Vibrio cholerae is the causative agent of cholera, a diarrheal disease with epidemic or pandemic potential. This enteric disease is characterized by profuse watery diarrhea and vomiting, resulting in dehydration, electrolytes loss and eventually hypovolemic shock and renal failure. Without prompt medical attention, death can occur within hours [1] in 30 to 40% of cases. Cholera remains a public health concern particularly in developing countries with lack or inadequate supply of potable water, poor hygiene and rudimentary sanitary facilities. Transmission is mainly by consumption of contaminated food and water. Thus, good hygiene, appropriate sanitation and safe water are of prime importance in the fight against cholera.

Vibrio cholerae is classified into more than 200 serogroups based on the somatic O antigen. These are grouped into three major groups: Vibrio cholerae O1, Vibrio cholerae O139 and Vibrio cholerae non-O1/non-O139. Epidemic cholera is caused by Vibrio cholerae O1 and O139 strains. Strains of the non-O1/non-O139 serogroup have been associated with sporadic cases of diarrhea disease [2, 3]. The epidemic potential of the organism is conferred by the production of a potent cholera toxin (CT) and an adhesion factor toxin-co-regulated pilus (TCP). However, ctxAB and tcp genes which code for these major virulence determinants have also been identified in environmental strains of some non-O1/non-O139 Vibrio cholerae [4, 5], and their presence have been attributed to serogroup conversion through gene transfer mechanisms such as natural transformation [6] and transduction [7]. Thus for an effective diarrheal disease prevention programme in cholera endemic localities, it is important to include the surveillance for non-O1/non-O139 strains in environmental samples.

V. cholerae is autochthonous to the aquatic environment explaining its detection in diverse aquatic environments (brackish water, marine and fresh water) and the occurrence of sporadic outbreaks of cholera in non-endemic localities has been associated with floods contaminating water sources or imported cases. It can survive in some aquatic environments for months to years, in association with zooplankton and other aquatic organisms [8]. Both laboratory and field studies [9, 10] have shown that its occurrence in the aquatic habitat is influenced by physico-chemical characteristics. Under adverse environmental conditions, it has been shown to exist in a viable but non-culturable state which could revert to a transmissible state when climatic conditions become favourable, suggesting that in endemic regions, the aquatic environment may serve as a reservoir for this pathogen in the absence of an outbreak of cholera. Thus a deep understanding of these factors would facilitate disease prediction and implementation of timely measures for prevention of outbreaks.

Treatment of cholera involves prompt rehydration to replace lost fluids and electrolytes, and administration of antibiotics. Although rehydration prevents death, antibiotics administration has been very crucial in reducing the shedding of the pathogen, preventing disease spread, treating severely ill patients by reducing the volume of diarrhea, and consequently, reducing the duration of illness and hospitalization. Like for most enteric pathogens, there are reports of changing antibiotic susceptibility pattern of Vibrio cholerae isolated from cholera cases in endemic regions due to the emergence and spread of multidrug resistant strains [11]. Worrisome also, are reports of Vibrio cholerae strains form environmental sources bearing R plasmids transferable by intraspecific and intergeneric matings [12]. These findings show that the organism is capable of acquiring or spreading resistance factors and thus the aquatic environment could serve as a reservoir for drug-resistant strains. Therefore in cholera endemic regions, surveillance of antibiotic susceptibility patterns of the organism must include isolates from aquatic sources to identify effective agents for treatment so as to avert the unpleasant consequences of treatment failure.

Cholera has been endemic in Douala, the economic capital of Cameroon since 1971 [13] when the on-going 7^th pandemic hit the African continent. Outbreaks have been occurring almost every two years and have often started in Bepanda [14], a slum area with poor hygiene and sanitary conditions. Thus, as inhabitants of Douala are constantly threatened by outbreaks of cholera, it is important to study the reservoirs (particularly during a period with no disease outbreak) and environmental factors maintaining the endemicity of Vibrio cholerae in this region. This study was aimed at isolating and characterizing V. cholerae strains from water sources in Bepanda, determining the susceptibility of isolates to antibiotics and evaluating some physico-chemical parameters that could be supporting the survival of the organism in this locality, so as to generate information that could be exploited for cholera prevention and management.

The city of Douala is well known as the main cholera-endemic area of Cameroon. It has experienced more than ten outbreaks of cholera since the arrival of the disease in Cameroon in 1971, most of which have originated from Bepanda [13, 14]. We analyzed water from various sources in Bepanda during a period with no known disease outbreak for the presence of V. cholerae, studied its susceptibility pattern to drugs used in cholera management and also determined the physico-chemical parameters associated with the presence of the organism in water in the study locality.

Eighty-seven samples (27.4%) were contaminated with Vibrio cholerae. V. cholerae is a normal inhabitant of natural aquatic environments such as rivers, estuaries and coastal waters [26] and has been detected in natural waters worldwide, including areas where clinical cases of cholera did not exist [27]. Samples from the stream were more contaminated. There was a significant difference in the prevalence of the organism in samples from various sources (Table 1). Most isolates were from wells. This is not surprising because recent reports [16] are showing deteriorating quality of water from dug-wells in study area. We could not access some wells and taps which were in premises with a gated fence since at times, there was no one within the premises to open the gate and let us in. The following wells were not sampled during the following months: March: W₁₃, W₁₆, W₂₅, W₃₇, W_40; April: W₁, W₃, W₉, W₁₈, W_21, W_45; May: W₁₀, W₁₅, W₂₄, W₂₈, W₃₃, W_41, W_42; June: W₁₂, W₇, W₈, W₁₄, W_32; July: W₁₁, W₂₂, W₃₀, W₃₉, W₄₃, W_44. Initially we had 6 sampling points on the stream. After the first visit, access way to three of those points was blocked and we never sampled from there again. We also had more taps on the first visit but subsequently we were not able to sample from some of them.

Serological characterization of V. cholerae isolates identified 32 (33.3%) and 64 (66.7%) isolates belonging to serogroups O1 and non-O1/non-O139 respectively (Table 2). Both serogroups were isolated in the dry and rainy seasons (Table 3) indicating the presence of the organism all year round in study site. However, the majority of isolates were obtained during the rainy season. In Douala, the rainy season is characterized by torrential rains, which result in flooding and runoffs loaded with contaminants from diverse origins to water sources. The majority of the wells in Bepanda are poorly constructed and not well protected [16]. During periods of flood water may fill pit latrines and re-distribute faecal matter resulting in contamination of wells and surface water sources. Improvement of sanitation and access to safe water sources are urgently needed in Bepanda. In addition, continuous monitoring of water sources in study area is necessary for a better preparedness and control of cholera and cholera-like disease.

V. cholerae non-O1/non-O139 was isolated from all sources. Although non-O1/non-O139 strains do not cause epidemic cholera, they are recognized to be of public health relevance because they have been associated with sporadic cases or outbreaks of cholera-like diseases [3, 28] and many extra-intestinal infections [2, 29, 30]. The emergence of serogroup O139 as a second etiologic agent of cholera epidemics, along with the possible conversion of non-O1 to O1 serotype [31, 32] has increased interest in non-O1/non-O139 V. cholerae strains. Members of non-O1/non-O139 serogroup do not produce major virulence factors but genes coding for virulence factors have been identified in certain strains of both clinical and environmental origins [4, 5]. Thus, non-O1/non-O139 serogroup in the study area may constitute a public health risk. Studies are underway in our laboratory investigating the virulence potential of non-O1/non-O139 Vibrio cholerae from the aquatic environment of parts of the city of Douala where devastating outbreaks of cholera had occurred to show their possible involvement in diarrheal disease in these areas. All isolates from tap water were serogroup non-O1/non-O139. Municipal water distributed in Douala undergoes adequate treatment prior to distribution. Contamination of tap water could have been through a leakage along the distribution mains which we observed during sample collection. Thus prompt repairs of leaking pipes his highly solicited to prevent re-contamination of treated water.

Serogroup O139 was not detected in our study. Since the detection of this strain in 1992 as another cause of epidemic cholera [32], cholera caused by serogroup O139 had been limited to the Indian sub-continent and thus was believed to be of no public health importance in Africa. However, du Preez et al. [33] detected V. cholerae O139 strain in the estuarine environment of Mozambique indicating its spread out of the Indian sub-continent. These findings call for the inclusion of V. cholerae O139 in cholera surveillance in cholera endemic areas where this strain has never been reported.

Serogroup O1 was isolated only from wells and stream samples (Table 2). This is of great public health importance as toxigenic strains are responsible of cholera outbreaks and epidemics. Although we did not investigate the potential of our isolates to cause disease, their detection in waters is expected to serve as warning against an impending outbreak of cholera if appropriate preventive measures are not enforced. Inhabitants of the locality studied, therefore run the risk of experiencing cholera outbreak in the future, since most of them use well water for drinking, cooking and washing. There was no outbreak of cholera during the period of our study.

Environmental factors such as temperature, salinity, rainfall, sunlight, pH, concentration of ferric ions, exogenous products of algal growth and chitin have been associated with V. cholerae dynamics [34]. A significant positive correlation was observed between temperature and occurrence of Vibrio cholerae in the various water sources in the dry season (+0.248, P = 0.030) (Table 4). Correlation was positive in the rainy season (+0.012, P = 0.910) though weaker and not significant. pH positively correlated with isolation of the organism in the dry season (+0.184) and rainy season (+0.09) but correlation was not significant in both seasons (P = 0.094 and P = 0.379 for dry and rainy season respectively). Salinity had a significant positive correlation with the occurrence of the organism in the dry season (+0.267, P = 0.015) as well as rainy season (+0.223, P = 0.028) showing salinity to be strongly associated with V. cholerae isolation from water sources in Bepanda. However, salinity did not score significantly in the logistic regression model. Salinity has been demonstrated to influence significantly the growth of V. cholerae in cholera endemic areas [35, 36]. We did not explore lagged relationships eventhough previous studies have shown significant associations. pH (Wald = 11.753, df = 1, P = 0.001) was the only significant predictor from the binary logistic regression. Comprehensive studies would be necessary to evaluate the influence of pH on the presence of V. cholerae in the aquatic environment of Bepanda.

Isolates exhibited 100% susceptibility to fluoroquinolones: ciprofloxacin and ofloxacin. Augmentin, doxycycline and chloramphenicol also showed high activities in the two serogroups while cotrimoxazole, tetracycline and amoxicillin exhibited the lowest activities (Table 5). Resistance of clinical isolates from Douala to these antibiotics has been reported by Garrigue et al. [37]. Such resistant strains might have been disseminated into water sources. Studies in other parts of the world [38, 39] have reported varying susceptibilities of the organism to these antibiotics. Changes in susceptibility pattern have been attributed to isolation time, source and geographical location. However, high activities of ofloxacin and ciprofloxacin observed in our study are similar to other studies [15, 38, 39] and corroborate the use of fluoroquinolones as new alternatives for cholera treatment in some cholera endemic regions [40]. Our data and reports from other cholera endemic localities in Douala [20] therefore suggest ciprofloxacin and ofloxacin could be very good alternatives for chemotherapy during cholera or cholera-like disease outbreaks in Douala.

High resistance of O1 isolates to amoxicillin corroborates the report of Ngandjio et al. [20] on the emergence of resistance to amoxicillin toward the end of the 2005 cholera epidemic in Douala. Doxycycline and amoxicillin were extensively used during this outbreak as drugs of choice in chemotherapy and chemoprophylaxis [23, 24]. Prolonged usage of antibiotics for curative and prophylactic purposes during an epidemic can provoke the emergence and spread of resistant strains into the environment [41]. Overuse of antibiotics could also exert a selective pressure on pathogenic organisms and the bowel flora, which once shed, become an important reservoir of resistance genes that can be transferred to other bacteria such as V. cholerae. Genetic exchange has been reported to occur between V. cholerae and enterobacteriaceae in the environment such as E. coli that possess transposon and plasmid of the same group [42].

Four multi-drug resistance (MDR) patterns emerged among V. cholerae O1 isolates (Table 6). Isolates that were resistant to more than 3 drugs were not found. The patterns AML^R TE^R SXT^R and TE^R SXT^R were the most common resistance patterns in this serogroup. Akoachere et al. [20] recently detected multi-drug resistant toxigenic V. cholerae O1 in Douala. Ndayo et al. [15] reported the isolation of MDR O1 strains from water sources in Douala that exhibited resistance to up to five drugs: AML^R AUG^R TIC^R S^R C^R SXT^R, a pattern similar to that obtained from clinical isolates during the 2004–2005 cholera epidemic. Like O1 isolates, multi-drug resistance was observed in the non-O1/non - O139 isolates. Thirteen MDR patterns were obtained (Table 6). Similar to serogroup O1, the pattern AML^R TE^R SXT^R predominated. We also observed that most non-O1/non-O139 isolates were simultaneously resistant to cotrimoxazole and tetracycline. The isolation of V. cholerae that were resistant to cotrimoxazole suggests the possible presence of the SXT element in the study area. The SXT element is a self-transmissible, chromosomally integrated genetic element which carries cross-resistance to cotrimoxazole, streptomycin and furazoloidine [43]. However, we did not test streptomycin and furazoloidine in our study. The SXT element has been demonstrated to acquire additional tetracycline and erythromycin resistance genes with high efficiency [44]. Thus, tetracycline and cotrimoxazole resistance observed in isolates could be due to the SXT element. From our results and other findings in Douala [15, 20] it would be necessary to investigate drug resistant V. cholerae from Douala for these resistance markers. However, the simultaneous resistance to tetracycline and cotrimoxazole could be rather related to the presence of both SXT element and resistance (R) plasmid known to carry tetracycline resistance genes [45]. The high prevalence of MDR V. cholerae in the water sources indicates the need for constant monitoring of the susceptibility pattern of the organism in cholera-endemic areas, in anticipation of an outbreak of cholera or gastroenteritis. It is necessary that preparedness activities for chemotherapy should include an efficient scheme for the management of antibiotics chosen, as this influences the emergence and prevalence of novel antibiotic resistant clones in cholera endemic areas.

Our study has shown that V. cholerae is present all year round in wells and streams in Bepanda. Since wells are a major source of water in study area and the stream is used for various activities which may expose residents to infection, cholera will continue to be a health threat in this locality if intervention measures to prevent outbreak are not implemented. Among the intervention measures, this locality should be considered a target for oral cholera vaccine as an immediate measure while resources are put together for WASH (water, sanitation and hygiene) improvement.