Absence of Escherichia coli O157:H7 Serotype in Small Indian Mongooses ( Herpestes auropunctatus ) in Grenada and Antimicrobial Drug Resistance of the non-O157 Isolates

Aim: To determine the occurrence of the E. coli including the O157:H7 serotype in mongooses and their antimicrobial drug resistance in Grenada . Study Design: Experimental based study of feces of mongooses captured from six parishes of Grenada from April 2011 to March 2013 during an active rabies surveillance program. Methodology: Fecal samples from 156 mongooses were cultured for E. coli and tested for O157:H7 serotype by the presence of non-sorbitol fermenting colonies and a positive reaction to Original Research Article Results: Of the 156 mongooses, 71 (46%) were culture positive for E. coli . A total of 213 E. coli isolates were recovered and examined for the presence of non-sorbitol fermenting colonies and O157-agglutination. Of the 213 E. coli isolates, only 8 (4%) were non-sorbitol fermenters. However, none of the 213 isolates gave a positive reaction (O157-agglutinating) to the two E. coli O157:H7 latex kits. Antimicrobial susceptibility tests against 12 drugs revealed a low resistance rates to ampicillin (8%), amoxicillin-clavulanic acid (0.5%), ciprofloxacin (1.4%), enrofloxacin (2.3%), gentamicin (0.5%), nalidixic acid (3.3%), and trimethoprim-sulfamethoxazole (5.6%). High resistance rates to streptomycin (38%) and tetracycline (36%) was observed among the E. coli isolates. The susceptibility rate to ampicillin, amoxicillin-clavulanic acid, cefotaxime, ceftazidime, ciprofloxacin, enrofloxacin, gentamicin, imipenem, nalidixic acid and trimethoprim-sulfamethoxazole ranged from 86.9 to 100%. Resistance to two or more antibiotics was observed in 57 (27%) of the E. coli isolates recovered. Conclusion: This study showed that presently, mongooses in Grenada are neither a reservoir for the E. coli O157:H7 serotype nor for multiple drug resistant E. coli strains. Among the 213 non-O157:H7 E. coli isolates, resistance drugs other streptomycin and tetracycline was very low.


Small
Indian mongooses (Herpestes auropunctatus) were introduced from Asia into the West Indies including Grenada in the late 1800s to control rats in sugar cane plantations [1,2]. Mongoose populations in the West Indies expanded rapidly due to their high reproductive capacity, year-around availability of food and the lack of natural predators [3]. In Grenada, mongooses are now among the most abundant mammals on the island with an estimated population density of 6.6 mongooses/ha [3]. They frequently den in proximity to human settlements and livestock farms [3] and tend to prey on poultry. Mongoose also feed on a variety of prey including small mammals, birds, reptiles, amphibians, insects and other invertebrates [3,4], many of which are known reservoirs for bacteria pathogenic to humans.
Worldwide, mongooses are known as reservoir hosts for a wide variety of human pathogens including those causing rabies [1,[5][6][7][8][9], hepatitis E [10] and tuberculosis [11]. In Barbados, Salmonella and Campylobacter species were isolated from the small Asian mongoose [12]. In Grenada, mongooses have been shown to harbor multiple human pathogens including those causing rabies [2,6,7], salmonellosis [13] and campylobacteriosis [14]. Fresh fruits, nuts, and vegetables grown in open fields and orchards in Grenada are vulnerable to pre-harvest microbial contamination; a wild animal such as mongoose can contaminate plants directly through fecal deposition or indirectly via fecal contamination of water or soil. E. coli serotype O157:H7 is the most important food-borne, zoonotic pathogen in humans [15,16]. It has also emerged as the most important enterohaemorrhagic (EHEC) serotype associated with human disease [16]. Its importance as a public health problem was recognized in 1982, following an outbreak in the  [35] and green iguanas [36].
The importance of mongooses for the epidemiology of E. coli O157:H7 is unknown in Grenada. In addition, there have been no published surveys on the prevalence and antimicrobial susceptibility of E. coli recovered from mongooses in Grenada. The present study was conducted to determine: the occurrence of the E. coli including the O157:H7 serotype in mongooses and their antimicrobial drug resistance in Grenada. John's. The methods used for anesthesia and euthanasia of the animals were previously described by Zieger et al. [6]. The animals were anesthetized within two hours of collection via intra-muscular injection with ketamine (10 mg/kg) (Ketamine hydrochloride, Rotexmedica ® , Trittau, Germany) and xylazine (0.2 mg/kg) (AnaSed®, Decatur, IL) and then euthanized via cardiac injection of potassium chloride (1-2 mmol/kg) (FisherScientific, Fair Lawn, NJ). A section of large intestine including luminal contents was removed between the ileo-cecal junction and the rectum and stored in a stomacher bag at −80°C till tested. All samples were processed within 18 months of collection.

MATERIALS AND METHODS
For the isolation of E. coli, each intestinal section (approximately 6-7 g including contents) was thawed and 10 mL tryptic soy broth (TSB) (BBL, Becton, Dickinson and Company, Sparks, MD, USA) were added to the stomacher bag. The contents were blended using a stomacher machine (Seward Limited, Worthing, West Sussex, UK). The entire suspension was transferred to a capped culture test tube (polystyrene with polyethylene cap, Fisher Scientific, Fair Lawn, NJ, USA) and incubated at 37°C for 24 hours. After incubation, an aliquot was streaked onto MacConkey (MAC) agar (Remel, Lenexa, KS, USA) and incubated at 37°C for 24 hours. To increase the chances of identifying E. coli O157:H7 in a sample, three (3) pink to red color colonies with or without a zone of precipitated bile morphologically representing E. coli were subcultured via streaking onto individual MAC agar and incubated at 37°C for 24 hours for isolation of pure colonies. Colonies from the second MAC agar plate were Gram stained and further tested using the API20E (Analytical Profile Index; BioM6rieux, Hazelwood, MO) bacterial identification strips for confirmation as E. coli. Non-lactose fermenting isolates identified as E. coli by API20E were also added in the study despite the fact that they were nonlactose fermenting variants.
For detection of E. coli serotype O157:H7, the pure colonies were first plated on sorbitol-MacConkey agar (Remel, Lenexa, KS, USA) and incubated at 37°C for 24 hours. After incubation, the sorbitol-MacConkey agar plates were examined for the presence of nonsorbitol fermenting colonies. All the colonies (both the sorbitol and the nonsorbitol fermenting colonies) were then subjected to slide agglutination using two (2) E. coli O157:H7 latex kits: Remel Wellcolex* E. coli O157 Rapid Latex Test (Remel Europe Ltd, Kent, UK) and Prolex TM E. coli O157 Latex Kit (Pro-lab Diagnostics, Toronto, Canada). Two latex kits were used to reduce the possibility of obtaining a false positive result. An isolate giving a positive reaction to both latex test kits was considered to be E. coli O157:H7 positive.
The antimicrobial susceptibility tests were carried out using the disc diffusion method as recommended by the Clinical and Laboratory Standards Institute (CLSI) using Mueller Hinton agar (BBL), and the inhibition zone sizes were

RESULTS
Of the 156 mongooses examined in this study, 71 (46%) were culture positive for E. coli. Of the 213 E. coli isolates, only 8 (4%) were nonsorbitol fermenters. However, none of the 213 isolates gave a positive reaction (O157agglutinating) to the two E. coli O157:H7 latex kits.
The results of the antimicrobial susceptibility test of the 213 non-O157 E. coli isolates are shown in Table 1. The results revealed a low resistance rate (0.5 to 8%) for all the 213 E. coli isolates to seven of the tested antibiotics except to streptomycin (38%) and tetracycline (36%). However, a number of the non-O157 E. coli isolates revealed a moderate to high intermediate resistance rate (1.4 to13.6%) to some of the tested antibiotics. The E. coli isolates recovered in this present study revealed a high susceptibility rate (86.9 to 100%) to ampicillin, amoxicillin-clavulanic acid, cefotaxime, ceftazidime, ciprofloxacin, enrofloxacin, gentamicin, imipenem, nalidixic acid and trimethoprim-sulfamethoxazole. Twenty seven percent (57 out of 213) of the E. coli isolates in this present study showed resistance to two or more antibiotics.
The results of our study showed that mongooses in Grenada harbor E. coli in their gastrointestinal tract, but do not harbor E. coli O157:H7 serotype. No information was found concerning the occurrence of E. coli O157:H7 in mongooses in other countries. Nevertheless, the observation in our study is not surprising as generally, wildlife with the exception of deer, does not constitute a significant source of E. coli O157:H7 serotype [21, 47,49]. Rice et al. [47] reported a low prevalence rate of E. coli O157 in white-tailed deer (<1%, 5/630), pooled bird of unknown species (<1%, 1/296), and pooled flies (3.3%, 2/60).
Studies have shown that pathogenic non-O157 E. coli serotypes such as the O26, O103, O145, O172, O174, O113 and O111, which are nonsorbitol fermenters that give a negative reaction to E. coli O157:H7 latex kits have been associated with infections in humans [50,51]. In this current study, 8 (3.8%) out of the 213 E. coli isolates were non-sorbitol fermenters that gave a negative reaction to the two E. coli O157:H7 latex kits. This study targeted only the E. coli O157:H7 serotypes which are typically nonsorbitol fermenters that give a positive reaction to the E. coli O157:H7 latex kits. In comparison with other studies, the survey carried out on wildlife in Trinidad and Tobago, showed the presence of non-sorbitol fermenting E. coli that were not agglutinated by O157 antiserum from: freeranging mammals (<1%), captive wild animals (2%) and animals in a zoo (0.5%) [38]. Sylvester et al. [36] reported a 29% (12 out of 42) prevalence rate of non-O157 E. coli serotypes in green iguanas in Grenada that were non-sorbitol fermenters that gave a negative reaction to the E. coli O157:H7 latex kits. This indicates that both wild and domestic animals in Grenada and other parts of the world may be potential reservoirs for the non-O157 group. Further studies that targets the non-O157 group is required to determine whether the 8 (3.8%) non-O157 E. coli isolates recovered in this study are among the pathogenic non-O157 group. This information will enable us to determine whether mongooses in Grenada are reservoirs for the pathogenic non-O157:H7 E. coli groups and may possibly be a source of environmental contamination.
The antibiotics susceptibility assays of this present study revealed low resistance rate ranging from 0.5% to 8% of all the 213 E. coli isolates to seven out of the 12 antibiotics tested (Table 1). This is somewhat similar to the findings of Sylvester et al. [36] who tested non-O157:H7 E. coli isolates from green iguanas to similar types of antibiotics. Their results revealed low resistance rates to ampicillin (7%), nalidixic acid (2%), and trimethoprim-sulfamethoxazole (2%) which are similar to the resistance rates to ampicillin (8%), nalidixic acid (3.3%), and trimethoprim-sulfamethoxazole (5.6%) observed in this current study. However, resistance rate of 12% to amoxicillin-clavulanic acid observed by these authors [36] was higher than the 0.5% resistance rate to amoxicillin-clavulanic acid observed in the current study. The authors [36] also observed a resistance rate of 2% to cefotaxime which is contrary to the zero resistance to cefotaxime observed in our study. A study on young healthy pigs in Grenada by Amadi et al. [52], revealed a low resistance rates to ampicillin (3%), amoxicillin-clavulanic acid (1%), trimethoprim-sulfamethoxazole (3%), and a zero resistance rate to cefotaxime, which closely corresponds to the findings of this current study.
The low resistance rates to ciprofloxacin (1.4%), enrofloxacin (2.3%), and gentamicin (0.5%) observed in this current study is contrary to the zero resistance rates to the same antibiotics reported by Sylvester et al. [36] and Amadi et al. [52]. In this current study, we observed a zero resistance rate to ceftazidime and imipenem which concurred to findings of these authors [36,52]. The resistance to streptomycin (38%) and tetracycline (36%) were the highest resistance rates observed in this present study. This is in contrast to the resistance rates to streptomycin (12%) and tetracycline (2%) reported by Sylvester et al. [36]. On the other hand, the high resistance rate to tetracycline (36%) observed in this present study somewhat concurred to the higher resistance rates of 100% and 96% to tetracycline observed in pigs in Grenada by Sabarinath et al. [53] and Amadi et al. [52], respectively. It is possible that the mongoose E. coli in our study acquired resistance to streptomycin and tetracycline via transmissible plasmids from human and animal strains because of the closeness of the habitats of mongooses to human settlements and livestock farms.
Some of the non-O157:H7 E. coli isolates recovered in this present study showed a moderate to high intermediate resistance rates (1.4 to 13.6%) to some of the tested antibiotics with the highest rates being to streptomycin (13.6%). This is similar to the 14% intermediate resistance rates to streptomycin observed in green iguanas [36]. Fifty seven (27%) out of the 213 non-O157:H7 E. coli isolates recovered in this present study were resistant to two or more antibiotics. This is of public health concern since these multiple antibiotic resistant organisms can be transmitted from animals to humans.

CONCLUSION
This current study showed that presently, mongooses in Grenada are not reservoirs for the E. coli O157:H7 serotype. It also indicates that the tested mongooses are presently not major reservoirs for multiple resistant E. coli strains. Among the 213 non-O157:H7 E. coli isolates, the resistance rate to drugs other than streptomycin and tetracycline was very low.