Shiga (Vero)-toxin producing Escherichia coli isolated from the hospital foods; virulence factors, o-serogroups and antimicrobial resistance properties

According to the presence of the weak, diabetic and immunosuppressive patients in hospitals, hospital foods should have a high quality and safety. Cooking a lot of foods higher than daily requirement, storage of cooked foods in an inappropriate condition and presence of nurses and servants in distribution of food to patients are the main reasons caused contamination of hospital foods. Shiga toxigenic Escherichia coli is one of the common cause of food poisoning in hospitals. The present research was carried out to study the distribution of virulence factors, O-serogroups and antibiotic resistance properties in STEC strains recovered from Iranian hospital food samples. Five-hundred and eighty raw and cooked food samples were collected and immediately transferred to the laboratory. E. coli-positive strains were subjected to PCR and disk diffusion method. Thirty-nine out of 580 (6.72%) hospital food samples were contaminated with E. coli. Raw (20%) and cooked meat (6%) were the most commonly contaminated samples. Raw samples had the higher prevalence of E. coli (P <0.01). Samples which were collected in the summer season had the highest prevalence of bacteria (64.10%). Significant difference was seen between the prevalence of EHEC and AEEC subtypes (P <0.01). The most commonly detected virulence factors in both EHEC and AEEC subtypes were stx1 and eae. The most commonly detected serogroups were O26 (43.75%) and O157 (25%) and there were no positive results for O103, O145, O91, O113 and O128 serogroups. Aac (3)-IV (100%), CITM (100%) and tetA (62.50%) were the most commonly detected antibiotic resistance genes. STEC strains harbored the highest levels of resistance against ampicillin (93.75%), gentamycin (93.75%), tetracycline (87.50%) and ciprofloxacin (81.25%). All of the STEC strains were resistant to at least 3 antibiotics, while the prevalence of resistance against more than 12 antibiotics were 12.50%. High presence of O157 serogroups, EHEC strains and animal-based antibiotics in cooked foods showed insufficiency of cooking time and temperature in the kitchens of hospitals. Judicious prescription of antibiotics and attentions to the principles of food safety can reduce the risk of resistant and virulent strains of STEC in hospital foods.


Background
Access to adequate amounts of safe and healthy foods is key to supporting life and promoting good health of people. Insecure food containing pathogenic bacteria, viruses, parasites or harmful chemical elements, causes more than two-hundred diseases and disordersranging from diarrhea to cancers [1,2]. Foodborne diseases cause approximately 76 million illnesses, 325,000 hospitalizations, and 5,000 deaths in the United States each year [2][3][4]. Observation of basic principles of food hygiene can control and eliminate the distribution of foodborne diseases. Attention to the principles of food hygiene is more urgent especially in public places such as hospitals, restaurants, kindergartens, prisons, barracks, elderly care centers and asylums.
Using from regular, nutritious, safe and accurate food diet based on the characters of patients is one of the main aspects of treatment in hospitals. Due to the general weakness and suppression of the immune system of majority of patients in hospitals, hospitals foods should have good sanitary quality. Among all pathogenic agents causing food-borne diseases and food poisoning in hospitals, Escherichia coli (E. coli) had a had a significant importance [5][6][7][8].
E. coli is a gram-negative, non-sporulating, flagellated, rod-shaped and facultative anaerobic bacterium which belongs to Enterobacteriaceae family. Shiga (vero) toxin (Stx)-producing E. coli (STEC) is a subdivision of an important pathogenic group of this bacterium named enterohemorrhagic E. coli (EHEC) [9][10][11]. STEC strains are responsible for intensive clinical syndromes like lethal hemolytic uremic syndrome (HUS), bloody and non-bloody diarrhea, thrombotic thrombocytopenic purpura (TTP) and hemorrhagic colitis (HC) [9][10][11]. E. coli strains that cause HC and HUS in humans, express high levels of Shiga toxins (stx), cause attaching-effacing (A/E) lesions in intestinal epithelial cells, and possess a specific 60-MDa EHEC plasmid are known as EHEC [9][10][11]. One feature EHEC and Enteropathogenic E. coli (EPEC) have in common is the causation of intestinal epithelial lesions known as attaching and effacing (A/E). Attaching-effacing E coli (AEEC) is a designation for those E. coli strains known to cause A/E lesions or at least carry the genes for this trait, and therefore include organisms that fall into either the EHEC or EPEC classes [9][10][11].
According to the uncertain role of STEC strains in hospital foods and lack of epidemiological investigations in this field in Iran, the present research was carried out to study the distribution of virulence factors, Oserogroups, antibiotic resistance genes and antibiotic resistance pattern of STEC strains isolated from various types of raw and cooked hospital food samples.

Samples and E. coli isolation
From September 2013 to September 2014, a total of 580 various types of raw and cooked hospital foods including raw meat (n = 60), raw chicken (n = 60) and raw fish (n = 70) and cooked meat (n = 100), cooked chicken (n = 100), cooked fish (n = 110), and soup (n = 80) were randomly collected and immediately transferred to the Food Hygiene Research Center of the Islamic Azad University of Shahrekord in cooler with icepacks. Samples were collected from the various hospitals in Isfahan province, Iran. All food samples showed normal physical characters including odor, color and consolidation. All samples were collected through 4 different seasons of the year including summer (n = 120), autumn (n = 150), winter (n = 160) and spring (n = 150).
Totally, 10-g of crushed food samples were homogenized for 2 min in 90 ml of Peptone Water (PW, Merck, Germany). Then the samples were cultured on 5% sheep blood and MacConkey agar (Merck, Germany) and incubated for 18 to 24 h at 37°C. Colonies with the typical color and appearance of E. coli were picked and streaked again on blood agar plates and re-streaked on EMB agar (Merck, Germany). All plates were further incubated for 24 h at 37°C. The green metallic sheen colonies were considered as E. coli. The presumptive colonies were biochemically tested for growth on triple sugar iron agar (TSI) and lysine iron agar (LIA), oxidative/fermentative degradation of glucose, citrate utilization, urease production, indol fermentation, tryptophan degradation, glucose degradation (methyl red test) and motility.
PCR amplification of virulence factors, O-serogroups and antibiotic resistance genes Table 1 shows the list of primers as well as program and condition of each reaction used for detection of Oserogroups, virulence genes and antimicrobial resistant genes [11,20]. Programmable DNA thermo-cycler (Eppendorf Flexrcycler 2 , Germany) was used in all PCR reactions. The PCR amplification products (15 μl) were subjected to electrophoresis in a 1.5% agarose gel in 1× TBE buffer at 80 V for 30 min, stained with SYBR Green (Fermentas, Germany). All runs included a negative DNA control consisting of PCR grade water and strains of E. coli O157:K88ac:H19, CAPM 5933 and E. coli O159:H20, CAPM 6006 were used as positive controls.

Statistical analysis
Statistical analysis was performed using SPSS/16.0 software for significant relationships. The incidences of serogroups, virulence factors, O-serogroups and antibiotics resistance properties of E. coli isolated from various types of hospital food samples were statistically analyzed. Statistical significance was regarded at a P value < 0.05. Table 2 represents the total distribution of E. coli in various types of hospital food samples. Thirty-nine out of 580 (6.72%) hospital food samples were contaminated with E. coli strains. Raw meat (20%), raw chicken (16.66%) and cooked meat (6%) had the highest prevalence of E. coli. Statistically significant difference was seen for the prevalence of E. coli between raw and cooked hospital food samples (P <0.01). Figure 1 shows the total distribution of E. coli strains in various seasons of the year based on the total numbers of 39 E. coli isolated. We found that hospital food samples of summer season had the highest prevalence of E. coli (25/39: 64.10%), while those of winter had the lowest (3/39: 7.69%). In the other hand, 25 out of 120 samples collected in summer (20.83%), 5 out of 150 samples collected in autumn (3.33%), 3 out of 160 samples collected in winter (1.87%) and 6 out of 150 samples collected in spring (4%) were positive for E. coli. Statistically significant difference was seen for the prevalence of E. coli between warm and cold season of the year (P <0.05). Table 3 represents the distribution of virulence factors in E. coli subtypes isolated from hospital food samples. Prevalence of EHEC and AEEC subtypes in the raw meat, raw chicken, cooked meat, cooked chicken, cooked fish and soup samples were 16.66% and 50%, 20% and 60%, 25% and 50%, 0% and 33.33%,0% and 50% and finally 33.33% and 66.66%, respectively. All of the     Table 4 shows the total distribution of O-serogroups in the STEC strains isolated from various types of hospital food samples. We found that O26 (43.75%) and O157 (25%) were the most commonly detected serogroups in the STEC strains of hospital food samples. There were no positive results for the O103, O145, O91, O113 and O128 serogroups. Statistically significant differences were seen for the prevalence of STEC O-serogroups between various types of samples (P <0.05). Totally, raw chicken and soup samples harbored the most variable types of Oserogroups.

Results
Total distribution of antibiotic resistance genes in the STEC strains isolated from various types of hospital food samples is shown in Table 5. We found that aac (3)-IV (100%), CITM (100%), tetA (62.50%), dfrA1 (56.25%) and sul1 (56.25%) were the most commonly detected antibiotic resistance genes. The less commonly detected antibiotic resistance genes were cmlA (12.50%), cat1 (25%) and tetB (31.25%). Statistically significant differences were seen for the prevalence of antibiotic resistance genes between various types of samples (P <0.05). Table 6 indicates the antibiotic resistance pattern of the STEC strains of various types of hospital food samples. STEC strains of our investigation harbored the highest levels of resistance against ampicillin (93.75%), gentamycin (93.75%), tetracycline (87.50%), ciprofloxacin (81.25%) and amikacin (75%). Resistance of STEC strains of cooked samples against human-based antibiotics was entirely higher than animal-based antibiotics. Statistically significant differences were seen for the prevalence of antibiotic resistance between various types of samples (P <0.05). Figure 2 represents the prevalence of multi-drug resistance in the STEC strains of hospital food samples. We found that all of the STEC strains of our investigation were resistant to at least 3 antibiotics, while prevalence of resistance against ten, eleven, twelve and more than twelve antibiotics were 37.50%, 25%, 18.75% and 12.50%, respectively.

Discussion
As far as we know the present investigation is the first prevalence report of the isolation of E. coli strains from raw and cocked hospital foods as well as molecular characterization and antimicrobial resistance properties of their STEC strains in the world. We found that 6.72% of hospital food samples were contaminated with E. coli strains and the prevalence of STEC strains were 41.02% which was considerable high. Gram-negative bacteria like E. coli are responsible for 30-70% of hospitalization in developed countries and E. coli was the most common etiologic gram-negative organism in the hospital environment [22,23]. There were some probable reasons for the high prevalence of E. coli and also STEC strains in the hospital food samples. At first. high-volume food production and long process of catering caused several problems such as lack of adequate precision and accuracy in the preparation and washing of raw materials and even their well cooking, lack of enough time and even temperature for cooking of raw materials, cooling of foods during processing, lack of reaching of sufficient heat to the center of meat and other food materials, lack of enough time to withdraw meat from the frozen state and finally cooking of meat and its products more than the daily requirement and then their storage at improper temperature and conditions. These mentioned circumstances maybe lead to survival and even growth of pathogenic microorganisms in food stuffs. At second, using from unsanitary and also contaminated equipment and dishes for production of Presence of powerful competitor microflora in raw fish and also their destruction and addition of E. coli strains derived from contaminated staffs of the hospital's kitchen are two main reasons for above findings. Because the large numbers of E. coli isolates recovered from raw meat, proper preparation of the raw meat can eliminate the distribution of bacteria. Unfortunately, basic principles of meat inspections were not observed in Iranian slaughterhouses. Therefore, close contact of animal carcasses with each other and even slaughterhouse floor, blood, content of the digestive tract and wool and skin of slaughtered animal caused transmission and distribution of pathogenic agents like E. coli to meat of slaughtered animals. Besides, the role of possible colonizers such as meat inspectors, butchers and miscellaneous people which mainly have come into the slaughterhouse for buying of meat and finally animals like rats, cats and birds which have been entered from outside the slaughterhouse as a sources of pathogenic E. coli should not be overlooked. Survival of STEC strains of raw food samples even after cooking procedure and occurrence of cross contamination after cooking procedure are two important routes of hospital foods contamination.
We found marked seasonality in distribution of E. coli strains in hospital food samples. High incidence of E. coli strains in summer season (64.10%) could be related to the low levels of individual hygiene in this season. The higher prevalence of STEC strains may be related to higher growth of them in hot seasons of the year. Of studies that have been conducted in this field [24][25][26], all have shown a seasonal distribution for E. coli with a higher prevalence of strains in warmer months of the year [24][25][26].
The results of the current study showed that resistant and virulence STEC strains had the high prevalence in the hospital food samples. Limit studies have been conducted in this field. Ifeadike et al., [27] revealed that the prevalence of E. coli strains in food handlers of Nigerian hospitals were 1.8% which was lower than our findings in hospital foods. They concluded that poor and faulty food-handling practices have been identified as the leading cause of the majority of food-borne disease in hospitals. Ha et al., [7] in a research which was conducted in order to study the contamination rate of raw foods in schools, factories, and hospitals of Vietnam reported that the prevalence of E. coli in raw poultry, meat, fish and vegetable samples were 45, 21.3, 6.6 and 18.5%, respectively which all were higher than our results. STEC strains of our investigation harbored resistance against multiple antibiotics. The results of disk diffusion  (4)   method obtained from STEC strains of our investigation were also confirmed by the presence of specific antibiotic resistance genes encode resistance against determined antibiotics. Stewardson et al., [5] reported that 92% of all food samples which were collected from hospitals in Switzerland were positive for Extended-Spectrum-Beta-lactamase-producing-Enterobacteriaceae (ESBL-PE). ESBL-producing E. coli was the most commonly detected (44.77%). They showed 6.45% of eligible food handlers were positive for ESBL-PE and prevalence of E. coli strains among ESBL-PE strains were 83.33%. ESBL-PE strains were generally not multidrug resistant, with 100, 90, 87, 79, and 98% of strains susceptible to meropenem, gentamicin, ciprofloxacin, cotrimoxazole, and fosfomycin, respectively. Antibiotic resistance-based finding of Stewardson et al., [5] study was in contrast with our results which showed the high prevalence of resistance and also antibiotic resistance genes. This part of our study was similar with those of India [28] (high prevalence of resistance against erythromycin, cephalothin, amikacin, kanamycin and gentamicin antibiotics), South Africa [29] (high presence of CITM, blaSHV and tetA antibiotic resistance genes), Korea [30] (high prevalence of resistance against ampicillin, tetracycline, streptomycin and amikacin antibiotics) and Mexico [31] (high presence of tet, blaSHV, qnr and aac (3)-IV antibiotic resistance genes and also high prevalence of resistance against ampicillin, trimethoprim-sulfamethoxazole, chloramphenicol and cephalotine antibiotics). Momtaz et al., [20] reported that aac (3)-IV (68.03%), sul1 (82.78%), blaSHV (56.55%), aadA1 (60.65%) and tetA (51.63%) and also resistance against tetracycline (86.88%), penicillin (100%), gentamycin (62.29%) and streptomycin (54.91%) were the most commonly reported antibiotic resistance-based finding of STEC strains of diarrheic patients which was similar to our results. STEC strains of hospital food samples and especially cooked samples harbored the high levels of resistance against human-based antibiotics such as meropenem, levofloxacillin, mezlocillin, cefipime, polymyxin B, cefotaxime, ciprofloxacin, cotrimoxazole, ceftazidime and imipenem which can indirectly confirm their anthropogenic origin of these strains. The prevalence of antibiotic resistance genes and especially those that were encode resistance against human-based antibiotics were also high among STEC strains recovered from cooked hospital food samples which also can indirectly approved the transmission of anthropogenic STEC strains probably from staffs of hospital kitchens to foods after cooking process. Prevalence of resistance against chloramphenicol antibiotic in the STEC strains of meat and chicken samples of our investigation were 25 and 50%, respectively. There were no positive results in other raw and also all of the cooked food samples. The Iranian Food and Drug Administration (FDA) listed chloramphenicol as a forbidden antibiotic for treatment of diseases of animals and poultries. It is because of prescription of this antibiotic may cause dangerous effects on animals and even humans who use from their edible sources like meat and milk. The presence of high chloramphenicol resistance showed its irregular and unauthorized use in veterinary treatment and especially field of poultry in Iran. Veterinary practitioners of the field of poultry use from this antibiotic as a primary choice for treatment of diseases. Therefore, in a very short period of time, antibiotic resistance appears. Similar results have been reported by Momtaz and Jamshidi [32], Ranjbar et al., [33] and Colello et al., [34]. Momtaz and Jamshidi [32] reported that 73.17% of STEC strains recovered from poultry meat samples were resistant against chloramphenicol which was higher than our findings. Ranjbar et al., [33] found that the prevalence of resistance against chloramphenicol in the STEC strains recovered from various types of food samples were 21.95% which was considerable. Colello et al. [34] showed that majority of STEC strains in the Argentina harboured the gene that encodes resistance against chloramphenicol.
The most commonly detected serogroups in the STEC strains of hospital food samples of our study were O26 and O157. Presence of O157 serogroup showed that some of our strains had animal-based origin. O157 serogroup is predominant in foods with animal origin. STEC strains of raw foods with animal origins like those of meat and chicken were remained even after cooking of foods. Presence of this serogroup. High prevalence of O157 and O26 serogroups have been reported previously from Iran (Hemmatinezhad et al., [35]), Nigeria (Mamza et al., [36]), Canada (Neher et al., [37]) and Turkey (Avaz et al., [38]). Dehkordi et al., [11] reported similar profile for distribution of STEC O-serogroups. They showed that total prevalence of O157, O145, O128, O121, O113, O111, O103, O91, O45 and O26 serogroups in the STEC strains of food stuffs were 26, 6, 8, 4, 6, 6, 6, 4, 4 and 12%, respectively, which was similar to our findings.
Another part of our investigation focused on the distribution of stx1, stx2, eaeA and ehly virulence factors in the E. coli strains of hospital foods. High presence of these factors in EHEC and AEEC subtypes showed their high pathogenicity for human and especially patients. Simultaneous presence of stx1 and eaeA and stx2 and eaeA virulence factors in some strains of E. coli of hospital foods indicated the important public health problem facing Iranian hospitals and health centers which can attributed to occurrence of dangerous food poisoning and STEC-based food-borne diseases in hospitals. Simultaneous presence of stx1 and eaeA and stx2 and eaeA virulence factors have been reported previously [39][40][41]. In a study which was conducted by Momtaz et al., [42] the prevalence of EHEC and AEEC subtypes in E. coli strains of food samples were 15.06 and 49.31%, respectively. They showed that all of the EHEC strains were positive for all stx1, eaeA and ehly virulence genes, while the prevalence of these genes in AEEC subtypes were 77.77, 13.88 and 55.55%, respectively. Higher prevalence of AEEC subtypes was also reported by various investigations [9-11, 20, 32, 33]. Prevalence of AEEC subtypes in the studies of Momtaz et al., [9] and Dehkordi et al., [11] were 79.41 and 62%, respectively which were higher than EHEC subtype in both studies. Momtaz et al., [10] reported that the prevalence of AEEC subtypes in the E. coli strains recovered from the meat samples of beef, sheep, goat and camel species were 45.54, 44.87, 39.02 and 50.00%, respectively which were higher than those of EHEC. Momtaz et al., [20] reported that the prevalence of AEEC subtypes among the E. coli strains of Iranian diarrheic patients had a range of 47 to 75% which was entirely high.

Conclusions
In conclusions, we identified a large number of Oserogroups, virulence factors, antibiotic resistance genes and pattern of antibiotic resistance in STEC strains recovered from hospital foods. Raw meat, raw chicken and cooked meat, summer season, stx1 and eaeA virulence factors, AEEC subtype, O26 and O157 serogroups, aac (3)-IV, CITM, tetA, dfrA1) and sul1 antibiotic resistance genes, resistance against ampicillin, gentamycin, tetracycline, ciprofloxacin and amikacin and presence of multi-drug resistant strains were the most commonly detected characters in the STEC strains of hospital foods. Presence of O157 serogroups, EHEC strains, animal-based antibiotics and even antibiotic resistance genes which encode resistance against animal-based antibiotics in cooked foods showed insufficiency of cooking time and temperature in the kitchens of hospitals. It seems that there were no strict supervisions on the principles of food hygiene in Iranian hospitals. Due to the low levels of STEC resistance against imipenem, meropenem, streptomycin and cefotaxim antibiotics, occurrence of food poisonings due to the STEC strains in tested Iranian hospitals can be treated with their regular prescription. Attentions to the results of disk diffusion method and principles of hazard analysis and critical control point (HACCP) system can reduce the risk of STEC strains in hospital food stuffs.