Khuri S.*, Nassif H., Al-Ali Merheby H., Khalaf K., Khachfe H., and Keyrouz W.
Computational Sciences and Bioinformatics Unit (HAM, HK) and the Departments of Computer Science (HN, WK), and Mechanical Engineering (KK), American University of Beirut, Lebanon; *The Dr. John T. Macdonald Foundation Center for Medical Genetics, University of Miami School of Medicine, Florida, USA.

          Monosaccharides are single-unit sugars which are essential players in many different biochemical pathways. Their main uses are in the release of cellular energy, in signaling pathways (eg in muscle and nerve tissue), as building blocks for more complex carbohydrates, as agents in the regulation of gene expression, and they are involved in many other cellular processes. DNA (deoxyribonucleic acid), for example, is a compound molecule between a monosaccharide (ribose), nucleic bases and phosphates. For each different function that a sugar molecule can play in the cell, there is a different type of protein that binds to it. These proteins belong to diverse protein families that have little, if any, overall sequence or structural similarity. However, certain groups of them do have structural similarity in the core conserved sugar binding active site. The mannose binding site of phosphomannose isomerase, for example, is very similar in 3D structure to the arabinose binding site of the bacterial transcription regulator AraC (Khuri, Bakker, and Dunwell, 2001). Both belong to the cupin superfamily of proteins, characterized by a small number of conserved residues spanning a diagnostic double stranded beta helix (cupin) domain. Sucrose binding proteins of unknown function, such as those found in the plasma membrane of legume leaves, also belong to this protein superfamily, even though sucrose is a disaccharide (Dunwell, Khuri, and Gane, 2000).
          A recent in-depth study on galactose binding proteins analyzed proteins from 7 nonhomologous protein families (Sujatha and Balaji, 2004). It was found that although there was conservation in amino acid composition and folding within proteins of the same family, there were no similarities across families. This allowed family-specific matrices to be developed for the prediction of galactose binding in proteins of unknown function. In this investigation, we concentrate on proteins that bind monosaccharides. We report on the amino acids present at the binding site, the way in which they fold, and whether there is a metal present within that same fold. The results are then discussed in the context of protein enzymic function and the possible prediction of sugar binding in proteins of unknown function.