Plants produce a plethora of structurally diverse natural products. The final step in their biosynthesis is often a glycosylation step catalyzed by a family 1 glycosyltransferase (GT). In biosynthesis of the cyanogenic glucoside dhurrin in Sorghum bicolor, the UDPglucosyltransferase UGT85B1 catalyzes the conversion of p-hydroxymandelonitrile into dhurrin. A structural model of UGT85B1 was built based on hydrophobic cluster analysis and the crystal structures of two bacterial GTs, GtfA and GtfB, which each showed approximately 15% overall amino acid sequence identity to UGT85B1. The model enabled predictions about amino acid residues important for catalysis and sugar donor specificity. p-Hydroxymandelonitrile and UDP-glucose (Glc) were predicted to be positioned within hydrogen-bonding distance to a glutamic acid residue in position 410 facilitating sugar transfer. The acceptor was packed within van der Waals distance to histidine H23. Serine S391 and arginine R201 form hydrogen bonds to the pyrophosphate part of UDP-Glc and hence stabilize binding of the sugar donor. Docking of UDP sugars predicted that UDP-Glc would serve as the sole donor sugar in UGT85B1. This was substantiated by biochemical analyses. The predictive power of the model was validated by site-directed mutagenesis of selected residues and using enzyme assays. The modeling approach has provided a tool to design GTs with new desired substrate specificities for use in biotechnological applications. The modeling identified a hypervariable loop (amino acid residues 156-188) that contained a hydrophobic patch. The involvement of this loop in mediating binding of UGT85B1 to cytochromes P450, CYP79A1, and CYP71E1 within a dhurrin metabolon is discussed.A glycosyltransferase (GT) is an enzyme that attaches a sugar molecule to a specific acceptor and thereby creates a glycosidic bond. GTs are found in all phylae. The nature of the acceptor molecules used is highly diverse, whereas the donor molecules typically are restricted to monosaccharides with either D-or L-configuration linked to a nucleotide (UDP/GDP/ CMP/(d)TDP; Leloir, 1964). Glycosylation reactions constitute an integral part of both primary and secondary metabolism. The final step in plant natural product synthesis is often a glycosylation reaction that serves to stabilize, solubilize, and provide a safe storage form of potentially toxic aglycones. In Sorghum bicolor, the GT UGT85B1 catalyzes glucosylation of Tyr-derived p-hydroxymandelonitrile to yield the cyanogenic glucoside dhurrin (Jones et al., 1999). In planta, neither p-hydroxymandelonitrile nor its degradation products are detectable, indicating that the glucosylation reaction is highly efficient and channeled (Møller and Conn, 1980). This observation has been substantiated by metabolome and transcriptome analyses of transgenic Arabidopsis (Arabidopsis thaliana) plants expressing the entire dhurrin pathway (Tattersall et al., 2001;Jørgensen et al., 2005;Kristensen et al., 2005), as well as by confocal laser-scanning microscopy studies ...

Through binding to and signaling via the insulin receptor (IR), insulin is involved in multiple effects on growth and metabolism. The current model for the insulin-IR binding process is one of a biphasic reaction. It is thought that the insulin peptide possesses two binding interfaces (sites 1 and 2), which allow it to bridge the two alpha-subunits of the insulin receptor during the biphasic binding reaction. The sequential order of the binding events involving sites 1 and 2, as well as the molecular interactions corresponding to the fast and slow binding events, is still unknown. In this study we examined the series of events that occur during the binding process with the help of three insulin analogues: insulin, an analogue mutated in site 2 (B17A insulin), and an analogue in which part of site 1 was deleted (Des A1-4 insulin), both with and without a fluorescent probe attached. The binding properties of these analogues were tested using two soluble Midi IR constructs representing the two naturally occurring isoforms of the IR, Midi IR-A and Midi IR-B. Our results showed that in the initial events leading to Midi IR-insulin complex formation, insulin site 2 binds to the IR in a very fast binding event. Subsequent to this initial fast phase, a slower rate-limiting phase occurs, consistent with a conformational change in the insulin-IR complex, which forms the final high-affinity complex. The terminal residues A1-A4 of the insulin A-chain are shown to be important for the slow binding phase, as insulin lacking these amino acids is unable to induce a conformational change of IR and has a severely impaired binding affinity. Moreover, differences in the second phase of the binding process involving insulin site 1 between the IR-A and IR-B isoforms suggest that the additional amino acids encoded by exon 11 in the IR-B isoform influence the binding process.