The P446L variant in GCKR associated with fasting plasma glucose and triglyceride levels exerts its effect through increased glucokinase activity in liver

Abstract

Genome-wide association studies have identified a number of signals for both Type 2 Diabetes and related quantitative traits. For the majority of loci, the transition from association signal to mutational mechanism has been difficult to establish. Glucokinase (GCK) regulates glucose storage and disposal in the liver where its activity is regulated by glucokinase regulatory protein (GKRP; gene name GCKR). Fructose-6 and fructose-1 phosphate (F6P and F1P) enhance or reduce GKRP-mediated inhibition, respectively. A common GCKR variant (P446L) is reproducibly associated with triglyceride and fasting plasma glucose levels in the general population. The aim of this study was to determine the mutational mechanism responsible for this genetic association. Recombinant human GCK and both human wild-type (WT) and P446L-GKRP proteins were generated. GCK kinetic activity was observed spectrophotometrically using an NADP(+)-coupled assay. WT and P446L-GKRP-mediated inhibition of GCK activity and subsequent regulation by phosphate esters were determined. Assays matched for GKRP activity demonstrated no difference in dose-dependent inhibition of GCK activity or F1P-mediated regulation. However, the response to physiologically relevant F6P levels was significantly attenuated with P446L-GKRP (n = 18; P <or= 0.03). Experiments using equimolar concentrations of both regulatory proteins confirmed these findings (n = 9; P < 0.001). In conclusion, P446L-GKRP has reduced regulation by physiological concentrations of F6P, resulting indirectly in increased GCK activity. Altered GCK regulation in liver is predicted to enhance glycolytic flux, promoting hepatic glucose metabolism and elevating concentrations of malonyl-CoA, a substrate for de novo lipogenesis, providing a mutational mechanism for the reported association of this variant with raised triglycerides and lower glucose levels.

Figure 1.

(A) Competitive inhibition of GCK using equivalent activities of WT and P446L-GKRP (n = 24). Competitive inhibition of 10 m U/ml GCK by one GKRP unit (WT=black circles, P446L=white circles, negative control=black triangles) was observed over a glucose concentration range of 0–100 mm. Each data point was plotted as a percentage of the negative control in which GKRP was absent and glucose concentration equaled 100 mm. Statistical analysis revealed no difference between the intrinsic inhibitory capacity of WT and P446L-GKRP. (B and C) Competitive inhibition of GCK using equimolar concentrations of WT and P446L-GKRP (n = 3). GCK activity was assayed in the presence of 0, 100 and 150 nm WT (black circles) and P446L (white circles) GKRP, over a glucose concentration range of 0–100 mm. Each data point was plotted as a percentage of the negative control in which GKRP was absent (black triangles) and glucose concentration equaled 100 mm. At 100 nm GKRP, there was statistically no difference between the responses of the two regulatory proteins (although it was noted that the variance of this data was larger than seen for other experiments and n = 3). At 150 nm GKRP, statistical analysis showed the intrinsic inhibitory capacity of P446L-GKRP to be significantly lower than that of the WT regulatory protein at [glucose] over 25 mm (P ≤ 0.03).

Figure 2.

(A and B) F1P negatively modulates both WT and P446L-GKRP-mediated inhibition of GCK, leading indirectly to an increase in enzyme activity. Inhibition of 10 m U/ml GCK by (A) one GKRP unit, 5 mm glucose (n = 18) and (B) 100 nm GKRP, 10 mm glucose (n = 3) was observed over 0–500 µm F1P (black circles=WT, white circles=P446L). GCK activity was plotted as a percentage of that obtained in the absence of either regulatory protein. Statistical analysis revealed no difference in response of WT and P446L-GKRP to F1P up to and including 500 µm F1P (P ≥ 0.1). Ki_apparent from Dixon plots of [F1P] versus 1/GCK activity did not significantly differ between the two regulatory proteins (A) 31.2 ± 4.3 µm for WT versus 24.9 ± 2.0 µm for P446L-GKRP, (B) 54.0 ± 10.8 µm for WT versus 46.3 ± 4.2 µm for P446L-GKRP; mean ± SEM; P ≥ 0.1.

Figure 3.

(A and B) F6P-mediated regulation of P446L-GKRP is significantly diminished compared with WT regulatory protein. Inhibition of 10 m U/ml GCK by (A) one unit GKRP, 5 mm glucose (n = 18) and (B) 100 nM GKRP, 10 mm glucose (n = 3) was observed over 0–500 µm F6P (black circles=WT, white circles=P446L). GCK activity was plotted as a percentage of that obtained in the absence of regulatory protein. Statistical analysis showed there to be a significant difference between the response of WT and P446L-GKRP (A) over the physiologically relevant range 25–500 µm F6P (31.9 ± 2.6 to 23.3 ± 1.0% GCK activity for P446L versus 28.5 ± 2.8 to 21.2 ± 0.8% GCK activity for WT; means ± SEM; P = 0.04, 0.0001, 0.006 and 0.03 for 25, 50, 100 and 500 µm F6P, respectively) and (B) over the entire 0–500 µm F6P range (P < 0.001). Dixon plots of [F6P] versus 1/GCK activity showed Ki_apparent was also significantly different between WT and variant regulatory protein in both (A) the comparable activity (31.0 ± 4.4 versus 21.1 ± 2.7 µm, respectively; P = 0.05) and (B) equimolar concentration experiments (16.6 ± 1.7 versus 9.2 ± 1.0 µm, respectively; P < 0.001).

Figure 4.

GCKR and GCK expression in human liver, pancreas, isolated islets, adipocytes, skeletal muscle and kidney. One microgram RNA was converted to cDNA via RT–PCR and expression of the human GCKR and GCK genes observed via Taqman gene expression analysis. Liver (n = 1), islets (n = 2), pancreas (n=1) and adipocytes (n = 1) were studied (as well as appropriate negative controls). These data show GCKR to be highly expressed in molar excess in the liver compared with GCK. However, in the pancreas and islets, GCK levels are much higher than that of the regulatory protein. In adipocytes, skeletal muscle and kidney both genes are only present at negligible levels.