|
|
||||||||
1Medical Endocrinological Department, Odense University Hospital, Odense, Denmark; 2Medical Department, Hanusch Hospital, Vienna, Austria; 3Division of Endocrinology and Metabolism, General Hospital of Vienna, Vienna, Austria; 4Department of Medical Physiology, Panum Institute, University of Copenhagen, Copenhagen, Denmark; and Departments of 5Epidemiology and Biostatistics and 6Medicine, Case Western Reserve University, Cleveland, Ohio
Submitted 10 August 2006 ; accepted in final form 25 December 2006
| ABSTRACT |
|---|
|
|
|---|
glucose 6-phosphate
glucose. Healthy men (n = 7) were fasted for 44 h. At 40 h, hepatic glycogen stores were depleted. GNG then contributed
90% to a GP of
8 µmol·kg1·min1. Galactose, 9 g/h, was infused over the next 4 h. The contribution of GNG to GP declined from
90% to 65%, i.e., by
2 µmol·kg1·min1. The rate of galactose conversion to blood glucose, measured by labeling the infused galactose with [1-2H]galactose (n = 4), was also
2 µmol·kg1·min1. The 41st h GP rose by
1.5 µmol·kg1·min1 and then returned to
9 µmol·kg1·min1, while plasma glucose concentration increased from
4.5 to 5.3 mM, accompanied by a rise in plasma insulin concentration. Over 50% of the galactose infused was accounted for in blood glucose and hepatic glycogen formation. Thus an increase in the rate of GP via the glycogenolytic pathway resulted in a concomitant decrease in the rate of GP via GNG. While the compensatory response to the galactose administration was not complete, since GP increased, hepatic autoregulation is operative in healthy humans during prolonged fasting. liver; galactose; deuterium oxide
The biochemical reactions in glycogenolysis are glycogen
glucose 1-phosphate (G-1-P)
glucose 6-phosphate (G-6-P)
glucose (Fig. 1). Initial reactions in the hepatic utilization of galactose are galactose
galactose 1-phosphate
UDP-galactose
UDP-glucose. The UDP-glucose can be converted to glycogen, and hence to glucose via glycogenolysis, or be directly converted to G-1-P, and hence glucose without glycogen, as an intermediate (6, 22).
|
G-6-P
glucose (Fig. 1), that conversion served as a surrogate for increasing glucose formation by glycogenolysis. | MATERIALS AND METHODS |
|---|
|
|
|---|
Procedure.
In seven of the subjects, glucose production (GP) was measured using [6,6-2H2]glucose, and the contribution of GNG to GP using 2H2O (25). The subjects were fasted for 44 h, beginning after dinner on the first day of study. They had been on their regular diets and were allowed to drink noncaloric fluids ad libitum during the fast. At 5 PM on the second day, i.e., 21 h into the fast, they were admitted to the Diabetes Research Centre of the Odense University Hospital. Drinking of 2H2O was then begun. The dose, 5 ml/kg body water, was intended to achieve a body water enrichment of
0.5%. Body water weight was calculated at 60% of body weight (25). The dose was drunk in four equal portions spaced 2 h apart, so that the last portion was ingested at 11 PM. Neither dizziness, nausea, nor any other side effect occurred. Fluids ingested after 11 PM were enriched to 0.5% with 2H2O.
At 8 AM on the 3rd day, 36 h into the fast, a catheter was inserted into a superficial vein of one hand for blood collection. Another catheter was placed in a vein of the other hand, and through it a prime of 750 mg of glucose composed of 166 mg of [6,6-2H2]glucose, 98% 2H enriched (Isotec, Miamisburg, OH), and 584 mg of unlabeled glucose in 15 ml of water, sterile and negative for pyrogen, was infused rapidly. Then 500 mg of glucose composed of 100 mg of the [6,6-2H2]glucose and 400 mg of unlabeled glucose, also at a concentration of 5% in water, were infused hourly at a constant rate for the next 8 h. A 5% water solution of D-galactose was infused at a rate of 9 g/h for the last 4 h. The study ended after the 44th h of fasting. During those last 8 h, blood was collected at 0.5- to 1-h intervals for measurements of plasma glucose, insulin, C-peptide, and glucagon concentrations and 2H enrichments at carbons 2, 5, and 6 of the glucose. Plasma was frozen until analyzed.
To measure the fraction of the infused galactose converted to blood glucose, three other subjects and one of the seven subjects given 2H2O and [6,6-2H]glucose were treated in the same way, except 2H2O and [6,6-2H]glucose were not given. Instead, the D-galactose infused was 2H enriched to
3% at its carbon 1 by adding D-[1-2H]galactose, 98% 2H enriched (Omicron Biochemicals, South Bend, IN). Blood was collected at hourly intervals from the 41st through the 44th h, for measurement of the 2H enrichment at carbon 1 of the blood glucose.
Analyses. Plasma glucose concentrations were determined using a glucose oxidase method (Beckman Glucose Analyzer I, Fullerton, CA). Enrichments of the hydrogens bound to carbons 2, 5, and 6 of blood glucose were determined as previously detailed (3, 25, 30). Briefly, the supernatant, obtained after deproteinizing a blood sample by ZnSO4 and Ba(OH)2 addition, was deionized by passage through a column of AG1-X8 in the formate form over AG50 W-X8 in the hydrogen form (Bio-Rad, Hercules, CA). The column was washed with water, and the effluent evaporated to dryness. The residue was applied to a Bio-Rad HPX-87P column in an HPLC system with water at 80°C as solvent and a flow rate of 0.5 ml/min. Glucose eluted between 15 and 17 min, and galactose between 18 and 20 min. The quantity of galactose was one-twentieth or less than that of glucose. The fraction containing the glucose peak was collected.
To determine 2H enrichments at carbon 6, an aliquot of the glucose was oxidized with periodate. The formaldehyde formed, which contained carbon 6 with its two hydrogens, was condensed with ammonia to form hexamethylenetetramine (HMT). The HMT was assayed by gas chromatography-mass spectrometry for mass m + 2. To determine the enrichment at carbon 2, carbon 1 of glucose in another aliquot was removed to form ribulose 5-phosphate, which was reduced to a mixture of arabitol 5-phosphate and ribitol 5-phosphate. These were oxidized with periodate, yielding formaldehyde containing carbon 2 with its hydrogen. Glucose from another aliquot was oxidized to remove its carbon 6. The resulting xylose was oxidized with periodate, yielding carbon 5 with its hydrogen, again in formaldehyde. These formaldehydes were converted to HMTs, which were assayed for mass m + 1. HMTs formed from [6,6-2H2]glucose and [1-2H]sorbitol of known enrichments served as standards in the assays.
The fraction of galactose converted to glucose was determined by reducing an aliquot of the infused galactose to galactitol and aliquots of blood glucoses to sorbitol using sodium borohydride (4). These polyols, purified using the HPLC system, were also oxidized with periodate. The formaldehydes formed, which contain carbons 1 and 6 of the polyols with their hydrogens, were converted to HMTs, which were also assayed for m + 1. Again, HMTs made from [1-2H]sorbitol with known enrichments served as standards. 2H enrichment of the hydrogen bound to carbon 1 was calculated, taking into account that the formaldehyde formed from the CH2OH group containing carbon 6 was unlabeled.
Plasma insulin and C-peptide concentrations were measured by a two-site time-resolved immunofluorometric assay (Wallac Dy, Tuku, Finland). Plasma glucagon concentration was measured by radioimmunoassay (13).
Calculations. The percent contribution of GNG to GP (%GNG) was set equal to 100 times the enrichment of the hydrogen bound to carbon 5 of blood glucose, divided by the enrichment of the hydrogen bound to carbon 2 (25). The percent contribution via the glycogenolytic pathway (GLY) then equals 100 %GNG.
The rate of appearance (Ra) of glucose in blood in micromoles per kilogram per minute was calculated using the equation (38):
![]() |
The fraction of GP derived from galactose, when [1-2H]galactose was infused from the 41st through the 44th h of fasting, was set equal to the 2H enrichment in the hydrogen bound to carbon 1 of blood glucose collected at the end of the last hour of the infusion, divided by the enrichment in the [1-2H]galactose infused (35).
Statistics. All results are given as means ± SDs. Changes in rates of GP with time were examined using a randomized block design in which the subject was the block. Comparison of GP in specified time intervals was accomplished using t-tests with contrasts, controlling for the mean GP of the subject (20).
| RESULTS |
|---|
|
|
|---|
4.5 to
5.3 mM and then remained at that concentration. GP (Table 2) was
8 µmol·kg1·min1 before galactose infusion, increased by
1.5 µmol·kg1·min1 in the first 2 h of infusion, and then declined to
9 µmol·kg1·min1. The increase in GP in the first 2 h of infusion was statistically significant (t = 5.12, df = 36, P < 0.0001), and GP during the next 2 h remained higher than that before infusion (t = 2.86, df = 36, P < 0.01).
|
|
90% before galactose infusion, decreasing to
65% with infusion (Table 3). The test for linear trend from before infusion to the 44th h was statistically significant (t = 21.54, df = 36, P < 0.0001), and an F-test for departure from linearity was not significant.
|
7 µmol·kg1·min1 before galactose infusion, rose transiently in the first hour of galactose infusion, and then declined to
6 µmol·kg1·min1. The decline in absolute GNG in the first 2 h of galactose infusion was not statistically significant, but the decline in the next 2 h was statistically significant (t = 5.96, df = 36, P < 0.0001). The absolute contribution of GLY to GP was
1 µmol·kg1·min1 before infusion and rose to
3 µmol·kg1·min1 on galactose infusion. The test for linear trend from before infusion to the 44th h was statistically significant (t = 19.49, df = 36, P < 0.0001), and an F-test for departure from linearity was not significant.
|
|
|
| DISCUSSION |
|---|
|
|
|---|
90% to GP and glycogenolysis
10%. Glucose is formed by GNG from gluconeogenic substrates via G-6-P, by glycogenolysis, and from galactose via G-1-P conversion G-6-P. Thus when galactose was infused, three sources were available for the production of glucose, all via G-6-P. The increase in GP on beginning the galactose infusion indicates the sum of the quantities of glucose produced from those sources, then exceeded that produced from GNG and glycogenolysis before infusion. During that time, there was conversion of galactose to glucose without an equivalent decrease in contribution from one or both of the other sources. The finding that GP returning toward the preinfusion rate means that, after that time, a compensatory reduction occurred. However, GP was still
1 µmol·kg1·min1 more than before galactose infusion, so the compensatory response was incomplete. Other conditions have been reported in which a breakdown has occurred in hepatic autoregulation (2, 21). The increase in GP presumably was needed to satisfy increased glucose utilization consequent to the
0.8 mM rise in plasma glucose concentration with the accompanying rise in insulin concentration.
The contributions of GNG and GLY to GP are estimated from the ratio of enrichments of the hydrogen bound to carbons 5 and 2 of blood glucose. [2-2H]glucose is produced in the presence of 2H2O from galactose and glycogen via GLY in the isomerization of G-6-P with fructose 6-phosphate (F-6-P), i.e., G-1-P
G-6-P
F-6-P
[2-2H]G-6-P
[2-2H]glucose (25). [2,5-2H]glucose is formed by GNG. Carbon 5 is labeled in the hydration of phosphoenolpyruvate and the isomerization of glycerol 3-phosphate with glyceraldehyde 3-phosphate (GAP), and again at carbon 2 in the isomerization of F-6-P with G-6-P. The decline in the percent contribution of GNG to GP (Table 3), measured by the ratio of 2H bound to carbon 5 to carbon 2 of the glucose, then reflects the mixing of that [2-2H]glucose formed via GLY with [2,5-2H]glucose produced by GNG.
Since the contribution of GNG to GP decreased by 25% (Table 3) at a GP of
8 µmol·kg1·min1, a decline of 2 µmol·kg1·min1 might have been expected. The decline of only a little more than 1 µmol may be explained by the increase in GP of
1 µmol on galactose infusion. Glucose from galactose contributed
25% to GP, since the 2H enrichment in blood glucose was
25% of that in the [1-2H]galactose infused (Table 5). Since GP was
9 µmol·kg1·min1 during galactose infusion, galactose then contributed
2 µmol·kg1·min1. In accord with that, the contribution of GLY to GP rose
2 µmol·kg1·min1 above that before galactose infusion. Not only did GNG decline on galactose infusion, but the small contribution glycogenolysis made from stored glycogen to GP before infusion,
1 µmol·kg1·min1, could also have declined.
Estimates of GNG and GLY contributions can also be affected by glycogen cycling (15, 23, 34) and transaldolase exchange (23, 24). The contribution of glycogen cycling is likely minimal, while the contribution of transaldolase may be significant. Also, to the extent, if any, that [2,5-2H]G-6-P formed via the gluconeogenic pathway was converted to glycogen, GNG could be underestimated. Galactose conversion to glucose via G-1-P and G-6-P could have inhibited that conversion, perhaps resulting in its infusion in the apparent transient increase in GNG. The transaldolase catalyzed exchange between F-6-P formed from the galactose with GAP, i.e., F-6-P + [2-2H]GAP
[5-2H]F-6-P + GAP, followed by its isomerization to G-6-P and hence to [2,5-2H]glucose, could result in some of the galactose converted to glucose appearing to originate via GNG.
The contribution to GP of glucose formed from the galactose could have been more than
25%, since the 2H enrichment in blood glucose appears to be still rising (Table 4). Also [1-2H]G-6-P formed from the [1-2H]galactose, before its conversion to glucose, could have lost 2H in the pentose phosphate cycle, i.e., 3[1-2H]G-6-P + 3NADP
3CO2 + 2 G-6-P + GAP + 3NADP3H (26), and in its equilibration with mannose 6-phosphate, i.e., [1-2H]G-6-P
[1-2H]F-6-P
mannose 6-phosphate
F-6-P
G-6-P (4).
That amount of glucose released into the circulation to fulfill the body's needs results from the hydrolysis of G-6-P to glucose, catalyzed by glucose 6-phosphatase (21, 28, 36). The quantity each source of the G-6-P contributes to forming that amount depends only on the relative contribution of each source to the G-6-P. Contributions of the sources to the glucose produced will then be in the same proportion as their contributions to the formation of the G-6-P. As those contributions change, so must the fluxes through the pathways by which the G-6-P is generated change. Thus, in response to the increased formation of G-6-P from galactose via GLY, there must have been a compensatory reduction in the flux of gluconeogenic substrates to G-6-P.
The mechanism by which the so-called hepatic autoregulation is achieved is not well understood (8, 18, 21, 28). The increase in insulin concentration upon galactose infusion was probably due to the small rise in glucose concentration, since galactose is reported not to stimulate insulin release from islets (11). A decline in free fatty acid concentration could have contributed to the decreased GNG (2, 5, 21), but under the study conditions the decrease in free fatty acid concentration was likely small (19). Also, while insulin can inhibit GNG, as well as glycogenolysis, its increase seems unlikely to explain the decrease in GNG's contribution because of the relative insensitivity of the gluconeogenic pathway to insulin (1, 7). Conceivably, an intermediate in galactose's metabolism could have inhibited GNG, but no such inhibition has been reported. Galactose 1-phosphate was reported to inhibit phosphoglucomutase in vitro, but that was in the absence of glucose 1,6-bisphosphate, and no inhibition was demonstrated in vivo (32). Furthermore, inhibition of that enzyme would be expected to decrease glycogenolysis and not GNG. An initial period of hepatic glycogenolysis was reported on intravenous injection of a bolus of galactose into men fasted overnight, perhaps due to an inhibition of UDP-glucose pyrophosphorylase by UDP-galactose (9).
Recently, our laboratory reported a 2H-NMR procedure for measuring enrichments of 2H from 2H2O at carbons 5 and 2 of glucose (19). Measurements, made in the postabsorptive state and after long-term fasting, were compared with measurements made by the chemical procedure used in this study (3, 30). Three men were treated the same way as in this study, i.e., galactose infused for 4 h beginning after 40 h of fasting and GP estimated using [6,6-2H2]glucose, except the 2H2O was given 12 h after fasting was begun. GNG again declined from 94 ± 5% to 69 ± 6%, following galactose infusion. GP was about the same at 40 h, 9.2 ± 0.6 µmol·kg1·min1, as 44 h of fasting, 8.8 ± 0.2 µmol·kg1·min1. The rate of glycogen synthesis, measured using nuclear mass resonance spectrometry, was 3.1 ± 0.6 µmol·kg1·min1. Thus, of the
10 µmol·kg1·min1 of galactose infused, over 50% can then be accounted for in glucose and glycogen formation. Most of the galactose infused would be expected to be taken up by liver (6).
Sunehag and Haymond reported giving galactose to healthy women fasted overnight (35). Doses of
7.5 and 22.5 g/h were ingested over 2-h periods. At the end of that time, at both doses, GP was
12 µmol·kg1·min1. At the higher dose, galactose contributed
75% to GP. After an overnight fast, before galactose infusion, GNG and glycogenolysis would be expected to have each contributed
50% to GP (e.g., Refs. 1, 3, 5, 15). Therefore, the results at the higher dose provide further support for a compensatory decline in contribution of GNG when glucose is produced from galactose. In that study, similar to the present study, plasma glucose concentration was 4.7 mM before and 5.1 mM after galactose ingestion, and there was a small increase, about a doubling, in insulin concentration. When men, fasted overnight, ingested 50 g of galactose,
20% appeared in circulating glucose over the next 8 h, in accord with our estimate of 25%, although under other conditions. There was only a transient increase in plasma glucose and insulin concentrations upon the galactose ingestion (10).
Inhibitors of hepatic phosphorylase have been considered for possible use in the treatment of diabetics. To be effective, GP could have to decline without a compensatory increase in GNG. The report of Fosgerau et al. (8) suggests that that would be the case. The present study suggests that there could be a compensatory increase. That assumes the mechanism resulting in a decrease in GNG when glycogenolysis is increased in normal subjects, as simulated by galactose's conversion to glucose, operates in diabetic patients to increase GNG when glycogenolysis is decreased.
In conclusion, galactose has been used as a surrogate for glycogen as a source of hepatic GP, since both glycogen and galactose are converted to glucose by the same reaction steps. The contribution of GNG to GP by liver of healthy subjects, depleted of glycogen by fasting, declined on galactose administration within a few hours. An increase in GLY, resulting in a compensatory decrease in GNG, is in accord with the existence of a mechanism in liver by which GP is regulated for the body's needs as the substrate for GP change.
| GRANTS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. A. Veldhorst, M. S Westerterp-Plantenga, and K. R Westerterp Gluconeogenesis and energy expenditure after a high-protein, carbohydrate-free diet Am. J. Clinical Nutrition, September 1, 2009; 90(3): 519 - 526. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |