Variant G57E of mannose binding lectin associated with protection against tuberculosis caused by Mycobacterium africanum but not by M. tuberculosis

Abstract

Structural variants of the Mannose Binding Lectin (MBL) cause quantitative and qualitative functional deficiencies, which are associated with various patterns of susceptibility to infectious diseases and other disorders. We determined genetic MBL variants in 2010 Ghanaian patients with pulmonary tuberculosis (TB) and 2346 controls and characterized the mycobacterial isolates of the patients. Assuming a recessive mode of inheritance, we found a protective association between TB and the MBL2 G57E variant (odds ratio 0.60, confidence interval 0.4-0.9, P 0.008) and the corresponding LYQC haplotype (P(corrected) 0.007) which applied, however, only to TB caused by M. africanum but not to TB caused by M. tuberculosis. In vitro, M. africanum isolates bound recombinant human MBL more efficiently than did isolates of M. tuberculosis. We conclude that MBL binding may facilitate the uptake of M. africanum by macrophages, thereby promoting infection and that selection by TB may have favoured the spread of functional MBL deficiencies in regions endemic for M. africanum.

Figure 1. MBL2 haplotype frequencies in different populations.

1, Argentina (Chiriguanos) ; 2, Argentina (Mapuche) ; 3, Greenland ; 4, The Netherlands ; 5, Denmark ; 6, Spain , 7; Ghana (present study); 8, Mozambique ; 9, Kenya ; 10, Iran ; 11, China ; 12, Korea ; 13, Japan ; 14, Australia .

Figure 2. Comparison of MBL binding to M. africanum vs. M. tuberculosis strains by flow cytometry.

Representative clinical isolates of M. africanum and M. tuberculosis strains were processed for flow-cytometric analysis by labeling with recombinant human MBL, biotinylated anti-human MBL antibody and Cy5-conjugated streptavidin). The percentage of positive cells for MBL binding was analyzed in triplicates for three different clinical isolates of M. africanum and M. tuberculosis along with the laboratory strain H37Rv. Calcium dependence of binding was examined by the addition of EDTA to the TBS/Ca²⁺ buffer. Negative controls were processed in the absence of MBL. The mean percentage of positive cells of these controls was subtracted from the corresponding percentages of positive cells of MBL-treated mycobacteria (A). Representative flow-cytometric profiles of MBL binding to clinical isolates (depicted in A) in the presence of Ca²⁺ (B). The Ghanaian MTBC strains used for the binding assays were M. africanum West African 2 strains 10415/02, 10476/01, 10514/01 and M. tuberculosis Cameroon strains 5390/02, 5400/02 and 1417/02.