Genomic characterization of Nontuberculous Mycobacteria

Abstract

Mycobacterium tuberculosis and Mycobacterium leprae have remained, for many years, the primary species of the genus Mycobacterium of clinical and microbiological interest. The other members of the genus, referred to as nontuberculous mycobacteria (NTM), have long been underinvestigated. In the last decades, however, the number of reports linking various NTM species with human diseases has steadily increased and treatment difficulties have emerged. Despite the availability of whole genome sequencing technologies, limited effort has been devoted to the genetic characterization of NTM species. As a consequence, the taxonomic and phylogenetic structure of the genus remains unsettled and genomic information is lacking to support the identification of these organisms in a clinical setting. In this work, we widen the knowledge of NTMs by reconstructing and analyzing the genomes of 41 previously uncharacterized NTM species. We provide the first comprehensive characterization of the genomic diversity of NTMs and open new venues for the clinical identification of opportunistic pathogens from this genus.

Figure 1. Whole-genome phylogeny of the Mycobacterium genus reconstructed using the newly sequenced genomes and the ones that were already available.

The tree is built using the concatenated alignments of the 243 fully conserved genes within the genus with the maximum-likelihood inference approach implemented in RAxML (see Methods) and displayed using GraPhlAn. Colored shades highlight the Mycobacteria groups/complexes including the newly inferred assignments supported by the phylogeny. External to the phylogeny, we annotate the original group assignments and, for each strain, whether it is a newly sequenced or already sequenced strain, whether it is a type strain, its growth rate, its average GC content and the number of identified ORFs. The lengths of the outer black bars are proportional to the total genome length. The abbreviations used in this figure are reported in Supplementary Table 2.

Figure 2. Comparison of genome relations as inferred by the sequence-based phylogeny and the gene presence/absence clustering and the size of the non-redundant gene catalog of the Mycobacterium genus.

(A) The phylogenetic tree (on the left) and the gene presence/absence tree (on the right) are contrasted to highlight the consistency of the complementary evolutionary signals and identifying taxa with potentially uncoupled genetic versus functional evolution. (B) The scatter-plot of the pair-wise distance of the strains in the phylogenetic versus gene presence/absence trees (color denotes the average number of ORFs between the compared strains). Both the arrangement of the points and the overall correlation support the consistency between the trees. (C) The increase of the pan-genome size as a function of the number of genomes included in the clustering (see Methods). The blue curve highlights the trend for all the genomes (newly sequenced and retrieved from NCBI), whereas the red curve refers to already available genomes only. This analysis suggests that genomes sequenced in this work roughly double the gene families available for the Mycobacterium genus as also confirmed by other clustering approaches (see Methods).

Figure 3. Heatmap representing the distribution of EggNOG functions within the newly sequenced genomes.

The heatmap reports the number of genes that are labeled with a particular function within each sample we sequenced. The horizontal top bar represents the functional category. Colors in the bar at the bottom indicate the genes specific for a particular complex (fisher test p < 0.05). The full names for the abbreviated species names are reported in Supplementary Table 2.

Figure 4. Distribution and abundance of virulence factors within the newly sequenced genomes.

Cell intensity represents the fraction of genes in the corresponding gene family present in a given genome.

Figure 5. Heatmap representing the presence of genes driving the biosynthesis of mycolic acids and other key components of the cell envelope within the newly sequenced genomes.

Genes are organized according to their role in MA or DIM biosynthesis pathways; species are grouped according to their assignment to complexes/groups. Only the genes that are absent from at least 2 of the genomes are shown (see Supplementary Table 10 for the complete list). On the left HPLC patterns of mycolic acids for each species are reported. Each HPLC pattern is identified by an acronym describing the number and the time of retention of peak clusters in the chromatogram as follows: two continuous sequences of peaks (C2), one early and two late clusters of peaks (E1L2), one late cluster of peaks (L1a, tuberculosis-like), one early and one late cluster of peaks (E1L1), two early and one late clusters of peaks (E2LI), one late cluster of peaks (L1b, kansasii-like), two late clusters of peaks (L2), three late clusters of peaks (L3).