Prevalence and antimicrobial resistance of diarrheagenic Escherichia coli and Shigella species associated with acute diarrhea in Tehran, Iran in La prévalence et la résistance antimicrobienne de l’ Escherichia coli diarrhéogène et des espèces de Shigella liés à la diarrhée aiguë à téhéran, en Iran

Prevalence and Antimicrobial resistance of diarrheagenic Escherichia coli and Shigella species associated with acute diarrhea A study was performed to determine the prevalence and antimicrobial resistance of Shigella species and diarrheagenic Escherichia coli isolates cultured from patients with acute diarrhea in Tehran, Iran. Between May 2003 and May 2005, 1120 diarrheal specimens were collected and assayed for bacterial enteropathogens by conventional and molecular methods. Etiological agents were isolated from 564 (50.3%) specimens, and included 305 (54%) E coli , 157 (27.8%) Shigella species, and 102 (18%) from other genera of bacteria. The predominant E coli was Shiga toxin-producing E coli (105 isolates [34.5%]) and the predominant Shigella serotype was Shigella sonnei (88 isolates [56.1%]). A high rate of antibiotic resistance was observed among E coli, with 40 of 53 (75.5%) Shiga toxin-producing E coli isolates resistant to amoxicillin and tetracycline, and eight (5.2%) E coli isolates resistant to more than six antibiotics. Most Shigella isolates were resistant to tetracycline (95%) and trimethoprim-sulfamethoxazole (91.7%), with greatest antibiotic resistance observed among S sonnei (53 of 88 [60.2%] isolates). Antibiotic resistance is widespread in diarrheagenic E coli and Shigella in children with acute diarrhea in Tehran, Iran; hence, updated strategies for appropriate use of antimicrobial agents in Iran are needed. a effectué une étude pour déterminer la prévalence et la résistance antimicrobienne des espèces et des isolats d’ Escherichia à partir de prélèvements de patients atteints Téhéran, en 2003 prélevé et titré 120 échantillons de diarrhée dépister des entéropathogènes bactériens moyen des classiques moléculaires. étiologiques auprès de échantillons bactéries E producteur de of the recent diarrheagenic E coli and Shigella isolates are resistant and our higher rates of resistance to commonly used and the of isolates resistant to more than six antibiotics with and an ongoing monitoring system for antimicrobial in Iran. multidrug

nausea, vomiting, abdominal cramps or fever of 38°C. Acute diarrhea was defined as diarrhea that lasted 14 days or less at the time of presentation. Isolates from children with persistent diarrhea were not included in this study. Persistent diarrhea was defined as diarrhea which lasted for more than 14 days at presentation. When diarrhea was present intermittently, it was considered persistent when diarrhea occurred during at least six days in a two-week period.

sample collection
The present study included children who presented with acute diarrhea from May 2003 to May 2005 from four hospitals (two of four were children's hospitals) in Tehran. The children were categorized into six different groups according to their age -zero to two months, three to five months, six to 11 months, 12 to 23 months, 24 to 35 months and 36 to 60 months. They were part of the routine 25% to 30% surveillance sampling of all diarrhea children seen at the children's hospitals and clinical centres in Iran. The study was endorsed by the Ethic Committee of Research Center for Gastroenterology and Liver Diseases. Informed consent was obtained from the children's parents.

Fecal specimens
All specimens were added to Cary-Blair and PBS transport media and transported on ice packs to the laboratory of the Research Department of Foodborne and Diarrheal Diseases of Shaheed Beheshti University of Medical Sciences, M.C. in Tehran. Stool samples were collected on the day of presentation at the outpatient department, immediately transported to the laboratory, stored at 4°C and processed within 24 h.

Bacterial isolation
Samples (swabs from Cary-Blair transport media) were cultured for E coli, Shigella and Salmonella species by streaking directly onto xylose-lysine deoxycholate agar, MacConkey agar and Salmonella-Shigella agar (Merck, Germany), and incubating for 24 h at 37°C, and by enriching in Selenite F broth at 37°C overnight. Enrichment cultures were subcultured on each of the media described above by incubating at 37°C overnight. Swabs from Cary-Blair transport media of samples were also cultured for Campylobacter species, on enrichment media including Preston enrichment broth base (HIMEDIA, Mumbai, India, M899) supplemented with Campylobacter selective supplement IV (HIMEDIA, Mumbai, India, FD158) with an antibiotic (Campylobacter Supplement 2, Blaser Wang, HIMEDIA, Mumbai, India, FD 006) and 5% defibrinated sheep blood and incubated for 48 h at 42°C under microaerophilic conditions. One presumptive Campylobacter colony from each selective agar plate was subcultured and tested for Gram stain, production of catalase and oxidase and hippurate hydrolysis. For Yersinia species, swabs from PBS transport media were cultured on selective agar, CIN agar, (Merck K GaA Darmstadt, Germany) and incubated at 22°C for 24 h to 72 h.

Molecular diagnosis of diarrheagenic E coli and
Shigella species E coli-like and other Gram-negative colonies growing on the primary culture medium were tested by polymerase chain reaction (PCR) assay for virulence genes. For DNA extraction, a loopful of bacterial growth obtained from the culture plate was suspended in 0.75 mL of sterile distilled water and boiled for 20 min.
The PCR product bands were separated by electrophoresis through 1.3% agarose gel in 1× TBE buffer. DNA fragments were visualized by ethidium bromide staining and photographed under ultraviolet light illumination.
Up to 15 to 30 suspected colonies for diarrheagenic E colipositive cases were tested for the virulence genes. E coli isolates that were eae-positive and stx-negative were identified as EPEC. Isolates positive for pCVD432 were identified as EAEC and those stx1-or stx2-positive, or both, were Shiga toxinproducing E coli (STEC). Those E coli isolates that were lt-or stpositive, or both, were ETEC. Positive controls with genomic DNA from E coli ATCC 35401 (lt-positive, st-positive), E coli ATCC 43894 (stx1-positive, stx2-positive, eae-positive), E coli RH 4620 (Pcvd432-positive) and Shigella ATCC 9290 (ipaH-positive) were included in each PCR run along with a nontemplate-containing reaction to detect false-positive results. For isolation the Shigella strains, PCR was done for ial gene by the ial lower GGAGGCCAACAATTATTTCC and ial upper CTGGATGGTATGGTGAGG primers (320 base pairs). Which sample was positive for both ial and ipaH genes presented as Shigella strains (17).

Biochemical identification and serological assay
Colonies morphologically resembling Shigella species were further identified by biochemical reactions according to standard methods (8 and confirmed by slide agglutination test using commercially available antisera from Mast Group Ltd (18). E coli isolates that were positive for both stx1 and stx2 genes were tested by O157 and H7 antiseras (MAST House, Merseyside, United Kingdom).
Shigella isolates were also assayed for the same 11 antibiotics in addition to erythromycin (10 μg), cefixime (5 μg) and ampicillin-sulbactam (20 μg). The resistance break points were those recommended by Clinical and Laboratory Standards Institute, (20) and was reported as sensitive or resistant based on break points. Bacteria used for controls included Staphylococcus aureus ATCC 25923 and E coli ATCC 25922.

statistical analysis
All data were entered into SPSS/PC 11.0 software (SPSS Inc, USA) and comparative statistics were calculated using the two-tailed χ 2 test and Fisher's exact test. P≤0.05 was considered to be significant.
The highest acute diarrhea rate was from children 12 to 24 months of age (n=266) during the dry seasons (from June 20 to September 20) (68.6%) ( Table 2). The 153 pure diarrheagenic E coli strains were isolated and selected for antibiotic sensivity testing. The number of isolates compared by PCR-positive samples show that the molecular technique is more sensitive than conventional tests.
Resistance to more than four antibiotics was observed in 40.6% (n=62) of strains of the diarrheagenic E coli isolates, and eight (5.2%) were resistant to more than six antibiotics ( Table 5). The predominant multidrug resistance pattern in these isolates (13.7%) was ampicillin, amoxicillin-clavulanic, chloramphenicol, tetracycline, trimethoprim-sulfamethoxazole  and streptomycin (P<0.05). Four E coli isolates were resistant to nine antibiotics (Table 5). According to the definition of multidrug resistance by Lockhart et al (21), five E coli isolates were multi-drug resistant. These isolates were resistant to at least eight antimicrobials from three classes. Of these isolated, four were ETEC and only one of them was among EAEC isolates (Table 6). Ninety-one (57.9%) Shigella species isolates were resistant to more than four antibiotics, with 16 resistant to six antibiotics (Table 7). S sonnei had the greatest percentage of multidrugresistant isolates, with 53 (58%) resistant to more than four antibiotics. Eight (16.6 %) S flexneri isolates were resistant to six antibiotics, and one was resistant to seven (Table 7).     All the Shigella boydii isolates were susceptible to ceftazidime, gentamicin, ceftriaxone, ciprofloxacin and ampicillin-sulbactam; however, two isolates were resistant to six antibiotics. Eleven (84.6%) S boydii and all eight Shigella dysenteriae isolates were resistant to tetracycline (Table 4). In addition, more than 90% of Shigella isolates were susceptible to cefixime, ceftriaxone, ceftazidime, nalidixic acid, gentamicin, ciprofloxacin and ampicillin-sulbactam, and all Shigella species isolates were susceptible to ciprofloxacin. Among the diarrheagenic isolates, 58% of Shigella species and 40% of E coli were resistant to more than four antibiotics. No Shigella species isolated categorized as multidrug-resistance by this definition.

dIsCussIon
The present study provides results of prevalence and antibiotic resistance patterns of diarrheagenic E coli and Shigella species isolates from children younger than five years of age with acute diarrhea in Iran. E coli (305 isolates [54%]) and Shigella species (157 isolates [27.8%]) were the most prevalent bacterial enteric pathogens isolated. These results are in the agreement with previous studies in which diarrheagenic E coli and Shigella species were identified as the most common bacterial enteric pathogens in Iran (18,22). Other isolated bacteria (Campylobacter [10.6%] and Salmonella [7.45%]) were less commonly associated with acute diarrheal illness in Iran than the diarrheagenic E coli and Shigella species (18,23,24) (P<0.05).
Determining the prevalence of diarrheal pathogens should save the way for better control of the disease in the country. Continued vigilance of the safety of food, health education of food handlers, and close attention to hygiene and sanitary conditions can provide an effective barrier against the spread of shigellosis (25). The antimicrobial resistance may be as a result of inappropriate and wide use of different antibiotics to treat infection.
Resistance to currently used antimicrobial agents among enteric pathogens has increased dramatically worldwide during the past decade (24,26,27). In developing countries, trimethoprim-sulfamethoxazole, ampicillin and tetracycline are widely used to treat diarrhea because of their low cost and availability (8). The widespread use of these antibiotics has resulted in an increased prevalence of resistance to these antibiotics by diarrheagenic bacteria, thereby raising concern among general practitioners and pediatricians, especially in developing countries (28 ERY, TET, AM, SXT  2  16  1  0  19  12   ERY, TET, AM, AMC  1  0  1  1  3  1.9   CEF, TET, AMP, SXT  1  1  1  0  3   In addition, a report from Iran cited by the World Health Organization indicates that sulfamethoxazole-trimethoprim, tetracycline and chloramphenicol were the least effective antibiotics since 112 (80.0%), 90 (64.3%) and 78 (55.7%) of the diarrheagenic E coli isolates were resistant to these antibiotics, respectively (30). Approximately 95% of Shigella species isolates in our study was resistant to tetracycline and 91.7% to trimethoprim-sulfamethoxazole. Comparing our data with previous reports from Iran reveals the same resistance pattern to commonly used antimicrobials such as tetracycline, sulfamethoxazole-trimethoprim and ampicillin, except resistance to chloramphenicol. Results from our study indicate the sensitivity of the most recent isolates to chloramphenicol has increased as all diarrheagenic E coli, S dysenteriae and S boydii isolates were sensitive to chloramphenicol previously.
Our findings are in agreement with reports of others in Iran indicating diarrheagenic E coli and Shigella isolates are resistant to trimethoprim-sulfamethoxazole, tetracycline and erythromycin (9,18). However, our results indicate higher rates of resistance to commonly used antibiotics and the emergency of isolates resistant to more than six antibiotics compared with reports from Iran and other countries. These findings illustrate the need for an ongoing monitoring system for antimicrobial resistance in Iran. Many studies have determined that multidrug resistance is common among Shigella and E coli isolates, especially to ampicillin and trimethoprim-sulfamethoxazole which are frequently used to treat shigellosis (31)(32)(33). In these cases, ampicillin, trimethoprim-sulfamethoxazole and tetracycline should be excluded for the treatment of diarrhea and quinolones, to which these isolates E coli and Shigella species were susceptible, can be used as an alternative treatment.
Our study revealed a high prevalence of multidrug antibiotic resistance among Shigella and E coli isolates to the antibiotics most frequently used to treat diarrheal illnesses in Iran (P<0.05). The emergence of multidrug resistance requires continuous monitoring of the resistance patterns and the development and implementation of best practices strategies for antibiotic usage to extend the efficacy of the few effective antibiotics that remain.
The pattern of antibiotic resistance was monitored for Shigella and diarrheagenic E coli isolates. According to the definition of multidrug resistance by Lockhart et al (21), five E coli isolates were multidrug resistant. These isolates were resistant to at least eight antibiotics from three classes (one extended-spectrum cephalosporin, one aminoglycoside and ciprofloxacin) (21). It is quite unique finding in that similar trend in antimicrobial resistance of diarrheagenic pathogens, and has not been observed in Iran. This may be due to inappropriate use of antimicrobials especially in blind cases. The changing patterns of resistance to common antimicrobial agents in Iran indicates that designing a surveillance system for antimicrobial resistance and the introduction of integrated guidelines for the appropriate use of antibiotics are urgently needed. The result of this study suggests antimicrobial resistance is widespread among diarrheagenic E coli strains.