Occurrence, Antimicrobial Susceptibility and Phylogroups of Escherichia coli O157:H7 Isolated from Food Outlets in Some Touristic Cities in Egypt

Foodborne illnesses are frequently caused by Escherichia coli ( E. coli ). E. coli O157 is regarded as a potentially harmful cause of gastrointestinal disorders associated with consumption of foods with animal origin. Therefore, this study was conducted to determine the presence of E. coli O157:H7 in food outlets in some touristic cities in Egypt. For this purpose, 648 samples including raw chicken meat, cooked chicken meat, raw beef meat, cooked beef meat, food handlers and equipment swabs were collected from 54 food outlets in some touristic cities in Egypt. E. coli O157 was 1.1% (7/648) and 1.2% (5/432) in all examined samples and food samples respectively. Cooked chicken samples were the most contaminated with E. coli O157:H7 with an overall prevalence of 1.9% (2/108). The highest prevalence of E. coli O157:H7 (8.3%) isolates was recovered from raw chicken and cooked beef meat in Hurghada Governorate followed by Luxor Governorate (6.3%). There is no E. coli O157:H7 isolates were identified in Sharm El Sheikh and Aswan governorates. All E. coli O157:H7 isolates (100%) showed resistance to ampicillin (AMP), cefixime, ciprofloxacin and cotrimoxazole. Multidrug resistance (MDR) was observed among all E. coli O157:H7 isolates. All E. coli O157:H7 isolates harbor the eae gene with complete absence of stx1 gene. The most prevalent phylogroup among the E. coli O157:H7 strains was B2 identified in raw and cooked beef and cooked chicken, collected from Luxor, Hurghada, and Alexandria governorates, respectively. Whereas, D phylogenetic group E. coli O157:H7 was only found in raw chicken sample collected from Hurghada Governorate. In conclusion, the detection of pathogenic MDR E. coli O157:H7 in food samples, food handlers and food equipment in some touristic cities in Egypt poses a serious risk to public health. Therefore, it is recommended to focus on identifying practices which increase the risk of food contamination, and on implementing measures to improve the sanitary conditions in the food outlets in touristic cities.


INTRODUCTION
E scherichia coli O157:H7 is one of the most serious foodborne pathogen strains which causes severe infections and significant fatality in humans (Blanco et al., 2003;Jo et al., 2004). More than 75,000 cases of foodborne illness attributed to E. coli O157:H7 occur annually (Perna et al., 2001).

O n l i n e F i r s t A r t i c l e
that, zoonotic transmission of E. coli 0157:H7 occurs after consumption of raw or under-cooked meat, inadequately pasteurized dairy products, or contact with contaminated fomites containing the Shiga toxin EHEC (Ameer et al., 2018). The contamination of beef may occur during slaughter, and the process of grinding beef may transfer pathogens from the surface of the meat to the interior. Additionally, the organism also could spread from one food item to another by hands, cooking utensils, cutting boards and unclean food preparation surfaces (Pal and Mahendra, 2016). Consequently, failure to implement appropriate food safety management system and applying sanitary conditions during the production process, handling and marketing of food products facilitates the transfer of E. coli O157 to the different food products (Reilly, 1998). It is known that early antimicrobial treatment can avoid Shiga toxin producing E. coli O157:H7 infection progression to the HUS (Schroeder et al., 2002;Amézquita-López et al., 2016;Mühlen and Dersch, 2020). However, studies have shown a significant increase in antimicrobial resistance in E. coli O157:H7 (Mühlen and Dersch, 2020). This in part may be related to the overuse and misuse of antibiotics by the people and food producing animals (Radostits et al., 2000). Moreover, the development of antibiotic resistance in E.coli O157: H7 is considered main challenge as it can spread the resistance determinants within the other commensals and pathogens (Ahmad et al., 2021). Clermont et al. (2000) earlier created a categorization method based on phylogenetic characterization of E. coli strains for monitoring the microbiological source, determining phylogenetic groups, and determining possible pathogenicity among E. coli strains. They revealed that E. coli isolates are divided into four main groups: A, B1, B2, and D, with seven subgroups: A0, A1, A2, B22, B23, D1 and D2. Clermont et al. (2013) then proposed a new phylogenetic grouping technique that comprised four new phylogroups: C, E, F, and Escherichia cryptic clade I.
There is a strong link between the virulence and phylogeny in E. coli infections (Pakbin et al., 2021), as phylogroup B2 E. coli strains, and to a lesser extent phylogroup D, are the most common causes of extraintestinal infections in humans (Bailey et al., 2010). Also, the strains belonging to the phylogroup A are typically commensal (Picard et al., 1999).
In Egypt, for many years, tourism has been the main source of economy, but it is now threatened due to food borne illness (Abdelhakim et al., 2020). Therefore, Egypt's main challenge is to ensure that it has the capacity to provide safe food for its own people. Several studies have been carried out in Egypt to investigate the incidence of E. coli O157:H7 within different food products, but, there are no sufficient reports about the food contamination with E. coli O157:H7 in touristic cities in Egypt. Therefore, within some of touristic cities in Egypt, the present study was conducted to investigate the occurrence, antimicrobial susceptibility, virulence genes and the phylogroup of E. coli O157:H7 isolated from food outlets.

Sampling
A grand total of 648 samples were collected from 54 food outlets in Egypt, including raw chicken meat, cooked chicken meat, raw beef meat, cooked beef meat, food handlers' hand swabs and equipment swabs (108 of each). All food samples were received in sterile bags, whereas hand and equipment swabs were placed in 5 ml liquid maximum recovery diluent (MRD) in a sterile screwcapped container (TS/5-31-UK). All samples were carried in ice box to be transferred with a minimum delay to the laboratory for bacteriological examination.

Isolation and identification of E. coli O157:H7
The isolation of E. coli were done according to procedure using enrichment methods, and then confirmed by PCR. Briefly, 25 g of the meat samples (beef and chicken) were transferred to a septic blender jar and 225 ml of 0.1% sterile peptone water was added aseptically (ISO, 2017). After that, each sample was homogenized in the stomacher for 1-2 min at 2000 rpm to produce a homogenate. Phenotype characterization of O157 strains was done using Sorbitol MacConkey Agar (Oxoid, England), 0.1 ml of the prepared samples, as well as hand and equipment swabs were incubated for 24 h at 35-37°C. The suspected colonies were sub-cultured and identified as E. coli through Gram's stain films and biochemical tests. Then, E. coli isolates were serotyped in the Serology Unit Animal Health Research Institute, Dokki, Giza Egypt, using commercial antisera anti-E. coli O157 (SIFIN) according to the manufacturer's instructions.

Antimicrobial susceptibility testing of E. coli O157:H7
Antimicrobial susceptibility of E. coli O157 isolates was determined by the disc diffusion method, according to the guidelines for the Clinical and Laboratory Standards Institute (CLSI, 2012) on trypticase soy agar (TSA) using commercially available discs. Zones of growth inhibition surrounding each antibiotic disc are measured to the closest millimeter after plates are incubated at 37°C for 16-24 h. The isolate's susceptibility and the speed at which the drug diffuses through the agar medium are both correlated with the zone's diameter. The zone diameters of each drug are interpreted using the criteria published by CLSI (2012).

Molecular detection of E. coli O157:H7
The EHEC O157 virulence genes stx1, stx2 and eae were assessed by PCR. Descriptions of the targeted genes and primer sequences are listed in Table I. QIAamp DNA mini Kit (catalogue no.51304) was used for extraction of DNA from the recovered strains of E. coli O157:H7, the PCR master mix was prepared using Emerald Amp GTPCR master mix (Takara, code No. RR310A).

Phylogenetic group of E. coli strains determination
According to Clermont et al. (2000) EPEC strains were divided into four main phylogenetic groups (A, B1, B2, and D) based on PCR detection of the chuA and yjaA genes and DNA fragment TSPE4.C2. Briefly, the primer pairs for chuA, yjaA and TspE4C2.1 (Table I), were added to the standard PCR mixture, PCR was performed under the following conditions: denaturation for 4 min at 94°C, 30 cycles of 5 s at 94°C and 10 s at 59°C, and a final extension step of 5 min at 72°C. Depending on whether a strain reacted positively or negatively with yjaA primers, group B2 or D was assigned to the strains that interacted with the chuA primers. Similar to this, the chuA-negative isolates were divided into groups B1 or A depending on whether the PCR for TspE4.C2 produced a positive or negative response, respectively.

Occurrence of E. coli O157:H7
A total of 7 (1.1%) E. coli O157:H7 strains were isolated and confirmed from 648 samples collected from food outlets located in some touristic governorates in Egypt (Table II). The occurrence of E. coli O157:H7 in food samples was 1.2% (5 out of 432) including cooked and raw chicken and beef meat, while the other 2 E. coli O157:H7 isolates were recovered from both of cutting knife and food handler hand swab from food outlet in Cairo Governorate. Among these samples examined, cooked chicken samples were the most contaminated with E. coli O157:H7 with an overall prevalence of 1.9% (2/108). With regard to the source of the samples, the highest prevalence of E. coli O157:H7 (8.3%) isolates was recovered from raw chicken and cooked beef meat in Hurghada Governorate followed by Luxor Governorate (6.3%); however, the occurrence of E. coli O157:H7 was 3.3 and 5.0 in Cairo and Alexandria Governorates, respectively. There is no E. coli O157:H7 isolate identified in Sharm El Sheikh and Aswan Governorates (Table II).   Antimicrobial susceptibility of E. coli O157:H7 isolates (n=7)

O n l i n e F i r s t A r t i c l e
The antimicrobial susceptibility investigation of 7 E. coli O157:H7 isolates against four different antibiotic classes and 12 commercially available antimicrobial discs revealed that all E. coli O157:H7 isolates (100%) showed resistance to ampicillin (AMP), cefixime, ciprofloxacin and cotrimoxazole. An overall resistance of 85.7% and 57.1% was recorded to pipracillin-Tazopactam, piperacillin and gentamicin. However, the lowest resistance was observed against aztreonam (14.3%). Furthermore, the isolates showed high susceptibility to Amikacin (85.7%) and Meropenem (71.4%) (Table III). Multidrug resistance (MDR) was observed among all E. coli O157:H7 isolates as 3 isolates showed resistance to three antimicrobial classes and 4 isolates evidenced resistance to four antimicrobial classes (Table IV). It can be also shown in Table IV that there were diverse patterns of antibiotic resistance among the isolates from each source.

E. coli O157:H7 virulence genes
Molecular identification of virulence genes revealed that all E. coli O157:H7 isolates harbor the eae gene with complete absence of stx1 gene. Whereas, stx2 was harbored by only one isolate obtained from cooked beef in combination with eae gene, in Hurghada Governorate ( Table V).

Phylogroup of E. coli O157:H7 isolates
The chuA gene was found in four strains from groups belonging to B2 and D, but not found in three strains belonging to group A, as a result of this, we were able to distinguish groups B2 and D from groups. Similarly, the yjaA gene allowed for complete discrimination between group B2 (42.9 % of the strains were positive) and group D (14.2 % of the strains were negative). Finally, clone TSPE4.C2 was found in four strains, three of which are B2 strains and one of which is a group D strain, whereas it was absent from all group A strains (Table V). In Cairo, phylogenetic group A was predominant among the three E. coli O157:H7 isolates recovered from different sources including; cooked chicken, cutting knife and hand swab. However, phylogenetic group B2 was the most prevalent phylogroup among the E. coli O157:H7 strains isolated from meat samples, including raw and cooked beef and cooked chicken, collected from Luxor, Hurghada, and Alexandria governorates, respectively. Whereas, D phylogenetic group E. coli O157:H7 was only found in raw chicken sample collected from Hurghada Governorate (Table V).

O n l i n e F i r s t A r t i c l e
Occurrence, Antimicrobial Susceptibility and Phylogroups of E. coli O157:H7

DISCUSSION
Several researches have suggested that animalderived foods could be a significant source of humanacquired MDR pathogenic E. coli (Rashid et al., 2013). Although various studies had been carried out in Egypt to investigate the incidence of E. coli O157:H7 within different food products (El-Alfy et al., 2013;Ahmed and Shimamoto, 2014;Khalil et al., 2015), information about the food contamination with E. coli O157:H7 in Egyptian touristic cities is scarce. Therefore, the aims of this study were to determine incidence rate, genotypes, phylogroups and antimicrobial susceptibility patterns in E. coli O157:H7 strains isolated from food products, as beef, chicken meat, and other sources, including food handlers and food equipment collected from some Egyptian touristic cities.
In the present study, 648 random samples of meat, food equipment and food handlers obtained from 54 food outlets in some touristic cities in Egypt were investigated for the presence of E. coli O157:H7. The total prevalence of E. coli O157:H7 was 1.1% (7 out of 648 samples), whereas, this prevalence in meat samples was 1.2% (5 out of 432), including cooked and raw chicken and beef samples. This result was in line with previous studies, which reported that the incidence of E. coli O157:H7 in UK was 1.1% of 2075 samples (Chapman et al., 2000) and 1.1% of 571 meat samples in the Netherlands (Heuvelink et al., 1999). In contrast, our finding as higher than that reported in minced beef samples in Antakya region (1.3%), in southern Turkey (Durmaz et al., 2007) as well as in Egypt (0.5%) (Hamed et al., 2017).
Regarding the geographical area, a higher incidence of E. coli O157:H7 was recorded in meat samples collected from food outlets in Hurghada (2/24, 8.3 %) followed by Luxor (1/32, 3.1%). Hurghada considered as one of the most popular resorts on the red sea coast that attracts tourist from all over the world. Therefore, unfortunately, such incidences of food-borne pathogens might negatively influence the tourism and hospitality industry in Egypt (Abdelhakim et al., 2020).
From the obtained results, it can be also noticed that in Cairo Governorate three isolates of E. coli O157 were O n l i n e

F i r s t A r t i c l e
T. Refaay et al.
recovered from cooked chicken, cutting knife (3.3%) and also from food handler's hand swabs (3.3%). However, none of the surface swabs from the cutting boards were positive. In a similar kind of study conducted in Ethiopia, E. coli O157:H7 was isolated from 3.6% (4/110) of the surface swabs of wooden cutting boards with complete absence in cutting knives and hand swabs (Beyi et al., 2017). However, in Pakistan, E. coli O157:H7 was not detected in surface swabs of cutting knives and wooden boards taken from 30 individual retail meat outlet markets (Ali et al., 2010). In the current study, in Cairo Governorate, E. coli O157 positive cooked chicken sample, hand swab and cutting knife swab were collected in the same visit from food outlet, indicating the possible contamination of the chicken meat from cutting knife and/or food handler or vice versa. Additionally, the presence of E. coli O157:H7 in asymptomatic food handler's hand swab may pose a significant public health risk as it increases the possibility of the transmission of this pathogen to tourists when the food handlers un-hygienically handle foods (Oundo et al., 2008). Therefore, food handlers must be trained effectively on food safety and hygiene. All E. coli O157 isolates show resistance to ampicillin (AMP), cefixime, ciprofloxacin and cotrimoxazole irrespective to their origin. High resistance was also found against pipracillin-tazopactam (85.7%), piperacillin and gentamicin (57.1% for each). Furthermore, the isolates showed high susceptibility to amikacin (85.7%) and meropenem (71.4%). Similar findings were reported by (Bhowmik et al., 2022), who showed that 100% (n=20) of their E. coli isolates exhibited resistance to ampicillin, and 41% against gentamicin. In contrast, they observed that their E. coli isolates highly sensitive to cotrimoxazole (83%) and ciprofloxacin (58%) and highly resistant to amikacin (66%). Additionally, previous studies in Egypt (Sobhy et al., 2020;Elmonir et al., 2021), Ethiopia (Haile et al., 2022) and Nigeria (Ojo et al., 2010) revealed high resistance among E. coli isolates to ampicillin, a finding similar to ours. Inadequate antimicrobial selection and abuse can lead to resistance in different bacteria and make it more difficult to treat bacterial infections (Kolář et al., 2001). Antimicrobial-resistant bacteria are one of the most serious public health issues, and are predicted to cause the death of 10 million people annually by 2050 (De Kraker et al., 2016).
Alarmingly, all the tested E. coli O157:H7 isolates (100%) expressed resistance to at least three different classes of antibiotics and were considered as MDR strains. Our finding was similar to that reported in China (100%) (Yu et al., 2020), and higher than those reported in Egypt (51.42%) (Elmonir et al., 2021), Iran (70.8%) (Pakbin et al., 2021) and Ethiopia (57.14%) (Haile et al., 2022). This result suggested high risk of transmission of MDR E. coli O157:H7 to consumers, including tourists, via food served in food outlets in touristic cities in Egypt. Therefore, MDR E. coli has been documented as one of the most significant challenges in food safety (Rashid et al., 2013). Furthermore, the transmission of MDR bacteria via the consumption of meat have been proposed as a potential source in Africa (Eibach et al., 2018).
The eae (encoding intimin) and stx (encoding Shiga toxin) harbored in foodborne pathogenic E. coli O157:H7 strains are central to the pathogenesis of HUS (Paton and Paton, 1998). Additionally, Shiga toxin produced by E. coli O157:H7 can enhance the adherence to epithelial cells and colonization in mice intestines (Robinson et al., 2006). In this study, molecular identification of virulence genes revealed that 100% of E. coli O157:H7 isolates harbor the eae gene with complete absence of stx1 gene. However, stx2 was harbored by only one isolate obtained from cooked beef in combination with eae gene. This finding was in agreement with Dambrosio et al. (2007), who stated that none of the meat STEC isolates harbored stx1 or stx2 genes and in contrast with Hessain et al. (2015), who reported that 45.45% of E. coli O157:H7 isolates recovered from meat samples harbored stx1 and stx2; while stx1 was present in only one isolate. Interestingly, it is believed that stx-negative E. coli O157:H7 strains that do not produce Shiga toxin may cause symptoms, such as diarrhea, they are not generally associated with HUS, even though they still carry virulence factors, such as eae and bfpA genes (Black et al., 2010;Ochoa and Contreras, 2011;Ferdous et al., 2015) and categorized as EPEC (Bentancor et al., 2010). It's interesting to note that the animal aEPEC serogroups O26, O103, O119, O128, O142, and O157 have been linked to human diarrhoea. Additionally, aEPEC has been linked to human infections through consumables such raw meats, pasteurized milk, meat samples, vegetables, and water (Kolenda et al., 2015). Therefore, further study is needed to examine whether our isolates carry more virulence genes rather than stx1, stx2 and eae.
To fully understand E. coli populations, the linkages between strains, their hosts, and disease, and the proven correlation between phylogenetic group and virulence, phylogenetic studies are crucial. Therefore, phylogroup PCR was conducted targeting different marker genes of 7 tested E. coli isolates. It was found that phylogroup A and B2 were the dominating groups (42.9% for each) followed by phylogroup D (14.2%). Interestingly, phylogenetic group B2 was the most prevalent phylogroup among the E. coli O157:H7 strains isolated from meat samples, including raw and cooked beef and cooked chicken, collected from Luxor, Hurghada, and Alexandria, respectively. However, D phylogenetic group E. coli O157:H7 was only O n l i n e

F i r s t A r t i c l e
Occurrence, Antimicrobial Susceptibility and Phylogroups of E. coli O157:H7 found in raw chicken sample collected from Hurghada. Similarly, a study conducted in India showed that the majority of E. coli strains obtained from different food samples belonged to phylogroup B2 (44%) followed by phylogroup B1 (29%), A (16%), and D (3%) (Godambe et al., 2017). In Iran, another study also conducted by Pakbin et al. (2021) revealed that the phylogenetic group A was the most prevalent (46%) among the E. coli isolates and phylogroup D was the least common. It is worth to mention that Phylogroups B2 and D include only virulent E. coli strains (Carlos et al., 2010). However, phylogroup A characterizes commensal E. coli strains (Picard et al., 1999), while E. coli isolates belonging to the phylogroups B2 most often contribute to extra-intestinal diseases, some strains included in other phylogroups (A and B1) have been identified as causes of diarrheal diseases in humans (Bailey et al., 2010). Additionally, in Cairo Governorate the 3 E. coli O157:H7 isolates recovered from different sources (cooked chicken, cutting knife and hand swab) belonged to the same phylogenetic group, which is the commensal phylogroup A. This finding might prove also the scenario mentioned above about the possibility of the cross contamination as well as these commensal strains may have gained virulence genes and turned pathogenic. However, further investigation is required to confirm our observation.

CONCLUSION
In conclusion, the detection of pathogenic MDR E. coli O157:H7 in food samples, food handler and food equipment in some touristic cities in Egypt poses a great public health problem. Therefore, it is recommended to focus on identifying practices which increase the risk of food contamination, and on implementing measures to improve the sanitary conditions in the food outlets in touristic cities. Further studies are required in comparative genomic analysis including genome sequencing, to recognize the epidemiological sources of E. coli O157:H7 and its points of contamination and to define appropriate risk mitigation strategies.

ACKNOWLEDGMENT
We appreciate Zoonoses Department, Faculty of Veterinary Medicine, Cairo University, Egypt for helping us carry out this research.

IRB approval
All food handlers provided oral consent after being told about the usage of hand swab samples. Ethical clearance to use respondents was obtained from the authorized health facility (National Research Centre, Giza, Egypt). The study was conducted in accordance with the ARRIVE recommendations.

Ethical statement
Samples collection protocol was carried out in compliance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Faculty of Veterinary Medicine, Cairo University, Egypt (VetCU-01102020212).

Statement conflict of interest
The authors have declared no conflict of interest.