Abstract: Background: Escherichia coli is one of the most common causative agent of bacterial diseases. The prevalence of E. coli infection and its outbreak has been reported in many countries but scanty information is available on the antibiogram of this bacterium in Cape Coast, a major tourist destination in West Africa. Materials and Methods: The antibiotic susceptibility and occurrence of Escherichia coli isolated from clinical and environmental samples from the Cape Coast Metropolis were investigated. Bacteria isolation and identification were carried out using various bacteriological media and Analytical Profile Index (API) 20 E kits, respectively. All the test E. coli isolates were screened for their susceptibility to 16 antibiotics. Results: In all, 389 E. coli isolates were obtained comprising 261 and 128 from clinical and environmental samples respectively. All E. coli isolates were 100% sensitive to Imipenem. The percentage sensitivities of clinical E. coli isolates to ampicillin (0-24.1%), tetracycline (16.0-28.4%), cotrimoxazole (16.8-22.0%), cefuroxime (27.6-43.2%) and nalidixic acid (22.1-47.8%) were found to be relatively low. The sensitivity of environmental isolates to the 16 antibiotics was higher than that of clinical isolates, except for nalidixic acid, aztreonam and amikacin to which isolates from environmental samples were less sensitive. Among clinical samples, stool samples had the highest percentage occurrence of E. coli (44.4%), followed by urine samples (36.4%); while blood had the lowest percentage (1.6%). Among environmental samples, the highest percentage occurrence of E. coli (44.2%) was recorded for the Fosu lagoon, followed by fresh beef samples (21.8%) and fresh chicken (4.7%). No isolates were obtained from seawater, cabbage and smoked fish. Conclusion: Escherichia coli isolated from environmental samples were more sensitive to most antibiotics used in this study compared to clinical isolates. This study has demonstrated that environmental samples may harbour some considerable level of antibiotic resistance and hence underscores the need to dispose off waste properly.
INTRODUCTION
Infectious agents, including bacteria and fungi have been the cause of numerous diseases and outbreaks worldwide1. Escherichia coli is known to be one of the most common causative agents of bacterial diseases2. The prevalence of E. coli infections and outbreaks has been reported in many countries including South Africa3, Trinidad and Tobago4 and Ghana5,6. Over the years, E. coli infections have been managed through rehydration and adequate oxygenation or by the administration of antibiotics. Antibiotics are useful in various ways. Aside the benefits derived from their use in managing many infectious diseases, they have been tremendously adopted as prophylaxis, especially in animal husbandry.
Globally, however, the antibiotic resistance among E. coli strains has been increasing since the 1970s7-10. Surveillance studies on antibiotic resistance of E. coli in Ghana are few and mostly focused on the major cities, such as Accra, Kumasi and Tamale11, leaving out equally important cities of the country especially Cape Coast, a very popular tourists destination in West Africa. The present study, therefore, aimed at identifying E. coli isolates from clinical and environmental samples from the Cape Coast Metropolis of the Central Region of Ghana and to provide data on the antibiogram and occurrence of these isolates as a means to generate valuable information that would help in the prevention, prognosis and treatment of E. coli related infections.
MATERIALS AND METHODS
Study area and time frame: The study was conducted in the Cape Coast Metropolis (5°6'0” N, 1°15'0” W) and its environs and involved the collection of clinical and environmental samples (Fig. 1) from June 2010-May 2012.
| Fig. 1: | Geographic location of sites where samples were collected for the study |
Ethical considerations: Ethical approval to undertake the present study was obtained from the Ghana Health Service Ethical Review Committee on Research Involving Human Subjects. Informed consent forms were signed or thumb-printed by every patient or by the parent/guardian of the patient before participation in the study.
Sample collection: Clinical samples were obtained from patients reporting or referred to the Microbiology Section of the Cape Coast Teaching Hospital with various infections. Clinical samples were collected in accordance with the guidelines provided by Cheesbrough12. Various clinical samples, namely: Stool, urine, wound aspirate and high vaginal swab samples were obtained from the patients.
Environmental samples comprised water and food samples. Briefly, 100 mL of water samples were collected in sterile wide-mouth bottles with dust-proof ground glass stoppers from each collection site (gutters/drainages around the Anaafo and Abura markets, the Kakum river and the Fosu lagoon). These were immediately placed in an ice-chest and transported to the laboratory for processing and microbial analysis.
In the case of food samples, cabbage heads were purchased from three different locations within the Cape Coast Metropolis, namely: The School of Agriculture Research Farm, University of Cape Coast and the Abura and Kotokuraba markets. About 2 g of outer leaves were transferred into 2 mL sterile normal saline and vigorously shaken to dislodge all bacteria into solution, which was later used for bacteriological analysis.
Smoked fish, fresh chicken, fresh beef were purchased from three markets of the Cape Coast Metropolis-the Abura, Kotokuraba and Anaafo markets and immediately placed into sterile polythene bags, which were subsequently deposited into an ice-chest and transported to the laboratory within 2 h of collection. A portion of each sample of smoked fish, thawed frozen chicken or fresh beef (1 g) was placed into a sterile translucent polythene bag and pulverized with 9 mL of sterile buffered peptone water for bacteriological analysis.
Bacteriological analysis: Depending on the type of clinical sample, bacterial isolations were made on four types of microbiological media using the streak method: Cystine Lactose Electrolyte Deficient (CLED) for urine samples, Xylose Lysine Deoxycholate agar (XLD) for stool samples, Tryptone soya (tryptic soy) diphasic medium for blood samples and MacConkey Agar for all other clinical samples. Incubation in all cases was done aerobically at 35°C for 12-18 h except for blood samples, where cultures were incubated for up to 7 days or earlier depending on how early growth was observed12.
In the case of water samples, serial decimal dilutions were made and suspensions were used as inocula for microbial analysis. The pour-plate method was adopted using MacConkey agar as the medium of choice and incubation was done aerobically at 35°C for 12-18 h.
A loopful of each suspension obtained from smoked fish, thawed frozen chicken and fresh beef were streaked on MacConkey agar plates which were also incubated aerobically at 35°C for 12-18 h.
For both clinical and environmental samples, presumptive E. coli colonies were selected and sub-cultured onto fresh sterile MacConkey agar plates. Presumptive E. coli are isolates that formed red or yellow colonies on MacConkey or CLED agar, respectively or isolates that formed yellow or white/cream colonies on XLD agar or Tryptic soy Diphasic medium, respectively. After overnight incubation, single colonies were transferred onto sterile nutrient agar slants and incubated at 35°C for 12-18 h. The pure cultures obtained after incubation were stored at 4°C for subsequent identification.
All presumptive E. coli isolates recovered were further identified and confirmed using strips of Analytical Profile Index (API) 20 E (Biomérieux, France) according to the manufacturer’s instructions. All confirmed E. coli isolates were then stored in 15% glycerol Nutrient broth (v/v) at-40°C for further analysis. To avoid duplication of clinical isolates, a single confirmed E. coli isolate per clinical sample was selected and stored. For environmental samples, however, three or four confirmed E. coli isolates per sample were selected and stored for further analysis.
Antibiotics susceptibility testing: The Modified Kirby-Bauer disk diffusion method12 was used to evaluate the susceptibility or resistance of E. coli isolates to 16 selected antibiotics. These were Ampicillin (10 μg), tetracycline (30 μg), cotrimoxazole (25 μg), gentamicin (10 μg), cefuroxime (30 μg), chloramphenicol (30 μg), ceftriaxone (30 μg), cefotaxime (30 μg) nitrofurantoin (300 μg), nalidixic acid (30 μg), aztreonam (30 μg), imipenem (10 μg), ciprofloxacin (5 μg), ceftazidime (30 μg), amikacin (30 μg), norfloxacin (10 μg).
Bacterial suspensions of each test E. coli isolate were then prepared using sterile normal saline and adjusted to the 0.5 McFarland turbidity standard.
Using a sterile cotton swab, each suspension was streaked unto the entire surface of a sterile Müeller Hinton agar plate thereby, providing bacterial lawns with confluent growth.
| Table 1: | Susceptibility profile of E. coli isolated from various samples to 16 antibiotics |
*Other clinical samples include blood, wound swab, aspirate and high vaginal swab samples, Amp: Ampicillin, Tet: Tetracycline, Cot: Co-trimoxazole, Gen: Gentamicin, Crx: Cefuroxime, Chl: Chloramphenicol, Ctr: Ceftriaxone, Ctx: Cefotaxime, Imp: Imipenem, Caz: Ceftazidime, Nit: Nitrofurantoin, Nal: Nalidixic acid, Azt: Aztreonam, Nor: Norfloxacin, Cpr: Ciprofloxacin, Amk: Amikacin | |
The inoculated plates were allowed to dry for approximately 10 min and the antibiotics disks placed on their surfaces after which the plates were incubated aerobically at 35°C for 16-18 h. Following the incubation period, the diameters of zones of inhibition (clear zones) that developed around each disk were measured using vernier calipers and recorded. The measurements were interpreted using guidelines provided by the Clinical Laboratory Standards Institute13. All isolates tested were classified as sensitive, intermediate or resistant to the various antibiotics used. All isolates that fell within the intermediate category were retested and subsequently, classified as sensitive or resistant.
Standard-typed strains of Escherichia coli ATCC 25922 and Klebsiella pneumoniae ATCC 700603 were screened at least once a week as a quality control measure.
Analysis of data: Data were analyzed using the SPSS version 16.0. The differences in proportions and statistical significance were assessed by chi-square tests and a p-value less than 0.05 was considered statistically significant.
RESULTS AND DISCUSSION
The sensitivity pattern of the E. coli isolates to the 16 antibiotics varied depending on the sample from which they were isolated. The percentage antibiotic sensitivity of E. coli isolates of environmental origin was higher than that for clinical samples (Table 1), except for the aztreonam antibiotic. The percentage sensitivity of environmental E. coli isolates to aztreonam was 43.8% compared to 47.4, 60.0 and 80.2%, respectively, for stool, urine and other clinical samples. Environmental E. coli isolates were also less sensitive to nalidixic acid and amikacin compared to E. coli isolates from clinical samples. In many African countries with high rates of infectious diseases, formal and informal health systems depend heavily on broad-spectrum orally administrable antibiotics14. Antimicrobial drug-resistant E. coli from human faeces and blood stream infections tend to be more similar to antimicrobial-resistant and susceptible E. coli from retail poultry meat sources15. These observations indicate that the selection of resistant E. coli is more likely to occur in animal food reservoirs than in humans16. In most developing countries, the use of sub-therapeutic doses that are often of substandard quality17 and the partial adherence of patients to antibiotic treatment doses are among factors that contribute to the emergence and fast spread of antibiotic-resistant strains18,19. The current study showed a low sensitivity of the isolates recovered from clinical samples to ampicillin (0-24.1%), tetracycline (16.0-28.4%), cotrimoxazole (16.8-22.0%), cefuroxime (27.6-43.2%) and nalidixic acid (22.1-47.4%). These findings are similar to the results of other studies11,20,21. However, Newman et al.22 found bacterial isolates, including E. coli, to be highly resistant to chloramphenicol instead of cefuroxime as observed in this study. Similar levels of resistance to chloramphenicol were recorded in studies conducted on children in Northern Ghana19 and the Accra Metropolis23.
| Table 2: | Prevalence of E. coli in clinical samples |
| *: p<0.01 | |
| Table 3: | Age and gender distribution of patients from whom clinical samples were obtained and E. coli recovered |
| *: p<0.01, **: p<0.01 | |
Ampicillin, tetracycline, cotrimoxazole and chloramphenicol have been abused over a long period of time in Ghana as they are easily accessible without prescription and are relatively cheap22.
The occurrence of E. coli isolates of clinical origin is shown in Table 2. Out of 382 clinical samples collected, 261 E. coli isolates were recovered with the stools samples having the highest percentage of isolates (44.4%) and blood samples with the least (1.6%). The 261 clinical E. coli isolates were recovered from 73 male and 188 female patients (Table 2). Among the different age groups, females of 20-49 years had significantly (p<0.05) higher frequencies of E. coli isolates than their male counterparts (Table 3). However, for those aged above 70 years, a higher frequency of E. coli isolates were found among the males compared to females (p<0.01). This could be explained by the fact that females aged 20-49 years are sexually active; thus, making them vulnerable to urinary tract infections, while males aged above 70 years are prone to urinary tract recurrent infections23,24, which could account for the differences observed in this study. However, sample collection was not linked to medical condition brought to the hospital. Nevertheless, the higher proportion (72%) of females, among the patients screened in this study was similar to that observed in other studies conducted in Ghana20, which attributed the difference to the fact that women report to hospitals more frequently than men.
In contrast to clinical samples, only 128 E. coli isolates were recovered from the 92 environmental samples screened with more than one isolate obtained from some samples, while no isolates was found in others (Table 4).
| Table 4: | Prevalence E. coli in environmental samples |
The highest percentage of E. coli isolates was obtained from the Fosu Lagoon (42.2%) while the lowest was obtained from the Kakum River and the chicken samples screened (4.7%). No E. coli isolate was obtained from seawater, cabbage and smoked fish.
In many Ghanaian and other West African markets, smoked fish is sold mainly in the open exposing it to dust, houseflies and other insects. In this study, smoked fish was found devoid of E. coli probably because the two types of smoked fish sampled (Atlantic horse mackerel and sardines) present an oily surface, which is not conducive for the successful growth and multiplication of most aerobic non-lipolytic bacteria. The current observations are similar to findings in earlier studies in Ghana and elsewhere, where other bacterial species were isolated from various fish samples but no E. coli was isolated25-27. Isolation of E. coli indicates faecal contamination linked to unhygienic handling practices as well as the length of the period the fish has been exposed to the open air. Hence, the smoked fish sampled in this study were either fresh or handled properly or both. The lack of E. coli isolates from seawater is not surprising, since most organisms that thrive in seawater are halophiles and thus, can tolerate the salinity and other conditions prevailing in seawater. Most enteric bacteria, however have a low tolerance to high salinity and when released into the sea, enteric bacteria are subjected to an immediate osmotic up-shock, which negatively affects their subsequent survival28,29. Although no E. coli isolates were obtained from cabbage samples in this study, other closely related bacterial species, including Klebsiella pneumoniae, Klebsiella oxytoca, Enterobacter spp. and Citrobacter freundii were isolated.
Among the environmental samples that tested positive for E. coli, the highest number of E. coli isolates (42.2%) was obtained from water samples from the Fosu Lagoon (Table 4). This suggests faecal contamination, which could be attributed to the high level of human activities around the lagoon. Moreover, the inadequate drainage system in the vicinity of the lagoon as domestic wastes from households are usually collected in gutters that are connected directly or indirectly to the Fosu Lagoon could partly contribute to this high prevalence of E. coli. The Kakum River, on the contrary had the lowest number of E. coli isolates (4.7%), probably, because of the low level of human activity around the site of collection. Indeed, the level of human and animal activities as well as human settlements around the point of sample collection was low compared to that of the Fosu Lagoon. The gutters around the Anaafo and Abura markets had a considerable number of E. coli isolates (Table 4), indicating some level of faecal contamination at these sites as well. A higher proportion of E. coli isolates (21.8%) was obtained from fresh beef compared to the chicken samples (4.7%). During the sample collection, it was observed that chicken was almost sold frozen, whereas fresh beef was usually sold at butchers’ shops where the sanitary conditions were usually unsatisfactory. Indeed, in most butchers’ shops in Cape Coast, the meat was usually displayed on tabletops, where they stayed throughout the day exposed to unsanitary conditions that could lead to contamination.
CONCLUSION
All clinical and environmental isolates were 100% sensitive to Imipenem. Escherichia coli isolated from clinical and environmental samples varied in proportions with some samples having higher percentage occurrence than others. However, for environmental samples, no E. coli was isolated from seawater, cabbage and smoked fish samples screened. The E. coli isolated from environmental samples were more sensitive to 12 out of the 16 antibiotics used in this study, compared to clinical isolates. Clinical isolates also were more sensitive to 3 antibiotics: Nalidixic acid, aztreonam and amikacin than environmental isolates. The widespread resistance of E. coli to commonly available and used antibiotics may have serious health implications for the treatment of E. coli-related infections in Ghana.
SIGNIFICANCE STATEMENTS
| • | This is one of the few studies where bacterial isolates from environmental samples are compared to isolates from clinical samples |
| • | This study provides baseline information on the antibiotic susceptibility profiles of E. coli isolated from clinical and environmental samples in the Cape Coast Metropolis of the Central Region of Ghana |
| • | Data obtained in this study provide valuable information that could help in the prevention, prognosis and treatment of E. coli-related infections in Cape Coast in particular, Ghana and elsewhere in the world |
| • | Findings of this study reiterate the need for proper handling of food and waste disposal as resistant E. coli isolates were found in both clinical and environmental samples |
ACKNOWLEDGEMENTS
Authors are thankful to Dr. Noah Obeng Nkrumah, Department of Microbiology, University of Ghana Medical School for providing the standard-typed strains used in the study. They are also grateful to all patients from whom clinical samples were obtained for the study. Funding for the study was provided by University of Cape Coast, Cape Coast, Ghana.