The human body has two main priorities for energy liberation during fasting.
The first priority is to maintain a stable supply of energy for CNS function. The brain has little stored energy in the form of glycogen or TAGs and therefore depends on the liver (and under some circumstances the kidney) for a constant supply of energy in the form of glucose or ketone bodies (see p. 1185). In the fed state and early in the fasting state, the brain derives essentially all its energy from oxidation of glucose because ketone bodies are not present and the blood-brain barrier is mostly impermeable to FAs. Because a continuous supply of glucose is required to meet the energy demands of the CNS, humans have evolved elaborate, redundant mechanisms to maintain plasma [glucose] within a very narrow range, between 60 and 140 mg/dL (3.3 to 7.8 mM), between fasting and fed states. Lower [glucose] impairs brain function, whereas elevated plasma [glucose] exceeds the renal glucose threshold, results in polyuria, and leads to the multiple complications associated with poorly controlled diabetes mellitus (retinopathy, neuropathy, and nephropathy).
Most other major organs of the body (liver, skeletal muscle, heart, kidney) fill their energy needs at this time by oxidizing FAs. In contrast to the well-regulated plasma [glucose], the concentrations of FAs and ketone bodies can vary by 10- and 100-fold, respectively, depending on the fed and fasted conditions. During prolonged fasting (>2 days), the liver metabolizes FAs to raise plasma levels of ketone bodies (see p. 1185) sufficiently to supply much of the brain's oxidative fuel needs and diminish the need for gluconeogenic substrate supply by proteolysis.
The second priority for the body is to maintain its protein reserves (i.e., contractile proteins, enzymes, nervous tissue, etc.) in times of fasting.
The body also has two main priorities for energy repletion after fasting. First, following a meal, liver and muscle replenish their limited glycogen reserves. Once these stores are full, liver and muscle convert any excess energy in the form of carbohydrate and protein to fat. Muscle glycogen is the most readily available form of energy for muscle contraction, especially when intense bursts of physical activity are required. Therefore, maintaining an adequate supply of muscle glycogen at all times has obvious survival benefits in times of fight or flight. The second priority during feeding is to replenish protein reserves.
During an overnight fast, glycogenolysis and gluconeogenesis maintain plasma glucose levels
The period after an overnight fast serves as a useful reference point because it represents the period before the transition from the fasted to the fed state. At this time, the concentrations of insulin, glucagon, and metabolic substrates that were altered by meal ingestion during the preceding day have returned to their basal levels. Moreover, the body is in a relative steady state in which the rate of release of endogenous fuels from storage depots closely matches fuel consumption.
Requirement for Glucose
After an overnight fast, the decline in circulating insulin leads to a marked decrease in glucose uptake by insulin-sensitive tissues (e.g., muscle) and a shift toward the use by these tissues of FAs mobilized from fat stores. Nevertheless, the average adult continues to metabolize glucose at a rate of 7 to 10 g/hr. Total-body stores of free glucose, which exists mostly in the extracellular space, amount to only 15 to 20 g or ~2 hours' worth of glucose fuel. However, the useful glucose store is even less if we consider that the plasma [glucose]—normally ~90 mg/dL (5.0 mM) after an overnight fast—may not fall to less than ~55 mg/dL (3.0 mM) before brain function becomes abnormal. Thus, maintaining plasma [glucose] in the presence of this ongoing glucose use, particularly by the brain, requires that the body produce glucose at rates sufficient to match its ongoing consumption.
Gluconeogenesis versus Glycogenolysis
Four to 5 hours after a meal (perhaps longer for a very large meal), a fall in plasma [insulin] (see pp. 1035–1050) and a rise in [glucagon] (see pp. 1050–1053) cause the liver to begin breaking down its stores of glycogen and releasing it as glucose. Moreover, both the liver and, to a lesser extent, the kidney generate glucose by gluconeogenesis. The release of glucose by these two organs is possible because they are the only two with significant amounts of pyruvate carboxylase (see p. 1176) and PEPCK (see p. 1176), as well as G6Pase (see p. 1182), which catalyzes the conversion of G6P to glucose. Net hepatic glycogenolysis and gluconeogenesis each contribute ~50% of whole-body glucose production during the first several hours of a fast.
Gluconeogenesis: The Cori Cycle
In the first several hours of a fast, the brain consumes glucose at the rate of 4 to 5 g/hr, which is two thirds the rate of hepatic glucose production (~180 g/day). Obligate anaerobic tissues also metabolize glucose but convert it primarily to lactate and pyruvate. The liver takes up these products and uses gluconeogenesis to regenerate glucose at the expense of energy. The liver releases the glucose for uptake by the glucose-requiring tissues, thus completing the Cori cycle (Fig. 58-13).
FIGURE 58-13 Metabolism during an overnight fast. αAA, α-amino acid; αKA, α-keto acid; αKG, α-ketoglutarate; AQP9, aquaporin 9; ECF, extracellular fluid; RBC, red blood cell.
Gluconeogenesis: The Glucose-Alanine Cycle
After an overnight fast, the body as a whole is in negative nitrogen balance, with muscle and splanchnic tissues being the principal sites of protein degradation and release of amino acids into the blood. Alanine and glutamine, which are particularly important, represent ~50% of total amino acid released by muscle, even though these amino acids represent only 10% to 13% of total amino acids in muscle protein. The reason that alanine and glutamine are overrepresented is that muscle synthesizes them (see Fig. 58-13). During fasting, breakdown of muscle protein yields amino acids, which subsequently transfer their amino groups to α-ketoglutarate (supplied by the citric acid cycle) to form glutamate. Glutamine synthase can then add a second amino group to glutamate, thereby producing glutamine. Alternatively, alanine aminotransferase can transfer the amino group of glutamate to pyruvate (the product of glucose breakdown), generating alanine and α-ketoglutarate. Both glutamine and alanine enter the blood. The intestine uses some of the glutamine as an oxidative fuel and releases the amino groups into portal blood as either alanine or ammonia.
The amino acids taken up by the liver provide carbon for gluconeogenesis. On a molar basis, alanine is the principal amino acid taken up by the liver. In the first several hours of fasting, the liver principally uses alanine for gluconeogenesis (see p. 1176). Because the carbon backbone of alanine came from glucose metabolism in muscle, and the liver regenerates glucose from this alanine, the net effect is a glucose-alanine cycle between muscle and liver, analogous to the Cori cycle.
In addition to playing a role in gluconeogenesis, the glucose-alanine cycle is critical for nitrogen metabolism and thus provides a nontoxic alternative to ammonia for transferring amino groups—derived from muscle amino-acid catabolism—to the liver (see Fig. 58-13). The hepatocytes now detoxify the amino groups on alanine and other amino acids by generating urea (see Fig. 46-14), which the kidney then excretes (see Fig. 36-1). Another key amino acid in nitrogen metabolism is glutamine, which muscle releases into the blood for uptake by the gut and liver as well as the kidney. The kidney uses the carbon skeleton of glutamine for renal gluconeogenesis and converts the amino group to ammonia, which it excretes (see pp. 829–831). This ammonia excretion is particularly important in maintaining body acid-base balance during fasting. Combined, alanine and glutamine account for >40% of the amino-acid carbon used by liver and kidneys in gluconeogenesis.
Neither the Cori cycle nor the glucose-alanine cycle in muscle yields new carbon skeletons. Rather, both cycles transfer energy—and the glucose-alanine cycle also transfers nitrogen—between muscle and liver. The energy for hepatic glucose synthesis comes from oxidation of fat in the liver.
Finally, the fall in plasma [insulin] after an overnight fast permits the release of FAs and glycerol from fat stores (see Fig. 51-10). This response appears to be more pronounced in visceral than peripheral fat depots. The decline in [insulin] and the ensuing lipolysis are sufficient to supply FAs to extracerebral tissues (e.g., muscle, heart, liver) for fuel and glycerol to the liver for gluconeogenesis. However, these changes are not sufficient to stimulate the hepatic conversion of FA to ketone bodies (see pp. 1185–1187).
The body never completely suppresses gluconeogenesis. When an individual ingests a meal, gluconeogenic flux provides glucose for hepatic glycogen stores (indirect pathway; see p. 1179). During fasting, the liver redirects the gluconeogenic flux to provide glucose for delivery to the circulation.
Starvation beyond an overnight fast enhances gluconeogenesis and lipolysis
We have just seen that, during an overnight fast, glycogenolysis and gluconeogenesis contribute about equally to maintain a fasting plasma [glucose] of ~90 mg/dL (5.0 mM). What happens if we extend our fast for 1 or 2 days? Because the glucose utilization rate is 7 to 10 g/hr, if half of this were provided by glycogenolysis (as is true for an overnight fast; see p. 1189), the hepatic glycogen stores of ~70 g that remain after an overnight fast would be sufficient to last only an additional day. However, in the early stages of starvation, the body compensates by accelerating gluconeogenesis.
Orchestrating the metabolic adaptations in the early stages of starvation—increased gluconeogenesis, but also increased proteolysis and lipolysis—are a decline in [insulin] to a level lower than that seen after an overnight fast and a modest increase in portal vein [glucagon]. Insulin deficiency promotes all aspects of the metabolic response, whereas the effect of glucagon is mostly confined to the liver (see pp. 1050–1053).
Adaptations in both liver and muscle are responsible for increasing gluconeogenesis (see Fig. 58-13). In muscle, acceleration of proteolysis leads to the release of alanine and other glycogenic amino acids, whereas the liver accelerates its conversion of gluconeogenic amino acids into glucose. This enhanced gluconeogenesis, however, is not the result of increased availability of substrates, because plasma levels of alanine and other glycogenic amino acids decline. Instead, fasting increases transport of alanine into the liver and upregulates key gluconeogenic enzymes (see p. 1176), which makes gluconeogenesis more efficient.
The dependence of gluconeogenesis on proteolysis is reflected by an increase in urinary nitrogen excretion in the early phase of starvation. During the first 24 hours of a fast, the average 70-kg person excretes 7 to 12 g of elemental nitrogen in the urine, equivalent to 50 to 75 g of protein. Because tissue protein content does not exceed 20% by weight for any tissue, 50 to 75 g of protein translates to 250 to 375 g of lean body mass lost on the first day of a fast.
The activation of HSL and ATGL (see p. 1182) increases release of FAs and glycerol from TAG stores in adipose tissue and muscle (see Fig. 58-13). The increased availability of glycerol provides the liver with an additional substrate for gluconeogenesis (see p. 1176) that contributes to glucose homeostasis. Moreover, the increased availability of FAs to muscle and other peripheral tissues limits their use of glucose, which preserves glucose for the CNS and other obligate glucose-utilizing tissues and diminishes the demands for gluconeogenesis and proteolysis.
Elevated levels of FAs cause insulin resistance in skeletal muscle by directly interfering with insulin activation of GLUT4 (see Fig. 58-13). Intracellular lipid intermediates (e.g., DAGs, ceramides) activate a serine/threonine kinase cascade that involves protein kinase C θ, which leads to increased serine phosphorylation of insulin-receptor substrate 1 (IRS-1; see p. 1042). This serine phosphorylation, in turn, leads to decreased tyrosine phosphorylation of IRS-1, and thus to a decrease in PI3K activity, which diminishes GLUT4 translocation to the plasma membrane of muscle. This FA-induced decrease in insulin-stimulated glucose uptake by muscle, and the parallel increased availability of FAs as a fuel for muscle, spare glucose for other tissues (e.g., brain, renal medulla, erythrocytes) under fasting conditions. However, this adaptation may play an important pathological role in mediating the insulin resistance associated with obesity and type 2 diabetes.
FAs not only have effects on muscle, but also enter the liver, where they undergo β-oxidation (see pp. 1183–1185) and generate energy. A fall in the insulin-glucagon ratio inhibits ACC (see pp. 1178–1179 and Fig. 58-7), thereby reducing levels of malonyl CoA and promoting mitochondrial FA oxidation. Thus, the hormonal changes both increase the supply of FAs and activate the enzymes necessary for FA oxidation. This β-oxidation furnishes the energy and reducing power required for gluconeogenesis. If the availability of FAs outstrips the ability of the citric acid cycle to oxidize the resulting acetyl CoA, the result may be the accumulation of ketone bodies (see p. 1185), which can serve as a fuel for the CNS as well as for cardiac and skeletal muscle.
Prolonged starvation moderates proteolysis but accelerates lipolysis, thereby releasing ketone bodies
As the duration of fasting increases, the body shifts from using its limited protein stores for gluconeogenesis to using its relatively large energy depots in fat for ketogenesis (Fig. 58-14). Moreover, the brain shifts from oxidizing glucose to oxidizing two ketone bodies (see p. 1185), β-hydroxybutyrate and acetoacetate, to meet most of its energy requirements.
FIGURE 58-14 Metabolism during prolonged starvation. aa, amino acids; AAc, acetoacetate; AQP9, aquaporin 9; ECF, extracellular fluid; βHB, β-hydroxybutyrate; RBC, red blood cell. N58-2
A fasting human could survive for only ~10 days if totally dependent on protein utilization to meet whole-body energy requirements. Thus, prolonged survival during starvation requires a major reduction in proteolysis. Indeed, urea excretion decreases from 10 to 15 g/day during the initial days of a fast to <1 g/day after 6 weeks of fasting. Because urea is the major obligatory osmolyte in the urine (see pp. 811–813), this reduced urea production lessens obligatory water excretion and therefore the daily water requirement. Ammonium excretion also decreases.
Decreased Hepatic Gluconeogenesis
The transition from protein to lipid degradation permits humans to extend their survival time during a prolonged fast from weeks to months, as long as fat stores are available and water intake is adequate. During this transition, hepatic gluconeogenesis decreases (see Fig. 58-14), mostly because of diminished substrate delivery. During the first few weeks of a fast, muscle releases less alanine, the principal substrate for hepatic gluconeogenesis, so that plasma [alanine] falls markedly, to less than one third of the concentrations seen after the absorption of a meal. Indeed, during a prolonged fast, infusing a small amount of alanine causes plasma [glucose] to rise.
Increased Renal Gluconeogenesis
Whereas hepatic gluconeogenesis falls, renal gluconeogenesis rises (see Fig. 58-14) to reach as much as 40% of whole-body glucose production. Renal gluconeogenesis, which consumes H+ (see Fig. 39-5A), most likely is an adaptation to the acidosis that accompanies ketogenesis (see p. 1185). Indeed, acidosis stimulates renal ammoniagenesis in parallel with renal gluconeogenesis.
Increased Lipolysis and Ketogenesis
During the first 3 to 7 days of fasting, hypoinsulinemia accelerates the mobilization of FAs from adipose tissue. As a result, plasma FA levels double and remain stable thereafter. The combination of low insulin and high glucagon levels also increases hepatic oxidation of FAs, leading to a marked increase of hepatic ketogenesis (see Fig. 58-14 and pp. 1051–1053) or ketogenic capacity. The liver achieves peak rates of ketone body production (~100 g/day) by the third day and maintains them thereafter. Low insulin levels also progressively reduce the extraction of ketone bodies by peripheral tissues. Thus, despite relatively stable rates of ketone body production, circulating levels of ketone bodies continue to rise throughout the next few weeks. As a result, the CNS receives an increasing supply of these water-soluble substrates, which eventually account for more than one half of the brain's energy requirements. In this way, ketone bodies ultimately supplant the brain's dependency on glucose. Thus, by limiting the brain's gluconeogenic demands, the body preserves protein stores. Besides the CNS, other body tissues, especially the heart and skeletal muscle, can use ketone bodies to cover a significant proportion of their energy demands.
As the fast progresses and fat stores are depleted, levels of leptin (see pp. 1001–1002) decrease. This decrease in leptin levels is a protective signal that profoundly affects the hypothalamic-pituitary-gonadal axis, reducing the oscillations of luteinizing hormone and follicle-stimulating hormone and causing anovulation. In times of famine, this mechanism protects fertile women from the additional nutritional demands associated with pregnancy.
In summary, the body has evolved powerful adaptive mechanisms that ensure adequate substrate supply in the form of glucose and ketone bodies during a prolonged fast to maintain adequate CNS function. Even during a prolonged fast, humans do not lose consciousness because of decreased substrate supply to the brain. Instead, death under these conditions typically occurs when fat stores are depleted and severe protein wasting causes failure of respiratory muscles, which in turn leads to atelectasis and terminal pneumonia.