Generation of mycobacterial lipoarabinomannan-specific monoclonal antibodies and their ability to identify mycobacterium isolates

Abstract Background/Purpose The World Health Organization has recommended commercial urine-sourced lipoarabinomannan (LAM) detection as a tool for screening HIV patients with suspected TB, but more sensitive immunodetection assays would help to identify HIV-negative TB patients. Here, we aimed to develop novel rabbit monoclonal antibodies (mAbs) against LAM for immunodetection purposes. Methods Rabbits were immunized with cell-wall components from the Mycobacterium tuberculosis (Mtb) H37Rv strain, an immune single-chain fragment variable (scFv) phage display library was generated, and the scFv mAbs produced were bound to purified LAM from H37Rv before isolation. The light and heavy chain variable region genes from the selected clones were sequenced, and the full-length light and heavy chains were cloned into vector pCMV3, and then co-expressed in 293 T cells to generate whole IgG antibodies. The performances and binding characteristics of the mAbs against purified LAM from M.tb H37Rv, multiple mycobacteria species (M.tb H37Rv, M. bovis and non-tuberculous mycobacteria (NTM) strains), and clinical mycobacteria isolates (Mtb and NTM isolates) were determined using various immunoassay methods. Results We obtained five rabbit mAbs against LAM, four of which had high sensitivities (100 pg/ml) and affinities (1.16 to 1.73×10-9 M) towards LAM, and reacted with M.tb H37Rv, M. bovis, and slow-growing NTM, but not with rapid-growing NTM. Similar results were obtained with mycobacterium isolates, where 96% of the Mtb isolates and 90% of the M. avium-intracellulare isolates were successfully identified. Conclusion Our novel rabbit LAM-specific mAbs performed well at recognizing LAM from slow-growing pathogenic mycobacteria, supporting their future diagnostic potential.