Proteomic identification of Mycobacterium tuberculosis glycoproteins as potential biomarkers for active tuberculosis: Development of a prototype competitive ELISA for sputum antigen detection

Background Protein glycosylation can influence the antigenicity of Mycobacterium tuberculosis (MTB) proteins, but glycoprotein antigens remain little explored for the development of antigen-detection assays for tuberculosis (TB). Here, MTB glycoproteins were enriched and identified by proteomics, and a prototype competitive enzyme-linked immunosorbent assay (ELISA) was developed for detection of TB-associated antigenic material in sputum. Methods Glycoproteins from MTB strain TB-1039 were enriched by concanavalin A (ConA) affinity chromatography and characterized by SDS-PAGE, periodic acid–Schiff (PAS) staining, and liquid chromatography–mass spectrometry (LC-MS). GlycoPP and the Immune Epitope Database (IEDB) were used to predict N- and O-glycosylation sites and linear B-cell epitopes, respectively. The purified glycoproteins were conjugated to horseradish peroxidase (HRP), and antibodies purified from pooled sera of 10 culture-confirmed active-TB patients were used to develop the assay. The assay was subsequently evaluated using 206 sputum specimens collected from 132 patients with culture-confirmed active TB and 74 culture-negative comparator participants without active TB. Results Thirteen glycoproteins (Rv3873, Rv0934, Rv1078, Rv0954, Rv0173, Rv0583c, Rv1270c, Rv1543, Rv0341, Rv1860, Rv2560, Rv0174, and Rv0838) were selected according to predicted glycosylation and epitope compatibility. Six candidates (Rv0954, Rv0174, Rv1078, Rv0838, Rv0583c, and Rv1860) exceeded 70% compatibility. The competitive ELISA yielded an apparent area under the receiver operating characteristic (ROC) curve of 0.951 (95% CI 0.92–0.98; P < 0.001). Using a cut-off of ≤0.65 chosen on the same specimens, apparent sensitivity was 90.15% (95% CI, 83.7–94.7%) and apparent specificity was 93.24% (95% CI, 84.9–97.8%). Conclusion ConA enrichment, LC-MS identification, and in silico analysis enabled the prioritization of MTB glycoprotein candidates and supported development of a prototype competitive ELISA for TB-associated antigen detection in sputum. These preliminary, single-center findings provide a basis for further optimization, but independent validation in larger cohorts, including appropriate disease-control groups, is required before clinical application.