof. Dr. Sameh Samir Ali (Biofuel Institute, Jiangsu University, China). ZnO 2 -NPs were synthesized, according to . Briefly, a mix of 10 ml of NH 4 OH and 20 ml of 0.1 M of zinc acetate dihydrate was added to homogenized acetone (70 ml) and glycerol (3 g). Forty ml of H 2 O 2 (40%) was added to the prepared solution with constant stirring at 25°C for 30 min. The precipitate was centrifuged and washed with distilled water.

anTimiCrobial aCTiviTy of zno 2 -nps
Antimicrobial activity of ZnO 2 -NPs was assessed by the preparation of various concentrations (50, 100, 150, and 200 μg/ml) of a stock solution (2 g/ml). Filter paper discs (6 mm) were dipped in the prepared solutions and placed on Muller Hinton agar plates that were previously inoculated with the bacterial isolates. The plates were incubated at 37 °C for 4 days. Distilled water was used as a control for the antibacterial activity. Triplicates were maintained, and the mean values and standard error were estimated for all the inoculated plates, as reported previously CLSI, 2017).

sTaTisTiCal analysis
The bacterial count was transformed using Log 10 before the statistical analysis. The obtained data were analyzed statistically using ANOVA (one-way) and t-test (using NCSS 2019 software. All analyses at p ≤ 0.05 were considered significant.

rESuLtS bioCHEmiCal idEnTifiCaTion of E. Coli
The biochemical identification for Gram-negative bacteria (E. coli) was performed for all isolates. The growth of suspected bacterial isolates on different media showed a good growth rate on MacConkey agar and blood agar. However, the growth was variable on EMB agar. The growth of the bacterial strains on EMB revealed green metallic sheen colonies. The selected bacterial isolates were subjected to biochemical identification which exhibited that the bacterial isolates were motile and showed a negative result when subjected to hydrolysis and H 2 S test. However, the gelatinase test showed variable results. The enzymatic reaction tests showed positive results for

miCrobiologiCal qualiTy assEssmEnT
The obtained data as shown in Table 4, revealed a higher count of Enterobacteriaceae in fresh meat (5.54 log 10 CFU/g) than chilled (3.03 log 10 CFU/g) and frozen (1.72 log 10 CFU/g). The total coliform count of fresh meat (3.81 log 10 CFU/g) was also significantly higher (P ≤ 0.05) than chilled meat (2.12 log 10 CFU/g) and frozen meat (1.23 log 10 CFU/g). Values were measured by log 10 CFU/g and represented as the mean of three replicates (mean±SD).

anTimiCrobial susCEpTibiliTy TEsT
The effectiveness of gentamicin and chloramphenicol was closely related to the maximum antibacterial activity against E. coli. However, showing minimum antibacterial activity, vancomycin, and tetracyclin were closely related.
One percent (3/262) of all isolates was susceptible to all antimicrobial agents, while 81% (212/262) isolates were MDR strains. The resistance rates were maximum for vancomycin (99%) and tetracycline (96%), while it was lowest for gentamicin (7.2%) and chloramphenicol (12.6%). Hence, gentamicin showed efficiency against E. coli isolates. The remaining antimicrobial agents showed varying resistance that ranged from 24-73% as shown in Table 5. and isolates were selected as PDR E. coli strains. The phenotypic characterization of the selected PDR E. coli strains revealed a100% positivity to β-hemolysins, while 86% of the isolates were positive to the cell surface hydrophobicity (CSH) test. However, 57.1% Of PDR strains were positive for MRHA and gelatinase enzymes.
Two E. coli isolates (ECDMGF6 and ECKFGF5) were positive for all of the virulence factors tested. However, stx1 and stx2 were absent in these isolates. The investigation of Shiga toxins producing E. coli strains revealed the presence of stx2 in two strains of E. coli (ECGHBF1 and ECMFLF4) ( Table 6).

anTimiCrobial aCTiviTy of zno 2 -nps
At a concentration of 50 μg/ml of ZnO 2 -NPs, E. coli showed lower inhibition zone that ranged from 0.0-20.0 mm. At 100 and 150 μg/ml, all E.coli strains showed susceptibility to ZnO 2 -NPs. A high dose of ZnO 2 -NPs (200 μg/ml) showed an outstanding susceptibility of all bacterial strains. The minimum inhibitory concentration (MIC 50 ) of E. coli isolates ranged from 6.25-150 μg/ml. However, minimum bactericidal concentration (MBC) ranged from 12.5-300 μg/ml (Table 7). Our data revealed an increased E. coli average count in fresh cuts-up samples (5.9 log 10 CFU/g) than chilled and frozen cuts-up (4.46 -3.49 log 10 CFU/g). We observed a higher count than previous reports. For example, Adu-Gyamfi et al. (2012) documented the microbiological quality of chicken at different retail stores in Accra, Ghana. They found varying counts of E. coli such as 1.27, 2.59, and 2.74 log 10 CFU/g in chicken meat, where all of them were lower than the ones observed in the current study. In another study, Pissol et al. (2013)  Shiga toxins were widely spread in the meat especially in minced meat. The capability of E. coli to induce infection in humans is based on the production of Shiga toxins (Leotta et al., 2008). Shiga toxin-producing E. coli can cause food-borne infection besides the severe fatal illnesses in humans, such as hemorrhagic colitis and hemolytic uremic syndrome which suggested being the major cause of acute renal failure in children (Prendergast et al., 2011 The antimicrobial activity of ZnO 2 -NPs was investigated against PDR E.coli isolates in our study and the obtained results showed an outstanding activity. Also, Ali et al. (2020) reported that the activity of metal oxide nanoparticles (ZnO 2 -NPs, ZnO-NPs, and TiO 2 -NPs) was tested against E. coli isolates and its efficiency estimated by 72.7, 52.7, and 43.6%, respectively. On the other hand, ZnO-NPs activity against S. aureus (Gram-positive bacteria) was demonstrated by Narasimha et al. (2014) and reported an excellent antibacterial effect. The nanoparticles destroy the bacterial cell wall membrane by changing their membrane permeability and stimulating oxidative stress. Reddy et al. (2014) demonstrated that the nanoparticles decline the invasion and internalization by non-phagocytic cells. Hsueh et al. (2015) reported the accumulation of nanoparticles in the cytoplasm or on the outer membranes of the bacteria, causing cell death. Also, Hoseinzadeh et al. (2017) found that the electrostatic attraction occurred when nanoparticles charges are positive, stimulate the accumulation into the negatively charged bacterial cell membrane.
concLuSIon Food containing multidrug-resistant and multivirulence producing E. coli especially poultry meat that could serve as a source of transmitting MDR genes to humans became a serious threat. ZnO 2 -NPs have displayed outstanding antibacterial activity against zoonotic PDR E. coli isolates, and therefore could be promising and potential metal oxide nanoparticles for food safety applications.