INTRODUCTION
Vibrio cholerae is the etiologic agent of cholera, the acute diarrheal illness that continues to be a public health threat worldwide [1]. The pathogen has been traditionally classified into two biotypes, the classical and El Tor, based on clear phenotypic and genotypic differences. It has caused seven pandemics. The fifth and sixth pandemics occurred due to infection by the classical biotype while the current seventh pandemic that started in 1961 is due to the El Tor biotype [Reference Chastel2]. The El Tor V. cholerae has undergone rapid genetic modifications due to the acquisition of different virulence gene cassettes by bacteriophage infection [Reference Faruque3] and genetic recombination [Reference Keymer and Boehm4]. The exchange of genetic material that takes place in either the aquatic environment [Reference Faruque5] or in the human GI system [Reference Regina6] has consequently led to the diversification of the seventh pandemic prototype El Tor strains into its different kinds of variants [Reference Grim7] among which the altered El Tor strains [Reference Nair8] have spread to different regions of the world from the point of its origin. With the advent of the hybrid and variants, identification of biotype based solely on phenotypic tests became impossible [Reference Nair9]. As a result molecular typing techniques became the reliable method of classification. This study was designed to determine the biotype of the strains by molecular methods and examine the genotypic characteristics of strains of V. cholerae O1 isolated in India between the years 2007 and 2008 from 10 different places.
METHODS
Bacterial strains included in the study
A total of 45 strains of V. cholerae from Bihar, Orissa, Chennai, Delhi, Guwahati, Kolkata, Pune, Mumbai, Rohtak and Surat as shown in Table 1 along with their year of isolation, were isolated from stool samples of cholera patients during the years 2007 and 2008 and sent to Vibrio Phage Laboratory at the National Institute of Cholera and Enteric Diseases in nutrient agar stabs for confirmation, biotyping, serotyping and phage typing. These served as representative strains causing cholera in India during 2007 and 2008.
Table 1. Places from where 45 representative strains from the years 2007 and 2008 were included in this study
Culture and phenotypic identification
Culture of V. cholerae was done by standard method by inoculating a loopful of culture from nutrient stab into 2 ml of APW (Alkaline Peptone Water) containing 1% NaCl and incubated for 6 hours at 37 °C with shaking. A loopful of the inoculum was transferred onto TCBS plate and streaked and the plates were incubated overnight at 37 °C and examined for typical round yellow V. cholerae colonies.
Further identification was done using standard conventional biochemical methods [Reference Kelly10]. The strains were serogrouped using polyvalent O1 antisera from Denka Seiken (Tokyo, Japan) and serotyped using monovalent Ogawa and Inaba antisera from Denka Seiken (Tokyo, Japan). Phenotyping for biotype identification was done using Voges–Proskauer test, Polymixin-B sulfate sensitivity test and phage typing using lytic phages Mukerjee classical phage IV and Mukerjee El Tor phage V according to standard diagnostic procedures [11].
Detection of virulence genes by polymerase chain reaction (PCR)
PCR analysis was used to detect the major genes associated with V. cholerae virulence using specific primers Table 2. Consequently, virulence genes rstR, tcpA, ctxA, tcpI, rtxA, rtxC, zot, toxR, VC1449, VC1450, VCAO417, VCAO316, VCAO728, VCAO729, VCAO730, nag, nan, rstC, intl, orfU, acfB, tlc were amplified, and these served as markers for the detection of the presence of various virulence regions of the V. cholerae genome. For screening the presence of VSP-I, VCO183 was amplified while for detecting the presence of VSP-II region, the ORFs VCO511, VCO513, VCO514, VCO515 and VCO516 were amplified. V. cholerae O1 El Tor strain, N16961 and the O1 classical strain O395 were used as standard controls strains.
Table 2. Sequence of PCR primers used in this study
For PCR template, 1 ml of overnight culture was taken and centrifuged at 6000 rpm (Biofuge, Heraeus, Germany) for 5 min. The cell pellet was collected and resuspended in 300 µl of sterile distilled water. The cell suspension was boiled for 10 min in a boiling water bath and immediately transferred to ice for 10 min followed by centrifugation at 12 000 rpm for 10 min. 2 µl of this supernatant was used as template for PCR reaction in 20 ul PCR mixture containing 1U of 10× Taq DNA Polymerase (Bangalore Genei). The same cycling conditions as described in previous studies were used for PCR reaction [Reference O'Shea12]. PCR products were run on 2% agarose gels and examined using a gel-documentation system from Bio-Rad.
Mismatch amplification mutation assay (MAMA) PCR and double-mismatch amplification mutation assay (D-MAMA) PCR
MAMA PCR was used to detect the presence of ctxB gene and the type of biotype specific gene for cholera toxin B subunit present in the 45 test strains by employing primers Fw-con and Rv-cla and Rv-elt [Reference Morita13]. The reaction conditions for MAMA PCR was followed according to standard technique described previously [Reference Morita13]. D-MAMA PCR was used to detect the presence of ctxB-7 allele in the strains tested by employing the same primers and cycling conditions as recently described [Reference Naha14].
Sequencing of the ctxB gene
Genomic DNA was extracted and purified using the method of Sambrook et al. [Reference Sambrook, Fritsch and Maniatis15] and used for PCR amplification of the 460 bp ctxB gene using the primer set ctxB-F 5′-(GGTTGCTTCTCATCATCGAACCAC)-3′ and ctxB-R 5′-(GATACACATAATAGAATTAAGGATG)-3′. PCR was carried out in 25 µl reaction mixture containing 2·5 µl of 10× PCR (Takara Shuzo, Otsu, Japan), 2 µl of dNTP mixture (concentration 2·5 mM each) (Takara), 1 µl of each primer (10 pmol/μl), 0·2 µl (1U) of rTaq DNA polymerase (Takara) and 40 ng of chromosomal DNA of V. cholerae as template.
The PCR product was purified using Qiaquick PCR purification kit from Qiagen, GmBH, Germany by loading into purification column (Qiaquick spin column, Qiagen). A 40 ng of the purified PCR product was used in the cycle sequencing reaction mixture comprising the sequencing premix obtained from Perkin Elmer, USA, that consisted of Big Dye Terminator 3.1 Sequencing Buffer (5×) and ready to use reaction mixture (2·5×) and Big Dye Terminators (Perkin Elmer, USA) with AmpliTaq FS polymerase. An automated thermal cycler (Perkin Elmer Thermocycler 2400, USA) was used to amplify the 460 bp gene using 25 cycles of denaturation at 94 °C for 10 s, annealing at 50 °C for 5 s and extension at 60 °C for 4 min [Reference Olsvik16]. The product was purified and dried and sequenced using the DNA Analyzer 3730 automated sequencer (Applied Biosystems, Hitachi). Sequence editing and analysis was done using SeqMan (DNASTAR Inc., USA) Sequences were compared with sequences of control strains O395 (classical) and N16961 (El Tor) with accession numbers NC_012582 and AE003852.1 respectively from GenBank, NCBI and multiple alignments were performed using ClustalW of MEGA 5.0 [Reference Kumar, Tamura and Nei17] with gap open penalty = 10 and gap extension penalty = 0·2. The phylogenetic tree was constructed using neighbour-joining method with the bootstrapping of 1000 replications and 70 000 random seeds.
Beads enzyme-linked immunosorbent assay (ELISA) for quantification of cholera toxin (CT) production
The amount of CT produced by the 45 V. cholerae strains was measured by Beads ELISA method, which uses polystyrene beads of 6·5 mm in diameter coated with anti-CT IgG as a solid phase. The coated bead was first incubated with the sample and then incubated with anti-CT IgG [F(ab)]-horseradish peroxidase conjugate. Peroxidase activity was determined colorimetrically with 3,3,5,5-tetramethylbenzidine as the substrate by measuring the absorbance value at 450 nm. For this assay V. cholerae strains were streaked onto 1% nutrient agar plates and left for overnight incubation at 37 °C. Overnight culture was transferred to AKI medium and incubated at 37 °C for 20 h without shaking. Assay of cholera toxin was carried out according to the method described by Oku et al. [Reference Oku18].
Pulsed-field gel electrophoresis
Seven of the V. cholerae strains were analysed by PFGE to detect polymorphisms and consequential changes in restriction sites across their genome. Bacterial strains were cultured and agarose-embedded DNA plugs were prepared and digested with the enzyme NotI [Reference Cooper19]. Salmonella serotype Braenderup H9812 was used as the molecular standard [Reference Hunter20]. Agarose plugs were loaded onto a 1% agarose gel (Bio-Rad) and electrophoresis was carried out by using a contour-clamped homogenous electrical field Mapper (CHEF-DR III) (Bio-Rad Laboratories, Richmond, CA, USA) PFGE System following the electrical parameters as described earlier [Reference Cooper19]. The PFGE gel was visualized and profile images were captured with Gel Doc XR (Bio-Rad Laboratories, Hercules, CA, USA) scanned and saved in the TIFF format using the Quantity One program (Bio-Rad Laboratories, Hercules, CA, USA) for analysis. The PFGE profiles were analysed and compared using the BioNumerics version 4.1 software (Applied Maths, Sint-Martens-Latem, Belgium). Similarity was based on the dice-coefficient and clustering analysis was performed and dendogram constructed with BioNumerics using the unweighted pair group-matching algorithm (UPGMA) with a band position tolerance of 1·5%. A profile that differed by at least one clear band was considered a distinct profile. 18–20 bands were considered for comparison. Patterns with less than a four-band difference were considered subtypes.
RESULTS
Phenotypic identification and detection of virulence genes
The 45 V. cholerae strains were confirmed to belong to the O1 serogroup and Ogawa was the dominant serotype. These strains showed phenotypic properties typical of the El Tor biotype.
PCR amplification for the detection of 28 major virulence genes (Table 3) showed that all of these strains had ctxA gene and El Tor specific tcpA gene proving that they were toxin producing strains of the El Tor biotype. The rstR gene amplified using biotype specific primers [Reference Nusrin21] successfully produced the 500 bp DNA fragment in all the strains with the rstR El Tor biotype specific primer pair while with rstR classical biotype specific primers there was no amplification of the rstR gene except in the classical standard control strain O395.
Table 3. Result of virulence gene profiling done with the aid of PCR
PCR for detection of CTXϕ in the small chromosome in these strains that was done using the set of primers CIIF and CIIR [Reference Maiti22] led to the amplification of 766 bp DNA fragment revealing the absence of a second copy of CTXϕ in the small chromosome.
PCR confirmed the presence of rstC, rtxA, rtxC genes and also a set of five genes (VCA0316, VCA0417, VCA0728, VCA0729, VCA0730) present only in El Tor strains and absent in classical strains [Reference Dziejman23].
Successful PCR amplification of tcpI, toxR, zot, ptlc, VC1449, VC1450, orfU, intl3, acf, nag, nan, rstA genes using previously described primer sequences [Reference O'Shea12] confirmed the presence of all these ORFs and their corresponding virulence regions in the genome. Likewise, amplification of the VC0183 gene confirmed the presence of VSP-I [Reference O'Shea12] while the successful amplification of VC0513, VC0514, VC0515 and VC0516 genes confirmed the presence of VSP-II region [Reference O'Shea12]. However, the absence of ORF VC0511 in 19 of these 45 strains was revealed by the failure to amplify the expected 385 bp amplicon using primer sequence as previously described [Reference O'Shea12].
Sequencing of ctxB to determine the genotype of CT in these strains
MAMA PCR revealed the presence of ctxB gene of the classical biotype in all the strains (Fig. 1). However, D-MAMA PCR showed that 16 of these strains failed to produce DNA product using primers ctxB-F4/Rv-cla [Reference Naha14] while they successfully produced 191 bp product (Fig. 2) with primer pair ctxB-F3/Rv-cla [Reference Naha14] that was exclusively designed to detect the single base change at the 58th nucleotide position [Reference Naha14] that was originally found in strains associated with a recent outbreak of cholera in Orissa [Reference Kumar24].
Fig. 1. Amplification of ctxB gene in 45 strains with biotype specific primers. (Upper panel) PCR with Re-cla primers showing 186 bp PCR product. From left O395 (classical control), N16961 El Tor control), 1-VOC1, 2-G1, 3-P3, 4-M24, 5-AFMC1, 6-IDH00563, 7-IDH00734, 8-IDH00725, 9-B1, 10-S4, 11-VOC27, 12-VOC30, 13-SRK1, 14-IDH00614, 15-4750. (Lower panel) PCR with Rv-elt primers. From left O395 (classical control), N16961 (El Tor control), 1-VOC1, 2-G1, 3-P3, 4-M24, 5-AFMC1, 6-IDH00563, 7-IDH00734, 8-IDH00725, 9-B1, 10-S4, 11-VOC27, 12-VOC30, 13-SRK1, 14-IDH00614, 15-4750. (Lower panel) PCR with Rv-elt primers.
Fig. 2. D-MAMA PCR with 7 representative strains. (Lanes from left): O395 (Classical control), N16961 (El Tor control), 2010EL-1786 (Haitian control), 1-IDH00563 (Kolkata, 2008), 2-IDH00725 (Kolkata, 2008), 3-IDH00734 (Kolkata, 2008), 4-VOC30 (Orissa, 2007), 5-VOC27 (Orissa, 2007), 6-P7 (Pune, 2007), 7-AFMC1 (Pune, 2008), Blank well, O395 (Classical control), N16961 (El Tor control), 2010EL-1786 (Haitian control), 1-IDH00563 (Kolkata, 2008), 2-IDH00725 (Kolkata, 2008), 3-IDH00734 (Kolkata, 2008), 4-VOC30 (Orissa, 2007), 5-VOC27 (Orissa, 2007), 6-P7 (Pune, 2007), 7-AFMC1 (Pune, 2008). Left side of the blank well PCR done, using ctxB F3/Rv-cla primer. Right side of the blank well PCR done, using ctxB F4/Rv-cla primer.
DNA sequencing of the 460 bp ctxB gene was carried out for comparing the sequence of nucleotides in these two groups of strains at the positions already designated to contain SNPs [Reference Olsvik16] and also for the confirmation of the results of D-MAMA PCR in this study. Sequence analysis revealed that all the strains had identical ctxB gene sequence like the classical standard control O395 with cytosine (C) at positions 115 and 203 [Reference Olsvik16] However, in the 16 strains that showed positive amplification result with D-MAMA PCR an expected base change was identified at the 58th nucleotide position wherein, the cytosine (C) at this position was found to be replaced with adenine (A) resulting in a change in the deduced amino acid sequence. Consequently, asparagine had replaced histidine at the 20th amino acid position (Table 4).
Table 4. Places in India from where ctxB genotype 7 has been isolated by ctxB sequencing of strains that caused cholera in these places in the years 2007 and 2008
Beads ELISA and comparison of CT production
Beads ELISA [Reference Oku18] showed that these strains produce CT in the range of 0 to 2774 ng. The replacement of histidine with asparagine at the 20th amino acid position in the CTB sequence of the 16 strains did not lead to any significant change associated with the amount of CT production as compared with the strains with amino acid histidine at this position.
Comparison of different clones by PFGE
PFGE was used to determine clonality of seven representative strains. From the dendogram that was constructed four distinct profiles, PI-PIV were observed among these seven strains (Fig. 3).
Fig. 3. PFGE gel showing four distinct NotI digestion profiles among 7 representative strains isolated from India in 2007 and 2008. Lanes-1-S. enterica serovar Braenderup strain H9812 (Marker), 2-VOC27, 3-VOC30, 4-P7, 5-AFMC1, 6-S.enterica serovar Braenderup strain H9812 (Marker), 7-IDHOO563, 8-IDH00725, 9-IDHOO734, 10-S.enterica serovar Braenderup strain H9812 (Marker).
These four major clusters were designated as PFI, PFII, PFIII and PFIV existed among the seven strains selected for the dendogram construction (Fig. 4). Approximately 97% similarity was found between PFI and PFII, 95% between PFII and PFIII and 92% between PFIII and PFIV.
Fig. 4. Dendogram showing clonal relationship among representative strains.
All of these seven strains had uniform phenotypic and genetic characteristics but still belonged to distinct PFGE profiles.
DISCUSSION
Cholera remains a public health threat in cholera endemic areas in India [25]. Among the 36 states and union territories almost every Indian state has witnessed outbreaks of cholera in the recent past. West Bengal, Orissa, Assam, Chattisgarh and the union territory of Andaman and Nicobar Islands reported the highest number of cases, accounting for 91%, according to a surveillance related study conducted over a period of 10 years [Reference Kanungo26]. In addition, underreporting overshadows actual number of cholera cases in India leading to discrepancies in the incidence of cholera [Reference Sarkar, Kanungo and Nair27].
At present, altered El Tor strains that were identified from isolates from 1993 onwards and replaced the prototype El Tor strains globally since 2001 [Reference Nair8] continue to cause cholera outbreaks in the country. Due to increasing mortality and morbidity caused by the pathogen quick epidemiological investigations based on molecular typing techniques is the need of the hour for rapid administration of control and preventive measures during an outbreak. Also, molecular typing methods are applied to resolve any ambiguity that may arise in the process of determining the biotype of the isolates. For example, previous studies showed the existence of strains like the Mozambique strains that are hybrids between the two biotypes and carry classical CTXϕ and also unique El Tor genes like VCA0728-VCA0730 and phenotypic properties of the El Tor biotype [Reference Faruque28]. In such contexts, molecular typing is highly advantageous.
This retrospective genetic analysis using molecular techniques is a valuable study that has compared genetic features of V. cholerae strains from diverse locations in India unlike other studies with Indian strains that have a more limited geographical coverage. It is a continuation of a previous study by Naha et al. [Reference Naha14] and has shown that the strains with ctxB-7 allele did not remain confined in terms of geographical distribution to Eastern India in Kolkata [Reference Naha14] or Orissa [Reference Kumar24] but were circulating in different cholera endemic regions in the country.
These strains have a single copy of CTXϕ in the genome. Difference in arrangement of the CTXϕ in the chromosome exists between classical and El Tor biotypes [Reference Faruque28]. Recent findings report the presence of two copies of the ctxB gene, one on each chromosome contrary to previous reports that ctxAB genes are located only in the large chromosome in El Tor biotype [Reference Son29]. Recent studies on El Tor variant strains from an outbreak in Hyderabad, India showed that one particular strain VCH35 carries two copies of the classical CTXϕ in tandem repeat in the small chromosome [Reference Goel30]. However, the organization and copy number of the CTXϕ in the strains included in this study is identical to that of prototype El Tor strains [Reference Ansaruzzaman31]. Also, these strains possess unique genes of the El Tor biotype. Classical and El Tor biotypes arose from different ancestors [Reference Salim, Lan and Reeves32] and vary distinctly in their variable genome [Reference O'Shea12]. Profiling of virulence genes can be used to trace the lineage of strains [Reference Safa33] thus helping in biotype identification. The genetic variation observed among these strains did not affect CT production nor had any correlation with PFGE profiles.
This study has also enabled one to get a glimpse into the genetic composition of Indian strains from the same period that is a milestone in the evolutionary history of the pathogen in several ways when the first reports of novel ctxB-7 allele [Reference Kumar24] and other crucial genetic changes had gradually started appearing in the contemporary V. cholerae strains worldwide as reported by several investigators [Reference Klinzing34, Reference Konstantin35].
ACKNOWLEDGEMENT
We sincerely thank Professor Yoshifumi Takeda for his constant support and advice during this work. The study was supported by Okayama University, Japan ; the Indian Council of Medical Research, New Delhi, India.
