Lineage-Specific Proteomic Signatures in the Mycobacterium tuberculosis Complex Reveal Differential Abundance of Proteins Involved in Virulence, DNA Repair, CRISPR-Cas, Bioenergetics and Lipid Metabolism

Abstract

Despite the discovery of the tubercle bacillus more than 130 years ago, its physiology and the mechanisms of virulence are still not fully understood. A comprehensive analysis of the proteomes of members of the human-adapted Mycobacterium tuberculosis complex (MTBC) lineages 3, 4, 5, and 7 was conducted to better understand the evolution of virulence and other physiological characteristics. Unique and shared proteomic signatures in these modern, pre-modern and ancient MTBC lineages, as deduced from quantitative bioinformatics analyses of high-resolution mass spectrometry data, were delineated. The main proteomic findings were verified by using immunoblotting. In addition, analysis of multiple genome alignment of members of the same lineages was performed. Label-free peptide quantification of whole cells from MTBC lineages 3, 4, 5, and 7 yielded a total of 38,346 unique peptides derived from 3092 proteins, representing 77% coverage of the predicted proteome. MTBC lineage-specific differential expression was observed for 539 proteins. Lineage 7 exhibited a markedly reduced abundance of proteins involved in DNA repair, type VII ESX-3 and ESX-1 secretion systems, lipid metabolism and inorganic phosphate uptake, and an increased abundance of proteins involved in alternative pathways of the TCA cycle and the CRISPR-Cas system as compared to the other lineages. Lineages 3 and 4 exhibited a higher abundance of proteins involved in virulence, DNA repair, drug resistance and other metabolic pathways. The high throughput analysis of the MTBC proteome by super-resolution mass spectrometry provided an insight into the differential expression of proteins between MTBC lineages 3, 4, 5, and 7 that may explain the slow growth and reduced virulence, metabolic flexibility, and the ability to survive under adverse growth conditions of lineage 7.

Keywords: DNA repair; ESX-3 secretion system; Ethiopia; Mycobacterium tuberculosis; lineage 7; proteomics; virulence.

FIGURE 1

(A–D) Proteomics analysis of M. tuberculosis complex (MTBC) lineages. (A) Experimental workflow used in this study includes lysis, protein precipitation, tryptic digestion, peptide desalting, and ultra-high-performance liquid chromatography followed by mass spectrometry on a Q exactive plus mass spectrometer (Thermo-Fischer, Inc.). Label free quantification and the downstream statistical analysis were performed using MaxQuant and Perseus software, respectively. (B) Overlap of differentially abundant proteins in MTBC Lineages 3, 4, 5, and 7. (C) iBAQ-based ranking of protein abundances in various lineages. (D) Principal component analysis for protein abundance data from MTBC lineages. Samples were separated by lineage (component 1, 43.9% variance) and by biological groups (component 2, 26.6% variance). Only proteins identified in all samples were considered for the analysis. Data from three biological replicates are represented.

FIGURE 2

Venn diagram showing protein overlap among the various MTBC lineages of 3, 4, 5, and 7.

FIGURE 3

Venn diagram showing protein overlap among the top 100 most abundant proteins identified in the MTBC lineages of 3, 4, 5, and 7.

FIGURE 4

Boxplot showing difference in EsxA and EsxB proteins abundance in the various MTBC lineages.

FIGURE 5

(A) Predicted sub-network interactions for differentially abundant proteins (RecA, FaS, FabD, PssD, and PssC) representing important protein hubs. (B) Predicted sub-network interactions for differentially abundant proteins involved in type VII ESX-3 secretion system (iron acquisition).

FIGURE 6

High expression of ESAT-6 in M. tuberculosis complex (MTBC) lineage 4. Immunoblotting demonstrated a very high abundance of ESAT-6 in lineage 4 strain H37Rv. Immunoblotting of 15 μg of total lysate from each strain using monoclonal antibody against ESAT-6 and GroEL2 confirmed the different amounts of ESAT-6 in protein samples from MTBC strains lineage 7 (L7), lineage 4 (L4), lineage 3 (L3) and lineage 5 (L5) detected by: Normalization control 1: The gel included a stain that stains tryptophan. When activated, the stain follows the protein over to the blot and can be used to detect the uniformity of transfer and that the amount of each sample loaded is similar (http://www.bio-rad.com/en-no/applications-technologies/stain-free-imaging-technology). This form of normalization control usually performs better than antibodies against housekeeping proteins for detection abnormalities. Normalization control 2: Immunoblotting with a monoclonal antibody against GroEl2. These data and the proteomics data (Supplementary file S13) indicated that the expression of GroEL2 was fairly even between the protein samples from strains of MTBC lineages 3, 4, 5, and 7.

FIGURE 7

(A) Predicted sub-network interactions for differentially abundant proteins involved in DNA repair. (B) Protein sub-network interaction among proteins involved in energy metabolism. (C) Predicted sub-network interaction for differentially abundant proteins involved in inorganic phosphate uptake. (D) Predicted sub-network interaction for differentially abundant proteins involved in CRISPR-Cas systems.

FIGURE 8

Shows multiple genome alignment of all lineages using Geneious Prime software. The analysis focuses on CRISPR-CAS locus in the following order: CRISPR-Cas region 2, IS6110 (mobile element), CRISPR-Cas region 1, Cas2, Cas1, Csm6, Csm5, Csm4, Csm3, Csm2, Cas10, and Cas6. The Rv2813 and Rv2825c that flank the CRISPR-Cas locus are included in the alignment as reference points. As shown, lineage 2 (ID no. DA 3 and DA 10) have deletion of most the CRISPR repeats. Both Cas 1 and 2, and Csm6 - Csm4 (half of Csm4 is deleted). Cas 10 and Cas 6 seem to be highly conserved among the lineages. SNPs have been observed (black horizontal lines). The size of SNPs in Lineage 7 strains (ID L28 and L35) seems smaller than in other lineages.