Characterization of Vibrio cholerae O1 El Tor biotype variant clinical isolates from Bangladesh and Haiti, including a molecular genetic analysis of virulence genes

Abstract

Vibrio cholerae serogroup O1, the causative agent of the diarrheal disease cholera, is divided into two biotypes: classical and El Tor. Both biotypes produce the major virulence factors toxin-coregulated pilus (TCP) and cholera toxin (CT). Although possessing genotypic and phenotypic differences, El Tor biotype strains displaying classical biotype traits have been reported and subsequently were dubbed El Tor variants. Of particular interest are reports of El Tor variants that produce various levels of CT, including levels typical of classical biotype strains. Here, we report the characterization of 10 clinical isolates from the International Centre for Diarrhoeal Disease Research, Bangladesh, and a representative strain from the 2010 Haiti cholera outbreak. We observed that all 11 strains produced increased CT (2- to 10-fold) compared to that of wild-type El Tor strains under in vitro inducing conditions, but they possessed various TcpA and ToxT expression profiles. Particularly, El Tor variant MQ1795, which produced the highest level of CT and very high levels of TcpA and ToxT, demonstrated hypervirulence compared to the virulence of El Tor wild-type strains in the infant mouse cholera model. Additional genotypic and phenotypic tests were conducted to characterize the variants, including an assessment of biotype-distinguishing characteristics. Notably, the sequencing of ctxB in some El Tor variants revealed two copies of classical ctxB, one per chromosome, contrary to previous reports that located ctxAB only on the large chromosome of El Tor biotype strains.

Fig. 1.

Voges-Proskauer (VP) results for acetoin production by the El Tor variants. V. cholerae classical biotype strain O395 is negative for the VP test. El Tor biotype strains C6706 and N16961 are bright pink, which is characteristic of a positive VP test result. The El Tor variants in this study all show various degrees of no pink to slight pink color, which is interpreted as a negative or weakly positive VP result, respectively. For comparison, VP-negative E. coli S17-1λpir also is shown.

Fig. 2.

Enzyme-linked immunosorbent assay (ELISA) for total cholera toxin production by El Tor variant strains. Total cholera toxin (CT) produced by wild-type classical O395, wild-type El Tor C6706, wild-type El Tor N16961, and the El Tor variants Bgd1, Bgd2, Bgd3, Bgd4, Bgd5, Bgd6, Bgd7, Bgd8, MQ1795, MQ04, and BAA-2163 under AKI-inducing conditions was determined by a colorimetric ELISA. The CT assay was conducted in triplicate, and the mean CT produced (ng CT ml⁻¹ OD₆₀₀⁻¹) is shown with standard deviations for each strain. A two-tailed standard t test yielded P values of <0.05 when the El Tor variants were compared to wild-type El Tor C6706 (*) and wild-type El Tor N16961 (**).

Fig. 3.

Immunoblot assays and densitometry for V. cholerae virulence factors TcpA and ToxT. Whole-cell extracts (WCE) were prepared, and 8 and 10 μg of total protein for TcpA and ToxT, respectively, was loaded onto a 16% Tris-glycine polyacrylamide gel. The resulting band intensities were quantitated and normalized against wild-type El Tor N16961. Anti-TcpA (A) or anti-ToxT (B) antibody was used to probe for TcpA and ToxT, respectively, in wild-type classical O395, wild-type El Tor C6706, wild-type El Tor N16961, and El Tor variants Bgd1, Bgd2, Bgd3, Bgd4, Bgd5, Bgd6, Bgd7, Bgd8, MQ1795, MQ04, and BAA-2163. O395ΔtcpA and O395ΔtoxT are shown as controls. Nonspecific bands serving as loading controls (LC) are shown below each respective immunoblot.

Fig. 4.

Determining the relative virulence of the El Tor variant strains using the infant mouse cholera model. The El Tor variant strains were individually inoculated (infectious dose of ∼7.5 × 10⁸) into infant mice (n = 6) to determine the length of time required before half the population succumbs to infection (LT₅₀) (inset). Wild-type El Tor C6706 and N16961 also were separately inoculated as controls and to determine the relative virulence (hypo-, equi-, or hypervirulent) of the El Tor variant strains MQ1795, MQ04, and BAA-2163 (A) and Bgd2, Bgd3, and Bgd8 (B).

Fig. 5.

Comparative schematic illustrating the hypothesized arrangement of the CTXφ integration in some of the El Tor variants. The orientation of the ctxAB operon in the large and small chromosomes of wild-type classical biotype O395 (A), wild-type El Tor biotype N16961 (B), and some of the El Tor biotype variants (MQ1795, MQ04, Bgd2, Bgd7 and Bgd8) (C) are shown. The ctxB gene was amplified in the El Tor variants by PCR using classical large and small chromosome-specific primers (primers 57 and 56, respectively). PCR and sequencing revealed that cxtB was present on both the large and small chromosomes of the variants, and both copies were that of the classical ctxB sequence (red asterisk, H39 in classical and El Tor variants and Y39 in El Tor N16961; blue asterisk, T68 in classical and El Tor variants and I68 in El Tor N16961). Relative attL and attR sequences (attB in El Tor N16961 small chromosome) are shown as shaded boxes. Identical sequences across the strains are color-coded and indicated with size (in bp). Positions of primers used for PCR and sequencing (primers 39, 51, 56, and 57) all are indicated with directional green arrows.

Fig. 6.

Phylogenetic analysis of the El Tor variants. (A) The four genes recA, pyrC, gyrB, and groEL were used to determine the phylogenetic relatedness of the El Tor variants MQ1795, MQ04, Bgd1, Bgd2, Bgd3, Bgd4, Bgd5, Bgd6, Bgd7, Bgd8, and BAA-2163 to wild-type El Tor biotype strain N16961 and classical biotype strain O395. The environmental strain, 12129, also was included in panel A for comparative purposes. Phylogenetic relatedness of the variants was compared to both O1 classical and El Tor wild-type strains with respect to ctxB (B) and tcpA (C and D) with the promoter region. (D) The clades generated in the tcpA branch showing the divergence of the variants and wild-type N16961 were further resolved. The bootstrap number at the node indicates the confidence for the clade grouping as determined by MEGA5 software. The scale bar represents the branch distance between the different strains.