Phylogenomic and genomic analysis reveals unique and shared genetic signatures of Mycobacterium kansasii complex species

Abstract

Species belonging to the Mycobacterium kansasii complex (MKC) are frequently isolated from humans and the environment and can cause serious diseases. The most common MKC infections are caused by the species M. kansasii (sensu stricto), leading to tuberculosis-like disease. However, a broad spectrum of virulence, antimicrobial resistance and pathogenicity of these non-tuberculous mycobacteria (NTM) are observed across the MKC. Many genomic aspects of the MKC that relate to these broad phenotypes are not well elucidated. Here, we performed genomic analyses from a collection of 665 MKC strains, isolated from environmental, animal and human sources. We inferred the MKC pangenome, mobilome, resistome, virulome and defence systems and show that the MKC species harbours unique and shared genomic signatures. High frequency of presence of prophages and different types of defence systems were observed. We found that the M. kansasii species splits into four lineages, of which three are lowly represented and mainly in Brazil, while one lineage is dominant and globally spread. Moreover, we show that four sub-lineages of this most distributed M. kansasii lineage emerged during the twentieth century. Further analysis of the M. kansasii genomes revealed almost 300 regions of difference contributing to genomic diversity, as well as fixed mutations that may explain the M. kansasii's increased virulence and drug resistance.

Keywords: Mycobacterium kansasii; lineages; pangenome; phylogenomics; virulence.

Fig. 1.. Midpoint rooted maximum likelihood phylogenetic tree of 665 M. kansasii complex genomes. Blue dots indicate nodes with bootstrap support above 75 % with shading of branches and nodes indicating the seven MKC species. The rings, from inside towards outside, designate: (i) the year of isolation (if available), (ii) the isolation source of the strains classified as either environmental, animal, human or unknown; (iii) the country of isolation. The outer red stars indicate clinical isolates with patient HIV information, being positive (solid) or negative (open).

Fig. 2.. The bipartite network of the MKC prophages and their gene content. The predicted genes of the prophages are shown as grey diamonds, whereas prophages genomes as coloured circles. Genomes circles are coloured by MKC species or mycobacteriophages genomes obtained from hosts other than MKC species. Genomes are connected to their predicted genes by edges.

Fig. 3.. Resfams belonging to MKC species-specific. Heatmap showing the presence or absence of Resfams profiles (n=10) species related. Numbers inside cells indicate the average copy number of resistance protein profiles found in each group of MKC species genomes.

Fig. 4.. Regions of difference M. kansasii lineage specific. Maximum-likelihood phylogeny tree of 492 M. kansasii genomes, inferred based on nonrecombinant core SNPs, shows the M. kansasii genetic groups defined by BAPS and their specific RDs. Four genetic groups (L1–L4) characterize monophyletic lineages and two genomes (MK14ES_S84 and KAN-020712) belonging to L5 genetic group are unclassified due to their polyphyletic origin. Genomes belonging to the lineage L1 (n=480) are collapsed in the tree. Seven regions of difference specific to four (L2–L5) small represented genetic groups are indicated in grey sticks. The RD227 and RD228 shared a genomic region of 4 922 bp (genomic coordinates: RD227 from 4 766 718 to 4 771 663; RD228 from 4 766 741 to 4 787 038), even though RDscan defined as distinct RDs.

Fig. 5.. Maximum likelihood dated phylogeny of M. kansasii genomes (n=460) belonging to four sub-lineages. The outer circle indicates the country of origin of the isolate, whereas the first inner circle is the sub-lineage predicted in the Bayesian population analysis. The tree branches are also coloured accordingly to sub-lineages, showing the ancestral scenario reconstructed by PastML. The date of the most likely root was 1815, belonging to the older sub-lineage (L1.1.1, n=37) (a), with an unknown geographic region. Three recent sub-lineages emerged in (b) the United States in 1929 (L1.1.2, n=104), (c) Brazil in 1932 (L1.1.3, n=186), and (d) Australia in 1966 (L1.1.4, n=133) (the full tree visualization is also available at iTOL: https://itol.embl.de/tree/179702462428511716343991).

Fig. 6.. Maximum likelihood phylogeny tree of M. kansasii strains isolated in Czech Republic. The tree inferred based on nonrecombinant core SNPs is rooted in the reference strain ATCC12478. The number of nucleotide substitutions per site is indicated by the scale bar and bootstrap values are indicated in branches with at least 75 % of support. Sample sub-lineage is indicated by colours at node names. Beside the tree, from left to right, the following characteristics are described: Roman numerals indicating clusters of genomes with 0 to 10 SNPs pairwise distance; the workplace (where available); source of the isolation as either human or environmental; clinical phenotype (human isolates) or detailed isolation source (environmental isolates); year of isolation; and geographical origin of the isolate (environmental isolates) or patient (human isolates).