Non mycobacterial virulence genes in the genome of the emerging pathogen Mycobacterium abscessus

Abstract

Mycobacterium abscessus is an emerging rapidly growing mycobacterium (RGM) causing a pseudotuberculous lung disease to which patients with cystic fibrosis (CF) are particularly susceptible. We report here its complete genome sequence. The genome of M. abscessus (CIP 104536T) consists of a 5,067,172-bp circular chromosome including 4920 predicted coding sequences (CDS), an 81-kb full-length prophage and 5 IS elements, and a 23-kb mercury resistance plasmid almost identical to pMM23 from Mycobacterium marinum. The chromosome encodes many virulence proteins and virulence protein families absent or present in only small numbers in the model RGM species Mycobacterium smegmatis. Many of these proteins are encoded by genes belonging to a "mycobacterial" gene pool (e.g. PE and PPE proteins, MCE and YrbE proteins, lipoprotein LpqH precursors). However, many others (e.g. phospholipase C, MgtC, MsrA, ABC Fe(3+) transporter) appear to have been horizontally acquired from distantly related environmental bacteria with a high G+C content, mostly actinobacteria (e.g. Rhodococcus sp., Streptomyces sp.) and pseudomonads. We also identified several metabolic regions acquired from actinobacteria and pseudomonads (relating to phenazine biosynthesis, homogentisate catabolism, phenylacetic acid degradation, DNA degradation) not present in the M. smegmatis genome. Many of the "non mycobacterial" factors detected in M. abscessus are also present in two of the pathogens most frequently isolated from CF patients, Pseudomonas aeruginosa and Burkholderia cepacia. This study elucidates the genetic basis of the unique pathogenicity of M. abscessus among RGM, and raises the question of similar mechanisms of pathogenicity shared by unrelated organisms in CF patients.

Figure 1. The M. abscessus CIP 104536T genome.

(A) Circular representation of the chromosome. The initiation codon for the dnaA gene was chosen as the starting point for numbering. The scale is in Mb. Moving inward, the first two circles show forward and reverse genes (blue lines); light-green lines indicate phage genes. The third circle shows tRNA genes (red) and rRNA operon (dark-green). The fourth circle shows genes presumably acquired “en bloc” from non mycobacterial organisms by HGT (purple) and IS elements (black). The inner black histogram represents the local G+C content (scale: 50% to 69%). (B) Circular representation of the 23-kb mercury resistance plasmid. The scale is in kb. Forward and reverse genes and the local G+C content are indicated with the same code as for the chromosome map. The plasmid carries a mercury resistance operon flanked by two genes encoding site-specific recombinases (MAB_p04c and MAB_p10, orange); it also encodes a relaxase/helicase that may function in conjugation or mobilization (MAB_p15c, yellow).

Figure 2. The M. abscessus full-length prophage.

Each arrow represents a predicted protein-coding gene (length approximately to scale). Orange, similar to other phage proteins; blue, similar to other bacterial proteins; green, hypothetical protein. The table shows homologs of bacterial proteins with identified functions (Uniprot Blast search).

Figure 3. Examples of gene blocks presumably inherited from non mycobacterial organisms.

(A) MAB_0295-0298 (phenazine biosynthesis). (B) MAB_0888c-0891c (homogentisate catabolism). (C) MAB_1093c-1098 (DNA degradation [dnd locus]). (D) MAB_1501-1504 (iron uptake). Genes are drawn approximately to scale and are indicated according to their names in the Embl-Ebi database.

Figure 4. plc, mgtC and msrA loci in M. abscessus: comparison with other mycobacteria.

a. plc: note that MAB_0557 (transcriptional regulatory protein AraC) has no homolog at the counterpart of the M. abscessus plc locus in the other mycobacterial species, suggesting an insertion; also note the substitution for a PE-PGRS gene in the corresponding M. tuberculosis region. b. mgtC: note that the gene encoding MgtC is located at other genomic sites in M. tuberculosis and M. avium; also note that MAB_3592c (probable chain fatty acid-CoA ligase, blue) has no homolog at the counterpart of the M. abscessus mgtC locus in the other mycobacterial species, also suggesting an insertion. c. msrA: note that MsrA-encoding genes are located at other genomic sites in other mycobacteria; and the presence of sodA (light blue) upstream of M. abscessus msrA; there is a substitution for a transposase gene (Rv3844) in the corresponding M. tuberculosis region. Out of scale.

Figure 5. Phylogenetic trees (maximum likelihood) of PlC (a), MgtC (b) and MsrA (c) proteins.

Branch supports values are indicated at the nodes. Branch colors indicate proteins from M. abscessus (red), Actinobacteria (purple), Proteobacteria (blue) and Firmicutes (green). Labels at the leaves show the Uniprot identifier of the proteins and the species they belong to.