Vibrio cholerae Classical Biotype Strains Reveal Distinct Signatures in Mexico

Abstract

Vibrio cholerae O1 classical (CL) biotype caused the fifth and sixth pandemics, and probably the earlier cholera pandemics, before the El Tor (ET) biotype initiated the seventh pandemic in Asia in the 1970s by completely displacing the CL biotype. Although the CL biotype was thought to be extinct in Asia and although it had never been reported from Latin America, V. cholerae CL and ET biotypes, including a hybrid ET, were found associated with areas of cholera endemicity in Mexico between 1991 and 1997. In this study, CL biotype strains isolated from areas of cholera endemicity in Mexico between 1983 and 1997 were characterized in terms of major phenotypic and genetic traits and compared with CL biotype strains isolated in Bangladesh between 1962 and 1989. According to sero- and biotyping data, all V. cholerae strains tested had the major phenotypic and genotypic characteristics specific for the CL biotype. Antibiograms revealed the majority of the Bangladeshi strains to be resistant to trimethoprim-sulfamethoxazole, furazolidone, ampicillin, and gentamicin, while the Mexican strains were sensitive to all of these drugs, as well as to ciprofloxacin, erythromycin, and tetracycline. Pulsed-field gel electrophoresis (PFGE) of NotI-digested genomic DNA revealed characteristic banding patterns for all of the CL biotype strains although the Mexican strains differed from the Bangladeshi strains in 1 to 2 DNA bands. The difference was subtle but consistent, as confirmed by the subclustering patterns in the PFGE-based dendrogram, and can serve as a regional signature, suggesting the pre-1991 existence and evolution of the CL biotype strains in the Americas, independent from Asia.

INTRODUCTION

Vibrio cholerae, the causative agent of cholera, is one of the most successful emerging and reemerging pathogens that has both human and environmental components in its life cycle (23). There are more than 200 distinct serogroups of V. cholerae, but only O1 and O139 are recognized as being responsible for epidemic and pandemic cholera (17). V. cholerae O1, which is the predominant pandemic serogroup to date, is further differentiated into two well-established biotypes, namely, the classical (CL) and El Tor (ET). This classification is based primarily on several phenotypic properties, including hemolysis of sheep red blood cells, agglutination of chicken red blood cells, Voges-Proskauer reaction, and susceptibility to polymyxin B (50 U) and biotype-specific phages (17). In addition to phenotypic traits, major genetic markers have also been used to determine the biotypes of V. cholerae strains, such as the major toxin coregulated pilus gene (tcpA), the gene encoding the cholera toxin subunit B (ctxB), and the repeat sequence transcriptional regulator (rstR) gene, all of which possess CL- and ET-specific alleles, and the presence or absence of the repeat in toxin (rtxC) gene (28). Recently, two genomic islands, Vibrio seventh pandemic island I (VSP-I) and VSP-II, were shown to be unique to the ET strains of the seventh pandemic, in that they were absent from both the pre-seventh pandemic ET strains and the CL biotype strains (11).

Since 1817, cholera has become pandemic, spreading into a multitude of countries. The sixth pandemic and presumably earlier cholera pandemics were caused by the V. cholerae CL biotype, while the most extensive, and ongoing, seventh pandemic, which started in 1961, is caused by the ET biotype (15). It was following the sixth pandemic that the V. cholerae CL biotype was gradually replaced by El Tor, and within a decade the CL biotype seemed to have disappeared (3, 4). However, the CL biotype reemerged as the predominant epidemic biotype in some areas of Bangladesh in 1982 and coexisted with El Tor strains for a few years until the late 1980s (29). In spite of having a mysterious history of disappearance and reappearance in Bangladesh, the CL biotype was thought to have become extinct, with its last isolation being recorded in 1992 in southern Bangladesh (13, 17).

When the cholera epidemic struck Peru in January 1991 and spread rapidly to other Latin American countries, reaching Mexico in June of the same year, the causal agent, V. cholerae O1 El Tor, was homogenous and considered to be an extension of the seventh pandemic (30), presumably arriving via ships from areas of cholera endemicity (20). This surmise was probably based on the observation that there had been no sign of cholera in South, Central, or North America for more than a century. Recently, however, V. cholerae El Tor hybrid variant strains were shown to be dominant among V. cholerae CL and ET biotype progenitor strains associated with endemic cholera in Mexico between 1991 and 1997 (2). Although the pre- and post-1991 existence of V. cholerae CL biotype strains along with ET hybrid and prototype ET strains up until 1997 was suggestive of a local environmental reservoir in Mexico (2), little is known about their phenotypic and genetic traits and, more specifically, their ancestry to discern whether the CL biotype strains isolated from endemic cholera in Mexico were evolving locally or were of Asian origin. In the present study, comparative phenotypic, molecular, and phylogenetic data related to V. cholerae CL biotype strains isolated in Mexico (n = 6; 1983 to 1997) and Bangladesh (n = 22; 1962 to 1989) have been analyzed, and the results show that the V. cholerae CL biotype strains may have evolved independently in the two geographically distant ecosystems.

MATERIALS AND METHODS

Bacterial strains.

V. cholerae O1 classical (CL) biotype strains (n = 28) involved in this study included 6 strains isolated in Mexico (1983 to 1997) and 22 strains isolated in Bangladesh (1962 to 1989). The V. cholerae reference CL strain O395 and reference El Tor (ET) strain N16961 were also included for comparison. The Mexican isolates were obtained from the Department of Public Health, Faculty of Medicine, National Autonomous University of Mexico (UNAM). These strains were isolated from cholera patients (n = 4) and surface water (n = 2) as part of the nationwide cholera surveillance program between 1983 and 1997 (6). Bangladeshi V. cholerae CL biotype strains were obtained from the culture collection at the International Centre for Diarrheal Disease Research, Bangladesh (ICDDR,B). The bacterial strains were shipped in soft agar, confirmed using standard culture methods, and identified by a combination of biochemical and molecular methods, as described previously (1).

Genomic DNA preparation.

Genomic DNA extraction was carried out following previously described methods (24). Briefly, cells were harvested by centrifugation, and the pellets were resuspended in TES buffer (10 mM Tris-HCl, pH 8.0, 10 mM EDTA, 100 mM NaCl, 10% SDS) and incubated at 65°C for 10 min. Following proteinase K treatment at 50°C for 18 h, DNA was extracted with phenol-chloroform-isoamyl alcohol (25:24:1), precipitated in ethanol, and dissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). RNase treatment was performed at 37°C for 1 h, and ethanol-precipitated DNA was dissolved in TE buffer and stored at −20°C.

Serogrouping.

The serogroups of the V. cholerae isolates that were identified using biochemical and molecular methods were confirmed serologically by a slide agglutination test using specific polyvalent antisera to V. cholerae O1 and O139, followed by a monoclonal antibody specific for each serogroup (1). The subtypes of all the strains were reconfirmed using a V. cholerae species-specific ompW PCR (22). The serogroups of these strains were reconfirmed using polyvalent O1 and monovalent Inaba and Ogawa antisera and by multiplex PCR targeted to identify genes encoding O1 (wbe)- and O139 (wbf)-specific O-biosynthetic genes and the cholera toxin (CTX) gene (ctxA) (16).

Confirmation of biotype.

Biotyping involved a number of phenotypic tests including sensitivity to polymyxin B and Mukherjee CL phage IV and Mukherjee ET phage V tests (17). Classical strains are typically sensitive to polymyxin B and type IV phage but resistant to type V phage.

In order to complement biotype characterization using phenotypic traits, PCR assays were also carried out targeting various biotype-determining genes. The major subunit protein type of the toxin-coregulated pilus (TCP) gene tcpA was determined using specific primers (26), which can differentiate between CL and ET alleles using an amplified product size of 620 bp and 451 bp, respectively. The repeat-sequence transcriptional regulator gene (rstR), which contains CL- and ET-specific alleles, was detected using primers described previously (18). The rtxC gene of the RTX toxin gene cluster has been considered to be a significant marker for differentiating biotypes and is present in ET but not in CL biotype strains. In this study, the rtxC gene was amplified using primers described by Chow et al. to produce a 203-bp PCR product (8). The cholera toxin (CTX) subunit B gene (ctxB) allele was detected by a mismatch amplification mutation assay (MAMA) PCR specific for the CL and ET biotypes involving a conserved forward primer and two allele-specific polymorphism detection primers under PCR conditions as described previously (21).

Antimicrobial susceptibility.

Bacterial susceptibility to antimicrobial agents was determined by standard disc diffusion methods (5, 9) using commercially available discs (Oxoid International). A total of seven antibiotics were used: erythromycin (E; 15 μg), tetracycline (TE; 30 μg), gentamicin (CN; 10 μg), ciprofloxacin (CIP; 5 μg), trimethoprim-sulfamethoxazole (SXT; 30 μg), ampicillin (AMP; 30 μg), and furazolidone (FR; 100 μg).

PFGE.

The whole agarose-embedded genomic DNA from V. cholerae was prepared, and pulsed-field gel electrophoresis (PFGE) was carried out with a contour-clamped homogeneous electrical field (CHEF-DRII) apparatus (Bio-Rad) according to procedures described elsewhere (7). The conditions used for separation were as follows: 2 to 10 s for 13 h, followed by 20 to 25 s for 6 h. An electrical field of 6 V/cm was applied at an included field angle of 120°. Genomic DNA of the test strains was digested by the NotI restriction enzyme (Gibco-BRL, Gaithersburg, MD), and Salmonella enterica serovar Braenderup was digested by XbaI, with the fragments being used as molecular size markers. The restriction fragments were separated in 1% pulsed-field-certified agarose in 0.5× TBE (Tris-borate-EDTA) buffer. In the post-electrophoresis gel treatment step, the gel was stained and destained. DNA was visualized using a UV transilluminator, and images were digitized by a one-dimensional gel documentation system (Bio-Rad).

Image analysis.

The fingerprinting pattern in the PFGE gel was analyzed using the computer software package BioNumerics (Applied Maths, Belgium). After background subtraction and gel normalization, the fingerprint patterns were subjected to typing on the basis of banding similarity and dissimilarity using the Dice similarity coefficient and clustering based on the unweighted-pair group method using average linkages (UPGMA), as recommended by the software manufacturer, and results are graphically represented as dendrograms.

Sequencing of the ctxB gene.

The ctxB gene was amplified from the strains using ctxF and ctxR primers as described previously (25). The PCR product was purified, and sequencing was carried out using the same primers on a 96-capillary model 3730xl system with a BigDye Terminator kit (Applied Biosystems). The sequences of the ctxB gene for the CL reference strains were retrieved from GenBank. The deduced amino acid sequences of the ctxB gene from all strains were aligned using Clustal W.

RESULTS AND DISCUSSION

The classical (CL) biotype, causing the sixth cholera pandemic, was believed to be extinct even from the Bengal region of south Asia, where cholera has been endemic for centuries and where V. cholerae has been established as autochthonous flora of the aquatic ecosystem. Together with El Tor (ET) and variant ET strains, the CL biotype strains were found to be associated with endemic cholera in Mexico between 1991 and 1997 (2). Since there had been no significant cholera outbreak for almost 100 years prior to the 1991 epidemic, little is known about the ancestry of these causative pathogens, especially if there was a local reservoir for pathogens in the Americas or if the pathogens were introduced from Asia, as has been proposed for ET cholera in 1991 (20, 30). In this study, we present comparative phenotypic and molecular data on V. cholerae CL biotype strains isolated in Mexico and Bangladesh and show evidence of the independent evolution of CL biotype strains in the two geographically distinct ecosystems.

The V. cholerae CL biotype strains isolated in Mexico and Bangladesh showed biochemical and molecular properties typical of the V. cholerae CL reference control O395. The strains were further confirmed serologically by a slide agglutination test using polyvalent antisera followed by a monoclonal antibody specific for V. cholerae serogroup O1 (1). Strain origin, year of isolation, and phenotypic results are shown in Table 1. All strains tested, irrespective of their origin, were sensitive to polymyxin B and phage IV and resistant to phage V, which are typical phenotypic properties of CL biotype strains although V. cholerae O1 El Tor and CL biotype strains have shown variable reactions to biotype-specific phages and polymyxin B in previously reported data (28).

Table 1.

Phenotypic and genotypic traits of V. cholerae O1 classical biotype strains isolated in Mexico and Bangladesh

Test strain origin or reference strain name (n)a	Year(s) of isolation	No. of isolates	Sourceb	Phenotypic response to:c			Genotypic traitd				Antibiotic resistance profilef	PFGE type (no. of isolates)
				Polymyxin B (50 U)	Phage IV	Phage V	tcpA type	ctxBe type	rstR type	rtxC
Bangladesh (22)	1962–1971	7	Clin	S	S	R	CL	CL	CL	−	Sensitive	I
	1982–1989	6	Clin	S	S	R	CL	CL	CL	−	AMP, CN, FR, SXT	I (5), Ib (1)
		3	Clin	S	S	R	CL	CL	CL	−	AMP, FR, SXT	I
		6	Clin	S	S	R	CL	CL	CL	−	Sensitive	I
Mexico (6)	1983–1995	2	Clin	S	S	R	CL	CL	CL	−	Sensitive	IIa
		2	Env	S	S	R	CL	CL	CL	−	Sensitive	IIa
	1997	2	Clin	S	S	R	CL	CL	CL	−	Sensitive	IIc (1), I (1)
O395	1965		Clin	S	S	R	CL	CL	CL	−
N16961	1971		Clin	R	R	S	ET	ET	ET	+

^an, number of strains.

^bClin, clinical; Env, environmental.

^cR, resistant; S, sensitive.

^dET, El Tor; CL, classical.

^eDetermined by MAMA PCR (21).

^fAMP, ampicillin; CN, gentamicin; FR, furazolidone; SXT, trimethoprim-sulfamethoxazole; sensitive, sensitive to all of the drugs in this group (AMP, CN, FR, and SXT).

Antimicrobial resistance of V. cholerae to effective drugs can hinder the treatment of the disease (27), affecting the recovery of patients and increasing the chance of death. In the present study, V. cholerae strains isolated in Mexico and Bangladesh were subjected to antimicrobial susceptibility assays (5) against ciprofloxacin (CIP), erythromycin (E), tetracycline (TE), trimethoprim-sulfamethoxazole (SXT), furazolidone (FR), ampicillin (AMP), and gentamicin (CN). As shown in Table 1, V. cholerae CL biotype strains isolated in Mexico were found to be sensitive to all seven antibiotics tested although the V. cholerae CL biotype strains isolated in Bangladesh varied in their responses toward the various drugs. Year-based data show that all V. cholerae CL biotype strains isolated in Bangladesh between 1962 and 1971 were sensitive to all seven drugs, while the majority of the strains isolated between 1982 and 1989 showed multidrug resistance (MDR); six strains were resistant to AMP, CN, FR, and SXT, and three were resistant to AMP, FR, and SXT. In a study carried out in Bangladesh during the early 1980s, V. cholerae O1 strains were shown to be MDR, having the resistance markers for AMP, kanamycin, streptomycin, spectinomycin, TE, sulfonamides, and trimethoprim (14). Although antibiotic resistance may not be a stable marker to be maintained and transferred vertically and spatiotemporally, the differences observed in the antibiotic resistance patterns of V. cholerae CL biotype strains isolated in Bangladesh (1982 to 1989) and Mexico (1983 to 1997) may be due to independent evolution of V. cholerae in two geographically distant ecosystems.

Results of simplex and multiplex PCR screening (17) for V. cholerae species-specific ompW and wbe genes encoding serogroup O1-specific surface antigen further confirmed all of the tested strains to be V. cholerae O1. These strains were toxigenic since all of them amplified primers for ctxA that encodes subunit A of the cholera toxin (CTX). The absence of the ET biotype-specific rtxC gene and presence of the CL biotype-specific alleles of the tcpA and rstR genes (Table 1) also confirmed that these V. cholerae O1 strains belonged to the CL biotype. The CL biotype-specific ctxB allele was confirmed by MAMA PCR (21) followed by sequencing of the ctxB amplicons (2). Overall results of phenotypic and genetic screening confirmed that all of the strains were the V. cholerae CL biotype of the sixth pandemic prototype.

Genomic characterization of the V. cholerae CL biotype strains isolated in Bangladesh was reported elsewhere using restriction endonuclease cleavage patterns of the rRNA gene (13). In the current study, PFGE of NotI restriction-digested genomic DNA was performed (7) to determine the genetic fingerprint of V. cholerae CL biotype strains isolated in two geographically distinct regions, Mexico and Bangladesh. The NotI restriction enzyme digested the genomic DNA of both the test and CL reference control strains into 20 to 23 fragments, the sizes of which ranged from 20 to 350 kb. All of the test strains presented overall banding patterns that were characteristic of V. cholerae CL biotype strains (O395), further confirming their CL biotype attribute. Overall PFGE analysis revealed the CL biotype strains to have two major banding patterns (pulsotypes), I and II, and pattern II had three different subtypes, IIa to IIc (Table 1). Four strains isolated in Mexico between 1991 and 1995, including the only strain isolated in 1983, showed an identical banding pattern (pulsotype IIa) that was different from the PFGE patterns (pulsotypes I and IIb) of strains isolated in Bangladesh. The predominant Mexican strains with pulsotype IIa had a unique DNA band of around 115 kb that was consistently absent from all of the V. cholerae CL biotype strains isolated in Bangladesh, including the reference CL strain O395 from India (Fig. 1). These results reflect genetic divergence that could serve as a distinct regional signature for the Mexican and Bangladeshi CL biotype strains. The two remaining Mexican strains isolated in 1997 produced slightly different patterns; one belonged to pulsotype I, which is the predominant banding pattern for Bangladeshi strains isolated between 1962 and 1989. The other strain, with an extra band of around 80 kb (designated pulsotype IIc), differed from all test strains including the CL reference strain, suggesting a unique molecular signature for this strain. The Mexican strain that matched in the PFGE pattern with the Bangladeshi strains cannot be clonal in the true sense because this strain had a distinctly different spatiotemporal origin and likely arose from genetically diverse V. cholerae populations involved with the 1991 epidemic and the subsequent endemic cholera in Mexico (2, 19).

Fig 1.

Dendrogram showing genomic fingerprints of patterns of V. cholerae O1 CL biotype strains isolated in Mexico (1983 to 1997) and Bangladesh (1962 to 1989). The dendrogram was prepared by Dice similarity coefficient and UPGMA clustering methods by using PFGE images of NotI-digested genomic DNA; for this two PFGE gels were run under the same conditions. The scale bar at the top (left) indicates the correlation coefficient (%). The two major clusters separated the Mexican V. cholerae CL biotype strains (cluster B) from V. cholerae CL biotype strains isolated in Bangladesh (cluster A), suggesting a regional signature and independent evolution in the two geographically distinct ecosystems. INCL, Inaba classical; OGCL, Ogawa classical; Clin, clinical; Env. environmental.

Cluster analysis by dendrogram (UPGMA clustering method) of the gel images separated the V. cholerae CL biotype strains into two major clusters, A and B. Four strains isolated from Mexico before 1997 showed similar banding patterns and belonged to cluster B, whereas all but one of the Bangladeshi V. cholerae CL biotype strains having a regional signature banding pattern belonged to a separate cluster A that also included the reference CL strain O395 from India and one Mexican V. cholerae CL biotype strain isolated in 1997. Although differences in the PFGE banding patterns were subtle, a clear dichotomy was observed in the clustering patterns, with the majority of the Mexican V. cholerae CL biotype strains forming a regional cluster separating them from the V. cholerae CL biotype strains isolated in Bangladesh. This further supports the hypothesis of independent strain evolution in the two geographically distinct ecosystems found in Mexico and Bangladesh.

Although cholera was said to be absent in Latin America for nearly more than a century before the 1991 epidemic and although the causal agent V. cholerae ET was thought to be introduced from regions of cholera endemicity in Asia and Africa (20, 30), a few sporadic cases reported in the Americas between 1973 and 1992, including the U.S. Gulf Coast and one case reported in 1983 in Cancun, Mexico (2, 6), do suggest a local origin. Endemic cholera caused by a genetically divergent population of V. cholerae strains in South (10, 12, 24) and Central America (2, 19) strongly suggests that V. cholerae has long been adapted to the aquatic ecosystems of these regions but remained unreported due to a lack of continuous surveillance studies as cholera cases were sporadic and considered to be unremarkable. The epidemics in Peru and Mexico seem to be independent events, as revealed by the pre-1991 presence of a V. cholerae CL biotype strain that was found to be associated with endemic cholera, together with the ET and ET variant strains in Mexico between 1991 and 1997 (2). It can be assumed that the pre- and post-1991 CL biotype strains circulating in the endemic cholera strains in Mexico may have contributed significantly to the dynamics of V. cholerae populations in this region. Even though the phenotypic and genetic differences observed between the CL biotype strains isolated in Mexico and Bangladesh were perhaps subtle, the differences were consistent and thus can serve as regional signatures, which supports the hypothesis of independent CL biotype strain evolution in Mexico.