Variants of ctxB alleles of Vibrio cholerae O1 caused sequential cholera outbreaks in the tribal areas of Odisha, India

Vibrio cholerae is an autochthonous environmental species existing in zooplankton, phytoplankton, shrimps, crabs, etc. It infects human beings causing cholera resulting in high morbidity and mortality in developing countries (Lacey 1995). V. cholerae has >200 serogroups based on the somatic O antigens, of which, only O1 and O139 have caused outbreaks and epidemics throughout the world. The V. cholerae O1 has two biotypes, namely, classical and El Tor. Several variants of El Tor biotypes of V. cholerae have been identified based on ctxB genotypes which differed by a few non-random point mutations that yielded various ctxB alleles from ctxB1 to ctxB9 (Faruque et al. 1998; Safa et al. 2010; Keymer & Boehm 2011). The current seventh pandemic of cholera that started in Indonesia during 1961 was due to El Tor biotype. This biotype spread to Africa in 1971 and the Americas in 1991. Subsequently, the altered El Tor biotypes of V. cholerae O1 that carried the classical ctxB1 allele emerged in 2002 and first reported from Bangladesh (Nair et al. 2006). Moreover, a new variant of ctxB genotype (ctxB7) of V. cholerae O1 was reported after the disastrous cholera outbreak in Haiti in 2010 (Piarroux et al. 2011). The appearance and the gradual dissemination of the ctxB7 allele of V. cholerae O1 were reported from Kolkata in 2006 (Naha et al. 2012). Later, the El Tor variant strains carrying ctxB7 allele of V. cholerae O1 were also reported from Nepal during 2012, eastern Africa (Kenya, Tanzania and Uganda) during 2015, and the ongoing cholera epidemic in Yemen (Dixit et al. 2014; Weill et al. 2019). The altered El Tor vibrios were also reported from cholera outbreaks from tribal areas of Odisha during 2007 and 2010 (Pal et al. 2010; Kar et al. 2015), whereas the ctxB7 alleles of V. cholerae O1 were first reported from Odisha during 1999 and caused a cholera outbreak in 2014 in the tribal areas of Odisha (Pal et al. 2017). In between, two sequential cholera outbreaks were reported in July to October 2012 accounting for high morbidity from Rayagada and Kalahandi districts of Odisha. However, the same cannot be reported to date due to the delay in molecular analysis and pulsed field gel electrophoresis (PFGE) profiling of isolated V. cholerae O1 strains from these outbreak areas. Thus, the present study was undertaken to find the antibiogram profiles and genomic diversities of the V. cholerae O1 strains isolated from the above cholera outbreaks reported in the tribal areas of Odisha.

Rayagada district is divided into 11 blocks with a population of 967,911 that consists mainly of tribes like the Kondhas and the Sauras. Kalahandi with a population of 1,576,869 is the neighbouring district bordered to the south-east by the Rayagada district (Figure 1). Its major tribes are Shabar, Kondhas and Gond. Kalahandi district has been sub-divided into 13 blocks under two sub-divisions, i.e., Bhawanipatna and Dharmagarh. The tribal population mostly lives in inaccessible hilly and mountainous areas on the forest outskirts. These tribes are mostly illiterate, poor and lack properly treated water supplies, roads and sanitation facilities. Open defecation practices are common in these areas which favour diarrhoeal pathogens to spread and cause the outbreaks reported in different time periods. Cholera-affected blocks of Rayagada district, namely, Kashipur, Kalyansingpur, Gudari, Gunupur (Jagannathpur village) and Rayagada were studied along with Jaipatna, Junagarh and Kalampur blocks from Kalahandi district.

The diarrhoea survey work was carried out in eight affected blocks of Rayagada and Kalahandi districts during the onset of disease in July until the disease subsided in October 2012. Line listing of all diarrhoea cases from hospital records, index cases, sources of drinking water, possible cause and spread of the disease, hygienic condition of the houses, social stigmas and misbeliefs existing among the villagers of each affected block were recorded.

The rectal swabs were collected from diarrhoea patients admitted to the Primary Health Centres (PHCs) and Community Health Centres (CHCs), and also from villages of the eight affected blocks of Rayagada and Kalahandi districts after obtaining the informed consent either from the patient or his/her attendant before the administration of any antibiotics. The rectal swabs from the diarrhoea patients were collected in Cary-Blair transport medium and transported to the laboratory for further analysis. Patients were subdivided into four age groups, i.e., >0–5 years, >5–14 years, >14–40 years and >40 years. Simultaneously, environmental water samples were also collected from the stream, river, chua (small pond in paddy field/pit on the river or stream bed) and nala (small stream) that were used by the local people and also the household water from different affected villages of Rayagada and Kalahandi districts.

The rectal swabs and environmental water samples were enriched in alkaline peptone water (APW) and streaked on thiosulfate-citrate-bile salts-sucrose (TCBS) agar (BD, USA) followed by 24 hr incubation at 37 °C for selective isolation of V. cholerae. Sucrose fermenting yellow colonies were selected from TCBS plates and tested using standard biochemical tests. Serological confirmation was performed using V. cholerae specific polyvalent O1 and monovalent Ogawa and Inaba antisera (WHO 1983; Pal et al. 2010). The sensitivity and resistance patterns of V. cholerae O1 strains were tested using antibiotic impregnated commercial discs (BD, USA) (Bauer et al. 1966). The antibiotics used were ampicillin (AMP, 10 μg), co-trimoxazole (COT, 25 μg), chloramphenicol (C, 30 μg), ciprofloxacin (CIP, 5 μg), furazolidone (FR, 50 μg), gentamicin (GEN, 10 μg), nalidixic acid (NA, 30 μg), norfloxacin (NX, 10 μg), streptomycin (S, 10 μg), tetracycline (TE, 30 μg), neomycin (N, 30 μg), azithromycin (AZM, 15 μg), ofloxacin (OF, 5 μg), doxycycline (DO, 30 μg) and erythromycin (E, 15 μg).

DMAMA (double mismatch amplification mutation assay) PCR assay was performed for the differentiation between classical ctxB and Haitian ctxB alleles of V. cholerae O1 strains using primers (ctxB-F3/ Rv-cla) for the Haitian ctxB allele and (ctxB-F4/Rv-cla) for the classical ctxB allele (Naha et al. 2012).

PFGE analysis of NotI-digested genomic DNA from representative isolates of V. cholerae O1 from both the districts was performed for subtyping the V. cholerae by the CHEF Mapper System (Bio-Rad, USA) according to the PulseNet standardized PFGE protocol (1% PFGE grade agarose in 0.5× TBE buffer). Run condition was generated by an auto-algorithm mode of the CHEF Mapper and the PFGE system was comparable to a size range of 20–300 kb markers for V. cholerae strains. The gel was stained in 10 μg/mL ethidium bromide solution for 30 min, de-stained in distilled water for 15 min and photographed under UV light in an Alpha Imager (Alpha InfoTech Corporation, USA). The analysis of the PFGE profiles of V. cholerae O1 isolates was compared using BioNumerics software (Applied Maths, Sint-Martens-Latem, Belgium). The similarities between the isolates were evaluated using the Dice coefficient method. Cluster analysis was carried out using the unweighted-pair group method using average linkages (UPGMA) (Cooper et al. 2006).

The cholera outbreak information was collected from villagers of the cholera-affected villages and, after analysing the hospital records, the index cases were traced back to Kalayansingpur block of Rayagada district, where the diarrhoeal outbreak started. One 65-year-old male from Singari village of Kalyansingpur block went to the paddy field on 20 July 2012 in the early morning, drank water from the small stream and returned home in the afternoon of the same day and suffered from loose motions and vomiting with fever. Gradually the cases increased in that block and the last case was noted on 6 October 2012. Then the disease spread to adjacent blocks, namely, Kashipur, Jagannathpur village of Gunupur block, Rayagada block with the outbreak subsiding in Gudari block on 15 September 2012. These blocks accounted for 263 cases (total population of five blocks was 351,532) with no deaths; the incidence rate was 0.07% with zero case fatality rate.

Similarly, three blocks of Kalahandi district, namely, Kalampur, Jaipatna and Junagarh were affected by diarrhoea. By analysing the hospital records and discussions among the villagers it was noted that one 50-year-old female from Jampara village of Kalampur block went to the field to work on the morning of 22 July 2012 and worked there up to the afternoon, drank water from the nearby river and suffered from loose motions and vomiting. Then, she washed her dirty clothes in the nearby river, which might have been the source and spread of infection. Gradually the diarrhoea cases spread to Jaipatna block on 28 July 2012 and then cases were reported from Junagarh block until 28 August 2012, accounting for 378 cases (total population of three blocks was 384,115) with two deaths. The incidence rate was 0.09% and the case fatality rate was 0.5%. The line listing of all diarrhoea cases is presented in Figure 2.

The line listing of all diarrhoea cases reported in both the districts indicated that the number of cases started increasing from the month of July and declined in October. The highest cases were reported on 12 September 2012 from Rayagada and on 9 August 2012 from Kalahandi district. The age-wise distribution of all cholera cases of both districts indicated that the >15–60 years age group was more affected than those of <5 years of age. The gender-wise distribution of cholera cases indicated that females were more infected than males in the ratio 1.3:1 (Figure 3).

Out of 221 rectal swabs collected from diarrhoea patients of Rayagada and Kalahandi districts, 93 (42%) were confirmed as V. cholerae O1 Ogawa. Thirteen (2.3%) out of 567 water samples collected from different contaminated sources from both the districts were also positive for V. cholerae O1 Ogawa, biotype El Tor (Table 1).

The results of antimicrobial susceptibility testing exhibited different resistance profiles of V. cholerae O1 Ogawa isolated from these two districts. The V. cholerae isolates obtained from cholera-affected blocks of Rayagada exhibited minor variations in antibiotic resistance profiles as noticed from different blocks. The common resistance profile of V. cholerae O1 in both the districts was ampicillin, nalidixic acid, furazolidone, co-trimoxazole, streptomycin and chloramphenicol. The common sensitivity profile to different antibiotics was ofloxacin, ciprofloxacin, norfloxacin, neomycin, gentamicin, azithromycin, tetracycline, and doxycycline with minor changes reported in Rayagada district. Details of the antibiogram resistance profiles are presented in Table 1.

The DMAMA PCR assay on all the V. cholerae O1 isolates from both the districts revealed that out of 53 isolates from Rayagada district, 43 (81.1%) were positive for the classical ctxB1 allele of V. cholerae O1 and 10 (18.9%) were ctxB7 allele of Haitian variants. Interestingly, out of 20 V. cholerae O1 isolates from Kalahandi district, 18 (90%) were positive for ctxB7, possessing the Haitian variant cholera toxin (Table 2).

Interesting results were obtained while comparing the PFGE analysis on V. cholerae O1 strains isolated from these districts, where three pulsotypes were noticed. The clinical and water isolates from Rayagada and Kalahandi districts were of one pulsotype (Bhawanipatna, Kalampur, Rayagada, Jagannathpur, Thuamulrampur, Gudari and Jaipatna) with different ctxB genotypes. The clinical and water isolates from Kalyansingpur block with the Junagarh isolates of V. cholerae O1 were of a different clone and belonged to ctxB1 genotype only (Figure 4). This indicated that the water isolates of V. cholerae O1 had a similar clone with the clinical isolates showing its source of infection.

The earlier cholera outbreaks in published studies from Rayagada district during 2007 and 2010 indicated that Kashipur, Kalyansingpur and Bissam Cuttack blocks were involved (Pal et al. 2010; Kar et al. 2015). However, the present cholera epidemic was reported from other blocks like Rayagada, Gudari, Gunupur (Jagannathapuur village) and also from Kashipur and Kalyansingpur blocks with 351,532 people at risk. Similarly, the cholera outbreak reported from Narla block of Kalahandi district during 2014 was different from the outbreak of 2012, where other blocks like Kalampur, Junagarh and Jaipatna were affected with 384,115 people at risk. The contamination of water sources is the main cause of transmission of cholera. During the rainy season, people go to the paddy field for farming, drink water in the field from nala, chua, stream, river, etc. and become infected this way. After returning home they suffer from loose motions, vomiting and fever. Unhygienic conditions of the houses, person-to-person contact and migration of people from one village to another were noticed among the cholera-affected villages from both the districts. This might be the source of the spread of infection from house to house, village to village and different blocks in these areas. This was also published in our previous studies. The water samples collected from nala, stream and river were found positive for V. cholerae O1, which was also reported in 2007, 2010 and 2014 from the same area (Pal et al. 2010, 2017; Kar et al. 2015).

From these cholera outbreaks, it was noticed that the female population in the age group >15–60 years was more prone to cholera than males with a male:female ratio 1:1.3. Females were more in contact with contaminated water in their day-to-day life, from cooking in the home to farming in the field, where they depended on nala, chua and stream water for their daily work and livelihood. This may be the reason for more females being infected than males in these tribal areas. The current findings have coincided with earlier reports from very different settings such as Glasgow (1832), Indonesia (2005), Kenya (1990) and developed countries like the UK, USA and Canada in 2005, where adult women and school-aged girls had consistently higher rates of diarrhoeal disease and cholera than their male counterparts (Tornheim et al. 1990; Fauveau et al. 1991; Agtini et al. 2005; Scallan et al. 2005).

The V. cholerae O1 isolates from the 2007 cholera epidemic in the Kashipur block of Rayagada district were due to ctxB1 genotype which was highly resistant to ciprofloxacin, norfloxacin, nalidixic acid and furazolidone (Pal et al. 2010). Further tetracycline-resistant strains of V. cholerae O1 were reported from Kashipur, Kalyansingpur and Bissam Cuttack blocks of the same district in 2010 (Kar et al. 2015). From the present study the reversal of sensitivity to ciprofloxacin, norfloxacin and tetracycline was observed. The present findings from Kalahandi show similar patterns of sensitivity to tetracycline, azithromycin, norfloxacin, gentamicin, ofloxacin, ciprofloxacin as observed in the 2014 cholera outbreak from the same district (Pal et al. 2017). The common resistance profile of V. cholerae O1 strains from both the districts was ampicillin, nalidixic acid, furazolidone and co-trimoxazole with little variation. Similar resistance patterns were also published in a study from Kolkata and Southern India (Kumar et al. 2014).

The bacteriological and DMAMA-PCR analysis indicated that the outbreak in Rayagada district was due to the predominance of classical ctxB1 allele (81.1%) of V. cholerae O1, whereas the epidemic in Kalahandi district was due to the HCT variant (90%) of V. cholerae O1 strains. The earlier reports from Rayagada district in 2007 and 2010 reported that ctxB1 variant of V. cholerae O1 was the causative agent of the cholera outbreaks (Pal et al. 2010; Kar et al. 2015), whereas the epidemic in Narla block in 2014 was completely due to the HCT variant of V. cholerae O1 strain (Pal et al. 2017). These findings indicated that the classical ctxB1 genotypes of V. cholerae O1 were spreading in Rayagada district causing the present outbreak in these areas. The cholera outbreak in Kalahandi district was predominantly due to HCT variant of V. cholerae O1, which might be the precursor of the 2014 epidemic in Narla block of the same district, demonstrating its correlation and spread to other blocks. Further PFGE analysis on V. cholerae O1 strains isolated in 2014 and the 2012 cholera outbreak from Kalahandi district will shed light on the spread of these variants in this district. Earlier studies also reported the spread of HCT-producing V. cholerae O1 to different parts of India during small cholera outbreaks from West Bengal, Bihar and Southern India (Naha et al. 2012; Kutar et al. 2013; Kumar et al. 2014). The cholera outbreaks in Baragarh district during 2018 and Kalyansingpur block of Rayagada district in 2019 were due to the HCT variant of V. cholerae O1 Ogawa (Nayak et al. 2021). The ongoing second wave of the cholera epidemic in Yemen which started in April 2017 was also caused by V. cholerae O1 serotype Ogawa harbouring the ctxB7 allele (Weill et al. 2019). The hybrid biotype of V. cholerae O1 El Tor biotype (carrying cholera toxin of classical type) is known for its evolutionary fitness and heightened virulence that enhances the infection level with increased fluid loss and a higher fatality rate (Kumar et al. 2014). This was evident from the present investigation, in that one diarrhoea patient from Kalampur block was given 114 bottles of intravenous fluids within 4 days of treatment and recovered during this outbreak. This confirmed the severity and toxicity of ctxB7 allele of V. cholerae O1 strains. Similar results were reported during the 2007 cholera epidemic in the tribal areas of Odisha (Pal et al. 2010).

The PFGE analysis on V. cholerae O1 strains isolated from both clinical and water samples from these districts exhibited three pulsotypes. The rectal swab and water isolates of V. cholerae O1 from Kalyansingpur, Jagannathpur of Rayagada district and Thuamulrampur of Kalahandi district were similar to each other, and this confirmed the source of infection. Although ctxB1 and ctxB7 genotypes were reported from both the districts, they belonged to one clone which differed from the pulsotypes of V. cholerae O1 isolated from Kalyansingpur and Junagarh block having ctxB1 genotypes. An interesting finding was observed from the PFGE analysis that there was 96% similarity between the clinical and water isolates of Kalyansingpur block of Rayagada district with the isolates of different places from Rayagada and Kalahandi districts (Figure 4). Similar reports have been published from Southern India during 2010 to 2014 from Belgaum, Karnataka (Bhattacharya et al. 2015). Different pulsotypes have been reported while comparing the normal biotype with hybrid biotypes having both El Tor and classical ctxB and El Tor variant strains of V. cholerae O1 of 1975 to 2017 isolates reported from Kenya (Bundi et al. 2019). Another PFGE analysis report from Kolkata showed a close relationship between classical and Haitian ctxB with >98% similarity (Naha et al. 2012).

In summary, we conclude that the outbreak strains of V. cholerae O1 shared a common resistance profile to antibiotics such as ampicillin, nalidixic acid, furazolidone and co-trimoxazole, whereas their sensitivity to tetracycline and ciprofloxacin again reappeared in both the districts. The water isolates of V. cholerae O1 were similar to the clinical isolates exhibited by PFGE analysis, which confirmed the source of infection. The ctxB1 and ctxB7 genotypes of V. cholerae O1 reported from these districts showed three pulsotypes with 96% similarity. Thus, continuous surveillance is highly essential to track the changing over of altered El Tor biotypes with antibiogram profiles of V. cholerae serogroups that caused cholera outbreaks/epidemics in the tribal areas of Odisha.

This work was supported by the Indian Council of Medical Research, New Delhi, India, Vide Project No. 5/8-1 (70)/2018-19-ECD-II that funded Dr B. B. Pal, and is highly acknowledged. The Chief District Medical Officer (CDMO) and medical staffs of different CHCs of Rayagada and Kalahandi districts are acknowledged for their help during field investigation and sample collection.

No ethical approval was done as there were cholera outbreaks reported in Rayagada and Kalahandi districts of Odisha. However, this was approved by the scientific advisory committee of the institute.

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