Simultaneous synthesis and breakdown of glycogen is called glycogen cycling. The extent of hyperglycemia and decreased glycogen stores in diabetes mellitus may relate in part to the extent cycling occurs. Four methods have been introduced to estimate its extent in liver in humans.1) In the fasted state, the rate of net hepatic glycogenolysis, i.e., glycogen breakdown minus synthesis, is estimated using NMR, and the rate of glycogenolysis is estimated from deuterium labeling of blood glucose on 2H2O ingestion.2) The rate of glycogen synthesis is estimated from the rate of labeling of carbon 1 of glycogen on [1-13C]glucose infusion, monitored by NMR, and the rate of breakdown from the rate of disappearance of that labeling on unlabeled glucose infusion.3) The rate of synthesis from glucose-1-P, formed by glycogenolysis, is measured by the decrease in the3H/14C ratio in acetaminophen glucuronide on acetaminophen and [2-3H,6-14C]galactose administration. 4) The rate of synthesis is estimated from the dilution of label from labeled galactose in its conversion to the acetaminophen glucuronide, and the rate of glycogenolysis is estimated from the amount of label in blood glucose. In the first method, the fate of glucose-6-P is assumed to be only to glycogen and glucose. In the second, only glucose-6-P molecules formed by breakdown that are not cycled back to glycogen are measured. In the third, 3H is assumed to be removed completely during cycling, and only the molecules cycled back to glycogen are measured. In the fourth, galactose conversion to glucose is assumed to be via glycogen. Quantitations in all four methods depend on assuming the order in which the molecules deposited in glycogen are released.
simultaneous synthesis and breakdown of glycogen has been called glycogen cycling. Recently, four methods have been introduced to quantitate hepatic glycogen cycling in humans. The first method to be examined combines rates of glycogenolysis and gluconeogenesis determined by use of2H2O with rates of net glycogenolysis determined by 13C NMR. Analysis of that method will serve as a basis for analyzing the other three methods and their applications. First, the definitions to be used of the pathways by which glucose and glycogen are produced and utilized are given.
Glucose-6-P is assumed to completely equilibrate with glucose-1-P (Fig. 1). The rate of gluconeogenesis is the rate of glucose synthesis via glucose-6-P from noncarbohydrate precursors, e.g., lactate, alanine, pyruvate, and glycerol (40). The rate of glucose-6-phosphoneogenesis is then the rate of synthesis of glucose-6-P from those precursors. The rate of glyconeogenesis is the rate of glycogen synthesis via glucose-6-P from those same precursors. The rate of glycogenolysis is the rate at which glucose is formed from glycogen. The rate of glycogen breakdown is the rate of conversion of glycogen to glucose-6-P 1. The rate of glycogenesis is the sum of the rates of glyconeogenesis and of glycogen formed from glycogen via glucose-6-P, i.e., glycogen → glucose-6-P → glycogen. Net glycogenolysis is the rate of glycogen breakdown minus the rate of glycogenesis. The rate of glucose production is the sum of the rates of glycogenolysis and gluconeogenesis. Of note, the rate of glucose-6-phosphoneogenesis is called the rate of gluconeogenesis by Shulman and associates (see Rothman et al., Ref. 37). Thus they use the definition of gluconeogenesis that was proposed by Krebs (22) and is generally accepted: the formation of carbohydrate, i.e., both glucose and glycogen, from noncarbohydrate precursors.
Combining 2H2O with 13C NMR.
Rates of gluconeogenesis and glycogenolysis in the fasted state are estimated from the enrichment of 2H in the hydrogens bound to carbons 5 and 2 of blood glucose on 2H2O ingestion (the 5/2 ratio) (26). The rate of net glycogenolysis is estimated from the decline in liver glycogen content measured by 13C NMR (37). Because net glycogenolysis is glycogen breakdown minus glycogenesis, by measuring glycogenolysis and net glycogenolysis, glycogenesis can be estimated.
Figure 2 depicts rates calculated from results of 2H2O-NMR measurements in normal subjects and in type 2 diabetics with fasting plasma glucose concentration ∼15 mM before and after their treatment with metformin (18). If we focus first on results before treatment, glucose production measured from dilution of a labeled glucose, i.e., [6,6-2H2]glucose, was 0.70 mmol · m body surface area−2 · min−1. The 5/2 ratio was 0.59. Therefore, 0.70 × 0.59 = 0.41 mmol · m−2 · min−1 was the rate of gluconeogenesis, and 0.70 (1 − 0.59) = 0.29 mmol · m−2 · min−1 was the rate of glycogenolysis. Net glycogenolysis, measured by 13C NMR, was 0.11 mmol · m−2 · min−1. Therefore, 0.70 − 0.11 = 0.59 mmol · m−2 · min−1 was the rate of synthesis of glucose-6-P from the gluconeogenic and glyconeogenic precursors, i.e., the rate of glucose-6-phosphoneogenesis. Let x equal the rate of glycogenesis. The fraction of the rate of glucose-6-phosphoneogenesis converted to glucose, i.e., the rate of gluconeogenesis, is then 0.70/(0.70 + x). Therefore, 0.59(0.70)/(0.70 +x) = 0.41 mmol · m−2 · min−1. Solving for x, the rate of glycogenesis was 0.30 mmol · m−2 · min−1. That is the rate of glycogen cycling2. Because net glycogenolysis was 0.11 mmol · m−2 · min−1, the rate of glycogen breakdown was 0.41 mmol · m−2 · min−1. Thus 0.41(30)/(30 + 70) = 0.12 mmol · m−2 · min−1 of glucose-6-P formed by glycogen breakdown was reconverted to glycogen, and 0.59(30)/(30 + 70) = 0.18 mmol · m−2 · min−1 of glucose-6-P formed by glucose-6-phosphoneogenesis was converted to glycogen, i.e., 0.12 + 0.18 = 0.30 mmol · m−2 · min−1.
Metformin decreased the rates of glucose-6-phosphoneogenesis, glucose production, and the extent of cycling. In normal subjects, net glycogenolysis was 0.18 and production 0.36, so that the rate of glucose-6-phosphoneogenesis was 0.18 mmol · m−2 · min−1. The 5/2 ratio was 0.55, giving a rate of gluconeogenesis of 0.36 × 0.55 = 0.20, not significantly different from the rate of glucose-6-phosphoneogenesis of 0.18 calculated from the NMR data. Thus, in the normal subjects, glycogenesis was not significant and gluconeogenesis contributed 0.18/0.36 = 50% to glucose production measured by NMR and 0.20/0.36 = 56% measured by2H2O.
The first conclusion from these results is that at least a major reason that the contribution of gluconeogenesis in the fasting state has been reported to be more by use of the NMR than by the2H2O method and other methods is the differing definition of gluconeogenesis and the occurrence of glycogenesis and, hence, glycogen cycling (18). Thus, in type 2 diabetics after an overnight fast, by NMR ∼85% of glucose production has been estimated to be from gluconeogenesis (29), and by the2H2O and other methods, ∼60% or less (6, 26, 43). Indeed, in the study (18) depicted in Fig. 2, by use of 2H2O, the contribution was 59% and with NMR, 0.59/0.70 = 84%. Gluconeogenesis was estimated to contribute more by the NMR than by the2H2O method in cirrhotic patients (87 ± 6 and 68 ± 3%) (33), but again not in normal subjects serving as controls (60 ± 10 and 54 ± 2%)3. Cycling of glycogen in the cirrhotics, and hence glycogenesis explaining that difference, is in accord with higher than normal concentrations of insulin and glucagon in their blood (33, 34).
Other reasons have been offered for why estimates in the diabetic subjects are higher by NMR, such as decline in glucose concentration during the measurements, incomplete meal absorption, and the presumed linearity in glycogen content decline (23). They may contribute, but the differing definition of gluconeogenesis was not recognized. Recently, because only changes in hepatic glycogen content are measured by NMR, glycogenolysis in the kidney was suggested to account for the higher estimates by NMR (6). Glycogen deposition in the kidney of uncontrolled diabetics was described over 100 years ago (20). Deposition was detected by electron microscopy in all types of renal cells from diabetic patients with high glucose concentrations (5). Deposition was observed histochemically in diabetic rats at blood glucose concentrations >19 mM, related to the extent of glucosuria and reversing with declines in blood glucose concentration (36). However, the content of glycogen in the kidney of rats infused with glucose to concentrations as high as 48 mM was ∼100-fold less per gram than in the liver (13). Glycogen content per gram was 5-fold less in kidney than in liver of diabetic rats at blood glucose concentrations ∼31 mM (21). Furthermore, glycogen accounted for <0.2% of the weight of the kidneys. Thus glycogen can be visualized in diabetic kidney, but its quantity appears too low to make a substantial contribution to glucose production. Murphy and Hellerstein (31) suggested incomplete visibility of glycogen to NMR as a potentially unifying explanation for what they concluded is the apparently systematic underestimations of hepatic glycogenolysis measured by NMR. However, the estimate by Shulman and colleagues (18, 29, 33, 37) is of net glycogenolysis. Thus the NMR method does not underestimate glycogenolysis or give a higher estimate of gluconeogenesis after an overnight fast in healthy normal subjects than other methods.
Barrett and Lui (2), from estimates of uptakes of gluconeogenic substrates by liver, concluded that gluconeogenesis could contribute about one-half of glucose production in postabsorptive type 2 diabetics. However, uptakes by the kidney should be taken into account, in particular that of glutamine, because production is from both liver and kidney. Although the contribution of kidney to production may be small in normal subjects after an overnight fast, it may be substantial in type 2 diabetics, as it is in normal subjects after long-term fasting (12, 30).
The second conclusion from these results is that metformin decreases the rate at which gluconeogenic and glycogenic substrates are converted to glucose-6-P. That may help reconcile the conclusion in some studies that metformin acts by decreasing gluconeogenesis (41) and in others that it acts by decreasing glycogenolysis (9, 10). Depending on the conditions of a study, decreased glucose-6-phosphoneogenesis could result in decreasing one more than the other.
Evidence has been provided to support the several assumptions made using the 5/2 ratio (7). Yet to be assessed is the possible contribution of the transaldolase exchange reaction (27). Thus, to the extent that fructose-6-P, formed from the glucose-6-P arising by glycogenolysis, experiences that exchange, glucose formed from it will have2H at carbon 5 as well as carbon 2: fructose-6-P + [2-2H]glyceraldehyde-3-P ↔ [5-2H]fructose-6-P + glyceraldehyde-3-P and [5-2H]fructose-6-P → [2,5-2H]glucose-6-P. Hence, glycogenolysis will be underestimated, and to an equivalent extent gluconeogenesis will be overestimated. [2-2H]glyceraldehyde-3-P will be regenerated from the glyceraldehyde-3-P, formed in the exchange, in its equilibration with dihydroxyacetone-3-P.
A major assumption in making the estimates is the order in which glucose-6-P molecules deposited in glycogen are reconverted to glucose-6-P. The estimates depicted in Fig. 2 are calculated with the assumption that the last deposited glucosyl units are not released. That is, only unlabeled glucosyl units are assumed released to form [2-2H]glucose-6-P (e.g., the 0.41 mmol · m−2 · min−1 in the untreated diabetic). Labeled glucose-6-P deposited in glycogen and then released is not included in the estimates, because that glycogenolysis will go undetected. Therefore, the estimates of cycling depicted in Fig. 2 are minimum estimates.
Figure 3 depicts rates set with the 5/2 ratio to be 0.59, net glycogenolysis to be 0.11, and glucose production to be 0.70 mmol · m−2 · min−1, as is the case for the untreated diabetics (Fig. 2) but with the assumption that 50% (Fig. 3 A) or 90% (Fig. 3 B) of the last deposited amount is released. The rates of cycling would be relatively large, requiring expenditure of significant amounts of energy. Also, to the extent the last deposited amount was released, gluconeogenesis by the 2H2O method would be overestimated. Thus, although with the last-deposited-not-released condition, 0.59 × 0.70 = 0.41 mmol · m−2 · min−1, i.e., 0.59[(0.70)/(0.30 + 0.70)] was the rate of gluconeogenesis; with 50% of that deposited being released of the 0.70 glucose produced, 0.59[(0.70)/(0.60 + 0.70)] = 0.31 μmol · m−2 · min−1 would be by definition gluconeogenesis, and 0.70 − 0.31 = 0.39 would be glycogenolysis. The data are not compatible with last-deposited-first-released being absolute, i.e., 100%; however, this will be discussed further. If the data were compatible, the 5/2 ratio found would have had to be 0.84, i.e., 0.59/0.70, the fraction gluconeogenesis contributes to glucose production by the NMR method.
With the assumption that some percentage of labeled glucosyl units deposited is released during cycling, further evidence for negligible cycling in fasted normal subjects can be found in the similar estimates of contributions of gluconeogenesis measured using2H2O (whether the 2H2O is ingested many or a few hours before blood is collected) (7,26). That is, if cycling were significant, the glucosyl units released in the multiple-hour period should have had more label at carbon 5, and hence estimates of gluconeogenesis would be higher.
Pulse-chase of glycogen monitored by 13C NMR.
In this method, [1-13C]glucose and then unlabeled glucose are continuously infused, and the intensity of the C1 resonance of the glucosyl units of glycogen is monitored (28). The rate of glycogen synthesis is estimated from the increase in that intensity during the [1-13C]glucose infusion and the rate of breakdown from its decline during the unlabeled glucose infusion.
When normal subjects were fasted for 5–10 h and maintained at a plasma glucose concentration of ∼9.5 mM, the rate of breakdown was estimated to be 57% of the rate of glycogen synthesis and 31% when the subjects were fasted for 12–14 h (28). Although the estimate of 31% has been referred to as an estimate in the fasted state, the determinations were then made when the subjects were hyperglycemic. Hepatic glycogen concentration in the subjects after the 5–10 h of fasting was estimated to be ∼400 mmol/l. It seems unlikely that those concentrations were significantly underestimated because of incomplete visualization by NMR (31), because higher concentrations would approach those found in glycogen storage disease.
Again, these estimates depend on an assumption of the extent to which the last deposited glucosyl units of glycogen are first released. If the extent of these units was absolute, only unlabeled glucosyl units from the unlabeled glucose infused would be released, and there would be no decline in the intensity of the C1 resonance. To the extent those unlabeled units were released, the rate of glycogen breakdown was therefore underestimated. Furthermore, it was underestimated to the extent that [1-13C]glucosyl units released from glycogen were reincorporated into glycogen. That is, what may be called true cycling, i.e., [1-13C]glycogen → [1-13C]glucose-1-P → [1-13C]glycogen, is not measured by the method.
The decline in the glycogen C1 resonance is also assumed, because of hydrolysis and glycolysis of glucose-6-P formed from the glucose-1-P (28). However, although phosphorylase catalyzes phosphorolysis of the 1,4-links in glycogen, hydrolysis of the 1,6-links releases free glucose. Indeed, although only about one-tenth of the glucosyl units in glycogen are 1,6-linked, theoretically, if the rate of glycogen → glucose-1-P → glycogen were rapid enough, the disappearance of the C1 resonance could have been due to the release of glucose by hydrolysis of the 1,6-links. Also, lysosomal acid α-glucosidase, when glycogen concentrations are high, may hydrolyze significant amounts of glycogen to glucose (17).
Petersen et al. (34) used the pulse-chase method to estimate rates of glycogen synthesis and breakdown in normal subjects, with glucose concentrations clamped at 5 and 10 mM and under hypoglucagonemic and hypo- and hyperinsulinemic conditions. Rates of glycogen synthesis and breakdown were also determined in normal subjects fasted overnight, who then had plasma glucose concentrations ∼5 mM. For those determinations, [1-13C]glucose was infused at a rate of only 0.2 mg · kg−1 · min−1. The rate of decline was the same in the C1 resonance and the C2 + C5 resonances of the glucosyl units of glycogen. If the [1-13C]glucose was being incorporated into glycogen, the decline of the C1 resonance should have been less than that of the C2 + C5 resonances. They calculated that a glycogen turnover of as little as 2% could be detected. That finding provides additional evidence that glycogenesis does not occur to a significant degree in normal subjects in the fasted state. Again, that conclusion rests on assuming the order of glucosyl unit release from glycogen. Thus, in the extreme, if [1-13C]glucosyl units were formed, but immediately released, declines of the C1 and C2 + C5 resonances in glycogen would have been the same, i.e., there would be no significant accumulation of C1 resonance in the glycogen.
[2-3H,6-14C]galactose is continuously infused, and acetaminophen is given (42). The3H/14C ratio in the excreted glucuronide compared with the ratio in the galactose is the measure of UDPglucose → glycogen → glucose-1-P→ UDPglucose. Thus the galactose is converted to UDPglucose, the immediate precursor of the glucosyl unit of glycogen and of the glucuronide moiety of the glucuronide, without loss or change in the position of the labels. With no glycogen cycling, the ratio in UDPglucose and hence in the glucuronide should then be the same as in the galactose. [2-3H,6-14C]glucose-1-P will be formed if there is glycogenolysis. Assuming extensive equilibrium between glucose-1-P and fructose-6-P, i.e., glucose-1-P ↔ glucose-6-P ↔ fructose-6-P, the 3H, but not the14C, is removed. The conversion of that [6-14C]glucose-1-P to UDPglucose will then lower the 3H/14C ratio in the glucuronide. Thus the method measures only what may be called true cycling. That is, it measures glucose-1-P that is reconverted to UDPglucose, and not glucose-1-P that is metabolized via glucose-6-P, as measured by the previous method.
Normal and type 2 diabetic subjects who fasted for 12 h were then infused with trace [2-3H,6-14C]galactose and unlabeled glucose, 4 mg · kg body wt−1 · min−1, for 5 h (42). Plasma glucose concentrations were ∼8 mM in the normal subjects and from 12 to 22 mM in the diabetic subjects. The3H/14C ratios in normal and diabetic subjects ranged from 0.88 to 0.98, giving estimates of cycling of 2–12%. A major assumption in making these estimates is that last deposited glucosyl units are first released. Thus, if [2-3H,6-14C]glucosyl units were deposited and unlabeled glucosyl units released, the ratio in the glucuronide would be the same as in the galactose, and cycling would be underestimated. Furthermore, in the pyrophosphorylase-catalyzed reaction (see[1-2H]galactose and glucuronide formation), i.e., UDPglucose + pyrophosphate (PP) ↔ glucose-1-P + UTP, 3H from UDP[2-3H,6-14C]glucose could be removed in the equilibration of glucose-1-P with fructose-6-P, without the conversion of the UDPglucose to glycogen. Cycling would then be overestimated.
[1-2H]galactose and glucuronide formation.
A trace amount of [1-2H]galactose, along with acetaminophen, is infused (16). The turnover of UDPglucose, equated with the rate of glycogen synthesis, is calculated from the enrichment of the [1-2H]galactose, the amount infused, and the enrichment in the excreted glucuronide, assumed to reflect the enrichment in hepatic UDPglucose. The rate of release of [1-2H]glucose into the systemic circulation is equated with the rate of glycogenolysis.
The method depends on the assumption that galactose has to be converted to glycogen before its conversion to glucose. That requires that the conversion of glucose-1-P to UDPglucose in the reaction glucose-1-P + UTP → UDPglucose + PP, the pyrophosphorylase-catalyzed reaction, be irreversible. Otherwise, labeled UDPglucose formed from galactose would be converted to glucose-1-P and, hence, further metabolized via glucose-6-P, without glycogen as an intermediate. Also, glucose-1-P formed by glycogenolysis would then equilibrate with the labeled UDPglucose, decreasing its enrichment without it being converted to glycogen. Evidence that the pyrophosphorylase reaction is reversible has been presented (14, 32). Also, the method requires that the labeled galactose not be converted to a significant extent to glucose via the Leloir pathway (39), because in that pathway, glycogen is not an intermediate, i.e. The extent that labeled galactose is converted to glycogen rather than glucose will then depend on the conditions, e.g., the UDPglucose will be converted more to glycogen than metabolized via glucose-1-P in the fed than in the fasted state. Furthermore, the rate of glycogenolysis will be underestimated by the method to the extent that labeled glucosyl units are deposited and unlabeled units are released. Also, 2H of [1-2H]galactose can be removed in the conversion of [1-2H]glucosyl units of glycogen to glucose (8), again resulting in an underestimation of glycogenolysis. Other investigators have applied this method by using carbon-labeled galactose rather than [1-2H]galactose (3, 38).
In response to the above concerns (24), Hellerstein (15) noted that incorporation of label from labeled PP into glucose-1-P has not been reported, an observation that he concluded would support the reversibility of the pyrophosphorylase reaction. However, labeling of glucose-1-Pcannot occur, because neither glucose-1-P nor PP is cleaved in the reaction. He has also calculated the role for the Leloir pathway to be negligible, because the enrichment of UDPglucose, measured in the glucuronide, is much less than that of the labeled galactose infused. But that calculation requires that the enrichment of hepatic galactose-1-P be that of the galactose infused. That would be the case if the conversion of galactose-1-P to UDPgal were irreversible. The Leloir pathway is reversible (39). Hellerstein et al. (16) estimated by using this method that there is extensive glycogen cycling in the fasted state. They also reported that, when labeled galactose was infused into subjects for the last 10 h of a prolonged fast, and then glucagon was infused, labeled glucose was released into the circulation. However, in addition to the galactose, a significant quantity in sum of glycerol, glucose, propylene glycol, and ethanol was infused (23).
During fasting, simultaneous glycogenesis and glycogenolysis occur in type 2 diabetics and liver cirrhotics. Evidence for that cycling includes measurements made of those rates by combining estimates of net glycogenolysis by the 13C NMR method with estimates of the contribution of gluconeogenesis to glucose production by the2H2O method. During fasting in normal subjects, glycogenesis is not significant. In normal subjects given glucose loads, glycogenesis and glycogen breakdown simultaneously occur.
The significance of glycogen cycling in normal individuals in the fed state is uncertain. In the conversion of a molecule of glucose-1-P→ glycogen →glucose-1-P, the equivalent of one high-energy bond is expended, but the complete oxidation of one molecule of glucose will generate 36–38 high-energy bonds. Only high rates of cycling would then impact significantly on energy expenditure. Also, that cycling was demonstrated at relatively high glycogen concentrations. Glycogen may regulate its concentration, such that, at high concentrations, glycogenolysis may be a physiological response not occurring at lower concentrations. Also, hydrolysis of glycogen to glucose, catalyzed by lysosomal acid α-glucosidase, may be a means of protecting the liver from high concentrations of glycogen.
The minimum rate of cycling in the diabetic subject after an overnight fast is estimated to be 40% of the rate of glucose production (Fig.2). Such a rate, if it occurred in the diabetics in the fed state, as suggested by the pulse-chase studies (34), could explain the lower than normal hepatic glycogen stores found in the diabetics (29). The hyperglycemia could then be attributable, at least in part, to decreased hepatic glycogen deposition. Therapy to inhibit cycling, e.g., the use of a phosphorylase inhibitor (4), would then be expected to decrease the hyperglycemia and increase the glycogen stores. As suggested by the results of administering metformin, inhibition of glucose-6-phosphoneogenesis could result not only in decreased glucose production but also a decrease in the rate of glyconeogenesis and, hence, cycling.
It should be emphasized that the rates of cycling depicted in Figs. 2and 3 are calculated by assuming that the differences between estimates of gluconeogenesis defined using the NMR and2H2O methods are due to glycogen cycling. Support for that is obtained from the significant difference also found in cirrhotics, but not normal subjects, as well as the other evidence that cycling does not occur in normal subjects. However, other explanations for at least some of those differences cannot be dismissed.
The quantitations of cycling by all four methods rely on the validity of the assumptions made. A major assumption in all four methods is the order in which glucosyl units in glycogen are removed. Experimental evidence supporting glucosyl units last incorporated being first removed rests on studies on the breakdown of labeled glycogen. Thus radioactive galactose has been infused to add labeled glucosyl units to unlabeled glycogen in livers of rats and dogs. When the glycogen was isolated from rat liver and degraded, the radioactive units incorporated last were liberated first (11). When glucagon was infused into dogs, the glucose first appearing in the circulation was also labeled (25).
However, there is evidence that last-deposited-first- removed can be far from absolute. Otherwise, after deposition of [1-13C]glucosyl units, the infusion of unlabeled glucose with continued synthesis of glycogen could not be accompanied by a large disappearance of the [1-13C]glucosyl unit of glycogen (28). Differential labeling of the glucosyl units at sites on the glycogen granules, and then their breakdown, remain a possible explanation for last-deposited not being first removed4. Synthesis and breakdown in different areas of the liver lobule could also be a possible explanation (1, 19). Thus the absolute order of release of the glucosyl units of glycogen in relationship to the time of their deposition remains uncertain. Hence, rates of cycling measured by any of the methods are uncertain.
This review was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-14507.
Address for reprint requests and other correspondence: B. R. Landau, Dept. of Medicine, Case Western Reserve Univ. School of Medicine, 10900 Euclid Ave., Cleveland, OH 44106–4951 (E-mail:).
↵1 Glycogenolysis in liver is defined as the formation of glucose from glycogen. However, the breakdown of glycogen to glucose-6-P in tissue where the glucose-6-P is not converted to glucose is also called glycogenolysis. Therefore, that rate of conversion of glycogen to glucose-6-P in liver is here called the rate of glycogen breakdown.
↵2 The rate of glycogen cycling in the fasted state, because glycogen breakdown exceeds synthesis, is therefore the rate of synthesis, i.e., the rate of glycogenesis. In the fed state, since synthesis exceeds breakdown, the rate of cycling is the rate of breakdown.
↵3 By the NMR method, Petersen et al. (35) estimated a contribution of gluconeogenesis of 55 ± 6% in normal subjects fasted for 12 h, Rothman et al. (37) 64 ± 5% in subjects fasted for 22 h, and Magnusson et al. (29) 70 ± 6% in subjects fasted for 23 h.
↵4 Enzymes catalyzing the synthesis and degradation of glycogen are bound to glycogen, although not in stoichiometric ratios (40). Thus synthesis could be occurring in portions of glycogen molecules (labeled units deposited), and in other portions, degradation could be occurring simultaneously (unlabeled units released).
- Copyright © 2001 the American Physiological Society