ИСТИНА

Protein-carbohydrate interactions are essential in many biologically and medically important processes. Ability to control them can potentially lead to new achievements in drug design. However manipulating these interactions is impossible without a complete understanding of the underlying molecular mechanism of the process. The importance of hydrogen bonds between the carbohydrate hydroxyl groups and polar moieties of amino acids in the binding of carbohydrates by proteins is well known. However, the role played by mostly hydrophobic aromatic moieties in binding of hydrophilic carbohydrates is less clear. One of the most common interactions for aromatic side chains discovered in proteins is СH-π stacking (stacking between CH-system of sugar and π-system of aromatic moieties) [1]. Previous research suggests it may be important for carbohydrate recognition by creating an unfavourable nonpolar environment for hydroxyl groups in case of interaction with non-specific ligand. Also, СH-π stacking is dependent on the interacting face of a sugar (α or β) and thus may participate in positioning of a ligand to improve the processive function of an enzyme [2]. Another assumption is that CH-π stacking is important for energetically stabilizing the complex since it is pivotal for carbohydrate-binding proteins [1]. However, most of the analyzed proteins were enzymes and the frequent presence of CH-π stacking could be explained with both of these hypotheses. We’ve decided to study the presence of CH-π stacking in non-catalytic carbohydrate-binding antibodies to find out whether the results for this group will differ from the enzymes. We have taken structures of carbohydrate-binding antibodies from RCSB PDB. Using PyMOL software we visualized the structures and analysed possible interactions, such as hydrogen bonds and CH-π stacking between carbohydrate and protein. Three optimal conformations of CH-π stacking have been previously described which also are supported by theoretical energy calculations [3]. Carbohydrates with additional carboxyl group have been presented in approximately half of the antibody structures. We have found that 82% of structures with such modified carbohydrates do not show CH-π stacking, and most of carboxyl groups of these carbohydrates were bound with a salt bridge by positively charged aminoacid residues (arginine or lysine). In other complexes, we have found that about 70% of antibody-carbohydrate complexes do not have CH-π stacking interaction, but those that do mostly have CH-π stacking with three interacting CH-systems. However, there are non-stacking CH-π interactions in some antibodies, possibly of a hydrophobic and electrostatic nature. Interestingly none of the complexes with mono- or disaccharides had CH-π stacking. On average, the number of hydrogen bonds between protein and ligand does not differ with and without the presence of CH-π stacking. These findings may support the hypothesis that CH-π stacking interaction is not essential for achieving high specificity or affinity and therefore possibly is important for another function of a protein (enzymatic, transport, etc). For example, in glycosyltransferases CH-π stacking may ensure proper orientation of carbohydrate for higher catalytic activity. References: 1. Hudson, Kieran L et al. “Carbohydrate-Aromatic Interactions in Proteins.” Journal of the American Chemical Society vol. 137,48 (2015): 15152-60. doi:10.1021/jacs.5b08424 2. “Carbohydrate–Aromatic Interactions” Juan Luis Asensio, Ana Ardá, Francisco Javier Cañada, and Jesús Jiménez-Barbero Accounts of Chemical Research (2013) 46 (4), 946-954. DOI: 10.1021/ar300024d 3. Chen, Wentao et al. “Structural and energetic basis of carbohydrate-aromatic packing interactions in proteins.” Journal of the American Chemical Society vol. 135,26 (2013): 9877-84. doi:10.1021/ja4040472