OUP user menu

Genomic diversity among Vibrio cholerae O139 strains isolated in Bangladesh and India between 1992 and 1998

Shah M. Faruque, Manujendra N. Saha, Asadulghani, Prasanta K. Bag, Rupak K. Bhadra, S.K. Bhattacharya, R.Bradley Sack, Yoshifumi Takeda, G.Balakrish Nair
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb09027.x 279-284 First published online: 1 March 2000


In order to assess the extent of genomic diversity among Vibrio cholerae O139 strains, restriction fragment length polymorphisms in two genetic loci, rrn and ctx, were studied. Analysis of 144 strains isolated from different regions of Bangladesh and India between 1992 and 1998 revealed the presence of at least six distinct ribotypes (B-I through B-VI) of which three were new ribotypes, and one of these was represented by a nontoxigenic O139 strain. Strains of ribotypes B-I through B-V shared 11 different CTX genotypes (A through K). Antimicrobial resistance patterns of the strains varied independently of their ribotypes and CTX genotypes. Results of this study suggest that V. cholerae O139 is undergoing rapid genetic changes leading to the origination of new variants, and temporal changes in antimicrobial resistance patterns may be contributing to the selection of different variants.

  • Vibrio cholerae O139
  • Ribotype
  • CTX genotype
  • Genetic diversity
  • Antibiotic resistance

1 Introduction

Prior to the emergence of Vibrio cholerae O139 Bengal in India [1] and Bangladesh [2] in late 1992, V. cholerae O1 was the only serogroup that could cause epidemic and pandemic cholera. V. cholerae O139 appeared in September 1992 in southern India and then spread rapidly to different cholera-endemic areas of this country, and to neighboring countries including Bangladesh as well as other Asian countries, and imported cases of O139-mediated cholera were reported from several countries across the globe [35]. In the beginning the new strain totally displaced the existing V. cholerae O1. However, a new clone of V. cholerae O1 biotype El Tor emerged in September 1993, and by February 1994 it had replaced the O139 vibrios and become the dominating serogroup causing cholera in India and in the northern and central regions of Bangladesh [6,7]. A reemergence of V. cholerae O139 was noted in Calcutta from August 1996, and subsequently the O139 vibrios reappeared in both India and Bangladesh [8]. Although the reemerged strains of V. cholerae O139 had identical biochemical traits as the strains isolated during 1992–1993, molecular characterization revealed unique changes in the organization of the CTX genetic element [9,10] and ribotype of reemerged strains [11]. All these studies demonstrated the changing epidemiology of cholera, and suggested that the O139 serogroup is likely to spread over a period of time to other cholera-endemic areas in the world and constitute the eighth pandemic of cholera. In view of this we decided to study the extent of genomic diversity undergone by this serogroup since its initial emergence in 1992. This will enable investigators to identify existing and new variants of V. cholerae O139, and detect the leading edge of the pandemic and the precise genotypes involved in the pandemic spread.

2 Materials and methods

2.1 Bacterial strains and growth conditions

From our collection, a total of 144 strains of V. cholerae O139 were included in this study: 56 clinical strains from Bangladesh and 86 clinical and two environmental strains from different parts of India, isolated between 1992 and 1998. A nontoxigenic O139 strain first isolated in Argentina in 1993 (Arg-03) was also included in this study for purposes of comparison. Before use, the identities of the strains were confirmed by standard biochemical and serological tests.

2.2 PCR assay

A multiplex PCR assay was used to confirm the presence of O139-specific rfb locus and the ctxA genes. The multiplex PCR assay was a modification of the assay developed by Hoshino et al. [12] and used two pairs of previously published primers for the ctxA gene and for the O139-specific rfb genes, respectively [12]. Thermocycle parameters were 94°C for 5 min followed by 35 cycles consisting of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C with a final round of extension at 72°C for 7 min. Amplified products were electrophoresed in 2% agarose gels in 1×TAE buffer along with standard molecular size markers, stained with ethidium bromide (1 μg ml−1) and visualized and documented using a Video Documentation System (Pharmacia Biotech Inc., San Francisco, CA, USA).

2.3 Antimicrobial susceptibility test

Strains of V. cholerae O139 included in this study were examined for resistance to ampicillin (A, 10 μg), chloramphenicol (C, 30 μg), co-trimoxazole (Co, 25 μg), ciprofloxacin (Cf, 5 μg), furazolidone (Fz, 100 μg), gentamicin (G, 10 μg), neomycin (N, 30 μg), nalidixic acid (Na, 30 μg), norfloxacin (Nx, 10 μg), streptomycin (S, 10 μg) and tetracycline (T, 30 μg) using commercial discs (Hi Media, Bombay, India) as described previously [3,13].

2.4 Ribotyping and CTX genotyping

Extraction and restriction endonuclease digestion of genomic DNA, as well as Southern blot hybridization were performed using standard techniques [14]. For ribotyping as well as CTX genotyping, restriction enzyme BglI (Takara Shuzo Japan or Bethesda Research Laboratories, Gaithersburg, MD, USA) was used to digest chromosomal DNA and the fragments were separated by agarose gel (0.8%) electrophoresis. The rRNA gene probe was a 7.5-kb BamHI fragment of pKK3535 described previously [9,11,15]. The gene probe used for CTX genotyping was a 0.5-kb EcoRI fragment of pCVD27 [16] representing 94% of the ctxA gene encoding the A subunit of cholera toxin (CT). At NICED, Calcutta, labeling of the probes, hybridization and detection of bands were performed using an ECL detection system (Amersham Life Science) in accordance with the manufacturer's recommendations. At ICDDR,B Dhaka, the probes were labeled by random priming [14] using a random primer DNA labeling kit (Bethesda Research Laboratories) and [α-32P]dCTP (3000 Ci/mmol, Amersham). Southern blots were hybridized with the labeled probes, and autoradiographs were developed as described by us previously [6,11].

3 Results

Of the 144 strains of V. cholerae O139 analyzed in this study, 56 strains isolated in Bangladesh and 88 strains isolated in India were initially screened independently at ICDDR,B and NICED, respectively. Subsequently, the patterns (ribotype and CTX genotype) were matched and representatives of each of the unique patterns were selected and reconfirmed at ICDDR,B. All strains included in this study agglutinated with O139-specific antiserum and the identity of the 88 strains from Calcutta was further confirmed by amplifying the O139-specific rfb locus. All 88 strains were positive for the 449-bp O139-specific amplicon and 87 of the 88 strains were also positive for the 308-bp ctxA amplicon (data not shown).

3.1 Ribotypes

Representative strains displaying a particular ribotype were examined at least three times in a blind fashion and for each experiment DNA was prepared separately from a single isolated colony. Ribotype patterns were stable and reproducible. The chromosomal fragments containing rRNA genes ranged from 18 to 1 kb. In total, we identified six different ribotypes among the 144 strains of V. cholerae O139 examined, and these ribotypes were designated B-I through B-VI (Fig. 1). Five of the ribotypes (B-I through B-V) were associated with toxigenic clinical strains while one ribotype (B-VI) was associated with a nontoxigenic clinical strain isolated in Calcutta (Fig. 1, lane 6). Interestingly, the pattern B-VI was altogether different from the other ribotypes obtained in this study and it was also found to be deviant from the ribotypes of El Tor genomes reported by Lan and Reeves [17]. The other nontoxigenic O139 strain included in this study for purposes of comparison was the Argentinean strain (Arg-03) which displayed a ribotype pattern different from B-I through B-VI and was designated NB-I (Fig. 1, lane 7). The ribotype nomenclature B-I to B-VI was given based on the chronology of appearance of each ribotype. The distribution of strains belonging to the different ribotypes in Bangladesh and India isolated on a yearly basis between 1992 and 1998 are presented in Table 1.

Figure 1

Southern hybridization analysis of rRNA genes in V. cholerae O139 strains. Genomic DNA was digested with BglI and probed with a 7.5-kb BamHI fragment of the Escherichia coli rRNA clone pKK3535. Restriction patterns corresponding to ribotypes B-I through B-VI, derived from strains isolated in Bangladesh or India between 1992 and 1998, are shown in lanes 1 through 6, respectively, and ribotype NB-I, derived from a nontoxigenic strain isolated in Argentina in 1993, is shown in lane 7. Size markers (in kb) are indicated on the right.

View this table:
Table 1

Yearly isolation of different ribotypes of V. cholerae O139 in Bangladesh and India between 1992 and 1998

Year of isolationCountry of isolationNumber of strains analyzedNumber (%) of strains belonging to different ribotypes
1992–199814436 (25%)82 (56.9%)21 (14.5%)2 (1.3%)1 (0.7%)1 (0.7%)
  • aThis included 18 strains from an outbreak in two north-central districts of Bangladesh in 1997 [11].

3.2 CTX genotypes

Southern blot analysis of BglI-digested chromosomal DNA of 144 strains of V. cholerae O139 isolated in Bangladesh and India examined in this study revealed 11 (A through K) distinct restriction patterns of the ctxA gene and its flanking chromosomal sequences (Fig. 2). The nontoxigenic strain CO853 did not contain sequences that hybridized with the ctxA probe. The restriction patterns representing the different CTX genotypes consisted of one to four bands, which ranged between 12 and 2.8 kb in size (Fig. 2). The CTX genotypes were shared by toxigenic strains belonging to different ribotypes as shown in Table 2.

Figure 2

Southern hybridization analysis of ctxA gene in V. cholerae O139 strains isolated between 1992 and 1998 in Bangladesh and India. Genomic DNA was digested with BglI and probed with the ctxA gene. Restriction patterns corresponding to CTX genotypes A through K (provided at the top of each lane) are shown. Size markers (in kb) are indicated on the right.

View this table:
Table 2

Ribotype, CTX genotype, and antibiogram of representative Vibrio cholerae O139 isolated in Bangladesh and India between 1992 and 1998

StrainCountry of isolationYearRibotypeCTX genotypeAntibiogram
CO 5India1993B-IAACCoFzNS
CO 298India1993B-IIACoFzS
CO 830India1994B-IIAACoFzNNaS
MO 579India1996B-IIGAFz
A: ampicillin; C: chloramphenicol; Co: co-trimoxazole; Cf: ciprofloxacin; Fz: furazolidone; G: gentamicin; Na: nalidixic acid; N: neomycin; Nx: norfloxacin; S: streptomycin; T: tetracycline.
ND, not done; NT, nontoxigenic.
  • aSusceptible: sensitive to ACCoCfFzGNaNxNST.

3.3 Antibiogram

O139 strains isolated in Bangladesh and India before 1995, including those isolated from the epidemics during 1992 and 1993, were resistant to co-trimoxazole and streptomycin whereas those isolated between 1996 and 1998 were mostly susceptible to these drugs. Strains resistant to one or more antibiotics in different combinations included ampicillin, chloramphenicol, furazolidone, neomycin, nalidixic acid, norfloxacin and ciprofloxacin. Antimicrobial sensitivity patterns of representative strains having different ribotypes and CTX genotypes are shown in Table 2.

4 Discussion

Restriction patterns representing ribotypes B-IV through B-VI (Fig. 1) have not been reported previously by us or other investigators who have analyzed a large number of V. cholerae strains from different countries [11,18,19]. This suggests that the O139 strains belonging to ribotypes IV through VI represent new clones of V. cholerae O139. During the initial outbreak of O139 vibrios in 1992 and 1993, only two ribotypes designated B-I and B-II were involved [19]. Ribotype B-III was reported by us recently and this ribotype was found to be associated with an outbreak of cholera in two north-central districts of Bangladesh in 1997 [11]. It is interesting to note that a single strain (MDO-2) belonging to this ribotype was isolated in a southern part of India (Madurai) as far back as 1993, but this strain was not associated with an epidemic outbreak. Ribotypes B-IV through B-VI are represented by either two strains or a single strain in the present study. However, the epidemic caused by ribotype B-III strains in Bangladesh, nearly 4 years after the initial isolation of a single strain belonging to this ribotype in India, emphasizes the importance of early detection of new ribotypes.

Ribotyping has been successfully used to study clonal diversity among V. cholerae strains, and use of the enzyme BglI for ribotyping has provided more discrimination among strains as compared to other enzymes tested [20]. Yearly isolation of different ribotypes of V. cholerae O139 in Bangladesh and India (Table 1) clearly shows a striking difference in the proportion of strains belonging to different ribotypes, as well as the distribution of the ribotypes between the two neighboring countries Bangladesh and India. Since 1993, a decline in the proportion of strains belonging to ribotype B-I and an increase in the proportion of strains belonging to ribotype B-II has been noticed both in India and in Bangladesh, and subsequently in some districts of Bangladesh strains belonging to ribotype B-III increased considerably and caused an outbreak in 1997 [11]. This shows the dynamic nature of the epidemiology of cholera and the different clones involved in the initiation of epidemics.

A total of 11 distinct CTX genotypes were identified among the six ribotypes (Table 1), and CTX genotypes were shared by more than one ribotype. This shows that CTX genotypes are likely to diversify more rapidly than the ribotypes. Although the structural gene for ctxA is identical in different strains, the observed diversity of CTX genotypes may result from duplication of the CTX element or its rearrangement due to continuous induction and reinfection of V. cholerae strains by CTXφ[21]. Thus, CTX genotypes may be more discriminatory but relatively less stable than the ribotypes. It may be noted that the strain belonging to ribotype B-III which caused the cholera outbreak in Bangladesh belonged to a different CTX genotype (J) from the ribotype B-III strain isolated in India, which belonged to CTX genotype C (Table 1).

In this study, it was observed that the antibiograms of strains with different CTX genotypes were different. Further, variation in antibiograms was also observed among strains having different ribotype patterns. Antibiotic resistance in V. cholerae has been shown to be encoded by horizontally transferable conjugative transposons, which reside in the chromosome of the host bacterium [22]. We have observed previously that the emergence of a new clone of V. cholerae O1 was preceded by a change in antibiogram [23]. It is thus possible that the temporal changes in antimicrobial resistance pattern observed in this study are contributing to the selection and enrichment of different variants of V. cholerae O139. This emphasizes the need for a continual monitoring of toxigenic V. cholerae strains to be able to detect the origination of new variants with epidemic potential.


We thank V.I. Mathan for useful discussions and suggestions. This research was funded, in part, by the United States National Institutes of Health under Grant RO1 AI39129-01A1 with the Department of International Health, Johns Hopkins University and the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B). The ICDDR,B is supported by countries and agencies which share its concern for the health problems of developing countries. The work conducted at the National Institute of Cholera and Enteric Diseases, Calcutta, India was supported, in part, by the Japan International Cooperation Agency (JICA/NICED Project 054-1061-E-O).


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
View Abstract