Tuberculosis (TB) is a global disease with an annual death toll of about two million people. To date, no diagnostic test is available, which enables the accurate diagnosis of the tuberculosis disease states. The serological responses of TB patients are heterogeneous, but there are efforts to identify antibody signatures to a number of M. tuberculosis antigens, whose abundance and presence could correlate to the disease state and may predict treatment outcome. Therefore, a whole TB proteome screen was recently performed, to identify the TB antigens which may be used in a serodiagnostic TB test.
The task of this thesis was to validate the results of the whole TB proteome screen, using a bead-based platform, and to test, whether a defined subset of these TB antigens could be used to build a serodiagnostic TB test.
Therefore, a serological assay for 112 purified TB antigens was developed and a technical validation was performed. Precision, linearity, antigen stability and coupling reproducibility were assessed.
To verify the performance of the assay and to enable a comparison between different runs reference control sera were generated by pooling samples from active tuberculosis patients.
Screening of more than 1000 human serum samples from non-TB, active TB and latent TB patients and subsequent statistical analysis were performed to identify a panel of antigens that might discriminate between different TB states.
In total, 41 differentially expressed proteins were identified, which enable active TB to be distinguished from healthy people. For the discrimination of patients with latent tuberculosis infection from healthy persons, a serological response pattern to 18 antigens could be identified. Multivariate statistical studies revealed that the heterogeneous response pattern to these 41 and 18 antigens did not correlate with clinical indices of disease status. The current results of this study do not support the hypothesis that it is easily possible to identify a serological response to defined TB antigen panels that correlate to certain disease states.