In addition to providing substantial insight into
In addition to providing substantial insight into substrate recognition, structural studies have revealed important details about mechanisms of glycosylation and HCF-1 cleavage [28,30,33]. Activation of the nucleotide-sugar is accomplished through interactions of the β-phosphate with the N-terminus of an α-helix as well as side chain contacts, including a contact with K842, an essential catalytic residue (Figure 2a middle panel). The previously mentioned contact from the α-phosphate to the amide of the glycosylated residue likely plays a crucial role in catalysis by positioning the side chain in the immediate vicinity of the anomeric carbon and providing additional activation of the leaving group [28, 29, 30]. The structures have also provided strong evidence for an electrophile migration mechanism in which the sugar rotates up and away from the leaving group and towards the nucleophile during the reaction. Surprisingly, the structures of substrate complexes did not show a proximal ‘catalytic base’ that could serve to deprotonate the nucleophile, and two hypotheses have been proposed for how the proton is removed during the reaction. In one, an α-phosphate oxygen serves as the Fisetin . In the other, the proton is translocated through two ordered water molecules to D554, a residue ˜7 Å from the site of reaction (Figure 2a middle and right panel) . While the issue of the catalytic base is unresolved, D554 is catalytically important for glycosylation, but not for HCF-1 cleavage, a difference consistent with its function as a base required for Ser/Thr glycosylation. The structures of OGT substrate complexes showed that peptide backbones of glycosylation and cleavage substrates are superimposable (Figure 2b right panel). Combined with biochemical evidence consistent with a reaction mechanism involving a glutamyl ester, the structural evidence suggested that HCF-1 cleavage initiates when the invariant glutamate in the cleavage motif (Gln in Figure 2b) attacks the anomeric carbon [28,33]. No base is required in this reaction as glutamate is deprotonated at the reaction pH. Other studies have led to the mechanism for cleavage shown in Figure 2e . It has been found that aspartate-containing substrates can also be glycosylated by OGT; this results in isomerization to isoaspartate via a succinimide intermediate . Whether aspartate to isoaspartate isomerization is physiologically relevant for OGT, either as a side reaction requiring repair or as a switch mechanism for activity is not yet known. OGT can also occasionally transfer GlcNAc to cysteines in cellular proteins . OGT has thus revealed that a range of post-translational modifications (O-glycosylation, S-glycosylation, peptide backbone cleavage, and peptide backbone isomerization) are possible for a glycosyltransferase only on the identity of the glycosylated residue. Small molecule inhibitors would also be useful for probing OGT cellular function. Until recently, the best cell permeable OGT inhibitor was peracetylated 5S-GlcNAc, which is converted to UDP-5S-GlcNAc in cells [39,]. Although useful, opportunities to modify this substrate analog to improve affinity or selectivity are limited. High-throughput screening followed by medicinal chemistry and iterative structure-based design has produced small molecule OGT inhibitors with low nanomolar Kds (Figure 3a) [41,]. Structures of a representative inhibitor, OSMI-1a, revealed that the quinolinone-6-sulfonamide (Q6S) moiety aligned perfectly with the uridine moiety of UDP-GlcNAc, lying directly over the imidazole of H901 and making hydrogen bonds with the same residues of OGT (Figure 3b middle panel). OSMI-1a adopts a U-shaped conformation (Figure 3b right panel), allowing the molecule to completely fill a space in the protein that accommodates two substrates stacked on top of each other (see Figure 2b right panel). While these inhibitors contain carboxylates, esterification makes them cell permeable. They should be useful for interrogating OGT’s biological function.