RG-II:xylosyltransferase (RG-II:XylT)

Recently, work from the Ulvskov and Geshi groups (274) has provided strong evi­dence that these investigators have identified two Arabidopsis thaliana RG-II-a-D-1,3- xylosyltransferases (RG-II-a1,3XylTs) (274). Following the identification of a novel family of 27 putative Arabidopsis thaliana glycosyltransferases (215) and through a series ofbioin- formatic analyses aimed at identifying novel plant cell wall biosynthetic glycosyltransferases with a predicted Type II membrane topology (359), two of these genes, named RGXT1 (At4g01770) and RGXT2 (At4g01750) were shown to encode proteins with characteris­tics consistent with a function as RG-II-a1,3XylTs. RGXT1 and RGXT2 encode proteins of 361 and 367 amino acids, respectively, share 90% sequence identity, and are members of GT-family 77 (138) (http://afmb. cnrs-mrs. fr/CAZY/). Two additional Arabidopsis genes, At4g01220 and At1g56550, are 68-75% identical to RGXT1 and RGXT2 (274). Expression of truncated soluble forms ofRGXT1 andRGXT2 inbaculovirus-transfected insect cells and enzyme assays using diverse radiolabeled nucleotide-sugars and free monosaccharide accep­tors demonstrated that the expressed proteins catalyze the transfer ofXyl from UDP-a-D-Xyl onto fucose. Biochemical analyses of the synthesized product using specific xylosidases and NMR spectroscopy indicated that the xylose was transferred onto the fucose in an a-1,3- linkage. Based on these results the authors hypothesized that RGXT1 and RGXT2 function in the synthesis of the RG-II side chain A that contains 2- O-methyl-D-Xyl attached in an a1,3-linkage to a-L-Fuc. Acceptor specificity studies demonstrated that both enzymes pre­ferred l-Fuc with an a-anomeric linkage and disaccharide acceptors with Fuc attached at the position 4, rather than at the 2 or 3 position, to another glycosyl residue; all characteris­tics consistent with the structure of RG-II (159). Importantly, RG-II isolated from RGXT1 and RGXT2 mutant walls, but not RG-II from wild type Arabidopsis walls, served as an acceptor for the enzyme, providing strong evidence that RGXT1 and RGXT2 function in RG-II synthesis (274) and providing strong support for the function of RGXT1 and RGXT2 as RG-II-a1,3XylTs. The lack of a clear difference in the structure of RG-II isolated from walls of RGXT1 and RGXT2 mutants compared to wild type walls, however, is perplexing and leaves open the question of whether there is gene redundancy, thus requiring a double (or more) gene knockout mutant to see a phenotype. Alternatively, the question remains as to whether RGXT1 and RGXT2 may have additional or alternative functions in the synthesis of some other, yet to be identified, wall polysaccharide structure. Further studies of RGXT1 and RGXT2 and related genes should clarify their role(s) in pectin synthesis.