Pattern Recognition in Pulmonary Tuberculosis Defined by High Content Peptide Microarray Chip Analysis Representing 61 Proteins from M. tuberculosis

BACKGROUND Serum antibody-based target identification has been used to identify tumor-associated antigens (TAAs) for development of anti-cancer vaccines. A similar approach can be helpful to identify biologically relevant and clinically meaningful targets in M. tuberculosis (MTB) infection for diagnosis or TB vaccine development in clinically well defined populations. METHOD We constructed a high-content peptide microarray with 61 M. tuberculosis proteins as linear 15 aa peptide stretches with 12 aa overlaps resulting in 7446 individual peptide epitopes. Antibody profiling was carried with serum from 34 individuals with active pulmonary TB and 35 healthy individuals in order to obtain an unbiased view of the MTB epitope pattern recognition pattern. Quality data extraction was performed, data sets were analyzed for significant differences and patterns predictive of TB+/-. FINDINGS Three distinct patterns of IgG reactivity were identified: 89/7446 peptides were differentially recognized (in 34/34 TB+ patients and in 35/35 healthy individuals) and are highly predictive of the division into TB+ and TB-, other targets were exclusively recognized in all patients with TB (e.g. sigmaF) but not in any of the healthy individuals, and a third peptide set was recognized exclusively in healthy individuals (35/35) but no in TB+ patients. The segregation between TB+ and TB- does not cluster into specific recognition of distinct MTB proteins, but into specific peptide epitope 'hotspots' at different locations within the same protein. Antigen recognition pattern profiles in serum from TB+ patients from Armenia vs. patients recruited in Sweden showed that IgG-defined MTB epitopes are very similar in individuals with different genetic background. CONCLUSIONS A uniform target MTB IgG-epitope recognition pattern exists in pulmonary tuberculosis. Unbiased, high-content peptide microarray chip-based testing of clinically well-defined populations allows to visualize biologically relevant targets useful for development of novel TB diagnostics and vaccines.