Mannan-binding lectin (MBL) is a collagen-related C-type lectin that recognizes and binds to carbohydrate structures on microbial surfaces. Relative deficiency (insufficiency) of MBL in humans confers an increased susceptibility to infection and adversely influences the course of numerous diseases that can be complicated by infections [1]. Associations with infections and autoimmune diseases are attributed to its ability to opsonize micro-organisms, both directly and via complement activation. Opsonization is believed to involve an interaction between the collagen region of the MBL molecule and collectin receptors on phagocytes. However, we have recently provided evidence that MBL can also bind to healthy autologous cells via its lectin domain and we have suggested such interactions could be of physiological importance under some circumstances [2, 3]. We speculate that MBL binding activates signalling pathways resulting in cells that are functionally different from MBL-naïve cells. If true, this has potentially important consequences for MBL replacement therapy: both plasma-derived [4] and recombinant [5] MBL have a short biological half-life, but their effects on cells might persist for much longer. Indeed, this could explain the surprising observation that after allogeneic stem cell transplantation, the mbl-2 status of the donor strongly influences susceptibility to serious infections [6]. An obvious alternative explanation, extra-hepatic synthesis of MBL by leukocytes, does not seem to be tenable [7, 8].

Our observations were based on the use of biotinylated MBL which was detected with a streptavidin conjugate. Taylor and Van den Berg [9] have demonstrated that biotinylated C-reactive protein (CRP), but not native CRP, bound to human endothelial cells. Although CRP and MBL are structurally unrelated, this raises the theoretical possibility that biotinylation might alter the binding properties of MBL. If so, our reported results could be due to an artefact and the inferences drawn from them invalid.

We therefore used an alternative strategy to try to confirm our previous results. Incubation of cells with native MBL was followed by incubation with a specific anti-MBL antibody (131-1) and any immune complex formed probed with caprine anti-murine IgG conjugated to fluorescein. The interaction was controlled by the absence of MBL and by the use of non-specific murine IgG1 instead of specific antibody. By this means, specific binding of MBL to human immature monocyte-derived dendritic cells was demonstrated to be of similar magnitude to that of biotinylated MBL (Fig. 1). Similar results were obtained with human mononuclear cells counter-stained with PE-anti-CD14 (monocytes) and PE-anti-CD19 (B lymphocytes). For each cell type, the profiles of the two negative controls were virtually superimposable upon the isotype controls. However, the profiles of cells incubated with MBL shifted significantly to the right. No binding of either MBL species to erythrocytes could be detected (negative cell control).


Figure 1.  Data were collected using a FacScan flow cytometer (Becton Dickinson Co, Oxford, UK) and analysed using Cell Quest software (Becton Dickinson Co). Results were visualized as fluorescence histograms as previously described [2, 3]. Illustrated above are the histograms with dendritic cells (DC), monocytes and B lymphocytes with the non-specific antibody control as a non-shaded overlay.

Download figure to PowerPoint

We conclude that labelling human MBL with biotin does not affect its binding to human cells. Furthermore, we have now confirmed by two independent methods of detection that MBL binds to dendritic cells, monocytes and B lymphocytes. It therefore remains a valid suggestion that cellular interactions of MBL at extra-vascular sites may be biologically meaningful.


  1. Top of page
  2. References