Characterization of V. cholerae O1 biotype El Tor serotype Ogawa possessing the ctxB gene of the classical biotype isolated from well water associated with the cholera outbreak in Kerala, South India

Cholera is an acute gastrointestinal illness characterized by severe watery diarrhoea and is caused by the bacterium Vibrio cholerae that is acquired by the ingestion of contaminated food or water. Globally, cholera alone causes 120,000 deaths annually (Sack et al. 2006). The Indian sub-continent has been an epicentre for cholera. However, cholera cases are hugely under-reported mainly because disease surveillance is limited. The incidence of cholera is estimated to be 1.6 cases/1,000 population per year or 40/1,000 cases of acute diarrhoea in India (Sharma et al. 2017). According to the World Health Organization (WHO), a cumulative total of 838,315 cases were notified in India for the time period of 2004 to 2008, as compared to 676,651 cases in the years 2000 to 2004. This represents an approximate 24% rise in the number of cholera cases identified (Kanugo et al. 2010). During a ten-year period (1997–2006) study in India, outbreaks were reported in Kerala in multiple years and the states having the highest number of reported outbreaks were West Bengal, Orissa, Maharashtra and Kerala (Kanugo et al. 2010). During 2010–2015, cholera was reported in three districts in Kerala during at least three out of five consecutive years, and the state was defined as a cholera endemic state (Ali et al. 2017).

The re-emergence of the cholera epidemic and the evolution of multidrug-resistant V. cholerae strains over the last decade, particularly in Asian countries, pose a great threat to the clinical diagnosis and treatment of this disease (Mahapatra et al. 2014). Kerala, a southwestern coastal state of the country, has recently experienced outbreaks of cholera. In the massive cholera outbreak reported from Malappuram and Palakkad districts in Kerala during June–July 2016, three deaths were recorded and more than a hundred people were affected (NCDC 2016). In these outbreaks, the causative organism was isolated from patients and characterized. However, isolation of pathogenic V. cholerae from the environment is often limited by the lack of a suitable technique to selectively enrich pathogenic strains from the vast majority of non-pathogenic strains normally found in the environment. Hence, very little information is available in the literature regarding the characterization of V. cholerae strains from drinking water sources implicated in the outbreak. In this context, the present study has been undertaken to isolate the toxigenic V. cholerae strains from well water samples from Palakkad district in Kerala (India) from where a cholera outbreak was reported and to identify their biotype and serogroup and to characterize them for virulence genes, antibiotic susceptibility profiles and genetic profiles.

Pattanchery is a village in the Palakkad district of Kerala, India and it is a densely populated area with little space between houses. People live in poor quality houses. Their habitats are characterized by overcrowding, lack of basic amenities and facilities such as drinking water and sanitation. Piped municipal water supply is intermittent and several households share public water taps. People depend on well water and they share water from common wells. The main source of income is from agriculture and fish farms.

A total of 28 samples including 22 water samples (17 well water and 5 pipe water (both chlorinated)) and 6 soil samples from the surroundings of wells were collected from the households of suspected patients in Pattanchery, Palakkad district (Figure 1) from where a cholera outbreak was reported during June–July 2016. All the well water samples collected were chlorinated, given that the health authorities chlorinated all the water sources in the area of the outbreak immediately after the first report. Samples were stored in insulated thermocole boxes containing flake ice and transported aseptically to the laboratory for further analysis within 2–3 hours of collection and care was taken to avoid cross contamination.

All samples collected were analysed for V. cholerae according to Standard Protocol (USFDA 2001). For isolation of V. cholerae, samples of water were directly plated (100 μL) onto thiosulfate citrate bile salts sucrose agar (Oxoid, UK), incubated at 37 °C for 18–24 h and characteristic colonies were picked, purified on tryptic soy agar (TSA) plates (Difco, USA) and identified. A part of the water (25 mL) was incubated in 225 mL alkaline peptone water (APW) at pH 7.5–8.5 (100 rpm) for 16–18 h and plated as described above. In order to prepare the soil samples for analysis, samples were added to 100 mL of distilled water and allowed to settle. An 8–10 mL amount of the slurry was centrifuged at 2,000 rpm for 8 min to remove particulate matter and 1 mL of slurry was added to 10 mL of APW (pH 8.5) for enrichment at 37 °C for 18–24 h, streaked onto TCBS agar plates and incubated overnight at 37 °C. Typical colonies were picked and purified on TSA. After purification, V. cholerae isolates were identified as per FDA method (Elliot et al. 2001) and confirmed using rapid diagnostic kit API 20E (BioMerieux SA, France).

A clinical strain, V. cholerae O1 El Tor Ogawa, supplied by Govt. Medical College Alappuzha, Kerala and V. cholerae O1 MTCC 3904 procured from Microbial Type Culture Collection (MTCC) were used as reference strains and positive controls.

All the 25 V. cholerae isolates were tested by slide agglutination using V. cholerae poly O1 antiserum and monospecific Ogawa, Inaba and Hikojima antisera (BD Difco, Maryland, USA). The V. cholerae isolates exhibiting a negative reaction with V. cholerae O1 polyvalent antiserum were typed using V. cholerae O139 ‘Bengal’ antiserum (Denka Seiken, Japan). The isolates that gave negative results with both polyvalent somatic O antiserum and O139 ‘Bengal’ antiserum were designated collectively as V. cholerae non-O1/non-O139.

Susceptibility to various antimicrobial agents was tested by the Kirby-Bauer disc diffusion method as described by the National Committee for Clinical Laboratory Standards (CLSI 2015, 2019) using commercially available discs (Icosa GII minus, Himedia, India) with 20 antibiotics: imipenem (10 μg), tobramycin (10 μg), ofloxacin (5 μg), ciprofloxacin (5 μg), co-trimoxazole (25 μg), gentamicin (10 μg), norfloxacin (10 μg), amikacin (30 μg), levofloxacin (5 μg), augmentin (30 μg), cefoxitin (30 μg), galifluxacin (5 μg), moxifloxacin (5 μg), colistin (10 μg), ceftriazone (30 μg), nalidixic acid (30 μg), ceftazidime (30 μg), azthreonam (30 μg), nitrofurantion (300 μg) and cefpodoxime (10 μg). Sensitivity towards vibriostatic agent 2,4-diamino-6,7-diisopropylpteridine, O/129 (10 μg and 150 μg) was also tested. Escherichia coli American Type Culture Collection (ATCC; Manassas, VA, USA) 25922, Pseudomonas aeruginosa ATCC 27853 and Staphylococcus aureus ATCC 29213 were used for the standardization of the Kirby-Bauer test for correct interpretation of the zone diameters. Isolates that are resistant to at least one agent in three or more antimicrobial categories are defined as multidrug resistant (MDR).

All the 25 V. cholerae isolates were analysed for the presence of ‘species specific’ outer membrane protein, ompW gene, virulence associated genes such as ctxA, ctxB, zot, ace and VPI genes and regulatory gene toxR. Serogroup-specific gene rfb was amplified to distinguish O1, O139, and non-O1/non-O139 V. cholerae strains. Genes tcpA, encoding the toxin coregulated pilus and hlyA were amplified for the discrimination of classical and El Tor biotypes. Clinical strain V. cholerae O1 El Tor Ogawa and V. cholerae O1 MTCC 3904 procured from MTCC were used as reference strains. The primers used in this study and the conditions employed are listed in Table 1.

Genomic DNA for PCR assay was extracted by the phenol–chloroform method (Ausubel et al. 1994) with minor modifications (Kumar et al. 2009). PCR assay was carried out in 25 μL of reaction buffer and was performed in AB Biosystems, Veriti 96 well thermal cycler. 50 and 100 bp molecular size markers (Gene Ruler™, Fermentas) were used for separation of the amplicons. PCR products were analysed by electrophoresis on 1.5% agarose gels, stained with ethidium bromide, visualized under UV light and recorded with the aid of a Gel Documentation system (SYGENE, Biorad).

Enterobacterial repetitive intergenic consensus (ERIC)-PCR was performed according to Versalovic et al. (1991), using ERIC primers ERIC1 R- 5′-ATGTAAG CTCCTGGGGATTCAC-3′ and ERIC2 5′- AAGTAAGTGACTGGGGTGAGCG-3′ ERIC-PCR and amplification in Mastercycler Personal (Eppendorf, Germany). Amplification was performed in 25 μL volume containing 2.5 μL 10x reaction buffer, 1 mM each of dNTP, 5 pmol each of the forward and reverse primers, 2.5 U of Taq polymerase, 2 mM MgCl₂ and 10 ng of genomic DNA. PCR amplification was done as follows: denaturation at 92 °C for 45 s, annealing at 52 °C for 1 min and elongation at 70 °C for 10 min. A final elongation step at 70 °C for 20 min at the end of 35 cycles was added. The PCR amplification products were separated by electrophoresis through 1.5% agarose gel and detected by staining with ethidium bromide.

PFGE of V. cholerae isolates was performed according to PulseNet standardized protocol using restriction enzyme NotI (WWW.cdc.gov/pulsenet/PDF/vibrio pfge protocol-508c.pdf) on a CHEF-DR XA system (Bio-Rad Laboratories, Richmond, CA, USA). Bacteriophage lambda DNA ladder (New England Biolabs) was used as size marker. Following electrophoresis, the gel was stained with ethidium bromide and DNA bands were visualized with UV transilluminator and photographed. ERIC-PCR and PFGE were repeated twice.

DNA fingerprinting pattern of ERIC-PCR and PFGE were analysed using Gel Compar II software (Version 5.1, Applied Maths, Belgium). Cluster analysis was performed using the unweighted pair group method with arithmetic averages (UPGMA), with Dice correlation method. The position tolerance was set at 1.0% and minimum profile for each band was set at 5.0%.

A total of 28 samples (17 well water, 5 pipe water, 6 soil samples) collected at the households of suspected cholera patients in Palakkad district, were analysed for the presence of V. cholerae. Among the 90 presumptive V. cholerae isolated from chlorinated water and soil and characterized biochemically as per FDA protocols (USFDA 2001), 25 isolates from three well waters were identified as V. cholerae and confirmed using rapid diagnostic kit API 20E. All the isolates were Voges–Proskauer (VP) test positive.

The presence of ompW gene confirmed the biochemical identification of all the 25 V. cholerae isolates. The rfb O1 gene (Figure 2(a)) was present in all the 25 isolates confirming that the isolates belong to O1 serogroup, and this is in agreement with the serological results. All the 25 isolates and the reference strains possessed tcpA El Tor (Figure 2(b)) and they were confirmed as V. cholerae O1 biotype El Tor and also confirmed in the phenotypic assay as VP positive, haemolysis positive and citrate positive. Serotyping of the isolates with monospecific antisera identified that the isolates belong to serotype Ogawa. V. cholerae O1 biotype El Tor serotype Ogawa isolates from well water samples in this study harboured ctxA (Figure 2(c)), ctxB (Figure 2(d)), hlyA, VPI, ace, zot and toxR genes. The presence of the CTX element and the Vibrio pathogenicity island in the isolates in this study indicates that they are pathogenic and epidemic strains of V. cholerae. The presence of ctxB gene of the classical biotype in the El Tor isolates indicates that they are new variants of V. cholerae O1 El Tor isolates that produce cholera toxin of the classical biotype. El Tor strains with classical ctxB gene were associated with the cholera outbreak reported from Chennai (Goel et al. 2010). In the earlier cholera outbreaks reported from Alappuzha and Palakkad districts in Kerala, ctxA and tcpA genes were detected in V. cholerae O1 EI Tor Ogawa clinical strains (Radhakutty et al. 1997). The detection of El Tor strains with classical ctxB gene in the environmental V. cholerae O1 in this study as well as in the outbreaks from Chennai and Odissa indicates their wide distribution in the aquatic environment of the Indian subcontinent (Goel et al. 2010; Pal et al. 2010). It has been reported that CT producing El Tor strains are now replacing the seventh pandemic El Tor strains in India and Bangladesh (Nair et al. 2002, 2006; Kumar et al. 2009). Cholera outbreaks reported in different parts of Kerala during 1996–2000 and during 2012–2016 were mainly caused by the serotype Ogawa (Radhakutty et al.1997; Geeta et al. 2005; NCDC 2016).

The phenotypic assays in this study confirmed that all the V. cholerae O1 isolates in this study are of the El Tor biotype background. The results of the VP test and citrate test that are used to distinguish between classical and El Tor biotypes (USFDA 2001) revealed isolates were VP and citrate positive. Genetic analysis of the tcpA gene that is routinely conducted to verify the biotype background of V. cholerae isolates (Keasler et al. 1993) also confirmed that the isolates belonged to El Tor biotype. The presence of ctxB gene of the classical biotype in El Tor isolates suggests that they are variants of El Tor biotype.

In the present study, 100% resistance to augmentin, nalidixic acid, ceftriazone, co-trimoxazoe, imipenem, nitrofurantion, and ceftazidime was found in V. cholerae O1 altered El Tor Ogawa strains and they were multidrug resistant, suggesting these drugs could not be used for clinical purposes. Similarly, 25 isolates were found to be resistant towards vibriostatic agent 2,4-diamino-6,7-diisopropylpteridine (O/129). Das et al. (2011) and Shrestha et al. (2015) also reported the resistance of V. cholerae O1 El Tor against nalidixic acid and cotrimoxazole.

However, all the isolates were sensitive to colistin, ciprofloxacin, moxifloxacin, cefoxitin, azthreonam, gentamicin, gallifluxacin, ofloxacin, tobramycin, norfloxacin, amikacin and levofloxacin. V. cholerae O1 El Tor Ogawa isolates possessing the ctxB gene of the classical biotype obtained in the present study were sensitive to ciprofloxacin and norfloxacin, as reported by Pal et al. (2010), for environmental strains of the El Tor variant with the ctxB gene of the classical biotype. However, a progressive increasing trend of antibiotic resistance in clinical strains of V. cholerae O1 El Tor towards common fluoroquinolone, i.e., ciprofloxacin and norfloxacin, has been reported since 1996 (Garg et al. 2001; Krishna et al. 2002; Goel et al. 2010; Pal et al. 2010). Similar to our observation, resistance to nalidixic acid, co-trimoxazole and furazolidone were also reported in V. cholerae O1 El Tor Ogawa isolates carrying the ctxB gene of classical biotype implicated in the cholera outbreak from Odissa and Solapur (Pal et al. 2010; Goel et al. 2011).

All the 25 V. cholerae O1 isolates were characterized by ERIC-PCR and PFGE to study their genetic relatedness. ERIC-PCR with genomic DNA of V. cholerae isolates resulted in amplification of multiple fragments of DNA ranging between 100 bp and 2,000 bp. ERIC-PCR finger prints grouped the isolates into two clusters (C1 and C2) (Figure 3). C1 formed the largest cluster with 23 isolates and the other 2 isolates belonged to C2 cluster. The clinical strains formed separate clusters. ERIC sequences are 126 bp imperfect palindromes that occur in multiple copies in the genomes of Vibrio, located near the haemolysin gene, and are less complex but more discriminative (Waturangi et al. 2012). In ERIC-PCR, toxigenic V. cholerae have shown multiple fragments of DNA ranging between 0.25 and 1.8 kb (Goel et al. 2010).

The analysis of PFGE fingerprints revealed that the isolates were grouped into two clusters (P1 and P2) (Figure 4). Seventeen isolates formed one cluster (P1) and eight isolates formed another cluster (P2). The clinical reference strains of O1 El Tor and O1 MTCC strain formed separate clusters different from P1 and P2. The results indicate that El Tor biotype isolates do not share similarities in their genetic traits with the clinical El Tor type strain used in the study. The distribution of isolates into two clusters suggest that two clones of V. cholerae might have been implicated in the Palakkad outbreak. This study indicates that continuous monitoring of V. cholerae strains associated with outbreaks is needed to understand the evolution of pathogen and disease patterns.

This study revealed the presence of a multidrug-resistant new variant of V. cholerae El Tor Ogawa in well water with CTX genes. To our knowledge, this is the first report regarding the isolation of a multidrug-resistant toxigenic new variant of V. cholerae O1 El Tor Ogawa from well water in Kerala. The study suggests that contaminated well water may be incriminated in the outbreak from Palakkad district in Kerala. The pathogens released from point and diffuse sources from the outbreak area get transported from upstream sources to coastal waters, thereby increasing the chances of spread of the disease to other areas. For enhanced control and to reduce the potential risks of the dreadful disease cholera, the sanitation process has to be improved in the outbreak area and, at the same time, the quality of drinking water need to be monitored frequently. The use of multiple typing methods would permit a more precise characterization of the genetic diversity to better understand the evolution of pathogen and disease patterns.

This work was carried out with financial assistance obtained from the National Agricultural Innovation Project (NAIP), the Indian Council of Agricultural Research (ICAR), New Delhi, India. The authors are grateful to the Director, CIFT, Cochin for providing the necessary facilities and kind permission to publish this paper. We are also very grateful to Dr Jayalakshmi, Head of the Microbiology Division, and Vice Principal Dr Sreerekha, Alappuzha Medical College, Kerala, India for providing clinical strains of V. cholerae. No conflict of interest is declared.

All relevant data are included in the paper or its Supplementary Information.