Identification of cellulose synthase(s) in higher plants: Sequence analysis of processive beta-glycosyltransferases with the common motif 'D, D, D35Q(R,Q)XRW'

More than ten beta-glycosyltransferases are now recognized that have limited similarity to the amino acid sequence of cellulose synthase from Acetobacter xylinum. Using hydrophobic cluster analysis (HCA), we recently identified two domains and putative catalytic residues in the processive beta-glycosyltransferases. In this study, we have found expressed sequence tags (ESTs) from higher plants (Arabidopsis thaliana, Brassica campestris, and Oryza sativa) that exhibit a limited sequence similarity to the A. xylinum cellulose synthase. These ESTs contain some of the conserved residues identified in the processive beta-glycosyltransferases. Complete sequencing of an EST clone (T88271) from A. thaliana led to the identification of all the conserved residues in the derived truncated polypeptide which appears to be part of a putative cellulose synthase. Sequence comparison of proteins with known function and several unidentified proteins have the 'D, D, D35Q(R,Q)XRW' motif which is considered a strong predictor for beta-glycosyltransferases that includes, among other proteins, cellulose and chitin synthases. The first two conserved aspartic acid residues in this motif were analysed by site-directed mutagenesis, and their replacement by another amino acid led to a loss of cellulose synthase activity in A. xylinum, suggesting that they are essential for enzyme activity. A correlation between the second residue (R or Q) in the Q(R,Q)XRW sequence and the synthesis of a long glucan chain (polysaccharide) or a short glucan chain (oligosaccharide) suggests that this residue may be involved in the degree of processivity.