Sugar Making Sugar: Gluconeogenesis Triggered by Fructose via a Hypothalamic-Adrenal-Corticosterone Circuit

Investigations of interconnecting physiological pathways are complex and await our complete understanding. One such pathway is the cellular use of hexose sugars as fuel to generate energy in the form of ATP and reducing equivalents. The complex balance of enzymes and substrates involved in this process is highly regulated and poised to affect multiple pathways in a dizzying network of biochemical reactions when the equilibrium is disturbed (1, 2). Although our bodies are primed to use glucose, consuming fructose, a molecule almost identical to glucose, adds an additional level of complexity to the system (3). In the current issue of Endocrinology, the work of Kinote et al. (4) have added insight to the effects of fructose metabolism by understanding one circuit in its network of circuits; the involvement of corticosterone.

Increased consumption of fructose as a sweetener in the form of high-fructose corn syrup (HFCS) has made it to the headlines in recent years due to its apparent association with increased obesity and type 2 diabetes mellitus in the general population, but more so in the younger generation (2, 3, 5). Indeed, the war of words wages on our TV screens at home as to its benefits or its detriments. HFCS contains a mixture of the monosaccharides glucose and fructose, artificially generated from corn starch (6) to mimic the natural ratio of these molecules found in sucrose. Consumption of fructose at these high levels can trigger a cascade of events originating in the liver, which include lipogenesis and gluconeogenesis that lead to triglyceridemia, hepatic steatosis, and insulin resistance (5, 7–11).

Glucose transport across the membrane is facilitated by the Na⁺ gradient, generated by the Na⁺-K⁺ ATPase, and is eventually converted to fructose 1,6-bisphosphate by phosphofructokinase, the rate-limiting step in glycolysis; however, fructose has relatively few obstacles for entry into its metabolic pathway because its transport by the GLUT5 transporter does not require ATP (12, 13) and it bypasses the initial regulated steps of glycolysis, entering at a later step as fructose 1-phosphate after phosphorylation by fructokinase (2). One consequence of this is the activation of the energy sensor of the cells, AMP-activated protein kinase (AMPK) (14). AMPK senses the level of AMP in relation to the end product, ATP, the energy molecule of the cell. When the ratio of AMP to ATP increases, as it does after the phosphorylation of the incoming flood of fructose, AMPK binds AMP and is phosphorylated by liver kinase B1 (15). The active AMPK then signals the cell and, indeed, the whole body, that energy is needed and in doing so helps to regain normal ATP levels by regulating cellular and whole body processes including catabolic and anabolic processes, eating behavior (16), and even cell cycle regulation (for review see Ref. 14).

However, it is not as simple as consuming fructose, activating AMPK in the liver, and making glucose; a gross oversimplification of what is going on anyway. There is more to it than just the direct local effect. The current study by Kinote et al. (4) has demonstrated a direct gluconeogenesis circuit involving corticosterone from the adrenal gland signaled by the hypothalamus (17). The authors show that, whether delivered ip or directly to the brain by intracerebroventricular injection, fructose and not glucose can induce the phosphorylation of the energy sensor AMPK in neurons of the paraventricular nucleus (PVN) of the hypothalamus. They showed that this activation, specifically of the α2 isoform of AMPK, triggered hepatic gluconeogenesis as measured by the production of circulating glucose in response to a pyruvate tolerance test. By pharmacological manipulation with activators (5-amino-1-[beta]-D-ribofuranosyl-imidazole-4-carboxamide and A769662) and an inhibitor (compound C) of AMPK and by small interfering RNA technology, the authors demonstrate nicely that fructose works through hypothalamic AMPKα2 to produce de novo synthesized glucose in the liver. But how does this happen?

Phosphoenolpyruvate carboxykinase (PEPCK) is the key regulatory enzyme in gluconeogenesis and has glucocorticoid receptor-binding sites on its promoter (17, 18). Also, under hypoglycemic conditions, counterregulatory responses emanate from the glucose-sensing region of the hypothalamus, one response being an increase in circulating corticosterone (19–21). Indeed, the increase of glucagon, catecholamines, and corticosterone in the circulation caused by hypoglycemia as a counterregulatory response, was prevented by intracerebroventricular injected compound C, an inhibitor of AMPK, or by expression of a dominant-negative form of AMPK (19), demonstrating that activation of AMPK in this region of the hypothalamus causes corticosterone secretion from the adrenal gland. Although it was not demonstrated in the current work by Kinote et al. (4), it is known that the PVN releases CRH, which acts to stimulate the release of ACTH from the anterior pituitary, which in turn acts to stimulate the release of corticosterone from the adrenal cortex. It was therefore a join-the-dots effort (not to make light of the current study) to close the circuit and definitively demonstrate that AMPK activation in the PVN by fructose induces corticosterone secretion, leading to increased expression of the PEPCK gene by the glucocorticoid receptor in the liver and consequential increase in de novo synthesis of glucose. The authors of the study did this convincingly, showing in this animal model that fructose-induced gluconeogenesis was mediated by glucocorticoid receptor activity because the increase in both the levels of PEPCK in the liver and gluconeogenesis after fructose treatment was prevented by RU486, a powerful glucocorticoid receptor antagonist.

The question arises, however, with hepatic clearance of consumed fructose being extremely efficient, of whether circulating fructose levels reach an amount sufficient to trigger these effects in the brains of humans. The authors consider this possible because the GLUT5 transporter is found in the blood-brain barrier (22), excess fructose in circulation can be found in urine after a high-fructose meal (23), and ip administered fructose can be metabolized to lactate in the hypothalamus (24); hence, consumed fructose at high levels has the possibility to have an effect in the brain. Future investigations are needed to confirm this in humans.

It appears therefore, that drinks, sweetened with HFCS, not only taste sweet and supply a ready source of glucose and fructose for energy, but from the current work by Kinote et al. (4) described here and by others, that high levels of fructose can cross the blood-brain barrier and cause a centrally directed glucocorticoid stress response, erroneously signaled to be in a state of an energy deficit, to make more glucose de novo as a counterregulatory response. Rhetorically, one may question the wisdom of using HFCS as a sweetener in this respect, especially when considering that naturally occurring monosaccharides are not abundant in common food sources. Adding them to our diet in high amounts likely plays havoc with the physically and metabolically balanced levels of enzymes that have evolved to digest and metabolize the more common polysaccharides at rates and mechanisms that are more appropriately controlled, hopefully avoiding the three for the price of two energy sources obtained from consuming high levels of HFCS.