The intracellular enzyme glucokinase (GK) is a key component of the body’s glucose-sensing network. GK activity is regulated by changes in blood glucose in the physiological range. Considerable effort has therefore been made to activate GK pharmaceutically and led to the discovery of Glucokinase Activators (GKAs). GKAs are orally available small molecules which bind to an allosteric activating site on GK and increase GK activity by increasing its affinity for glucose and/or by increasing its maximal velocity. Several GKAs have been in clinical development as a potential new therapeutic approach for diabetes. GKAs work primarily by (i) increasing insulin secretion from pancreatic β-cells and (ii) in hepatocytes by increasing glucose uptake and consequently decreasing glucose output.
Different GKAs have different effects on GK’s kinetic properties and have different impacts on target tissues. In preclinical models and in several clinical trials on type 2 diabetic patients, GKAs have been shown to increase insulin secretion and decrease blood glucose and HbA1c. However, there are side-effect concerns associated with GKAs, particularly hypoglycaemia on a daily basis, and over longer treatment periods the possibility of liver steatosis and dyslipidaemia. Moreover, GKA’s positive effects may diminish over the longer term. These concerns bring challenges to the development of GKAs as a novel type 2 diabetes therapy.
Glucokinase and “glucose sensing”
Glucokinase (GK) (EC 18.104.22.168; ATP:D-glucose 6-phosphotransferase) is a monomer of ~50kDa also known as hexokinase D or type IV, and is one of four hexokinases catalysing the phosphorylation of glucose (using ATP) and some other hexoses in the cell cytoplasm .
GK differs markedly from other hexokinases through:
(i) its kinetic properties, having a much higher Km (binding affinity) for glucose (and within the blood physiological range), exhibiting positive co-operativity towards glucose, and not being inhibited by glucose-6-phosphate
(ii) its tissue distribution, with expression in pancreatic β-cells and liver hepatocytes (>99% of GK is in liver), and in a few other cell types, including incretin-secreting entero-endocrine gut L- and K-cells, pancreatic α-cells, and some CNS glucose-sensing cells (including hypothalamic neurones and pituitary gonadotropes) all of which also respond to changes in blood glucose. These “glucose-sensing” cells take up glucose concentration-dependently via the GLUT2 glucose transporter.
In liver GK is also regulated by GK Regulatory Protein (GKRP) which reduces GK activity (e.g. in starvation) by sequestering GK into the nucleus.
These key features enable GK’s role as a key component of the body’s “glucose-sensing” network and in controlling blood glucose homeostasis, with glucose-induced increased GK activity in pancreatic β-cells (leading to increased insulin biosynthesis and release) and in parenchymal liver cells (leading to increased glucose uptake and metabolism, and reduced glucose output) being paramount. Thus GK activation has long been considered as a potential anti-diabetic approach.
Glucokinase mutations and consequences
More than 600 mutations have been described for the human GK gene which is inherited in an autosomal dominant manner .
De-activating mutations cause the monogenic trait MODY due to glucokinase mutations, or MODY 2, a condition characterised by elevated blood glucose resulting from GK haplo-insufficiency; full insufficiency causes Permanent neonatal diabetes (PNDM)
Activating mutations are rarer but several have been described, with affected individuals presenting with Persistent Hyperinsulinemic Hypoglycemia of Infancy (PHHI) of varying degree. This is largely attributed to a decreased threshold for glucose-stimulated insulin secretion, although effects due to liver GK activation may also be involved; GKRP mutations are also known which lead to GK activation.
These naturally occurring activating (and de-activating) mutations reinforced the idea that GK activation could be a viable therapeutic approach for hyperglycaemia in type 2 diabetes.
Glucokinase activators (GKAs)
Hoffman-La-Roche first reported (2003) that GK could be activated directly with small molecules that had the potential to be developed into orally-available pharmaceuticals . Their discovery of Ro-28-1675 followed a high-throughput screening campaign with human GK and chemical-optimisation.
Figure 1. Schematic of the dual mechanism underlying anti-hyperglycaemic efficacy of GKAsRo-28-1675 was shown to activate GK in a mixed manner by increasing its affinity for glucose (i.e. decreasing its Km) and increasing its maximal velocity (Vmax). Ro-28-1675 enhanced glucose-stimulated insulin release from rat pancreatic islets and increased hepatocyte glucose uptake, and had acute and chronic anti-hyperglycaemic effects in rodent models (Figure 1).
This led to much pharma interest, and over 150 patents (June 2014) filed for GKAs, and several compounds have entered clinical development. All these novel GKAs share some overall common pharmacophoric features and bind within an allosteric site on GK, distal to the active site and in the same region as activating mutations. X-ray crystallographic studies have described in detail how binding and GK activation is brought about via stabilisation of its closed form .
GKAs have also been shown to increase GK activity of some MODY2 mutations. Different GKAs have differing degrees of effects on glucose Km and Vmax although what mix is optimal is open to debate. They also have different pharmacokinetics (e.g. some suitable for post-prandial regimes) and distribution (e.g. several are reported as more liver selective). GKAs also disrupt binding of GK to GKRP in liver leading to further GK activation.
Figure 1. Schematic of the dual mechanism underlying anti-hyperglycaemic efficacy of GKAs
Preclinical efficacy of GKAs
Ro-28-1675  already indicated the potential anti-diabetic properties of GKAs, and oral single doses in diabetic rodent models, including oral glucose tolerance tests, demonstrated increases in plasma insulin accompanied by marked blood glucose reductions. Clamp studies in rodents with controlled insulin and glucose levels also revealed the effects of Ro-28-1675 on liver to suppress endogenous glucose production and promote liver glucose uptake.
There are now many reports on GKAs showing anti-hyperglycaemic efficacy acutely and sub-chronically in preclinical diabetic models; e.g. PSN-GK1  showed impressive anti-hyperglycaemic efficacy when given orally for several days to the diabetic models, db/db mouse, ob/ob mouse, and high fat fed female Zucker Diabetic Fatty rat. HOMA-IR and HbA1c indices of longer-term blood glucose control have also been shown to be improved in preclinical studies and there are indications that GKAs may increase β-cell mass.
GKAs in these hyperglycaemic models appeared to normalise blood glucose but not cause hypoglycaemia. However, when GKAs were given at suitably high doses (giving efficacy in diabetic models) to normal animals then this could result in hypoglycaemia, which is not surprising given their mechanism.
Clinical efficacy of GKAs
Several GKAs have entered clinical trials since 2008, and good efficacy indications have been seen with all that have had data reported, with clear findings of insulinotrophic, anti-hyperglycaemic efficacy and reductions in HbA1c, and with indications of both pancreatic β-cell and hepatic effects. Thus far (June 2014) however, none seems to have progressed beyond Phase 2. At least 18 companies have had one or more GKA in clinical trials, although many of the GKAs trialled in the clinic have now been discontinued and many (including most of the larger) companies now seem to have left the field.
However, GKAs continue to be developed, with a current bias towards GKAs which are liver selective (June 2014): these include TransTech Pharma (TTP-399; Phase 2), Advinus (GKM-001, Phase 2), Teijin Pharma (TMG-123; Phase 1), Hua Medicine (HMS-5552, Phase 1). The “partial” GKA PF-04937319 was in Pfizer’s May 2014 pipeline, however no trials appear to be ongoing (June 2014).
Concerns and challenges in developing GKAs as novel anti-diabetic therapies
A first concern, as with other insulin secretagogues, is the potential for GKAs to cause hypoglycaemia. This is thought primarily due to their potent insulinotrophic properties in pancreatic β-cells via their mechanistic reduction in GK’s glucose Km resulting in a lower blood glucose threshold for insulin secretion, although a concomitant GKA-induced reduction in the liver’s capability to enhance glucose production in hypoglycaemia should also be considered. Hypoglycaemic efficacy is of course desirable in patients with hyperglycaemia, and risk of hypoglycaemia can be managed as with several currently used diabetes therapies, and also in individuals with activating GK mutations.
This has, however, led to a search for more liver specific GKAs such as TTP399, which was recently reported (June 2014) to normalise HbA1c in type 2 diabetics on stable metformin after 6 weeks dosing, and with no increases in hypoglycaemia or in fasting plasma triglycerides, cholesterol, lactate, insulin or C-peptide .
Nevertheless, a second concern is the potential for fatty liver induction and dyslipidemia with GKAs. GKA stimulation of liver glucose metabolism (Figure 1) has the potential to increase rates of glycogenesis and glycolysis (and hence pyruvate, lactate and alanine production) – and also of de novo fatty acid synthesis which can lead to enhanced lipid levels in liver and plasma, and this undesired potential side effect of GKAs has been observed in both preclinical  and clinical  studies.
This concern had already been indicated by some GK over-expression studies , although generally humans with activating GK mutations do not seem to have lipid problems. The comprehensive preclinical study by De Ceunick et al.  is worth noting in this regard: 9 different GKAs belonging to three distinct chemical classes and having a range of_ in vitro_ and in vivo activities were studied in different rodent models. Significant increases in liver triglyceride were observed as early as after four days treatment, and most importantly there was a strong correlation between GKA efficacy in lowering HbA1c and the degree of liver triglyceride accumulation.
A third concern, which is perhaps the most critical, is that the HbA1c and glucose lowering effects of GKAs appear to diminish over longer treatment periods which would clearly be a disadvantage to their development for the chronic disease type 2 diabetes. This possibility had already been indicated in some preclinical studies, however it has been the finding in two separate longer-term clinical studies with two different GKAs that their effects appeared to diminish over time that has caused most concern  . This could be related to an induction of insulin resistance due to possible liver lipid accumulation noted above. The study with MK0941  also reported some hyperlipidemia in treated patients, as well as increases in vascular hypertension.
These concerns, some of which have only recently emerged, represent challenges that will need to be addressed before GKAs can be successfully advanced as a novel therapeutic approach for diabetes.
Dr. Jim McCormack held stock or stock options in (OSI)Pharmaceuticals Inc. until 2009 and in Novo Nordisk until 2010, however he has no conflicting interests relevant to this Diapedia page written in June 2014.
^ Matchinsky FM. Assessing the potential of glucokinase activators in diabetes therapy. Nat Rev Drug Discov 2009;8:399-416
^ Osbak KK et al. Update on mutations in glucokinase (GCK), which cause Maturity Onset Diabetes of the Young, Permanent Neonatal Diabetes, and Hyperinsulinemic Hypoglycemia. Hum Mutat 2009;30:1512-26
^ Grimsby J et al. Allosteric activators of glucokinase: potential role in diabetes therapy. Science 2003;301:370-3
^ Fyfe MCT et al. Glucokinase activator PSN-GK1 displays enhanced antihyperglycaemic and insulinotropic actions Diabetologia 2007;50:1277-87
^ Valcarce C et al. TTP399, a liver selective glucose kinase activator (GKA) lowers glucose and does NOT increase lipids in subjects with type 2 diabetes mellitus (T2DM). American Diabetes Association (ADA) Scientific meeting 2014, abstract number 122-OR; www.diabetes.org
^ De Ceuninck F et al. Small molecule glucokinase activators disturb lipid homeostasis and induce fatty liver in rodents: a warning for therapeutic applications in humans. Br J Pharmacol 2013;168:339-53
^ Meininger GE et al. Effects of MK-0941, a novel glucokinase activator, on glycemic control in insulin-treated patients with type 2 diabetes. Diabetes Care 2011;34:2560-6
^ Ferre T et al. Long-term overexpression of glucokinase in the liver of transgenic mice leads to insulin resistance. Diabetologia 2003;46:1662-8
^ Wilding JPH et al. Dose-ranging study with the glucokinase activator AZD1656 in patients with type 2 diabetes mellitus on metformin. Diab Ob Metab 2013;15:750-9