Predominantly Orphan Secretome in the Lung Pathogen Mycobacterium abscessus Revealed by a Multipronged Growth-Phase-Driven Strategy

Abstract

The emerging lung pathogen Mycobacterium abscessus is understudied for its virulence determinants and molecular targets for diagnosis and therapeutics. Here, we report a comprehensive secretome (600 proteins) of this species, which was identified using a multipronged strategy based on genetic/genomic, proteomic, and bioinformatic approaches. In-solution digested bottom-up proteomics from various growth phases identified a total of 517 proteins, while 2D-GE proteomics identified 33 proteins. A reporter-gene-fusion-based genomic library that was custom-generated in this study enabled the detection of 23 secretory proteins. A genome-wide survey for N-terminal signal sequences using bioinformatic tools (Psortb 2.0 and SignalP 3.0) combined with a strategy of the subtraction of lipoproteins and proteins containing multiple transmembrane domains yielded 116 secretory proteins. A homology search against the M. tuberculosis database identified nine additional secretory protein homologs that lacked a secretory signal sequence. Considering the little overlap (80 proteins) among the different approaches used, this study emphasized the importance of using a multipronged strategy for a comprehensive understanding of the secretome. Notably, the majority of the secreted proteins identified (over 50%) turned out to be "orphans" (those with no known functional homologs). The revelation of these species-specific orphan proteins offers a hitherto unexplored repertoire of potential targets for diagnostic, therapeutic, and vaccine research in this emerging lung pathogen.

Keywords: Mycobacterium abscessus; bottom-up proteomics; hypersensitivity pneumonitis; lung infections; metalworking fluid; secretome.

Figure 1

Differential secretory proteomes of M. abscessus during various growth phases. Secretory proteins precipitated from supernatants harvested from the control, 5 min, early-log (EL), mid-log (ML), and late-log (LL) cultures of M. abscessus were subjected to in-solution digestion and bottom-up proteomic analysis. Differential protein abundances were analyzed via spectral counting. (A) Heat map of differential secretory proteomes during various growth phases. Relative abundances (scaled total spectral counts) are shown, with proteins being sorted by p-values per the following order: (1) EL vs. 5 min, (2) ML vs. 5 min, and (3) LL vs. 5 min. A total of 517 proteins were identified (<1% false discovery rate) in all growth phases. (B) Proteins with an increased abundance in the secretomes of the EL, ML, and LL phases versus the 5 min time point are highlighted in red in the volcano plots. The fold change (FC) value cut-off at 2 (log2 −2 = −1 and log2 2 = 1) and p-value cut-off at 0.05 (−log10 0.05 = 1.301) are indicated by a gray line(s) on the x-axis and y-axis, respectively. (C) Proteins that were differentially secreted in the ML phase and LL phase compared to the EL phase. (D) Proteins secreted that were during the early phase and continued to be secreted in either the ML or LL phase. A total of 46 differentially expressed proteins that were common to EL vs. 5 min (FC > 2, *: p < 0.05), ML vs. EL (FC > 2, *: p < 0.05), and LL vs. EL (FC > 2, *: p < 0.05) were identified. The numerical data for the plot(s) included in each panel ((A) through (D)) are provided in the Supplementary Excel File.

Figure 2

Proteomic profiling of the secretory proteins of M. abscessus during the log phases of growth. (Panels A and B) The 2D-GE profiles of filtrate proteins from cultures grown to the mid-log phase (Panel A) and late-log phase (Panel B). Individual protein spots (red color) are shown with an arrow and number (green color). The cultures were grown in Sauton’s medium at 37 °C by shaking at 225 rpm for 24 h (180 Klett Reading) and 48 h (400 Klett Reading). (Panel C) A Venn diagram showing the differential and overlapping distributions of the secreted proteins between the mid-log and late-log phases.

Figure 3

Functional distribution of the identified secretome proteins of M. abscessus. (Panel A) The proteins identified in silico. (Panel B) Proteins identified with 2D-GE. (Panel C) Proteins identified with the in-solution-digested shotgun proteomic method.

Figure 4

Venn diagram showing the differential distribution of the secretome of M. abscessus across different methods of identification. Secretome identification was conducted with four different approaches: the genomic reporter gene fusion library screening method, in silico analysis, 2D-GE proteomic analysis, and bottom-up shotgun proteomic method.