Characterization of the epitope of anti-lipoarabinomannan antibodies as the terminal hexaarabinofuranosyl motif of mycobacterial arabinans.

mAb CS-35 is representative of a large group of antibodies with similar binding specificities that were generated against the Mycobacterium leprae lipopolysaccharide, lipoarabinomannan (LAM), and which cross-reacted extensively with LAMs from Mycobacterium tuberculosis and other mycobacteria. That this antibody also cross-reacts with the arabinogalactan (AG) of the mycobacterial cell wall, suggesting that it recognizes a common arabinofuranosyl (Araf)-containing sequence in AG and LAM, is demonstrated. The antibody reacted more avidly with 'AraLAM' (LAM with naked Araf termini) compared to 'ManLAM' (in which many Araf termini are capped with mannose residues) and mycolylarabinogalactan-peptidoglycan complex (in which the terminal Araf units are substituted with mycolic acids). Neither did the antibody bind to AG from emb knock-out mutants deficient in the branched hexa-Araf termini of AG. These results indicate that the terminal Araf residues of mycobacterial arabinan are essential for binding. Competitive ELISA using synthetic oligosaccharides showed that the branched hexa-Araf methyl glycoside [beta-D-Araf-(1-->2)-alpha-D-Araf-(1-)(2)-(3 and 5)-alpha-D-Araf-(1-->5)-alpha-D-Araf-OCH(3)] was the best competitor among those tested. The related linear methyl glycoside, beta-D-Araf-(1-->2)-alpha-D-Araf-(1-->5)-alpha-D-Araf-(1-->5)-alpha-D-Araf-OCH(3), representing one linear segment of the branched hexa-Araf, was less effective and the other linear tetrasaccharide, beta-D-Araf-(1-->2)-alpha-D-Araf-(1-->3)-alpha-D-Araf-(1-->5)-alpha-D-Araf-OCH(3), was ineffective. The combined results suggest that the minimal epitope recognized by antibody CS-35 encompasses the beta-D-Araf-(1-->2)-alpha-D-Araf-(1-->5)-alpha-D-Araf-(1-->5)-alpha-D-Araf within the branched hexa-Araf motif of mycobacterial arabinans, whether present in LAM or AG.