A Fresh View of Glycolysis and Glucokinase Regulation: History and Current Status^*

History of Glycolysis Research

Research during the last 25 years has provided a fresh view of glycolysis. This minireview focuses on the importance of the upper part of the glycolytic pathway, the fructose esters fructose 6-phosphate (Fru-6-P)² and fructose 1,6-bisphosphate (Fru-1,6-P₂) (see Fig. 1), for metabolic flux regulation, with special emphasis on the role of glucokinase.

In 1913, Carl Neuberg (born July 29, 1877, in Hannover, Germany; died May 30, 1956, in New York) considered the glycolytic breakdown of the glucose molecule to be achieved through the formation of a methylglyoxal intermediate (1). Although this concept ultimately turned out to be incorrect, it dominated the thinking of the research community for many years (2).

In 1933, Gustav Embden (born November 10, 1874, in Hamburg, Germany; died July 25, 1933, in Nassau, Germany) proposed a new scheme for glycolysis that comprised the crucial step for splitting Fru-1,6-P₂ into the triose esters dihydroxyacetone phosphate and glyceraldehyde 3-phosphate (3). Publishing this new scheme in the journal Klinische Wochenschrift (3) allowed Embden to bypass the journal Biochemische Zeitschrift, for which Neuberg was the editor. Embden was probably the only one at that time who believed this scheme (2), but it ended up being correct. Nevertheless, as Carl Cori mentioned later, progress in the research field was delayed by the “persistence of many wrong leads” (2).

After Embden's sudden death in 1933, other researchers, in particular Otto Meyerhoff (born April 12, 1884, in Hannover; died October 6, 1951, in Philadelphia) and Jakub Parnas (born January 16, 1884, in Mokrinay/Drohobych, Ukraine, near Lviv; died January 29, 1949, in Moscow), proceeded with experimental work on glycolysis, and within a few years, the steps involved in the phosphorylating degradation of glucose as known today were fully elucidated (1, 2). This is why the glycolytic pathway is also known as the Embden-Meyerhof or Embden-Meyerhof-Parnas pathway.

Glycolysis was the first important metabolic pathway elucidated between 1913 and 1938 (1). Metabolic flux is typically regulated during the initial steps of a metabolic pathway (for review, see Ref. 4). 6-Phosphofructo-1-kinase (PFK1)-mediated phosphorylation is the regulated step in glycolysis. In tissues such as pancreatic beta cells and liver, it is the phosphorylation of glucose by the low affinity glucose-phosphorylating enzyme glucokinase. In some tissues, glucose uptake into the cell can be rate-limiting, such as is the case for GLUT4 glucose transporter-mediated glucose uptake in muscle.

Being of Jewish descent, Neuberg and Meyerhof were dismissed from their academic positions at the Kaiser Wilhelm Institute in Berlin by the National Socialist regime in 1933. Embden, who was professor of physiology at Frankfurt University at that time, died soon after the Nazi regime came into power in Germany. Although born in Galicia, Parnas had studied in Berlin before becoming a professor at the universities of Strasbourg, Warsaw, Lviv (Lemberg), and finally Moscow, where he died in 1949 from a heart attack during an interrogation by the KGB in the infamous Lubyanka prison.

Embden and Meyerhof were medically qualified; Neuberg and Parnas obtained their first degrees in chemistry during their university educations. In a sense, they were all life scientists who worked at the interface between medicine and biochemistry. Their lives were much affected by the political geography of the twentieth century in Europe. Embden and Parnas died of sudden cardiac death triggered by exceptional emotional stress in difficult times of their lives. Neuberg and Meyerhof emigrated to the United States in the late 1930s. Thus, the political events of the twentieth century with the cruel dictatorships in Germany and the Soviet Union had a severe impact upon research in this field of biochemistry (for further historical information, see Refs. 5–7).

Role of Glucokinase in Regulation of Physiological Insulin Secretion from Pancreatic Beta Cells

The glycolytic pathway is characterized in some cell types (8–10), in pancreatic beta cells (4, 11–13) and alpha cells (14), hepatocytes (15–19), certain specific pituitary cells (20), and brain neurons (21) by a specific feature. This is the expression of a fourth hexokinase isoenzyme, glucokinase, which has a K_m value in the physiological range of blood glucose concentration (15–17, 22–25). These kinetic properties provide the glycolytic pathway in these cell types with additional options for the regulation of metabolic flux in both the healthy and diabetic states (4, 11, 13).

In pancreatic beta cells of the islets of Langerhans, metabolism has a signal-generating function for glucose-induced insulin secretion. Glucose transporters do not limit access of glucose to beta cells (4, 13), even in human beta cells, where the expression of the GLUT2 glucose transporter is lower than in rodent beta cells (26, 27). Together with the GLUT1 glucose transporter, the expression of these two transporter isoforms supports sufficient glucose uptake into the beta cell in the postprandial state, reaching millimolar concentrations that are in the range of the K_m of glucokinase. Glucokinase thus equips the insulin-secreting cells with the capability to connect changes in the physiological blood glucose concentration to changes in glycolytic flux (4). Two pathways act in concert (for review, see Ref. 28) in beta cell glucose metabolism; glucokinase serves as the glucose sensor (11) and, by recognizing glucose, initiates glucose-induced insulin secretion (4, 13). When blood glucose increases during the postprandial phase, glucokinase is the dominating glucose phosphorylator because, in contrast to glucokinase, high affinity hexokinase isoenzymes are inhibited by glucose 6-phosphate (Glc-6-P) (4, 29, 30). Thus, the regulation of glucokinase activity is the crucial determinant for the flux rate through glycolysis during the postprandial state.

The first signaling pathway of glucose-induced insulin secretion comprises a number of steps involving glucose uptake and metabolism in beta cells, depolarization of the plasma membrane through the closure of a K⁺_ATP channel induced by an increase in the ATP/ADP ratio, and the subsequent opening of voltage-sensitive Ca²⁺ channels, initiating the exocytosis of insulin via an increase in the free cytosolic Ca²⁺ concentration. The last crucial component of this pathway to be identified was the K⁺_ATP channel in the beta cell membrane (31–33). This pathway is well established and called the “triggering pathway” (28).

Metabolic flux is increased through glycolysis, the citric acid cycle, and the respiratory chain to achieve changes in the citrate concentration and the NAD(P)/NAD(P)H and ATP/ADP ratios (34–36). The increase in citrate and the ATP/ADP ratio can also inhibit glycolytic flux allosterically through inhibition of the glycolytic enzymes PFK1 and pyruvate kinase and also indirectly through inhibition of 6-phosphofructo-2-kinase (PFK2), thereby providing an opportunity to link the pulsatile activities of glycolysis, K⁺_ATP channels, and Ca²⁺ channels to the rate of insulin exocytosis.

A second K⁺_ATP channel-independent pathway was also described (37), called the “amplifying pathway” of glucose-stimulated insulin secretion (28, 38). This latter pathway amplifies the signal generated by the triggering pathway through metabolic amplification of the action of increased cytosolic Ca²⁺ (28). Although not yet fully elucidated, the amplifying metabolic signals are most likely generated in the tricarboxylic acid cycle. These metabolites, in particular citrate, are exported into the beta cell cytosol (39). In quantitative terms, the amplifying pathway provides at least as much insulin to the organism as the triggering pathway (28).

Role of Glucokinase in Glycogen Storage in the Liver

In the liver, glucokinase regulation is important for synthesis and subsequent storage of glycogen in the postprandial state, but not for energy, as the liver does not rely on glucose as its primary energy source. During starvation, when liver glycogen stores decrease, glucose supply to the brain is maintained by an increasing rate of gluconeogenesis. In this context, glucose phosphorylation by low affinity enzymes is minimal due to the low blood glucose concentration and because glucokinase is translocated into the nucleus and inhibited by its binding to glucokinase regulatory protein (GRP) (40–45).

Differential Regulation of Glucokinase in Pancreatic Beta Cells and the Liver

Not only is glucose the crucial nutrient for beta cell survival, but it also fuels insulin biosynthesis and exocytosis. In the liver, insulin and glucagon together regulate gene expression, whereas glucose plays this role in beta cells through changes in its concentration during feeding-starvation-refeeding cycles. Glucose is the crucial regulatory molecule for pancreatic beta cell glucokinase (46, 47). The regulation of glycolytic flux through enzyme phosphorylation and dephosphorylation does not play the prominent role in beta cells that it does in the liver. Glucokinase enables glycolysis to generate the metabolic signals needed for initiation and amplification of insulin secretion (4, 28).

Glucokinase Regulation

For many years, glucokinase has been considered to be only weakly regulated (15–17). Early on, glucokinase enzyme activity was known to be dependent on the nutritional status of the cell. Only later did it became evident that glucokinase regulation is different in hepatocytes and pancreatic beta cells. Following 48 h of starvation, glucokinase enzyme activity was reduced in the liver by nearly two-thirds (47, 48), and gene expression was reduced to nearly zero (47, 49), whereas glucokinase activity and gene expression were reduced by only 50% in pancreatic beta cells (47, 48). Refeeding experiments indicated that the restoration of starvation-reduced glucokinase enzyme activity, as well as glucokinase gene expression, is insulin-dependent in the liver but glucose-dependent in beta cells (47). These findings were subsequently confirmed in in vitro experiments (46).

A teleological explanation can be provided for the smaller reduction in glucokinase expression and activity seen in beta cells. In the liver, glucokinase mediates the postprandial phosphorylation of glucose needed for the synthesis and storage of glycogen, whereas in beta cells, glucokinase is involved in the generation of the metabolic signals necessary for physiological glucose-induced insulin secretion. Thus, even after a long starvation period, at least some glucokinase enzyme activity is instantaneously available in beta cells to link increased blood glucose to glucose phosphorylation. After a meal, the metabolically stimulated insulin secretion can then immediately supply the organism with insulin, and the insulin-sensitive organs will gain abundant energy via insulin-regulated glucose uptake through the GLUT4 glucose transporter. Thereafter, superfluous glucose is stored as glycogen in the liver. Therefore, a delayed restoration of glucokinase expression and enzyme activity during energy replenishment after starvation in the liver is physiologically favorable.

Liver Glucokinase Regulation through Inhibitory Proteins

Hepatocyte glucokinase is inhibited through a specific inhibitory protein, GRP, which was detected by Van Schaftingen in 1989 (42). GRP is a nuclear protein that binds glucokinase, thereby inactivating the enzyme during starvation (40–45). Fru-6-P strengthens the inhibitory interaction between glucokinase and GRP, whereas fructose 1-phosphate, which is an intermediate of fructose degradation but not of glycolysis, weakens this interaction (40–43). After re-exposure to food, glucokinase is released from being bound to GRP in the nucleus, returning it to its active form in the cytosol (40–43). Fructose 1-phosphate, which is generated by fructokinase, particularly after fructose ingestion (50), acts antagonistically to Fru-6-P to cause release of glucokinase from binding to GRP in the nucleus (40–43). This increases the availability of active enzyme for glucose phosphorylation and subsequent glycogen synthesis in the cytosol.

Pancreatic Beta Cell Glucokinase Regulation through Inhibitory Proteins

Although GRP is not expressed in pancreatic beta cells (46), beta cells express proteins that bind to glucokinase and inhibit its enzyme activity. This occurs through ubiquitination (51) and proteins (midnolin and parkin) having ubiquitin-like domains (52). These interactions are weaker than the interaction between glucokinase and GRP. Glucokinase ubiquitination can also occur in the liver, but it is less important there than in beta cells because the strong glucokinase inhibitor GRP localizes glucokinase to the nucleus and therefore reduces the likelihood of glucokinase being ubiquitinated in the cytosol.

Pancreatic Beta Cell Glucokinase and Liver Glucokinase Are Regulated by the Same Stimulatory Protein

In 2001, the enzyme PFK2/fructose-2,6-bisphosphatase (FBPase) was identified as a potent glucokinase activator in both pancreatic beta cells and liver (53). Activation is achieved through the binding of glucokinase to the bisphosphatase domain of the enzyme. This interaction is physiologically crucial for maintaining glucokinase activity after food ingestion during a state of maximal activity. This activation increases the capacity for glucose phosphorylation at any given glucose concentration in both organs but does not decrease the affinity of glucokinase for glucose, avoiding the risk of hypoglycemia (54). This is in contrast to the action of small molecule chemical glucokinase activators (8). These chemical compounds have not yet found their way into antihyperglycemic therapy of diabetic patients with defective beta cell responsiveness to physiological glucose stimulation, as none of the candidate molecules synthesized thus far are able to activate glucokinase enzyme activity without concomitantly reducing the K_m value of the enzyme (54), the prerequisite for a safe drug. Although a large number of compounds with different chemical structures have been synthesized and patented and all having glucokinase-activating potency (55), it is an open question whether a synthetic glucokinase activator with an optimized chemical structure (56) and without serious hypoglycemic potential will be developed. Interestingly, nature has already achieved this with the endogenous glucokinase activator PFK2/FBPase-2 (54).

A Fresh View of Glycolysis: The Importance of the Fructose Ester Steps for Its Regulation

The classical view is that Glc-6-P has to be converted first in the glycolytic pathway into Fru-6-P because only the fructose molecule can be cleaved into trioses. However, this view is too simple because the role of the fructose esters (Fru-6-P and Fru-1,6-P₂) (Fig. 1) in glycolysis is more complex.

The glycolytic pathway contains a regulatory unit composed of the fructose ester steps at the interface between the initial step of glucose trapping in the cell through phosphorylation and the conditioning of the glucose molecule for catabolism (Fig. 1). Several molecules in these fructose phosphorylation steps are regulatory in nature and determine the glycolytic flux rate.

As regulator of PFK1, Fru-1,6-P₂ is weak, and fructose 2,6-bisphosphate (Fru-2,6-P₂) is strong. Fru-6-P is an activator of PFK2, as well as of GRP in the liver. These fructose esters (Fru-1,6-P₂ and Fru-2,6-P₂), as well as Fru-6-P, can also inhibit the respective phosphatases FBPase-1 and FBPase-2. However, only the inhibition by Fru-2,6-P₂, as well as that by Fru-6-P, takes place at physiologically relevant concentrations. Through this inhibition of phosphatases, glucose catabolism is not hampered in the postprandial state when glucose supply is plentiful. This is the case in both tissues.

PFK1/FBPase-1 and PFK2/FBPase-2 are bifunctional enzyme complexes, and different isoforms exist (53, 57, 58). It is the muscle-type PFK1 isoform and not the liver-type isoform that is expressed predominantly in beta cells, and it is also this isoform that is activated by micromolar Fru-2,6-P₂ concentrations under near-physiological conditions (57). Conversely, FBPase-2 is inhibited.

The liver-type PFK2/FBPase-2 isoform is regulated by glucagon and insulin through phosphorylation and dephosphorylation via cAMP, whereas the brain isoenzyme, which is expressed in pancreatic beta cells (53), is not regulated through this mechanism. The activation of glucokinase through binding of the FBPase-2 domain is antagonized in the liver by cAMP and mediated through a regulatory site for phosphorylation by a cAMP-dependent protein kinase (45), whereas this interaction is glucose-regulated in pancreatic beta cells (59).

In the liver, where these phosphatases are activated during starvation through phosphorylation via a cAMP-mediated mechanism involving the action of glucagon, they provide Glc-6-P for gluconeogenesis by glucose-6-phosphatase (60), allowing glucose neoformation during periods of food deprivation. As pancreatic beta cells do not express glucose-6-phosphatase (61), gluconeogenesis is not functional in this cell type.

Conclusion

Whereas glycolysis research in the first decades of the twentieth century was devoted to the elucidation of intermediates and enzymes in the pathway, research in the last decades has sought to identify crucial regulatory components of the pathway. This research allowed the definition of a regulatory unit in the glycolytic pathway that comprises the two fructose steps, in which a variety of regulatory molecules are generated (Fig. 1). The phosphofructokinase-mediated enzymatic steps confer the pulsatile nature to glycolytic flux and physiological insulin secretion, whereas the fructose (bis)phosphate esters confer the potential to regulate metabolic flux through glycolysis. With its bisphosphatase domain, PFK2/FBPase-2 is the endogenous activator of glucokinase. It endows the glucokinase-expressing tissues with the ability to activate glucokinase in the postprandial phase after a starvation period and fulfill its task as a signaling enzyme after carbohydrate ingestion, inducing glycogen storage in the liver and glucose-induced insulin secretion from pancreatic beta cells. Thus, although glucose phosphorylation by hexokinase and in particular glucokinase initiates glycolytic flux, the PFK/FBPase isoenzymes and their (bis)phosphate ester products are the crucial regulatory molecules that control metabolic flux through glycolysis.