Origins of the current seventh cholera pandemic

Abstract

Vibrio cholerae has caused seven cholera pandemics since 1817, imposing terror on much of the world, but bacterial strains are currently only available for the sixth and seventh pandemics. The El Tor biotype seventh pandemic began in 1961 in Indonesia, but did not originate directly from the classical biotype sixth-pandemic strain. Previous studies focused mainly on the spread of the seventh pandemic after 1970. Here, we analyze in unprecedented detail the origin, evolution, and transition to pandemicity of the seventh-pandemic strain. We used high-resolution comparative genomic analysis of strains collected from 1930 to 1964, covering the evolution from the first available El Tor biotype strain to the start of the seventh pandemic. We define six stages leading to the pandemic strain and reveal all key events. The seventh pandemic originated from a nonpathogenic strain in the Middle East, first observed in 1897. It subsequently underwent explosive diversification, including the spawning of the pandemic lineage. This rapid diversification suggests that, when first observed, the strain had only recently arrived in the Middle East, possibly from the Asian homeland of cholera. The lineage migrated to Makassar, Indonesia, where it gained the important virulence-associated elements Vibrio seventh pandemic island I (VSP-I), VSP-II, and El Tor type cholera toxin prophage by 1954, and it then became pandemic in 1961 after only 12 additional mutations. Our data indicate that specific niches in the Middle East and Makassar were important in generating the pandemic strain by providing gene sources and the driving forces for genetic events.

Keywords: Vibrio cholerae; comparative genomics; evolution; pandemic.

Fig. 1.

The maximum-likelihood phylogenetic tree of 21 V. cholerae strains based on mutational differences after excluding recombination events. This is a traditional tree that was rooted using the sixth-pandemic strain O395. The branches corresponding to the six stages in the evolution of the seventh pandemic are colored in violet. The branching data were estimated by BEAST (median value) and are shown at the nodes. The lengths of the branches correspond to the times between divergence from MRCA and isolation. Genetic events (mutation, insertion, deletion, and recombination) were allocated to and shown on specific branches. The MAK757, NCTC8457, BX330286, and 2740-80 branches are shown by dotted lines and were not included in BEAST for analysis because of the lack of suitable-quality sequence data, and the dates for their divergence from MRCA were predicted based on the SNP allocations of strains M66-2, NCTC5395, M2140, and E506, respectively. The strain NCTC9420 carried a mutator gene, and there is no SNP on branch 0, which is the outermost branch of the tree; therefore, both of these branches are also shown by dotted lines. The branches representing the evolutionary stages that occurred in the Middle East and Makassar are enclosed in boxes. See Fig. S1 for a tree with genetic events allocated to branches.

Fig. 2.

Details of the six stages leading to the seventh-pandemic strain. The stages on the path to the seventh-pandemic strain are shown in the horizontal bar, which comprises the branches shown in violet in Fig. 1. The branches to the prepandemic and prepandemic-related strains that demarcate the stages are shown below the bar, with details of genetic events given on each branch.

Fig. S1.

The details of mutations, recombination events, and indels on each branch of the phylogenetic tree shown in Fig. 1.

Fig. S2.

Phylogenetic tree for eight V. cholerae strains on the main lineage with complete genome sequences. The branches are numbered according to Fig. 1, with many corresponding to two or three branches in Fig. 1. The node dates are also taken from Fig. 1. The terminal branches’ ends relate to the isolation dates. All genetic events allocated to a specific branch are shown in curved brackets, with the data as recorded in Fig. S1 shown in square brackets. For branch 22, the genetic events include those found by comparison of the C5 and E9120 sequences, as discussed in the text.

Fig. S3.

Original timeframe tree generated by BEAST. The strains in Fig. S3 are the same as those in Fig. 1, except for MAK757, NCTC8457, BX330286, and 27400-80, which were not included in the BEAST analysis and are not shown here. The branching dates estimated by BEAST (median values) are shown at the nodes, and the branch ends relate to the isolation dates.

Fig. 3.

Migration of prepandemic strains and the locations and stages in the evolution of the seventh-pandemic strain. Three important locations in the origin and history of the seventh-pandemic strain are circled. The yellow line indicates the migration of the ancestral form; blue lines indicate migrations from the Middle East; red lines indicate migrations from Makassar; green lines indicate the early spread of the seventh pandemic from Bengal; dotted lines indicate the pathways of prepandemic-related strains to the US Gulf, Australia, and China. The six stages in the evolution of the seventh-pandemic strain are shown in the labels. The date ranges given for transmission events, which are taken from BEAST analysis, are also shown, as well as the locations of the prepandemic-related strains.

Fig. S4.

CTX regions and associated elements. The open arrows represent gene locations and orientations. The types of genes or elements were determined by nucleotide sequence comparisons. We have not included A6 here, as it is probably the same strain as our C5. However, it is important to note that, using the data as deposited, we found no support for the presence of two copies of the CTX gene in A6 or for the presence of the RS1 genes, as reported (18). Indeed, we found no convincing evidence for any difference in sequence from our C5 sequence.

Fig. S5.

Summary of genetic events on each branch of the phylogenetic tree shown in Fig. 1. The numbers in brackets are the numbers of events (SNPs, insertions, and deletions) that occurred in the large and small chromosomes.

Fig. 4.

The chromosomal locations of all mutation and recombination events on the major branches related to the six evolutionary stages of the seventh-pandemic strain. Mutations are marked as blue vertical bars above thick black lines, which represent the two chromosomes. The red vertical bars below show the locations of recombination events. The branch numbering is the same as in Fig. 1, but the mutations and recombination events shown include those back to the nodes. The additional events are shown on their own branches in Fig. S6, which includes all branches. Note that there are recombination events on all stage branches except for stage 6, which is short and has only 12 mutations, whereas the branch from the start of the pandemic to strain B33 has no recombination events. Additionally, for the early part of the 1937 outbreak, branch 12 has recombination events, but there are none on the terminal branches 13 and 14.

Fig. S6.

The chromosomal locations of all mutation and recombination events on each branch of the phylogenetic tree shown in Fig. 1.