Comprehensive global genome dynamics of Chlamydia trachomatis show ancient diversification followed by contemporary mixing and recent lineage expansion

Abstract

Chlamydia trachomatis is the world's most prevalent bacterial sexually transmitted infection and leading infectious cause of blindness, yet it is one of the least understood human pathogens, in part due to the difficulties of in vitro culturing and the lack of available tools for genetic manipulation. Genome sequencing has reinvigorated this field, shedding light on the contemporary history of this pathogen. Here, we analyze 563 full genomes, 455 of which are novel, to show that the history of the species comprises two phases, and conclude that the currently circulating lineages are the result of evolution in different genomic ecotypes. Temporal analysis indicates these lineages have recently expanded in the space of thousands of years, rather than the millions of years as previously thought, a finding that dramatically changes our understanding of this pathogen's history. Finally, at a time when almost every pathogen is becoming increasingly resistant to antimicrobials, we show that there is no evidence of circulating genomic resistance in C. trachomatis.

Figure 1.

Global phylogeny and recombination landscape of 563 C. trachomatis genomes. The phylogeny (left), with associated genotype and geographical data, is displayed alongside the linearized chromosome (right). Lineage labels are as per previous publications. Line graph (right, top) shows the number of recombination events affecting individual genes and are colored according to lineage. Black blocks below this graph show previously identified “hotspots” of recombination (Harris et al. 2012). Colored blocks (right, bottom) indicate inferred recombination events each affecting selected taxa and genomic location, with color indicating the number of inferred events. Gene annotations correspond to strain D/UW3. (PZ) Plasticity zone.

Figure 2.

Proportion of pairwise isolates sharing a given trait (country or genotype) as a function of genomic divergence for the well-sampled urogenital lineages T1 (top) and T2 (bottom). (Top, left) For instance any pair of isolates less than approximately 400 mutations apart contain the same trait (country of isolation) in 50% of cases. (Red) Observed data; (blue) 100 permutations; (left) country of isolation; (right) genotype.

Figure 3.

Clustering of genes by recombination frequency in each lineage reveals lineage-specific profiles. Each horizontal line represents a gene with colors corresponding to standard deviations from the clade-specific mean. (Right) Expansion of the seven most actively recombining clusters. Clusters do not necessarily represent order of genes along the genome.

Figure 4.

(A) Temporal analysis of the LGV clade indicates a most recent common ancestor (MRCA) between 200 CE and 1430 CE (95% highest posterior density [HPD]). Dates along the x-axis are in years (CE), and blue bars show the 95% posterior probability. Posterior probabilities of node positions are indicated by closed circles (P = 1) or open circles (P > 0.8). (B) The Chlamydia trachomatis LGV mutation rate is shown in the context of other viruses, bacteria and eukaryotes. Error bars differ per species according to methodology, but for the case of C. trachomatis represent 95% posterior probability. Data sources shown in Supplemental Table S3. Buchnera aphidicola, another intracellular bacterium, has a similar genome size and mutation rate.

Figure 5.

Diversity of the major outer membrane protein (MOMP) gene ompA. (A) Phylogenetic relationship between all 563 Chlamydia trachomatis isolates (ompA gene) shows separation into three clades labeled α, β, and γ. (B) Phylogenetic tree of 1003 MOMP-encoding genes across the Chlamydiaceae. (C) Divergence (proportion of differing nucleotides) of the ompA gene in three C. trachomatis clades compared with that of a reconstructed ancestor, shading indicates 10th and 90th percentiles.