Multilocus sequence typing system for the endosymbiont Wolbachia pipientis

Abstract

The eubacterial genus Wolbachia comprises one of the most abundant groups of obligate intracellular bacteria, and it has a host range that spans the phyla Arthropoda and Nematoda. Here we developed a multilocus sequence typing (MLST) scheme as a universal genotyping tool for Wolbachia. Internal fragments of five ubiquitous genes (gatB, coxA, hcpA, fbpA, and ftsZ) were chosen, and primers that amplified across the major Wolbachia supergroups found in arthropods, as well as other divergent lineages, were designed. A supplemental typing system using the hypervariable regions of the Wolbachia surface protein (WSP) was also developed. Thirty-seven strains belonging to supergroups A, B, D, and F obtained from singly infected hosts were characterized by using MLST and WSP. The number of alleles per MLST locus ranged from 25 to 31, and the average levels of genetic diversity among alleles were 6.5% to 9.2%. A total of 35 unique allelic profiles were found. The results confirmed that there is a high level of recombination in chromosomal genes. MLST was shown to be effective for detecting diversity among strains within a single host species, as well as for identifying closely related strains found in different arthropod hosts. Identical or similar allelic profiles were obtained for strains harbored by different insect species and causing distinct reproductive phenotypes. Strains with similar WSP sequences can have very different MLST allelic profiles and vice versa, indicating the importance of the MLST approach for strain identification. The MLST system provides a universal and unambiguous tool for strain typing, population genetics, and molecular evolutionary studies. The central database for storing and organizing Wolbachia bacterial and host information can be accessed at http://pubmlst.org/wolbachia/.

FIG. 1.

Map of the wMel chromosome, showing the locations of the five MLST loci and wsp (inner circle). The outer and middle circles show the open reading frames on the plus and minus chromosomal strands, respectively.

FIG. 2.

Pairwise divergence at MLST loci. Thirty-three strain pairs identical at one or two loci were used to show the variation in similarity across loci. The most striking pattern was found for strains identical at gatB (vertical bar); some of the strains showed remarkable divergence in the fbpA and coxA alleles, likely associated with recombination events. The arrows indicate examples of contradictory relationships inferred from the results for the five genes involving N. vitripenni (B), C. alternans, and A. encedon strains (values are circled; see text for explanation).

FIG. 3.

Bayesian likelihood inference phylogeny based on the concatenated data set for the five MLST loci (37 strains, 2,079 bp). Groups of strains sharing at least three MLST alleles are highlighted. Arrows and asterisks indicate two examples of strain pairs whose predicted relationships are highly discordant with the wsp phylogeny (Fig. 4). Posterior probability (left) and parsimony bootstrap (right) values are indicated at major nodes if they were supported by both clustering algorithms.

FIG. 4.

Bayesian likelihood inference phylogeny based on wsp sequences of the 36 strains analyzed by MLST (without the T. deion strain). ST complexes and strains with identical STs identified by MLST are highlighted. Arrows and asterisks indicate two examples of strain pairs whose predicted relationships are highly discordant with the MLST phylogeny (Fig. 4). The supergroup placement of the A. sparsa and S. invicta strains based on MLST phylogeny (in parentheses) (Fig. 3) is not supported by the wsp-based phylogeny. Posterior probability (left) and parsimony bootstrap (right) values are shown at major nodes if they were supported by both clustering algorithms.

FIG. 5.

Neighbor-joining tree based on the MLST allelic profiles (right columns) of 37 Wolbachia strains. ST complexes and strains with identical STs are highlighted (STs shared by strains are indicated by arrowheads). Only closely related strains predicted by MLST were also predicted by the concatenated MLST phylogeny (Fig. 3), WSP profiles (Fig. 6), and wsp phylogeny (Fig. 4).

FIG. 6.

Unweighted pair group method with arithmetic mean dendrogram based on the WSP profiles (right columns) of 36 Wolbachia strains, corresponding to the same sample typed by MLST (without the T. deion strain). Groups of strains sharing at least three HVR haplotypes are highlighted. In some cases groups of strains identified by MLST (Fig. 5, highlighted strains) were also closely related based on their WSP profiles. In strains belonging to the ST-13 complex there was a clear recombination signature in HVR4, which resulted in remarkable genetic divergence in this region among the five strains.