Serum proteomics of active tuberculosis patients and contacts reveals unique processes activated during Mycobacterium tuberculosis infection

Abstract

Tuberculosis (TB) is the most lethal infection among infectious diseases. The specific aim of this study was to establish panels of serum protein biomarkers representative of active TB patients and their household contacts who were either infected (LTBI) or uninfected (EMI-TB Discovery Cohort, Pontevedra Region, Spain). A TMT (Tamdem mass tags) 10plex-based quantitative proteomics study was performed in quintuplicate containing a total of 15 individual serum samples per group. Peptides were analyzed in an LC-Orbitrap Elite platform, and raw data were processed using Proteome Discoverer 2.1. A total of 418 proteins were quantified. The specific protein signature of active TB patients was characterized by an accumulation of proteins related to complement activation, inflammation and modulation of immune response and also by a decrease of a small subset of proteins, including apolipoprotein A and serotransferrin, indicating the importance of lipid transport and iron assimilation in the progression of the disease. This signature was verified by the targeted measurement of selected candidates in a second cohort (EMI-TB Verification Cohort, Maputo Region, Mozambique) by ELISA and nephelometry techniques. These findings will aid our understanding of the complex metabolic processes associated with TB progression from LTBI to active disease.

Figure 1

Summary of the quantitative shotgun proteomic study. Venn diagram representation of the five TMT experiments (A). A total of 418 proteins were identified and quantified with at least one unique peptide in the whole study. 154 of them were common for the five TMT experiments. Dispersion diagrams for the 154 common proteins (B–D) representing the mean of the log2 ratio and the standard error of the mean (SEM).

Figure 2

Summary of the number of unique peptides quantified per protein (A). Ninety-five percent of the proteins were quantified with at least two unique peptides. Volcano-plot representations of the statistical analysis of the quantification ratios. A specific proteomic signature is detected when comparing active TB patients versus both LTBI (B) and uninfected (C) contacts. No apparent specific signature is detected when comparing uninfected versus LTBI contacts (D). Significance is considered when p-value ≤ 0.001.

Figure 3

Interaction and pathway analysis of the proteins detected as modulated in active TB patients. String 10.1 analysis show a strong interaction network (A) between those proteins. Statistical pathway analysis (B–D) show that most of the proteins play roles in defense against pathogens, complement activation and inflammation.

Figure 4

Box-plot representation of selected proteins that are detected as over-represented in the serum of active TB patients versus both LTBI and uninfected contacts but are not modulated when comparing uninfected versus LTBI contacts. ****p-value ≤ 0.00001; ***p-value ≤ 0.001.

Figure 5

Box-plot representation of selected proteins that are detected as decreased in the serum of active TB patients versus both LTBI and uninfected contacts but are not modulated when comparing uninfected versus LTBI contacts. ****p-value ≤ 0.00001; ***p-value ≤ 0.001.

Figure 6

Targeted measurement of the serum levels of selected candidates (three over-represented and three decreased in active TB patients) in an independent cohort (verification cohort, Maputo region, Mozambique). Significant (p-val≤ 0.05) modulation is detected for five out of the six proteins (A). ROC (B) and Youden index (C) analysis was done for the six targets to investigate their discriminatory potential between active TB patients and latently infected contacts.

Figure 7

Schematic representation of the processes and proteins detected as modulated during the progression of the TB from a latent asymptomatic infection to an active disease. Proteins in blue are decreased in active TB patients whereas proteins in red are increased in this group versus both LTBI and uninfected contacts. Proteins in bold were also detected as significantly modulated in an independent cohort by targeted antibody-based techniques.