A Computational Approach for Exploring Carbohydrate Recognition by Lectins in Innate Immunity

Abstract

Recognition of pathogen-associated carbohydrates by a broad range of carbohydrate-binding proteins is central to both adaptive and innate immunity. A large functionally diverse group of mammalian carbohydrate-binding proteins are lectins, which often display calcium-dependent carbohydrate interactions mediated by one or more carbohydrate recognition domains. We report here the application of molecular docking and site mapping to study carbohydrate recognition by several lectins involved in innate immunity or in modulating adaptive immune responses. It was found that molecular docking programs can identify the correct carbohydrate-binding mode, but often have difficulty in ranking it as the best pose. This is largely attributed to the broad and shallow nature of lectin binding sites, and the high flexibility of carbohydrates. Site mapping is very effective at identifying lectin residues involved in carbohydrate recognition, especially with cases that were found to be particularly difficult to characterize via molecular docking. This study highlights the need for alternative strategies to examine carbohydrate–lectin interactions, and specifically demonstrates the potential for mapping methods to extract additional and relevant information from the ensembles of binding poses generated by molecular docking.

Antibody Recognition of Cancer-Related Gangliosides and Their Mimics Investigated Using in silico Site Mapping

“…Furthermore, this optimized cutoff affords site map quality comparable in reproduction and correctness to previously studied antibody- and lectin-carbohydrate systems [32], [34].…”

“…= 0.08). However, the use of the 90% cutoff resulted in relatively poor map correctness (∼0.5–0.6) compared to previous cases [32], [34], [36]. This low level of map correctness can be attributed to the inclusion of many erroneous van der Waals contacts.…”

“…The settings used for these programs are detailed elsewhere [34], [45]. Briefly, the ligands were treated flexibly by each docking program, with the exception of the pyranose rings, which were kept in chair conformations.…”

Latest developments in molecular docking: 2010–2011 in review

“…Another method, based on NMR-derived proximities of ligand protons to the target surface, was proposed, which improved docking accuracy from~50% to~70%. [132] A similar methodology was developed by Agostino et al [133][134][135] However, in their approach, a set of docked poses, rather than experimental structures, is used to derive the interactions required for pose filtering. The advantage of this method, compared with the one above, is that it is applicable to systems with no or only limited experimental structural data.…”

“…Also, because the method was developed for carbohydrates, particularly flexible ligands, the poses are also parsed by conformational filters, reflecting the conformational composition of the docked pose ensemble. The method was tested against a set of 20 antibody-carbohydrate systems [135,137] and 15 lectin-carbohydrate systems [134] and demonstrated a superior performance compared with considering the top-docked pose alone. It has also been successfully applied to peptide ligands.…”