Multi locus sequence typing of Chlamydiales: clonal groupings within the obligate intracellular bacteria Chlamydia trachomatis

Abstract

Background: The obligate intracellular growing bacterium Chlamydia trachomatis causes diseases like trachoma, urogenital infection and lymphogranuloma venereum with severe morbidity. Several serovars and genotypes have been identified, but these could not be linked to clinical disease or outcome. The related Chlamydophila pneumoniae, of which no subtypes are recognized, causes respiratory infections worldwide. We developed a multi locus sequence typing (MLST) scheme to understand the population genetic structure and diversity of these species and to evaluate the association between genotype and disease.

Results: A collection of 26 strains of C. trachomatis of different serovars and clinical presentation and 18 strains of C. pneumoniae were included in the study. For comparison, sequences of C. abortus, C. psittaci, C. caviae, C. felis, C. pecorum (Chlamydophila), C. muridarum (Chlamydia) and of Candidatus protochlamydia and Simkania negevensis were also included. Sequences of fragments (400 - 500 base pairs) from seven housekeeping genes (enoA, fumC, gatA, gidA, hemN, hlfX, oppA) were analysed. Analysis of allelic profiles by eBurst revealed three non-overlapping clonal complexes among the C. trachomatis strains, while the C. pneumoniae strains formed a single group. An UPGMA tree produced from the allelic profiles resulted in three groups of sequence types. The LGV strains grouped in a single cluster, while the urogenital strains were distributed over two separated groups, one consisted solely of strains with frequent occurring serovars (E, D and F). The distribution of the different serovars over the three groups was not consistent, suggesting exchange of serovar encoding ompA sequences. In one instance, exchange of fumC sequences between strains of different groups was observed. Cluster analyses of concatenated sequences of the Chlamydophila and Chlamydia species together with those of Candidatus Protochlamydia amoebophila and Simkania negevensis resulted in a tree identical to that obtained with 23S RNA gene sequences.

Conclusion: These data show that C. trachomatis and C. pneumoniae are highly uniform. The difference in genetic diversity between C. trachomatis and C. pneumoniae is in concordance with a later assimilation to the human host of the latter. Our data supports the taxonomy of the order of Chlamydiales.

Figure 1

Clonal groupings among C. trachomatis strains. Allelic profiles were analysed by eBurst and groups were defined as sets of related strains sharing identical alleles at six of the seven loci loci with at least one other member of the group. Blue dot in group I indicate the putative founder, yellow dot that of a subgroup.

Figure 2

Phylogenetic analyses of seven housekeeping gene fragments of C. trachomatis strains. A) The tree was constructed using UPGMA algorithm in SplitsTree4 using MLST allelic profiles. Distance matrix was obtained from allelic profiles using the SplitsTree program. B) UPGMA cluster analyses, with Jukes-Cantor correction, using concatenated sequences. Bold numbers indicate bootstrap values over 50%. Horizontal lines are scale for genetic distance.

Figure 3

SplitsTree decomposition analyses of MLST data of C. trachomatis strains. A) SplitsTree decomposition network was obtained using distance matrix obtained from allelelic profiles in SplitsTree4. B) SplitsTree decomposition network was obtained using distance matrix obtained using concatenated sequences.

Figure 4

Nucleotide substitutions among the three fumC alleles of C. trachomatis. Sequences between position 268 and 364 with the three nucleotide substitutions are shown. Allele 1 (ST5) has a nucleotide substitution at position 137 (not shown).

Figure 5

Phylogenetic analyses of concatenated sequences of 6 housekeeping gene fragments. A) Concatenated sequences of six housekeeping gene fragments were aligned and analysed in MEGA 3.1. Phylogenetic tree was constructed using the Neighbour-Joining algorithm with Kimura-2 parameter. Bootstrap test was for 1000 repetitions. Bold numbers indicate bootstrap values over 50%. Horizontal lines are scale for genetic distance. B) Phylogenetic tree based 23S rRNA gene sequences (Adapted from Everett et al [10]).