Molecular and cellular regulation of human glucokinase

Abstract

Glucose metabolism in humans is tightly controlled by the activity of glucokinase (GCK). GCK is predominantly produced in the pancreas, where it catalyzes the rate-limiting step of insulin secretion, and in the liver, where it participates in glycogen synthesis. A multitude of disease-causing mutations within the gck gene have been identified. Activating mutations manifest themselves in the clinic as congenital hyperinsulinism, while loss-of-function mutations produce several diabetic conditions. Indeed, pharmaceutical companies have shown great interest in developing GCK-associated treatments for diabetic patients. Due to its essential role in maintaining whole-body glucose homeostasis, GCK activity is extensively regulated at multiple levels. GCK possesses a unique ability to self-regulate its own activity via slow conformational dynamics, which allows for a cooperative response to glucose. GCK is also subject to a number of protein-protein interactions and post-translational modification events that produce a broad range of physiological consequences. While significant advances in our understanding of these individual regulatory mechanisms have been recently achieved, how these strategies are integrated and coordinated within the cell is less clear. This review serves to synthesize the relevant findings and offer insights into the connections between molecular and cellular control of GCK.

Keywords: Conformational dynamics; Diabetes; Glucokinase; Protein-protein interaction; Regulation.

Figure 1.

Regulation of GCK activity is essential for glucose homeostasis. Under physiological conditions (black), GCK displays a sigmoidal kinetic response to glucose that is characterized by a Hill coefficient = 1.7. The midpoint of GCK’s glucose responsiveness (K_0.5 = 7 mM) correlates with circulating glucose concentrations and the threshold for glucose-stimulated insulin secretion, providing maximum sensitivity in this region. Activated PHHI-causing GCK variants (red) display lower cooperativity and require lower glucose for insulin secretion. MODY-2 associated, reduced activity GCK variants require higher glucose levels to induce insulin release.

Figure 2.

GCK undergoes large-scale conformational changes. In the absence of glucose, GCK exists as an ensemble of conformations. X-ray crystallography has been used to elucidate one of these conformations (left), which has been coined the super-open conformation due to the sizable opening angle between the small domain (gold) and large domain (gray). In this conformation, the intrinsically disordered active site loop displays no electron density. Glucose (green) binding promotes folding of the small domain towards the large domain and the interdomain angle is reduced to 65° (middle). In this intermediary open structure, the active site mobile loop is still invisible and the C-terminal α13 helix (red) shifts from a solvent exposed position to a more structurally compact conformation between both domains. The closed conformation (right) is the most compact conformation and is characterized by a small opening angle of 40°. In this conformation, the α13 helix is buried and the mobile loop forms an antiparallel β-hairpin (cyan) and folds over the bound glucose molecule.

Figure 3.

GKRP regulates GCK by localizing it in the hepatocyte nucleus. GKRP binds to the super-open conformation, thereby locking GCK in an inactive state and reducing glycolysis. This interaction is inhibited by fructose 1-phosphate (F1P) and promoted by fructose 6-phosphate (F6P). Glucose levels strongly influence nuclear-to-cytosolic translocation of the GCK-GKRP complex. GKRP can be acetylated near its N-terminus by p300, which enhances it inhibitory and nuclear localization potential. Conversely, Sirtuin 2 deacetylates GKRP, subsequently promoting glucose uptake via glucose transporter 2 and favoring cytoplasmic GK-GKRP localization. SUMO-1 conjugation to GCK stabilizes and activates the enzyme and impairs GKRP-mediated nuclear sequestration of GCK. Finally, mitochondrial association of GKRP has been observed, but the influence this has on GCK physiology has yet to be determined.

Figure 4.

GCK is activated and stabilized by interaction with the bifunctional enzyme PFK-2/FBPase-2. PFK-2/FBPase-2 is tightly regulated by phosphorylation of S32, which is promoted by glucagon, AMP and several kinases. Phosphorylated PFK-2/FBPase-2 contributes to high glycolytic activity and low gluconeogenic activity. Insulin modulates liver hepatocyte metabolism by activating protein phosphatases that dephosphorylate PFK-2/FBPase-2, thereby increasing the ratio of glycolysis to gluconeogenesis. A similar regulatory strategy exists in pancreatic β-cells but is instead modulated by glucose. It should be noted however, that gluconeogenesis is of minimal importance in β-cells as they store virtually no glycogen. Data suggest that the phosphatase domain of PFK-2/FBPase-2 weakly interacts with the closed conformation of GCK, a process which correlates with intracellular glucose levels. This interaction increases GCK’s V_max and impedes nuclear localization of GCK by GKRP. In addition, a stabilizing role for the interaction has been described by which PFK-2/FBPase-2 protects GCK from oxidative stress (reactive oxygen species, ROS). PFK-2/FBPase-2 interconverts fructose 6-phosphate and fructose-2,6-bisphosphate, the latter of which promotes phosphofructokinase-1 (PFK-1) activity. PFK-1 has been observed to interact with BAD, but the impact this has on mitochondrial GCK association remains to be seen.

Figure 5.

BAD integrates glycolysis and apoptosis by anchoring GCK to the mitochondrial membrane. The mitochondrial complex consists of BAD, GCK, PP1, Wave-1 and PKA. Phosphorylation of BAD promotes activation of GCK and enhances glycolysis and GSIS. Glucose and insulin promote BAD phosphorylation and BAD-GCK interaction via feedback mechanisms. Additionally, phosphorylation of BAD serves to reduce its interaction with apoptotic proteins and promote β-cell survival. β-cell specific insulin receptor knockout impedes the GCK-BAD interaction and results in a reduction of GCK activity. The BAD-GCK interaction is disrupted in patients with type 2 diabetes, indicating a need for drugs that promote the interaction. Finally, BAD has been shown to interact with PFK-1. Another interaction partner of GCK, PFK-2/FBPase-2, produces the primary regulatory molecule of PFK-1, suggesting a link exists between these important metabolic interactions.

Figure 6.

GCK activity is influenced by the UPS. The proteins of the UPS act as molecular guardians when newly synthesized GCK is misfolded or when GCK aggregates form over time. Ubiquitin (Ub) is covalently linked to the improperly folded or aggregated GCK polypeptides to mark them for degradation, which is carried out by other UPS proteins. MG132 inhibits the UPS and results in an increase in cytoplasmic GCK aggregation. E1 proteins, which are part of the UPS, also link Ub molecules together. Ub₅ chains interact with GCK to significantly increase its activity. Presumably, the increase in GCK activity would result in glycolytic ATP production, which in turn, would allow for further Ub chain formation, thereby enhancing GCK activity to an even greater extent.

Figure 7.

SUMOylation augments GCK’s activity and stability. SUMOylation of GCK occurs primarily at lysine residues that are in close spatial proximity to the six-residue NES. SUMOylation serves to mask the NES and promote nuclear translocation of GCK, in contrast to GKRP-mediated nuclear sequestration of GCK, which impedes GCK activity. It is not clear how these two nuclear localization interactions are connected, but metabolic cues such as glucose or insulin likely determine the fate of SUMOylation, as seen for GKRP-mediated GCK regulation.

Figure 8.

Localization of GCK to insulin granules results in reduced activity and is partially mediated by interaction with NOS dimers. In addition to interaction with GCK, NOS can S-nitrosylate C371 of GCK, which serves two primary functions. It improves GCK’s activity and promotes release of GCK from granules. In addition, a model in which insulin post-translationally modulates GCK granular localization and activity has been postulated. Inhibitors of insulin secretion impede GCK granular localization and the synthesis, packaging and secretion of insulin likely significantly influences granular GCK localization.