Studies on Human Glucokinase Regulation

Over the last decades, the enzyme Glucokinase (GCK) has interested biochemists due to its physiological and kinetic characteristics. GCK is vital for glucose homeostasis, as it controls the levels of insulin and glycogen synthesis in the human body by catalyzing the phosphorylation of glucose. GCK displays a positive kinetic cooperativity towards glucose despite being a monomeric enzyme with only one binding site for this substrate. Previous crystallography and NMR experiments showed that cooperativity in GCK originates from order-disorder transitions within the enzyme’s intrinsically disordered small domain that happen in the same timescale as catalysis. A very detailed description of protein structure has been gained from both techniques in the presence of glucose. Unfortunately, the dynamic nature of the unbound GCK state prevents these techniques from revealing its structure. The first two chapters of this dissertation are focused on biochemical and biophysical investigations into the cooperative regulation of GCK. First, we investigated the role of a connecting region between the two protein domains on GCK cooperativity. Using steady state kinetics and equilibrium fluorescence we found the connecting region must possess a minimal length to enable catalytic turnover. Also, we found that the sequence identity of the connecting region is directly linked to GCK cooperativity. Second, a non-natural fluorescent amino acid (fUAA) was site specifically incorporated into a single tryptophan version of GCK to perform Förster resonance energy transfer (FRET) experiments. Steady state FRET, lifetime, and FRET-lifetime measurements on wild-type GCK demonstrate that the enzyme in the unliganded-state exists in at least three uniformly distributed conformations originating from a broad conformational ensemble. A slightly more abundant state is postulated to correspond to the "super open" conformation previously observed in the crystal structure of the unliganded GCK. Upon glucose binding, the conformational ensemble is narrowed to two conformations, one of which is dominant and postulated to be the closed conformation of the enzyme. Similar experiments conducted on GCK variants that activate GCK by abolishing cooperativity, support the hypothesis that these variants function via two distinct mechanisms, one (α-type) that produces a shift in the conformational ensemble to resemble the wild-type glucose bound state, while the other (β-type) does not substantially perturb the conformational ensemble. GCK is specifically regulated in the liver by the Glucokinase Regulatory Protein (GKRP), which binds and inhibits GCK. GKRP inhibition is in turn modulated by sugar phosphates; fructose 1-phosphate weakens the GCK-GKRP binding while sugar 6-phosphates (fructose 6-phosphate or sorbitol 6-phospate) strengthen the binding. As an alternative therapeutic approach to treat diabetes, Amgen has optimized a group of piperazines that bind to GKRP and disrupt the GCK-GKRP complex. Interestingly, the binding sites of sugar phosphates and piperazines are 30Å removed from the GCK-GKRP binding interface. In the third part of this dissertation, we aim to understand the mechanism for the ligand-mediated control of the GCK-GKRP interaction. Using mutagenesis, GCK inhibition assays, stopped-flow and fluorescence techniques we were able to propose a model in which GKRP exists in two conformations with different affinities towards GCK, and GKRP ligands modulate the distribution between the two conformations. The difference in GCK affinity is due to the N-terminus-mediated repositioning of two arginines located at the GCK-GKRP binding interface; in the high-affinity GKRP state, the two arginines are positioned to electrostatically interact with GCK, while in the low-affinity state they are stabilized in different rotameric positions. According to our model, sugar 6-phosphates shifts the GKRP population towards the high-affinity state, while piperazine binding produces the low-affinity state.