Ligand-mediated dimerization of a carbohydrate-binding molecule reveals a novel mechanism for protein-carbohydrate recognition

J Mol Biol. 2004 Mar 19;337(2):417-26. doi: 10.1016/j.jmb.2003.12.081.

Abstract

The structural and thermodynamic basis for carbohydrate-protein recognition is of considerable importance. NCP-1, which is a component of the Piromyces equi cellulase/hemicellulase complex, presents a provocative model for analyzing how structural and mutational changes can influence the ligand specificity of carbohydrate-binding proteins. NCP-1 contains two "family 29" carbohydrate-binding modules designated CBM29-1 and CBM29-2, respectively, that display unusually broad specificity; the proteins interact weakly with xylan, exhibit moderate affinity for cellulose and mannan, and bind tightly to the beta-1,4-linked glucose-mannose heteropolymer glucomannan. The crystal structure of CBM29-2 in complex with cellohexaose and mannohexaose identified key residues involved in ligand recognition. By exploiting this structural information and the broad specificity of CBM29-2, we have used this protein as a template to explore the evolutionary mechanisms that can lead to significant changes in ligand specificity. Here, we report the properties of the E78R mutant of CBM29-2, which displays ligand specificity that is different from that of wild-type CBM29-2; the protein retains significant affinity for cellulose but does not bind to mannan or glucomannan. Significantly, E78R exhibits a stoichiometry of 0.5 when binding to cellohexaose, and both calorimetry and ultracentrifugation show that the mutant protein displays ligand-mediated dimerization in solution. The three-dimensional structure of E78R in complex with cellohexaose reveals the intriguing molecular basis for this "dimeric" binding mode that involves the lamination of the oligosaccharide between two CBM molecules. The 2-fold screw axis of the ligand is mirrored in the orientation of the two protein domains with adjacent sugar rings stacking against the equivalent aromatic residues in the binding site of each protein molecule of the molecular sandwich. The sandwiching of an oligosaccharide chain between two protein modules, leading to ligand-induced formation of the binding site, represents a completely novel mechanism for protein-carbohydrate recognition that may mimic that displayed by naturally dimeric protein-carbohydrate interactions.

MeSH terms

  • Base Sequence
  • Binding Sites
  • Carbohydrate Metabolism*
  • Carbohydrates / chemistry
  • Cellulase / chemistry
  • Cellulase / genetics
  • Cellulase / metabolism
  • Crystallography, X-Ray
  • DNA, Fungal / genetics
  • Dimerization
  • Fungal Proteins / chemistry*
  • Fungal Proteins / genetics
  • Fungal Proteins / metabolism*
  • Glycoside Hydrolases / chemistry
  • Glycoside Hydrolases / genetics
  • Glycoside Hydrolases / metabolism
  • Ligands
  • Macromolecular Substances
  • Models, Molecular
  • Multienzyme Complexes / chemistry
  • Multienzyme Complexes / genetics
  • Multienzyme Complexes / metabolism
  • Mutagenesis, Site-Directed
  • Oligosaccharides / chemistry
  • Oligosaccharides / metabolism
  • Piromyces / genetics
  • Piromyces / metabolism
  • Thermodynamics

Substances

  • Carbohydrates
  • DNA, Fungal
  • Fungal Proteins
  • Ligands
  • Macromolecular Substances
  • Multienzyme Complexes
  • NCP1 protein, Piromyces equi
  • Oligosaccharides
  • cellohexaose
  • Glycoside Hydrolases
  • hemicellulase
  • Cellulase

Associated data

  • PDB/1OH3