Enteroaggregative Escherichia coli in diarrheic children in Egypt : molecular characterization and antimicrobial susceptibility

Introduction: Little information is available regarding the significance of enteroaggregative Escherichia coli (EAEC) in pediatric diarrhea in Egypt. Methodology: Escherichia coli was isolated from stool samples of 62 diarrheic and 43 non-diarrheic (control) Egyptian children. Samples were screened for genes specific for enteroaggregative E. coli (EAEC), enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), Shiga toxin-producing E. coli (STEC), and enteroinvasive E. coli (EIEC) using polymerase chain reaction (PCR). Diarrheagenic E. coli were grouped phylogenetically using PCR and tested for their susceptibility to antibiotics using the disk diffusion method. Isolates designated as EAEC were examined for eight virulence factors (VFs) using PCR. Results: EAEC was detected in 19 (30.7%) and 4 (9.3%), EPEC in 2 (3.2%) and 1 (2.3%), and ETEC in 2 (3.2%) and 0 (0.0%) diarrheic and control children, respectively; STEC and EIEC were not detected. Only EAEC was significantly isolated from diarrheic children compared with controls (p < 0.01, OR = 4.31).Three or more VFs (multivirulent isolates) were found in 52.6% and 50% of EAEC isolated from diarrheic children and controls, respectively. More than 73% (17/23) of EAEC isolates were identified as belonging to phylogenetic group D. Multiple-antibiotic resistance (resistance to three or more drugs) was observed in more than 91% of EAEC. Conclusions: Multivirulent EAEC is a significant causative agent of pediatric diarrhea in Egypt, with the majority of isolated EAEC belong to phylogenetic group D. Multiple-antibiotic resistance among EAEC has the potential to be a serious public health problem for the country.


Introduction
Diarrheagenic Escherichia coli (DEC) are important intestinal pathogens causing a wide variety of gastrointestinal diseases, particularly among children in developing countries.These organisms are differentiated into enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), Shiga toxin-producing E. coli (STEC), enteroaggregative E. coli (EAEC), and diffusely adherent E. coli (DAEC) [1].Person-to-person transmission and consumption of foods and drinking water contaminated with human waste are the main modes of acquiring DEC infections.
Generally, acute infectious diarrhea is self-limiting and responsive to therapy by oral rehydration.However, antimicrobial therapy is recommended for severely ill patients and for those with risk factors for invasive infection after obtaining appropriate blood and fecal cultures [2].High frequency of resistance to antimicrobial agents among DEC isolated from children with diarrhea has been reported from several developing countries [3,4].With the exception of ETEC [5], there little information published on the susceptibility of DEC from Egypt to antibiotics.
The epidemiological significance of different DEC pathotypes in childhood diarrhea varies geographically.Few studies have examined the clinical relevance of ETEC-associated diarrhea in Egypt [5,6].However, there is little information on the role of EAEC in pediatric diarrhea in the country.The objectives of the present study were to determine the prevalence of DEC pathotypes in diarrheic and nondiarrheic Egyptian children and to determine the susceptibility of the isolated DEC to antimicrobial agents.In addition, the phylogenetic grouping of the isolated DEC and determination of virulence-encoding genes in EAEC isolates were determined by PCR.

Methodology
Stool and bacterial samples E. coli was isolated from stool samples obtained from 62 children with acute diarrhea and 43 control children without diarrhea admitted to Ain Shams University and Boulaq hospitals in Cairo, Egypt, during the summer of 2007.Acute diarrhea was defined as three or more loose or liquid stools per day lasting fewer than 14 days.Children were ≤ 5 years of age; > 90% were < 3 years of age.After receiving informed consent from a parent or guardian, a clinical history for each child was obtained.Histories included results of the attending doctor's physical examination and clinical symptoms, including fever, vomiting, abdominal pain, dysentery, convulsions, and dehydration using standard charts.Feeding practices (breastfeeding, artificial supplements, or both [mixed]), history of antibiotic usage, source of drinking water (treated or untreated), and history of travel abroad in the 30 days prior to stool collection were also recorded.Control children were admitted to either hospital with non-gastrointestinal illnesses and had not received antibiotics during the 30 days prior to stool collection.The majority of children sampled were from low-income families.Stool specimens from enrolled children were collected using wide-mouthed sterile plastic containers and transported within two hours of collection to the microbiology laboratory for analysis.
Stool specimens were directly streaked onto MacConkey agar (Difco, Detroit, MI, USA) for isolation of E. coli.After overnight incubation at 37°C, three lactose-fermenting colonies and a representative non-lactose fermenting colony with a different morphological appearance were picked, and their identities were determined using a panel of biochemical tests interpreted as previously described [7] and the API 20E system (bioMerieux, Marcy l'Etoile, France).

Statistical analysis
Epi Info version 3.5.1 software (Centers for Disease Control and Prevention, Atlanta, GA, USA) was used to analyze the data.Pvalues were calculated using the χ 2 test; pvalues < 0.05 were considered statistically significant.
Acute diarrhea lasting for more than one day was recorded for 91.3% of the cases; vomiting and fever (87%), dehydration (82.6%), and abdominal pain (43.5%) were recorded for DEC-positive children with diarrhea.Distribution of DEC isolated from diarrheic children according to age, gender, type of feeding practices, and source of drinking water is shown in Table 2.No significant differences were found in the isolation rates of DEC in relation to gender, type of feeding, or source of drinking water.However, EAEC was significantly associated with diarrheic children one year of age or younger compared with diarrheic children > 12 months of age (p < 0.03, OR = 3.93).None of the children included in the present work had traveled abroad in the 30 days prior to stool collection.

Distribution of virulence genes among the isolated DEC
Table 3 shows the distribution of virulence factors in DEC isolated from children with diarrhea and from controls.No significant differences were observed among EAEC from diarrheic children compared with controls for the presence of any of the eight genes examined.More than 95% (22/23) of EAEC from diarrheic and control children carried the aggR gene.Three or more VF genes (multivirulent isolates) were found in 47.4% and 50% of EAEC isolated from diarrheic children and controls, respectively.Typical EPEC (eae+ve and bfpA+ve) were found in 50% (1/2) of EPEC from children with diarrhea.Only genes encoding heat-labile enterotoxins were detected in the recovered ETEC isolates.

Determination of DEC phylogenetic groups
Of the 23 EAEC isolates recovered (including the 4 from controls) a single isolate (4.4%) was identified as belonging to phylogenetic group A, 2 (8.7%) to group B1, 3 (13%) to group B2, and 17 (73.9%) to group D. Of the 3 EPEC isolates (1 from a control) 1 (33.3%) was identified as belonging to group B1 and 2 (66.7%) to group B2.Of the 2 ETEC isolates, 1 (50%) was identified as belonging to group A, and 1 (50%) belonged to group D.

Susceptibility to antimicrobial agents
More than 95% (22/23) of EAEC isolates were resistant to ampicillin, trimethoprimsulfamethoxazole, and tetracycline.Multidrug resistance (MDR, resistant to three classes of antimicrobial agents or more) was observed in more than 91% (21/23) of EAEC isolates.In addition, EAEC isolates were significantly more resistant to cefotaxime, ceftriaxone, and chloramphenicol (p < 0.006, OR = undefined;p < 0.0002, OR = undefined; and p < 0.02, OR = undefined, respectively) compared with EPEC and ETEC isolates.Furthermore, ESBLs were detected in 78.3% (18/23) of EAEC isolates but in none of the EPEC or ETEC isolates.All EAEC, EPEC, and ETEC isolates recovered in the present study were susceptible to imipenem and ciprofloxacin.The antimicrobial susceptibility profile of DEC isolated from children in Egypt with diarrhea and controls is shown in Table 4.

Discussion
Reports on the prevalence of DEC using molecular methods from North Africa and the Middle East countries are few.Albert et al. investigated the etiology of DEC in hospitalized children with diarrhea in Kuwait [24].They found no significant difference in the detection of different DEC pathotypes from diarrheic children compared with controls.In the present work, only EPEC, ETEC, and EAEC were detected in stool samples, with only EAEC being significantly found in diarrheic children (30.7%) compared with controls (9.3%) (p < 0.01).Compared with our findings, lower prevalence rates of EAEC from diarrheic children have been reported from Iraq (8%), Kuwait (2.6%), Libya (4.1%), and Tunisia (11.3%) [24][25][26][27].
Prevalence rates between 1% and 24% for EPEC from diarrheic children using PCR techniques have been reported from several countries in North Africa and the Middle East [24][25][26].In the present investigation, EPEC was detected in 3.2% of diarrheic children in Egypt.Previous studies observed that ETEC was a significant cause of severe childhood diarrheal disease in Egypt, with prevalence rates of 10%-20% [6,28].In this study, only two patients were positive for ETEC (3.2%).It should be noted that previous studies from Egypt used the GM1 enzymelinked immunosorbent assay for detection of LT and STa [29,30], while we used PCR techniques in the present investigation.Although it is not easy to explain the differences in prevalence rates of DEC pathotypes found in the present and previously mentioned investigations, several factors may contribute to such differences, including geographical locations, populations studied, and quality of sanitation.
In the present study, the three main symptoms (fever, dehydration, and vomiting) were observed in more than 80% of DEC-positive children.The high rates of these clinical symptoms among affected children supports the etiological importance of DEC, particularly EAEC, in Egypt, a region where diarrhea is endemic.A significant association of EAEC with diarrheic children one year of age or younger compared with diarrheic children older than one year of age (p < 0.03) was observed as part of this study.Al-Gallas et al. in Tunisia reported similar findings [27].Substantial proportions of DEC were isolated from breastfed children (76.5%) in this study.Although breastfeeding protects from the morbidity and mortality caused by diarrhea in the first few months of life [31], lack of hygiene, sanitation, and other factors may undermine the role of breastfeeding in reducing the prevalence of infectious diarrhea among children, particularly in low socio-economic status communities.
EAEC isolates with three or more VF genes (multivirulent isolates) were found in 47.4% of EAEC isolated from Egypt (Table 3).A recent study from Libya found similar results, with multivirulence being detected in 50% of EAEC from diarrheic children [26].Sarantuya et al. reported a lower rate (< 30%) of multivirulence among EAEC from children with diarrhea in Mongolia [32].
We found the majority (> 73%) of EAEC (from both diarrheic children and controls) to be associated with phylogenetic group D. Okeke et al. reported that EAEC strains (> 80% isolated from diarrheic and control children) from Nigeria were almost evenly distributed among A, B1, and D phylogenetic groups and that only 4.7% of them belonged to group B2 [33].On the other hand, Pérez et al. did not find phylogenetic group D among eight EAEC strains from Costa Rican diarrheic children [34].The differences in phylogenetic groups among our EAEC isolates and those reported by the above-mentioned studies could be due to different ancestral origins of EAEC in each location.
Knowledge of recent local antibiotic susceptibility patterns of pathogens can assist in choosing the proper treatment when required.We observed high rates of resistance to the commonly used drugs ampicillin, trimethoprim-sulfamethoxazole, and tetracycline among the isolated DEC, which is concerning as treatment may not be suitable with these drugs.High rates of MDR (resistance to three drugs or more) among enteric bacteria are not uncommon in developing countries.The misuse of antimicrobials and their availability over the counter without a prescription is common in developing countries, including Egypt [35,36].Therefore, it is not unexpected to find high rates (more than 91%) of MDR among our EAEC isolates.Nguyen et al. reported similar observations (90% MDR) among DEC (including EAEC) isolated from children [3].In addition, we found that more than 78% of EAEC and none of the EPEC and ETEC isolates produced ESBL.A study from Kuwait reported a lower rate for ESBL (28.6%) among EAEC from diarrheic children [24].
In conclusion, multivirulent EAEC is a significant causative agent of pediatric diarrhea in Egypt, with the majority of the isolates belonging to phylogenetic group D. Multiple-antibiotic resistance among EAEC is a serious health problem in the country.More studies are needed to determine if EAEC may be prevalent in other cities in Egypt and to address the problem of antimicrobial resistance associated with these organisms.

Table 1 .
PCR primer sequences for detection of different virulence-associated genes in E. coli a For all PCR conditions, a final extension step of 7 min at 72°C was performed.

Table 2 .
Distribution of diarrheagenic Escherichia coli (DEC) from diarrheic children according to age, gender, type of feeding, and source of drinking water *Only children ≤ 2 years of age positive for DEC were included.1EAECwas significantly associated with diarrheic children ≤ 12 months of age compared with diarrheic children > 12 months of age (p< 0.03, OR = 3.93).

Table 4 .
Antimicrobial susceptibility profile of diarrheagenic Escherichia coli isolated from children in Egypt with diarrhea or from non-diarrheal controls