Glycogen Phosphorylase (PDB ID = 8GPB)
Created by John Hatch
Glycogen is one of the principle energy-storage molecules for living organisms and is nearly ubiquitous in species ranging from yeast to humans (3). A single glycogen molecule can contain thousands of glucose residues. However, these carbohydrates are only metabolically useful after degradation catalyzed by glycogen phosphorylase, according to the reaction (1,5):
(α-1,4-glucoside)n + Pi → (α-1,4-glucoside)n-1 + α-glucose-1-P
These glucose molecules are the primary source of cellular energy via production of adenine triphosphate (ATP) through cellular respiration. Because so much of an organism's energy is only accessible in the presence of glycogen phosphorylase, the enzyme is crucial to normal metabolic function.
A
single chain of rabbit (Oryctolagus cuniculus) glycogen phosphorylase (PDB ID = 8GPB)contains 842 amino acids and weighs 97.158 kDa (2,3). The chain consists of two domains folded back on each other: the N-terminal domain containing amino acids 1-482, and the C-terminal domain of amino acids 483-842 (4). The N-terminal domain has 9 parallel beta-sheets surrounded by 16 alpha-helices. The helices make up about 200 of the domain's amino acids, and primarily serve to connect the sheets of the protein's core. Additionally, this domain contains the only random coil (AA 509-525). The C-terminal domain has 7 beta-sheets surrounded by 18 helices. The C-terminal domain includes a nucleotide binding fold, flanked by 2 bundles of alpha-helices, as well as the binding site for the cofactor pyridoxal 5'-phosphate.
While the protein may exist in vivo as both a monomer and a tetramer, the predominant form is the
biologically active homodimer. Although a significant portion of each subunit's surface area is hidden between the chains, the dimer is stabilized by a relatively small number of significant loci (1). A tower helix (AA 261-274; red) and the carboxyl termini of the Dali loops (AA 184-185; purple) from the catalytic subunit
interact via Van der Waals forces with the alpha2 helix (AA 261-274; green) of the regulatory subunit. The
strongest bonds form between a beta strand of the catalytic unit (AA 193-195; purple) with the cap loop (AA 40-45; green) that connects the alpha1 and alpha2 helices of the regulatory unit. Arg193 forms hydrogen bonds with the carbonyl groups of Leu39' and Val40', while Glu195 forms a salt bridge with Lys41' (prime denotes residues from the regulatory subunit). These are the primary conserved interactions during allosteric-induced conformational changes (1).
The catalytic activity of the enzyme depends entirely on the presence of a
pyridoxal 5'-phosphate (PLP) cofactor. Phosphorolysis of glycogen is catalyzed by acid attack of the glucose phosphate, which can only take place in the presence of the cofactor. In a concerted reaction step, the substrate phosphate donates a proton while immediately receiving one from the cofactor phosphate group. Therefore, the residues binding PLP and stabilizing its interaction with glycogen are essential to normal enzyme function. The cofactor aldehyde group binds as a Schiff base to Lys680 (red). The binding conformation is fixed by a large number of noncovalent interactions with residues from both the C- and N-terminal domains. Of these, the most significant are the interactions of the cofactor pyridine ring with the aromatic rings of Tyr90, Trp491, and Tyr 648, and the ring's Van der Waals interactions with Gly134, Gly135, and Leu136. The substrate-cofactor interface is stabilized by Arg569 and Lys568 (5,7).
Glycogen phosphorylase was the first known enzyme that was controlled by allosteric effectors. Early studies of glycogen phosphorylase led to the discovery of reversible phosphorylation, which has subsequently been shown to be a ubiquitous regulatory mechanism for protein activation and inhibition (2,3,5). Activation via phosphorylation at
serine-14 is sensitive to concentrations of insulin and glucagon, which monitor intracellular sugar levels (3). When activated, both chains convert from the low-affinity T state (above) to the
active R state (PDB ID: 1GPA) (1). The same active state can also be reached solely through
binding of adenine monophosphate (AMP) via salt bridges with Arg309 and Arg310 (red) in the alpha8 helix (1,4). AMP is a product of ATP hydrolysis and is most plentiful when energy is needed. Together, these mechanisms allow glycogen phosphorylase to function only when glucose or ATP levels are low (1,2,3).
The T conformation is unable to bind to glycogen primarily because
the catalytic site is blocked by the 280s loop. Upon Ser14 phosphorylation, the 10-18 loop reorganizes entirely. It changes from a disordered chain to a roughly helical structure, while unwinding the first half of the alpha1 helix. Here, Ser14P forms a hydrogen bond with Arg16, disrupting the T state hydrogen bonds between Arg16 with Glu105, Gln96, and Val493. Further, the phosphate of Ser14P forms salt bridges with Arg43', Arg69, Asn32', and His36'. His36' then breaks its salt bridge with Asp838 and rotates, stabilizing the new conformation by avoiding steric conflict with Ile68. Additionally, this allows Phe37' to pack between Asp61 and Leu64, unwinding part of the alpha2 helix. These shifts improve the ability of Arg309/Arg310 to bond with AMP, since the AMP binding site lies between the alpha2 and alpha8 helices. Additionally, these changes in noncovalent interactions rotate the two subunits 10º with respect to each other (2). This rotation slightly displaces the two tower helices, which are associated with the catalytic site as well as the interface between the subunits. Their movement results in a
reordered 280s loop, which moves the blocking Asp283 (orange) out of the catalytic site and replaces it with Arg569 (purple) (7). The change from an acidic to basic residue in the catalytic site results in a high-affinity site for glycogen, allowing it to contact PLP (blue) (1,2,3,7).
Interestingly, while rabbit glycogen phosphorylase only shares 49% sequence similarity with yeast glycogen phosphorylase (PDB ID = 1YGP), the two enzymes have 100% sequence similarity at the catalytic site. Although yeast phosphorylase cannot be activated by AMP, the residues involved with stabilization of the cofactor-substrate interface and activation of the catalytic site are entirely conserved (3). This similarity suggests that these residues are highly sensitive to mutation, and minor changes result in an inactive form of the enzyme.
Recently, glycogen phosphorylase has been proposed as a novel target for treatment of type II diabetes. Regulation of glycogen phosphorylase may be a useful mechanism for controlling sugar concentrations in the blood because over 70% of blood glucose is produced by glycogen degradation in the liver. Efforts have focused primarily on competitive inhibition of the active site by glucose analogs. However, the high specificity of the active site has made discovery and production of analogs difficult, and none are currently effective treatments (6).