Advanced molecular characterization of enteropathogenic Escherichia coli isolated from diarrheic camel neonates in Egypt

Background and Aim: Camels are important livestock in Egypt on cultural and economic bases, but studies of etiological agents of camelid diseases are limited. The enteropathogen Escherichia coli is a cause of broad spectrum gastrointestinal infections among humans and animals, especially in developing countries. Severe infections can lead to death. The current study aimed to identify pathogenic E. coli strains that cause diarrhea in camel calves and characterize their virulence and drug resistance at a molecular level. Materials and Methods: Seventy fecal samples were collected from diarrheic neonatal camel calves in Giza Governorate during 2018-2019. Samples were cultured on a selective medium for E. coli, and positive colonies were confirmed biochemically, serotyped, and tested for antibiotic susceptibility. E. coli isolates were further confirmed through detection of the housekeeping gene, yaiO, and examined for the presence of virulence genes; traT and fimH and for genes responsible for antibiotic resistance, ampC, aadB, and mphA. The isolates in the important isolated serotype, E. coli O26, were examined for toxigenic genes and sequenced. Results: The bacteriological and biochemical examination identified 12 E. coli isolates from 70 fecal samples (17.1%). Serotyping of these isolates showed four types: O26, four isolates, 33.3%; O103, O111, three isolates each, 25%; and O45, two isolates, 16.7%. The isolates showed resistance to vancomycin (75%) and ampicillin (66.6%), but were highly susceptible to ciprofloxacin, norfloxacin, and tetracycline (100%). The structural gene, yaiO (115 bp), was amplified from all 12 E. coli isolates and traT and fimH genes were amplified from 10 and 8 isolates, respectively. Antibiotic resistance genes, ampC, mphA, and aadB, were harbored in 9 (75%), 8 (66.6%), and 5 (41.7%), respectively. Seven isolates (58.3%) were MDR. Real-time-polymerase chain reaction of the O26 isolates identified one isolate harboring vt1, two with vt2, and one isolate with neither gene. Sequencing of the isolates revealed similarities to E. coli O157 strains. Conclusion: Camels and other livestock suffer various diseases, including diarrhea often caused by microbial pathogens. Enteropathogenic E. coli serotypes were isolated from diarrheic neonatal camel calves. These isolates exhibited virulence and multiple drug resistance genes.


Introduction
One-humped Arabian camel (Camelus dromedarius) livestock is widely distributed in the Middle East; the animal is well adapted to semi-arid and arid environments. In Egypt, camel herds are important semi-domestic animals heavily used in tourist safaris and transportation and for milk and meat [1].
Escherichia coli is a Gram-negative rod in Enterobacteriaceae. The bacterium is considered an important constituent of gut microbiota in birds and mammals. However, various strains of this species are pathogenic and cause various gastrointestinal disorders, such as diarrhea [2]. In Egypt, diarrheagenic E. coli (DEC), especially Shiga toxin-producing E. coli, are described in livestock, including calves, sheep, goat, and buffalo calves [3][4][5]. However, few studies of DEC in camels are available.
Some strains of DEC are public health concerns and may be transmitted from livestock (e.g., camels) to humans. Gastrointestinal infection may cause nonbloody or bloody diarrhea, hemorrhagic colitis (HC), and hemolytic uremic syndrome. These illnesses threaten millions of peoples worldwide, especially in developing countries [6].
Pathogenic E. coli diagnosis requires differentiation from non-pathogenic strains that are constitutive components of normal intestinal flora. Such pathotypes are defined by the presence of one or more definable E. coli virulence factors and can be identified by advanced molecular and certain conventional Available at www.veterinaryworld.org/Vol.14/January-2021/11.pdf methods [7]. Molecular techniques are more rapid and precise for the identification of bacteria compared with conventional phenotypic methods. Multiplex polymerase chain reaction (mPCR) and quantitative real-time PCR (qRT-PCR) are typically used for identification of bacterial species. Further, sequence analysis is an accurate approach for species delineation and epidemiological tracking [8].
Detection of specific genes is fundamental for diagnosis, confirmation, and/or investigation of virulence and antibiotic resistance of E. coli isolates; yaiO is a housekeeping gene in E. coli orphan open reading frame that encodes a protein originally expressed and localized in the outer membrane [9]. traT is a virulence factor and a constituent of conjugative plasmids that encodes a surface protein responsible for entry exclusion [10]. An essential step for initiation and development of enteritis is bacterial attachment to intestinal epithelial cells. E. coli attachment is mediated by fimbriae (adhesins) that bind to host cell receptors; the FimH gene is representative [11].
Resistance of microorganism to multiple antimicrobial drugs is a global threat that interferes with and prolongs therapy. Pathogenic E. coli strains characteristically harbor antibiotic resistance genes. MphA is responsible for production of macrolide 2′-phosphotransferase I that inactivates 14-ring macrolides, such as erythromycin and oleandomycin [12]. Another drug resistance gene, AmpC, encodes AmpC β-lactamase that degrades penicillins [13]. aadB is responsible for the production of 2'-aminoglycoside nucleotidyltransferase enzyme that inactivates aminoglycoside antibiotics [14].
Studies of molecular characteristics of E. coli isolates from camel herds in Egypt are limited. The current study explores pathogenic E. coli among diarrheic camel calves using advanced molecular techniques.

Ethical approval
As per CPCSEA guidelines, a study involving clinical and postmortem samples does not require the approval of the Institute Animal Ethics Committee.

Sampling
Seventy fecal samples were collected from neonatal camel calves raised for meat production. Calves were under 6 months of age and suffered from diarrhea. Calves were raised in the Giza Governorate during 2018-2019. Swabs were immediately placed in cooled boxes, transported to the laboratory, and incubated overnight at 37°C in trypticase soy broth (Merck KgaA, Darmstadt, Germany).

Isolation and identification of E. coli
Each sample was streaked onto MacConkey's agar (Merck, Germany) and incubated for 24 h at 37°C. Lactose-positive colonies were similarly incubated on eosin methylene blue agar (EMB, Merck, Germany). We used two selective media because EMB is used for selection because it is specific for E. coli, and MacConkey is for lactose fermentation.
Green colonies with a metallic luster were considered E. coli; such colonies were further identified using the API 20E system (BioMérieux, France) following the manufacturer's instructions to obtain biochemical profiles of isolates.
The isolates were serologically identified by "Seikin" E. coli O-antigen diagnostic antisera (polyvalent and monovalent vials, product code; 312001) following the manufacture's instructions.

Molecular diagnosis of E. coli isolates
Polymerase chain reaction (PCR) DNA extraction DNA was extracted from E. coli isolates using a QIAamp DNA Mini kit (Qiagen, Germany, GmbH) with modifications from the manufacturer's recommendations. Briefly, 200 μL of sample suspension was incubated with 10 μL of proteinase K and 200 μL of lysis buffer at 56°C for 10 min. After incubation, 200 μL of 100% ethanol was added to the lysate. The sample was then washed and centrifuged following the manufacturer's instructions. Nucleic acid was eluted with 100 μL of elution buffer provided in the kit.

Analysis of the PCR products
The products of PCR were separated by electrophoresis in 1.5% agarose gels (Applichem, Germany, GmbH) in ×1 Tris/Borate/ethylenediaminetetraacetic acid buffer at room temperature using gradients of 5 V/cm. Thirty microliters of PCR product was loaded in each gel slot. A 100 bp ladder (Qiagen, Germany, GmbH) was used to determine fragment sizes. The gel Available at www.veterinaryworld.org/Vol.14/January-2021/11.pdf was photographed using a gel documentation system (Alpha Innotech, Biometra, Germany), and data were analyzed with computer software. Reference E. coli strains ATCC35150 and HB 101 were used as a positive control and Staphylococcus aureus ATCC29737 as a negative control.

RT-PCR
Positive E. coli O26 samples were extracted using a Sigma kit following kit instructions. RT-PCR using a MTplexdtec-RT-qPCR Test (Edifici-Quόrum3, Spain) that includes vt1 and vt2 genes with vt specific primers and probes (Table-2) [18]. Signal was collected, and reactions run in an Applied Biosystem StepOne RT-PCR System. FAM fluorogenic cycle threshold of reactions was identified using StepOne TM software (Life Technology, USA). The qualitative detection module analyzed the samples for the presence of vt1 and vt2 based on fluorescence levels above the background.

E. coli DNA sequencing
The amplified traT PCR fragments (307 bp) were excised from gels and the DNA extracted using ExoSAP-IT PCR Product Cleanup kit (Affymetrix, USA) following the manufacturerʼs instructions. Purified amplicons were sequenced with an ABI PRISM ® BigDyedye™ terminator cycle sequencing kit with AmpliTaq ® DNA polymerase on an MJ Research PTC-225 Peltier Thermal Cycler (Applied Biosystems, USA) following manufacturer's recommendations. DNA sequences were identified by comparison with established sequences in GenBank.

Results and Discussion
The bacteriological and biochemical examination of the 70 fecal samples showed 12 E. coli isolate (17.1%). These results are similar to the findings of Hussni et al. [19], who reported a prevalence of enteropathogenic and enterotoxigenic E. coli in camel fecal samples in Qatar of 21.1%. This incidence was slightly higher than 12.3% of pathogenic E. coli found in Kenya [20]. Conversely, the prevalence in our study was much higher than among 140 fecal samples collected from camels in the United Arab Emirates (4.3%) [21] and 3.8% among 600 fecal samples collected from camels in Nigeria [22].
E. coli isolates were more frequently resistant to vancomycin (75%) and ampicillin (66.6%). However, significant resistance was also observed to amoxicillin (58.3%) and streptomycin (50%). In contrast, isolates were found to be highly sensitive to ciprofloxacin, norfloxacin, and tetracycline (100%) and somewhat less sensitive to gentamycin and trimethoprim-sulfamethoxazole (83.3%). Previous evaluation of pathogenic E. coli isolates for antibiotic resistance varies.
Bessalah et al. [24] indicated that E. coli isolates were sensitive to ciprofloxacin, gentamicin, amikacin, chloramphenicol, and ceftiofur and resistant to ampicillin and tetracycline. Further, E. coli isolates were 100% susceptible to ciprofloxacin, norfloxacin, cefotaxime, chloramphenicol, and polymyxin B [21]. yaiO is a housekeeping gene of E. coli, and its detection confirms that isolates are E. coli strains. The intended 115 bp band was amplified in all 12 isolates to confirm E. coli (Figure-1). Molina et al. [9] reported that yaiO amplification is highly specific for E. coli, with superior detection ability and non-redundancy with enzymatic methods.
The presence of the virulence genes is indicative of the pathogenicity of E. coli strains and is used to distinguish such strains from non-pathogenic bacteria. traT is important for conjugal transfer in E. coli, encodes a complement resistance protein, and is responsible for preventing unproductive conjugation between bacteria carrying plasmids [10]. Moreover, adherence of E. coli to host receptors confers resistance to mechanical elimination and increases persistence. fimH is a major determinant encoding adhesins that target epithelial receptors; thus, the gene is crucial for E. coli colonization [11]. Results obtained from diplex PCR (Figure-2) recognized traT and fimH genes in 8 and 10 isolates (75%, 83.3%), respectively. Moreover, all four isolates of serotype O26 harbored both virulence genes. The three isolates of serotype O103 carried fimH, but only one carried the traT gene. Two of three isolates of serotype O111 carried traT and fimH and the two O45 isolates harbored one gene each. Similar data were reported by Díaz-Jiménez et al. [25], who indicated that about 85% of E. coli isolates implicated diarrheic conditions carried traT and fimH genes. Staji et al. [26] showed that the traT gene is common enteropathogenic (EPEC) and enterohemorrhagic E. coli (EHEC) ink serotypes O26 and O111. Further, the fimH gene was expressed in enteropathogenic E. coli strains of zoonotic importance [27] and in multidrug-resistant diarrheagenic strains [28].
The emergence of multidrug-resistant strains is important for economic concerns as well as a public health [29]. The 12 E. coli isolates were checked for the presence of three prevalent antibiotic resistance genes (Figure-3). All 12 isolates harbored at least one gene. Nine carried ampC (75%), eight mphA (66.6%), and five aadB (41.7%). Two isolates harbored all three genes (25%), showed the O26 serotype. Five isolates carried two genes (41.7%). These data indicate that 7 isolates (58.3%) could be considered MDR. Interestingly, genotypic results are harmonized with the phenotype of one isolate; high and intermediate resistance to vancomycin, ampicillin, amoxicillin, and streptomycin. Multidrug resistance for this combination among DEC strains is frequently reported (80%, 94%, and 78.1%) [30][31][32], respectively.  Pathogenic and toxigenic E. coli are implicated in many disorders in animals, such as diarrhea and HC. The O26 serotype is the second most significant pathogenic serotype worldwide after O157:H7, since it produces toxins and causes zoonotic disease. The four O26 isolates were given more attention in the present study. These isolates were positive for vt1 and vt2, genes responsible for verocytotoxin expression. One isolate was positive for targeting vt1 at cycle 32, and three isolates were positive for targeting vt2. Previous reports also show an association between enterotoxigenic E. coli O26 strains and the presence of verotoxigenic genes [18,33].
DNA sequence was initially performed to establish identity to sequences in GenBank. E. coli O26 GH1, 2, 3, and 4 were selected for similarity to virulence traT gene with genes in GenBank [34][35][36].

Conclusion
In spite of camel has unique circumstances, but as other livestock suffer from different diseases,  including diarrhea, caused by microbial agents. Enteropathogenic E. coli serotypes isolated from diarrheic neonatal camel calves displayed virulence and multidrug resistance, the ability to produce toxins, and possible importance for zoonotic disease transmission.

Authors' Contributions
MAS and AND designed the study. HAA and EAK collected samples and then performed the bacterial isolation and biochemical typing. ASH conducted serological typing. HAA and EAK performed DNA extraction and PCR. SFA and AND fulfilled the qRT-PCR, while EK and SFA implemented sequencing. MAS and AND analyzed the data, ASH and AND drafted the manuscript, revised, and finalized the manuscript for submission. All authors read and approved the final manuscript.