Increased Isolation Frequency of Toxigenic Vibrio cholerae O1 from Environmental Monitoring Sites in Haiti

Since the identification of the first cholera case in 2010, the disease has spread in epidemic form throughout the island nation of Haiti; as of 2014, about 700,000 cholera cases have been reported, with over 8,000 deaths. While case numbers have declined, the more fundamental question of whether the causative bacterium, Vibrio cholerae has established an environmental reservoir in the surface waters of Haiti remains to be elucidated. In a previous study conducted between April 2012 and March 2013, we reported the isolation of toxigenic V. cholerae O1 from surface waters in the Ouest Department. After a second year of surveillance (April 2013 to March 2014) using identical methodology, we observed a more than five-fold increase in the number of water samples containing culturable V. cholerae O1 compared to the previous year (1.7% vs 8.6%), with double the number of sites having at least one positive sample (58% vs 20%). Both seasonal water temperatures and precipitation were significantly related to the frequency of isolation. Our data suggest that toxigenic V. cholerae O1 are becoming more common in surface waters in Haiti; while the basis for this increase is uncertain, our findings raise concerns that environmental reservoirs are being established.


Introduction
Epidemic cholera was first identified in Haiti in October of 2010, with an altered toxigenic V. cholerae O1 El Tor biotype responsible for the sudden outbreak [1,2]. At the beginning of this study in April 2013, over 650,000 cholera cases and 8,000 deaths had been reported to the Haitian Ministry of Public Health and Population (MSPP) [3]. Although the number of reported cholera cases continues to decline, cholera remains an important cause of morbidity and mortality, with almost 50,000 cases reported in Haiti between April 2013 and March 2014 [4]. The government of Haiti, in collaboration with the Dominican Republic and the Pan American Health Organization (PAHO) has developed a plan for eradication of cholera from Haiti by 2022, which relies mainly on strategic immunization and improvements in sanitation [5]. A recent model reported by Bertuzzo et al [6], which did not incorporate the ability of toxigenic V. cholerae to establish environmental reservoirs, suggests that while eradication is possible, the likelihood of eradication will remain low without carefully targeted and effective interventions. However, if V. cholerae were able to establish environmental reservoirs as it has in other tropical countries [7,8], eradication of the disease from Haiti would become much less likely, if not impossible. To assess the possible role of surface water in cholera transmission, and to explore whether V. cholerae has established environmental reservoirs in Haiti, our research group at University of Florida has monitored 15 fixed environmental sites in the Ouest Department of Haiti on a monthly basis since April 2012. The observations made during the initial year of surveillance from April 2012 through March 2013 documented the isolation of toxigenic V. cholerae O1 from 3 of 179 (1.7%) water samples [9]. In the current report, the results of a second year of environmental surveillance performed during April 2013 to March 2014 are presented.

Water sampling sites
As a component of our ongoing surveillance for the isolation of V. cholerae strains in the Gressier/Leogane regions in Haiti, we selected 15 environmental sentinel sites as described previously [9]. The original 15 sites in the Leogane flood basin included locations along transects of the Momance and Gressier Rivers (4 up-river sites and one estuarine site at the mouth of each river), the Tapion River (4 river sites), and an independent estuarine site at Four-a-chaux. In addition to the original 15 sites, two sites were added that include a populated market in Gressier (R2S5) directly after Gressier (R2S4), and the Reserved (Res), a site downstream from Bellvue that receives water from the Taipon River and a nearby spring. Water samples were collected from the 17 environmental sites during the last week of every month from April 2013 to March 2014. Water sampling (Field work) was conducted with the knowledge of health authorities at the Haitian National Public Health Laboratories. All collection sites were at locations that were open to the general public (rivers and shoreline areas); specific permissions were not required for collection of water samples from these sites. Field studies did not involve collection of any animals, endangered or protective species.

Isolation and characterization of V. cholerae from water samples
Water samples were collected from 17 environmental sites on a monthly basis from April 2013, through March 2014; a sterile Nalgene bottle (500 ml) was used to collect water from each site before transportation to the University of Florida field laboratory in Gressier for analysis. All samples were transported at ambient temperature and arrived within three hours of collection. Because of difficulty in accessing some sites at various time points due to seasonal fluctuations in water levels or road conditions, 197 water samples were available for analysis.
To isolate V. cholerae from collected water sample, a modified enrichment technique was used as described previously [9]. Briefly, three 1.5 ml aliquots from each water sample were enriched with equal volumes of 2X alkaline peptone water (APW) and incubated at 37°C for 6-8 h, 18-24 h at 37°C, and at 40°C for 6-8 h. Following enrichment, a sterile loop was used to streak cultures onto thiosulfate citrate bile salts sucrose (TCBS) agar (Becton, Dickinson and Company, NJ, USA) and the cultures were incubated at 37°C for 16-24 h. Six to eight yellow colonies from TCBS agar were transferred to L-agar plates and the plates were incubated at 37°C for 16-24 h. Colonies were screened for oxidase activity and oxidase-positive colonies were subjected to further screening using polyvalent V. cholerae O1-specific and O139 specific antisera (DENKA SEIKEN Co., Ltd., Tokyo, Japan) using slide agglutination assay.

Genetic characterization
For further characterization, each isolate was examined using polymerase chain reaction (PCR) to amplify ompW and toxR using V. cholerae species specific primers as previously described [10]. Chromosomal DNA was extracted, using a GenElute Bacterial Genomic DNA kit (Sigma-Aldrich, St. Louis, MO), from each isolate exhibited positive PCR results for ompW and toxR. DNA was then used as a template for PCR amplification for key virulence genes, including ctxA, ctxB, rstR, rstC, and tcpA for both classical and El Tor genotypes as described previously [10,11]. Convergent PCR primers used in this analysis are listed in S1 Table. A standard membrane filtration technique was used to enumerate the concentrations (colony forming units per 100 ml) of fecal coliform bacteria in the water samples [12]. During the collection of water samples, water temperature of each site was also recorded.

Antibiotic susceptibility test (AST)
Antimicrobial susceptibility test (AST) was performed for each isolate against a battery of antimicrobial agents using disc diffusion (Kirby-Bauer) assay and following the recommended guidelines by Clinical and Laboratory Standards Institute (CLSI, formally the National Committee for Clinical Laboratory Standards, NCLLS) as previously described [13]. Briefly, a small inoculum of each bacterial isolate was emulsified in 3 ml of sterile normal saline in 10 ml glass tube and the density of the bacterial suspension was compared with a barium chloride standard (0.5 Mcfarland). Subsequently, a sterile cotton swab stick was dipped into the standardized bacterial culture and the culture was evenly spread on the Mueller-Hinton plates and the plates were allowed to dry. Antibiotic discs (Oxoid, Basingstoke, England) with the following drug concentrations, including nalidixic acid (30 μg), tetracycline (30 μg), doxycycline (30 μg), amoxicillin (30 μg), ampicllin (30 μg), trimethroprim-sulfamethoxazole (cotrimoxazole) (25 μg), ciprofloxacin (5 μg), and chloramphenicol (30 μg) were placed on the plates. Discs were placed at least 15 mm apart and from the edge of the plates to prevent the overlapping of zones of inhibition. Plates were incubated at 37°C for 18-24 h, and the diameters of zones of inhibition of each antibiotic were compared with the recorded diameters of the control organism E. coli ATCC 25922 to determine susceptibility, intermediate or resistance pattern of the V. cholerae isolate to that antibiotic.

Statistical analysis
The water temperatures from all of the collection sites were averaged to obtain monthly estimates of the water temperature in degrees Celsius (°C). Daily estimates of rainfall in millimeters per day collected by the Tropical Rainfall Measuring Mission (TRMM) satellite were obtained from the National Aeronautics and Space Administration (NASA) for the area directly proximal to the environmental sampling sites (18.2-18.5°N, 17.1-17.4°W) and aggregated into weekly accumulated precipitation (mm/week) [14]. The association between the isolation of toxigenic V. cholerae O1 from individual water samples (presence/absence) and fecal coliform concentrations was assessed using conditional logistic regression with the site of collection (n = 17) as the grouping variable. A binomial regression model with time-lagged covariates was used to determine the association between the frequency of isolation of toxigenic V. cholerae O1 from all 17 collection sites and the monthly average water temperature and the weekly accumulated precipitation. The raw data used for the statistical analysis is available in S1 Dataset.

Results and Discussion
Between April 2013 and March 2014, toxigenic V. cholerae O1was isolated from 17 (8.6%) of 197 surface water samples and from 10 (59%) of the 17 sample collection sites. The date of isolation, unique strain identification number associated with the isolate, the abbreviated site name, the full name of the collection site, and the concentration of fecal coliform bacteria isolated from the same water sample are presented in Table 1   ( Table 2). Interestingly and in contrast to our first year survey [9], we observed that 12 (70.6%) of 17 isolates exhibited rstC ET genotype ( Table 2). As described in Methods section, we used diverse APW enrichment techniques in order to enhance increased V. cholerae isolation from water samples collected in Haiti. Data, presented in Table 3, demonstrated that if we were to use only conventional APW enrichment technique [15], we would have missed 6 (35%) of the 17 toxigenic V. cholerae strains reported in this investigation. Given the difficulty of the isolation of culturable V. cholerae from aquatic reservoirs [16], our modified APW enrichment approach played a crucial role for the increased isolation of toxigenic V. cholerae in Haiti.
In order to determine the susceptibility of V. cholerae isolates to antibiotics, each V. cholerae strain was subjected to antibiotic susceptibility test to a battery of antibiotics (Table 4). All 17 isolates exhibited identical response to each antibiotic examined. Environmental Haitian V. cholerae O1 strains were resistant to nalidixic acid and SXT (cotrimazole) and showed reduced susceptibility to ampicillin and chloramphenicol. However, they were sensitive to doxycycline, tetracycline and ciprofloxacin. Our data are similar to results presented in a previous study where Haitian clinical strains were subjected to antibiotic susceptibility assay [17].
In contrast to the previous year, where V. cholerae O1 was exclusively isolated in estuarine sites at the mouths of the Taipon and Momance Rivers (Lassale, Jeffra, and Gressier Beach); in the current study V. cholerae O1 was only isolated from upstream collection sites, which included a transect of the Momance River approaching Mount Chandelle (Jean-Jean). The geographic locations of the fixed collection sites where toxigenic V. cholerae O1 were isolated are presented with respect to isolation frequency in Fig 1. The likelihood of isolating V. cholerae from the environment was not increased with higher concentration of fecal coliform bacteria in the water samples (P > 0.5). The isolation frequency of V. cholerae O1 and non-O1/O139 from all 17 collection sites is presented by month along with the average water temperature in Fig 2A. In contrast to the year-round isolation of non-toxigenic V. cholerae non-O1/O139 from the environment, the isolation of V. cholerae O1 only occurred during the early summer and late fall (June to November), corresponding with increases in water temperatures. The average water temperature was highly correlated with the isolation frequency of V. cholerae O1 both during the same month (P = 0.002) and the following month (P < 0.001). For each additional one degree Celsius increase (°C) in the average water temperature, the frequency of isolation was almost twice as high (RR = 1.87; 95% CI: 1.27, 2.76) during the same month and 2.2 times as high (RR = 2.23; 95% CI: 1.46, 3.39) the following month. The isolation frequency of V. cholerae O1 and Non-O1/O139 from all 17 collection sites is presented by week along with accumulated rainfall in Fig 2B. As shown, both toxigenic V. cholerae O1 and V. cholerae non-O1/O139 were isolated more frequently during the rainy season (approximately from mid-April to mid-September), however toxigenic V. cholerae O1 was isolated exclusively after the onset of the rainy season until November when the precipitation decreased. The isolation frequency of V. cholerae O1 from the environment increased significantly four to five weeks after increases in accumulated rainfall (P = 0.001, 0.002) and also showed a smaller, less significant increase during the same week (P = 0.036). For each additional 10 mm of accumulated rainfall, there was a 27% increase in isolation frequency during the same week (RR = 1.27, 95% CI: 1.02, 1.58), a 48%  , the isolation rate of toxigenic V. cholerae O1 from the environment has increased more than five-fold from 1.7% (3/179) to 8.6% (17/179) during the same period of time [4]. The geographic area where toxigenic V. cholerae O1 strains were isolated also appears to have expanded to include over half of the collection sites, with isolations taking place further up-river. Furthermore, the temporal pattern of isolation has begun to demonstrate seasonal correlations with increases in precipitation and water temperature. This corresponds with observations from countries with endemic V. cholerae O1, where seasonal increases in water temperature or the onset of the rainy season serve as triggers for the proliferation of V. cholerae O1 in the environment that precede seasonal cholera epidemics [15,[18][19]. The additional inclusion of fecal coliform bacteria has also indicated that though fecal coliforms were present in all surface water samples, there was not an increased likelihood of isolating toxigenic V. cholerae at higher fecal coliform concentrations, and that V. cholerae was isolated in water samples with relatively low concentrations of fecal coliforms (< 1 CFU/ml). Regardless of whether toxigenic V. cholerae O1 have established true environmental reservoirs or the surface water was contaminated by those infected with cholera, the presence of toxigenic V. cholerae O1 in the surface water represents a potential ongoing source of exposure. Supporting Information S1 Table. S1 Table contains a list of the oligonucleotide primers used in polymerase chain reactions (PCR) for detection and characterization of toxigenic Vibrio cholerae O1. (DOCX) S1 Dataset. S1 Dataset contains a the sample collection date, epidemiological week from the beginning of the sampling period, the number of weekly reported cases, the frequency of V. cholerae O1 and non O1 isolations per week, the average monthly water temperature (°C), and the accumulated weekly precipitation (mm/week). (XLSX)