Genomic insights into the plasmidome of non-tuberculous mycobacteria

Abstract

Background: Non-tuberculous mycobacteria (NTM) are a diverse group of environmental bacteria that are increasingly associated with human infections and difficult to treat. Plasmids, which might carry resistance and virulence factors, remain largely unexplored in NTM.

Methods: We used publicly available complete genome sequence data of 328 NTM isolates belonging to 125 species to study gene content, genomic diversity, and clusters of 196 annotated NTM plasmids. Furthermore, we analyzed 3755 draft genome assemblies from over 200 NTM species and 5415 short-read sequence datasets from six clinically relevant NTM species or complexes including M. abscessus, M. avium complex, M. ulcerans complex and M. kansasii complex, for the presence of these plasmids.

Results: Between one and five plasmids were present in approximately one-third of the complete NTM genomes. The annotated plasmids varied widely in length (most between 10 and 400 kbp) and gene content, with many genes having an unknown function. Predicted gene functions primarily involved plasmid replication, segregation, maintenance, and mobility. Only a few plasmids contained predicted genes that are known to confer resistance to antibiotics commonly used to treat NTM infections. Out of 196 annotated plasmid sequences, 116 could be grouped into 31 clusters of closely related sequences, and about one-third were found across multiple NTM species. Among clinically relevant NTM, the presence of NTM plasmids showed significant variation between species, within (sub)species, and even among strains within (sub)lineages, such as dominant circulating clones of Mycobacterium abscessus.

Conclusions: Our analysis demonstrates that plasmids are a diverse and heterogeneously distributed feature in NTM bacteria. The frequent occurrence of closely related putative plasmid sequences across different NTM species suggests they may play a significant role in NTM evolution through horizontal gene transfer at least in some groups of NTM. However, further in vitro investigations and access to more complete genomes are necessary to validate our findings, elucidate gene functions, identify novel plasmids, and comprehensively assess the role of plasmids in NTM.

Keywords: Antimicrobial resistance; Genomics; Non-tuberculous mycobacteria; Plasmids.

Fig. 1

Phylogeny and plasmid content of 328 complete genomes from non-tuberculous mycobacteria. The phylogeny was reconstructed from an alignment of 291 single-copy gene families present on the chromosomes of all strains. The tree scale is in substitutions per site. Branches having a low bootstrap support (< 70%) are colored red. The species most clinically relevant to humans are shaded grey. The number of plasmids in each genome is shown in the form of green bars. SGM: slowly growing mycobacteria, RGM: rapidly growing mycobacteria

Fig. 2

Phylogeny and distribution of plasmid clusters across 98 plasmid-bearing complete genomes from non-tuberculous mycobacteria. The phylogeny was reconstructed from 858 single-copy genes and midpoint rooted. Clinically relevant species are indicated with red dots: MAB = M. abscessus, KAN = M. kansasii, MUC = M. ulcerans complex, MAC = M. avium complex. For each plasmid cluster, a different color was used

Fig. 3

Predicted plasmid distribution in 1486 M. abscessus isolates. Illumina short-read sequence data were screened for the presence of 112 annotated plasmid sequences from non-tuberculous mycobacteria. Potential novel plasmids were not predicted. If a plasmid sequence was also identified in draft genomes of species other than M. abscessus, those species are indicated in brackets. Hits with annotated plasmid sequences from SMC-4 are not shown. The BRA-100 clade, including isolates belonging to the surgery-related outbreak in Brazil, is shaded light orange. DCC = dominant circulating clone

Fig. 4

Predicted plasmid distribution in 2451 M. avium complex isolates. Illumina short-read sequence data were screened for the presence of 112 annotated plasmid sequences from non-tuberculous mycobacteria. Potential novel plasmids were not predicted. If a plasmid sequence was also identified in draft genomes of species other than M. avium complex, those species are indicated in brackets. Only plasmids found in more than six isolates are visualized. Hits with annotated plasmid sequences from SMC-4 are not shown

Fig. 5

Predicted plasmid distribution in 983 M. ulcerans complex isolates. Illumina short-read sequence data were screened for the presence of 112 annotated plasmid sequences from non-tuberculous mycobacteria. Potential novel plasmids were not predicted. If a plasmid sequence was also identified in draft genomes of species other than M. ulcerans complex, those species are indicated in brackets

Fig. 6

Predicted plasmid distribution of annotated NTM plasmids in 495 M. kansasii complex isolates. Illumina short-read sequence data were screened for the presence of 112 annotated plasmid sequences from non-tuberculous mycobacteria. Potential novel plasmids were not predicted. If a plasmid sequence was also identified in draft genomes of species other than M. kansasii complex, those species are indicated in brackets

Fig. 7

Heatmap of putative resistance, stress, and virulence genes found in 197 annotated plasmid sequences from non-tuberculous mycobacteria using AMRfinder + . Only genes with no internal stopcodons, > 30% amino acid identity and > 70% coverage compared to markers in the AMRfinder + reference database are shown. Plasmid accession numbers are preceded by the plasmid cluster number. U = unclustered, AMR = antimicrobial resistance, V = virulence. Sul = sulfonamide, Amg = aminoglycoside, MAC = macrolide, Rif = rifamycin, Tmy = Tetracenomycin, Phe = phenicol, Ql = quinolone, Li = lincosamide, Tmp = trimethoprim, As = arsenic, Cd = cadmium, Cu = copper, Au = gold, Ag = silver, Hg = mercury, Ni = nickel, QAC = quaternary ammonium compound, Te = tellurium, Tet = tetracycline, Fos = Fosfomycin, Gly = glycopeptide