As more than two billion people possibly have a latent tuberculosis (LTB) infection, early LTB diagnosis is crucial for the efficient control and elimination of tuberculosis (TB). The aim of this study is to detect the serum antibody responses to dormancy-related DosR regulon antigens of Mycobacterium tuberculosis for the diagnosis of active and latent TB infections. A membrane array with 25 latency antigens detected by silver-enhanced gold nanoparticles was used to determine the corresponding cognate antibody levels in clinical serum samples from healthy controls, TB patients, and individuals with LTB. The array is sized to fit into a 24-well ELISA plate and follows an ELISA-like experimental procedure without expensive instrumentation. Linear discriminant analysis (LDA) of the resulting antibody profiling data set identified a panel of nine DosR antigens with significant discriminatory capability among different subjects with ≥90% sensitivity, specificity, and overall accuracy. Furthermore, the high predictive performance validated by an independent test sample set reflects the robustness and reliability of the LDA classification model. Our current data demonstrate that the nine DosR antigen combination associated with the proposed membrane array platform is a clinically feasible approach for distinguishing different TB infection statuses. The proposed methodology in this study could be further developed for multiple disease serodiagnoses with high sensitivity and specificity.
Keywords: DosR antigen; latent tuberculosis; protein membrane array; serodiagnosis; tuberculosis.