|
|
Chaikuad, A., Froese, D.S., Yue, W. W., Krysztofinska, E., von Delft, F., Lee, W.H., Marsden, B.D., Liu. Q., Weigelt, J., Arrowsmith, C.H., Edwards, A.M., Bountra, C., Oppermann, O.
Glycogen is a polymer of α-1,4-linked glucose branched by α-1,6-glycosidic linkages, and serves as a major storage reserve of glucose for organisms from bacteria to human. The biosynthesis of glycogen is initiated by the priming protein glycogenin (EC 2.4.1.186) [1], which synthesizes a linear α-1,4-linked glucan chain of up to ~10 glucose units attached to a protein tyrosyl residue. Subsequently, the major biosynthesis enzyme glycogen synthase will take over the bulk elongation of the glucan chain, in concert with the branching enzyme that catalyzes the α-1,6-glycosidic bond formation every 10-13 glucose units. In mammals there are two glycogenin isoforms, namely muscle GYG1 and liver GYG2 [2], sharing in common catalytic core and GS binding site (~350 aa), while GYG2 contains a unique ~110 aa extension of unknown function at the C-terminus. GYG1/GYG2 catalyzes a self-glucosylation reaction using UDP-glucose as sugar donor, and as sugar acceptor a tyrosine residue (in the first step) or the C4-hydroxyl of the terminal glucose unit from the nascent glucan chain (in subsequent steps). The disease relevance of glycogenin is highlighted by the recent identification of missense and nonsense patient mutations in the human gyg1 gene that caused muscle weakness and cardiac arrhythmia, linking glycogenin deficiency to a new form of glycogen storage disorder [3].
Structural Features
Overall fold
We have determined the crystal structure of human glycogenin-1 catalytic domain (hGYG1) in the apo form at 1.98 Å resolution. hGYG1 adopts the classic architecture of the glycosyltransferase type A (GT-A) fold, featuring a central six-stranded core β-sheet (yellow) encompassing the active site for the glucosyl transfer reaction, flanked by helices. In addition hGYG1 contains three extension regions that are important for its glucosyl transfer and substrate specificity. These include a coil-helix segment ('lid', blue), a βαα subdomain ('arm', purple) that harbours the acceptor residue Tyr195, as well as a neighbouring C-terminal hairpin loop (green). In the apo form, part of the arm and hairpin are disordered suggesting their intrinsic flexibility. hGYG1 crystallizes as a homodimer, with the dimer interface formed by arm subdomain as well as a loop segment connect β5-β6 of the core (red). In this dimeric arrangement, the Tyr195 side-chain is situated ~8 Å from its dimeric counterpart. The structure of hGYG1 is highly homologous to that of rabbit GYG1 (PDB code 1ll0), with a Cα-RMSD of 1.5 Å and 92% sequence identity.
(TIP 1: the background colour can be changed to [WHITE] or [BLACK])
(TIP 2: reset the view)
Binding of Mn and UDP
We have further determined the structure of hGYG1 in complex with cofactor Mn2+ and by-product UDP at 2.26 Å resolution, revealing the same conformation (purple) as in the apo form (pink). UDP, which represents the by-product of the sugar donor UDP-glucose after the glucosyl transfer reaction, is positioned in the active site pocket via a number of interactions with the protein. The uracil ring is sandwiched between two hydrophobic residues Tyr15 and Val82. The ribose 2' and 3' hydroxyl groups form hydrogen bonds with the main-chain peptide bond of Leu9 and Asp102, respectively. The UDP pyrophosphates are stabilized by the active site Mn2+ ion, which in turn is coordinated by a conserved sequence motif containing Asp102, Asp104, and His212.
Ligand-induced conformational change
To provide snapshots of hGYG1 along its reaction path, and identiy possible conformational changes during the glycosyl transfer reaction, we co-crystallized wild-type hGYG1 with Mn2+-UDP glucose (PDB code 3T7O) or Mn2+-UDP (PDB code 3T7M, 3T7N), and determined three structures from different crystal forms. They reveal a drastically different conformation of the enzyme compared to the apo form, involving structural rearrangements in three regions of the protein that influence active site accessibility. These include a ‘lid’ segment (aa 60-91), a helix-turn-helix ‘acceptor arm’ (aa189-207) harbouring the Tyr195 acceptor residue, and the C-terminal loop (‘C-loop’; aa 233-243) located close to the acceptor arm.
Superimposition of the two hGYG1 conformers reveals a maximum motion in the lid segment amongst the three flexible regions. In the apo enzyme, the lid adopts an open conformation (blue) positioned away from the active site, leaving it accessible to the exterior. Upon binding the cofactor (UDP glucose), the lid adopts a closed conformation (orange) resulting from a ~19 Å positional shift and ~77° rotation, thereby shielding the active site pocket and reducing its accessibility. In addition to the lid movement, other less dramatic structural alterations between the two conformers including the concomitant ordering of the acceptor arm and C-loop, which lie in different polypeptide stretches but are topologically juxtaposed, with Pro238 from the C-loop packing against helix α7 of the acceptor arm that harbours the Tyr195 acceptor residue.
T83M disease mutation
Recently, a missense mutation Thr83Met is associated with a new form of glycogen storage disorder type 15 (GSD15). We have determined the structures of hGYG1-T83M mutant in complex with Mn2+-UDP, and in complex with Mn2+-UDPG. both adopting the same protein conformation (cyan and yellow) as in the wildtype apoenzyme (pink).
The sugar donor UDPG binds to the active site pocket in a similar manner as UDP. UDPG adopts the classic ' folded back' geometry as observed in many glycosyltransferases, where the donor glucose moiety is tucked below the pyophosphates.
The Thr83 residue, which is mutated to Met in the disease variant, resides in the lid segment close to the active site pocket. It is intimately locked in a hydrophobic pocket surround by residues in the lid (Leu80, Gly81, Leu87) as well as a neighbouring loop segment (Gly157, Ser158, Gly161, Gly162, Asp163).
Reference
Updates: All SGC datapacks are frequently updated, according to new findings and information. To check for updates for this specific datapack, please
click here. Internet connection is required.After download, the new datapack will be automatically launched and a copy will be saved in your local drive.
Note: The target annotations and structure descriptions within this datapack are compiled by our Principal Investigators and are not peer-reviewed. If you find anything in the annotations that is not accurate, please notify us using the our on-line feedback page or send an e-mail to isee@sgc.ox.ac.uk.
Datapack created using Molsoft ICM and Molsoft Browser technologies (patent pending).