Immunochemistry of mycolic acid antigens in tuberculosis

Abstract

Tuberculosis (TB) is a collective name for the bacterial infection, which is caused by members of the Mycobacterium tuberculosis (M. tb) complex and can infect the lungs (pulmonary) as well as the kidneys, lymph nodes, bones and joints (extra-pulmonary). The re-emergence of drug-resistant strains and the HIV epidemic are among the main reasons for the resurgence of TB and there is a need for new drugs and diagnostic assays which are rapid and sensitive. Serodiagnostic assays have the potential of being rapid, inexpensive and relatively non-invasive. The most abundant antigen in the cell wall of M. tb, which has been analysed with ELISA and resonant mirror biosensor assays for use in serodiagnosis, is mycolic acid (MA). The sensitivity previously obtained in the ELISA assay was however inadequate for serodiagnostic purposes. It was believed that MA mimicked the structure of cholesterol, thereby causing anti-cholesterol human antibodies from TB negative sera to bind to MA and result in a large number of false positives. Within this work the apparent molecular mimicry between MA and cholesterol was investigated using a competitive enzyme linked inhibition assay (CELIA) assay. The results suggested that MA in liposomes resembled the liquid ordered arrangement of cholesterol in liposomes, rather than a direct mimicry of individual molecules. The nature of the antibody from TB negative patient sera binding to MA coated onto ELISA plates was also investigated. The results obtained from this study have not disproved the hypothesis of a cross-reactive anti-cholesterol antibody, but it would appear that the MA signal from TB negative serum was partially due to the binding of anti-MA antibodies. The presence of anti-MA antibodies in TB negative serum could have been the result of prior BCG vaccination, latent infection or due to constant immune stimulation from saprophytic mycobacteria. This creates the potential of using antibodies to MA to distinguish between latent TB infection and active disease. Furthermore, in order to overcome the low sensitivity of the ELISA assay due to high background signals from TB negative serum, members of our group previously developed a resonant mirror biosensor inhibition assay based on MA contained in liposomes. The biosensor measured mass accumulation and the identity of the binding molecules were unknown. It was shown here that one of the serum components binding to the immobilised MA liposomes in the biosensor inhibition assay was immunoglobulin G antibodies. The specificity of both the ELISA and biosensor assays previously analysed using a natural mixture of MA however, remained poor, and in the search for a more specific antigen, this study investigated the potential of MA subclasses for TB serodiagnosis using ELISA. It was observed that the antibody binding signal to the MA subclasses depended on the polarity of the coating solution, for which hexane was the preferred solvent. Both the alpha- and keto-MA subclasses could better distinguish between a range of TB positive patient and TB negative sera compared with the natural mixture of MA. These results suggested that a particular subclass applied in the biosensor inhibition assay could enhance the test to reach the required sensitivity and specificity required for the serodiagnosis TB.