In a search for molecules that disrupt the GK-GKRP complex, compounds that directly activate GK by binding to an allosteric site of the enzyme were se-rendipitously discovered. The GKAs were shown to increase the affinity of GK for glucose (lower S0.5), and to increase the Vmax, although not all GKAs share the latter characteristic. The degree of positive cooperativity with respect to glucose decreased in the presence of a GKA, as reflected by a decrease in Hill slope from around 1.5-1.7 to approximately 1.2. The glucose saturation curve is also altered from being slightly sigmoidal to being very close to hyperbolic in the presence of an activator .
GKAs bind to the open form of GK in a small pocket remote from the active site. In the super-open form, this allosteric site does not exist. Upon binding of a GKA in the allosteric pocket, GK is prevented from relaxing to the lower energy super-open form, thereby locking the enzyme into the fast catalytic cycle.
In hepatocytes, where the inhibitory regulatory protein GKRP is involved in the regulation of GK activity, the binding of a GKA prevents relaxation to the super-open form to which GKRP binds (vide supra). Since the binding of GKRP is necessary for the sequestration of GK into the nucleus, GKAs have the effect of restricting the localization of GK to the cytoplasm .
To date, the X-ray co-crystal structures of GK with GKAs RO-27-5145 (1) , 2  and with LY-2121260 (3)  have been reported. This work revealed a palm-shaped structure, consisting of a small and a large domain, separated by an interdomain cleft. Cleft residues Asn204 and Asp205, in combination with Glu256 and Glu290 of the large domain and small domain residues Thr168 and Lys169, are implicated in glucose binding. In addition, the allosteric site was identified at the interface connecting the two domains and is surrounded by linking region-I, the large domain (p1 strand and a5 helix), and the small domain (a13 helix) . In each structure, the GKA binds at the allosteric site, which is 20 A removed from the glucose binding site. In the case of 1, the allosteric site is comprised of a ceiling created by residues 65-68, which are part of the first connection linking the two domains of GK, and a floor created by the hydrophobic residues Met235, Met210, Ile211, Val62, Val452, and Ile159. The crucial hydrogen-bond interactions of 1 with GK occur between the amide NH and the thiazole N of 1 to the Arg63 backbone carbonyl O and NH, respectively. Similarly 2 and 3 bind at the same location, and achieve analogous hydrogen-bonding interactions with Arg63. Moreover, 2 and 3 appear to make equivalent interactions with the other residues in the allosteric site.
Regulation of GK activity via the allosteric site facilitates the explanation of the activated kinetic properties of certain GK mutations that have been identified in patients with persistent hyperinsulinemic hypoglycemia of infancy (PHHI). For example, in patients with mutations in one GK allele in which methionine was substituted for valine at residue 455 (V455M), tyrosine for cysteine at residue 214 (Y214C), or valine for alanine at residue 456 (A456V), improved GK activity was exhibited . These residues, which seem to be included in the allosteric site regulatory domain, are important for the activity of GK and also may aid the design of specific GK activators.
In vivo, GKAs increase glucose phosphorylation in the liver, thereby increasing glucose disposal from the blood, increasing hepatic glycogen production, and reducing hepatic gluconeogenesis. GKAs have been shown to be effective in reducing blood glucose levels in several animal models of T2D . In pancreatic p-cells, both in vitro and in diabetic animal models, GKAs lowered the threshold of GSIR and increased the amount of secreted insulin. [1,7,14]. The fact that GKAs can favorably influence glucose homeostasis in the liver and pancreatic p-cells, two sites of action critical to diabetes, makes GK activation an attractive target for the potential treatment of T2D. Despite their potential, GKAs are still in the early stages of clinical development. This is probably because while historically the target appeared attractive for manipulation of liver glucose metabolism, the requirement of a kinase activator, and what was originally perceived as the necessity for the small molecule to interfere with a protein-protein interaction (GK-GKRP) seemed insurmountable.
Was this article helpful?