| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Divisions of General Internal Medicine (M.-J.P., J.R.J., J.A.L., P.S., C.J.T.) and Endocrinology (M.-J.M.P.), Department of Medicine, and Department of Chemical Endocrinology (P.N.S.), University Medical Center, 6500 HB Nijmegen, The Netherlands
Address all correspondence and requests for reprints to: Marie-Jose Pouwels, M.D., Division of Endocrinology, University Medical Center, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: m.pouwels{at}endo.azn.nl
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
GlcN infusion raised the plasma GlcN concentration in the infusion arms to 0.42 ± 0.14 and 0.81 ± 0.46 mmol/L; plasma GlcN remained very low (<0.07 mmol/L) in the control arms and in the placebo group. GlcN infusion did not change forearm blood flow. During insulin, FGU increased more than 10-fold. At all time points, FGU was similar in the GlcN-infused arm compared with the control arm and was not different from FGU in the placebo experiments. Similar results were obtained for forearm arteriovenous glucose differences or extraction and for whole body glucose uptake. Thus, despite relevant GlcN concentrations for 5 h in the infused forearm, GlcN had no effect on insulin-induced glucose uptake. These results do not support involvement of the hexosamine pathway in the regulation of insulin sensitivity in humans, at least not in the short-term setting.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Animal studies suggest that overactivity of the hexosamine pathway represents an important mechanism by which hyperglycemia causes insulin resistance (2, 3). Impaired glucose transport in adipocytes, as observed after prolonged exposure to high levels of glucose, insulin, and glutamine can be explained by increased glucose flux into the hexosamine pathway (4, 5, 6). Glutamine fructose-6-phosphate amidotransferase (GFAT) is the key enzyme of this pathway, using the amide group of glutamine to convert fructose-6-phosphate to glucosamine-6-phosphate (5, 6). Glucosamine (GlcN) is transported into the cell by glucose transporters and phosphorylated to glucosamine-6-phosphate by hexokinase, thereby entering the hexosamine pathway and bypassing GFAT (5). Incubation with GlcN reduces basal and insulin-stimulated glucose uptake in both adipocytes (6) and rat skeletal muscles (7). Additional in vivo studies in animals support the hypothesis that recruitment of this pathway is a major mechanism by which hyperglycemia (8, 9, 10, 11) and GlcN (8, 9, 10, 11, 12, 13, 14, 15, 16) cause insulin resistance. The observation that GlcN infusion did not cause a further impairment in glucose uptake in diabetic rats suggests that the effects of chronic hyperglycemia and GlcN infusions on insulin resistance are not additional and may operate through the same pathway (8).
To date, studies in humans concerning the involvement of the hexosamine pathway in the effects of hyperglycemia have been limited and controversial. In one study GFAT activity in human muscle cells from diabetic patients was significantly and positively correlated with glucose disposal rate measured by clamp experiments (17), but in another study in muscle cells from diabetic patients a negative correlation existed (18). If activation of the hexosamine pathway is involved in hyperglycemia-induced insulin resistance in humans, GlcN infusion should evoke insulin resistance in vivo. To test this hypothesis, we assessed the effect of intrabrachial GlcN infusion on insulin-stimulated forearm skeletal muscle glucose uptake (FGU) and whole body glucose uptake.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Twenty healthy volunteers (10 men and 10 women; mean ± SD age, 24 ± 4 yr; body mass index, 22.3 ± 1.9 kg/m2) gave informed consent before participating in the study, which had been approved by the hospital ethics committee. They were recruited by advertisement and received a payment. All participants were nonsmoking, used no medication (oral contraceptives excepted), and had a negative family history for diabetes mellitus. All subjects were normotensive (blood pressure, 114 ± 13/72 ± 9 mm Hg) and normoglycemic.
Procedures
The experiments were performed with the subjects in the supine position after an overnight fast in a quiet, temperature-controlled room (24-25 C). Under local anesthesia (<0.4 mL 2% xylocaine), a 20-gauge, 2-in. catheter (Angiocath, Becton Dickinson and Co., Mountain View, CA) was inserted into the brachial artery of the nondominant arm and connected with an arterial pressure monitoring line (Viggo Spectramed, 5269–129) to a monitor (78353B, Hewlett-Packard Co., Palo Alto, CA). In the same arm and on the contralateral side, a catheter (Venflon; 20-gauge, 32 mm; Becton Dickinson and Co., Helsingborg, Sweden) was inserted retrogradely into a deep forearm vein to obtain venous blood samples. Hereafter, we will refer to the arm with the intraarterial line as the infusion arm and to the contralateral arm as the control arm. In the foot another venous catheter was inserted for infusion of insulin and glucose. Subjects were clamped at fasting arterial glucose levels. Euglycemia was maintained by a variable infusion of 20% glucose solution, adjusted by arterial glucose measurements at 5-min intervals (19). Forearm blood flow (FBF) was measured in both arms using mercury-in-SILASTIC (Dow Corning Corp., Midland, MI) strain-gauge venous occlusion plethysmography as previously described (20, 21). Arterial and venous blood was sampled with inflated wrist cuffs at both sides simultaneously at relevant time points.
Protocols
In all three different protocols, we used the hyperinsulinemic,
euglycemic clamp technique during the entire experiment. For a
schematic representation see Fig. 1
. In
protocol 1, saline was infused intraarterially (placebo; n = 6).
In protocol 2, 4 µmol/dL·min GlcN were infused in the brachial
artery from 90–240 min (GlcN150; n = 6) and in protocol 3, 4
µmol/dL·min GlcN were infused from 0–300 min (GlcN300; n =
6). In one subject in the GlcN300 protocol, the GlcN infusion was
discontinued after 190 min.
|
Insulin (Actrapid, Novo-Nordisk, Copenhagen, Denmark) was infused at a dose of 60 mU/m2·min (430 pmol/m2·min) during the whole experiment. Insulin was diluted in 50 mL 0.9% NaCl with addition of 2 mL albumin to a concentration of 1 U/mL. Glucosamine sulfate (Rotta Research Group, Monza, Italy) or placebo (0.9% saline) was administered (per dL forearm tissue, measured by the water displacement method). In two pilot experiments, GlcN was infused in a dose of 1 µmol/dL forearm·min. Assuming a mean basal FBF of 2 mL/dL·min, it was calculated that the plasma GlcN concentration in the efferent venous blood of the infused forearm would be raised to approximately 0.5 mmol/L. However, the mean plasma GlcN concentration in the infused arms was only approximately 0.15 mmol/L. This discrepancy may be explained by rapid diffusion into the body compartments, because GlcN is a small molecule (22), by ready entry into the cell through the glucose transporters (5), or by a sampling error. Significant binding of GlcN to red blood cells was excluded by in vitro studies. For further experiments we used 4 µmol/dL·min.
Laboratory assays
Plasma glucose was measured using the glucose oxidation method (Glucose Analyzer II, Beckman Coulter, Inc.). Plasma insulin was determined using a double antibody in-house RIA (interassay coefficient of variation, 6%). Serum GlcN concentrations were measured as described previously (18), in which tissue preparation and reverse phase high performance liquid chromatography were as described for the quantification of glucosamine-6-phosphate in the assay for GFAT. Some modifications were applied. GlcN eluted with a retention time of 18.8 min. The correlation between GlcN additions to blank serum and the peak was 0.99. The mean GlcN recovery in six samples was 111%, with a coefficient of variation of 7.3%.
Calculations and statistical analysis
Whole body glucose uptake was calculated from glucose
infusion rate during steady state and expressed as micromoles per
kg/min. FBF (expressed as milliliters per dL/min) was measured every 20
min and averaged to mean hourly values. In four subjects, slightly
increased blood flow data related to urinary urgency were excluded. The
ratio FBFinfused/FBFcontrol
arm was calculated for each measurement to correct for potential
systemic changes related to arousal or to systemic effects of drugs.
Forearm glucose uptake (FGU) was calculated as:
glucosearterial-venous x (1 - 0.3
x hematocrit) x FBF, assuming that the whole blood glucose
= (1 - 0.3 x hematocrit) x plasma blood glucose
(23). Statistical analyses of differences were performed
by Student’s t tests (two-sided) or by repeated measures
ANOVA as appropriate, using the SPSS statistical package (SPSS, Inc., Chicago, IL). P < 0.05 was considered
statistically significant. Results are given as the mean ±
SEM.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Baseline characteristics were similar among the three study groups. Arterial plasma glucose levels during the clamp procedures were stable in all subjects and similar in the placebo, GlcN150, and GlcN300 group (mean plasma glucose levels, 5.04 ± 0.08, 5.05 ± 0.08, and 4.96 ± 0.03 mmol/L, respectively). Coefficients of variation were 5.4 ± 2.7% (placebo), 5.4 ± 2.6% (GlcN150), and 4.2 ± 1.2% (GlcN300). The plasma insulin concentration increased in all subjects throughout the insulin infusion and remained stable during the test. Baseline and steady state (210 min) plasma insulin concentrations during the clamp were similar in the three groups (placebo, 56 ± 19 to 678 ± 72; GlcN150, 70 ± 27 to 710 ± 90; GlcN300, 48 ± 14 to 651 ± 91 pmol/L; P = NS).
GlcN concentrations
During the GlcN infusion (4 µmol/dL·min), venous GlcN concentrations increased in the infusion arm to 0.42 ± 0.14 mmol/L in GlcN150 and to 0.81 ± 0.46 mmol/L in GlcN300 (mean ± SD; n = 6 in each group). The GlcN concentration in the control arms remained undetectable (<0.07 mmol/L); no systemic effects of GlcN were observed during the study. The GlcN concentration in the infusion arm was strongly, inversely, correlated with FBF at that side (r = -0.53). After the infusion was stopped, the plasma GlcN concentration dropped rapidly; 30 min after discontinuation, GlcN was undetectable in all arms.
Effects of GlcN infusion on total body insulin sensitivity
Forearm GlcN infusion did not affect total body insulin
sensitivity. During the clamp, whole body glucose infusion rates were
similar in both GlcN infusion groups vs. the placebo group.
For instance, glucose infusion rates during the last 90 min of the
clamp were 57 ± 4 (placebo), 50 ± 7 (GlcN150), and 50
± 4 µmol/kg·min (GlcN300; P = NS; see Fig. 2).
|
Baseline FBF values were similar among the three groups and were
not different between the infusion and control sides. In the placebo
protocol, FBF tended to increase in both arms over time, but these
differences did not attain statistical significance [FBF from
1.94 ± 0.27 to 2.35 ± 0.92 mL/dL·min (infusion arm) and
from 2.39 ± 0.41 to 2.98 ± 1.02 mL/dL·min at 5 h
(control arm); P = NS; Fig 3, top left panel]. GlcN
infusion had no effect on FBF. During the GlcN infusion, FBF at the
infusion arm was similar to FBF at the control arm (Fig. 3, top
panel). Results were similar when data were expressed as forearm
vascular resistance or as the FBF infusion arm/FBF control arm
ratio.
|
Baseline arteriovenous (glucoseAV)
differences were low and were not different between groups. During
insulin infusion, the glucoseAV difference
increased approximately 5- to 10-fold. GlcN had no effect on
glucoseAV; during GlcN infusion,
glucoseAV in the infusion arm at the various
time points was similar to glucoseAV in the
control arm and was not different from values observed in the placebo
group [for instance, GlcN150 at 4 h, 0.88 ± 0.28 mmol/L
(infusion arm) vs. 1.17 ± 0.21 mmol/L (control arm);
GlcN300 at 5 h, 1.23 ± 0.23 mmol/L (infusion arm)
vs. 1.09 ± 0.25 mmol/L (control arm);
P = NS; Fig. 3
, middle panels].
Effects of GlcN infusion on FGU
Baseline FGU was equal in all three groups. Insulin infusion
increased FGU more than 10-fold in both arms [placebo protocol,
0.23 ± 0.18 at baseline to 4.34 ± 1.60 µmol/dL·min
after 5 h (infusion arm) and from 0.34 ± 0.09 to 5.73
± 1.86 µmol/dL·min (control arm); P < 0.01; Fig. 3, lower left panel]. GlcN had no effect on FGU. During
GlcN infusion, FGU in the infusion arm at the various time points was
similar to FGU in the control arm and did not differ from FGU in the
placebo group [for instance, FGU at 4 h in GlcN150, 2.88 ±
0.83 (infusion arm) vs. 3.35 ± 1.66 µmol/dL·min
(control arm); P = NS; FGU at 5 h in GlcN300,
2.77 ± 0.70 (infusion arm) vs. 2.08 ± 0.33
µmol/dL·min-1 (control
arm); P = NS; Fig. 3, lower panels].
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Studies in adipocytes have suggested that GFAT is under rapid transcriptional control, and half-maximal inhibition of GFAT activity was observed within 4 h (6). Preexposure to GlcN inhibited subsequent glucose uptake in rat skeletal muscles significantly after 2 h (7), and in GlcN treated animals, the rate of glucose infusion decreased over the course of a 4-h clamp (8). During 7-h clamp experiments the maximal reduction in glucose uptake was achieved within 5 h, and the calculated time required to achieve 50% of the maximal effect of GlcN was 174 ± 26 min (8). Several other studies reported a significant GlcN-induced resistance to insulin-stimulated glucose uptake after 150 min (14) or even earlier (12, 13, 15). Based on these reports, we conclude that the duration of GlcN infusion in our study should be more than sufficient to show any effect on glucose uptake.
In animals, a GlcN dose as low as 0.1 mg/kg·min ( = 0.5 µmol/kg·min) caused a reduction of glucose uptake by 50% (12). In one study, the lowest GlcN dose decreasing glucose uptake was 3.3 mg/kg·min (16.7 µmol/kg·min), although the decrease in glucose uptake with 6.5 mg/kg·min (36.3 µmol/kg·min) was more effective (35% vs. 13% reduction) (11). Several other studies have applied a GlcN dose of 30 µmol/kg·min (8, 11, 12, 13, 14, 15, 16). In the current study, GlcN was administered in a dose of 4 µmol/dL forearm·min, equaling a forearm tissue dose of 40 µmol/kg forearm·min. To our knowledge this is considerably more than the applied dosages in animal studies.
In vitro, GFAT activity was already inhibited at a GlcN
concentration of 0.21 mmol/L (6). In several animal
studies adverse effects of GlcN on glucose homeostasis were observed at
GlcN concentrations from 0.15–2.0 mmol/L (8, 10, 14). In
our study, the plasma GlcN concentration in the effluent venous blood
of the forearm was approximately 0.5 mmol/L in the GlcN150 experiments
and 0.8 mmol/L in the GlcN300 experiments. This difference is partly
caused by variations in blood flow and by differences in GlcN
extraction among subjects; the GlcN concentration in the infusion arm
was inversely correlated with FBF at that side. Similar to the reported
insulin-induced 5- to 10-fold increase in FGU and arteriovenous glucose
difference, one may expect insulin-induced forearm GlcN uptake and
arteriovenous GlcN difference. Finally, the measured venous GlcN blood
levels are almost certainly an underestimation of the real GlcN blood
levels obtained at the arterial capillary bed, as some admixture of
venous blood not drained from the forearm will hamper blood sampling
despite our intention to obtain optimal blood samples. Therefore,
according to findings in the literature, GlcN concentrations in the
infused arm, obtained in this study, should be more than sufficient to
affect insulin sensitivity. Also in subjects who reached higher GlcN
concentrations (
1.8 mmol/L), no effect of GlcN on glucose uptake was
detected. We conclude that the GlcN concentration in this study cannot
explain the lack of effect.
Application of the double forearm technique in combination with the hyperinsulinemic clamp (24, 25) allowed us to study the acute effects of a high plasma GlcN concentration on skeletal muscle glucose metabolism in humans while avoiding systemic adverse effects. As GlcN levels were undetectable in the control arm, we were able to compare the effect of GlcN on FGU and FBF in the infused arm (high GlcN level) to that in the control arm (negligible GlcN level) for the total duration of the clamp. Unlike previous studies (24, 25) applying the reliable double forearm technique, we inserted the cannula for insulin and glucose infusion in a foot vein, further excluding interference between measurement of the forearm glucose concentration and glucose infusion as well as avoiding direct, local, vascular effects of insulin (26). We used an insulin dose within the high physiological range, approximating the ED50 for stimulation of peripheral glucose uptake, to detect changes in sensitivity. In rats both a physiological insulin dose and a maximally stimulating insulin dose have been used, resulting in clear changes in insulin sensitivity induced by GlcN (11). Therefore, we believe that the lack of effect of GlcN on glucose uptake and FBF cannot be explained by the investigation method used.
A few reports suggest an effect of GlcN on blood flow (14, 27). In this study we did not find any effect of GlcN on the FBF in the infused arm vs. the control arm or vs. placebo.
Although this study was performed in 20 volunteers, and the possibility of a type 2 error cannot be disregarded, there was not a trend toward an inhibiting effect of GlcN on glucose uptake; if anything, findings would suggest the opposite trend.
As in this study the GlcN infusion was of sufficient duration and caused effective concentrations and as the applied investigation methods were appropriate, we believe that our results do not support a role for the hexosamine pathway in the regulation of insulin sensitivity in humans. Also, Monauni et al. reported no effects of systemic GlcN infusion on insulin secretion and action (28). Systemic GlcN levels reported in their study compare well to those measured in our infusion arm; the total systemic GlcN infusion (1.6 and 5 µmol/kg·min) was a factor 2.5–8 times higher than our total local forearm infusion. Because the forearm represents approximately 1% of total body weight, the local forearm concentration in our study may have been higher than the systemic concentration in their study (28). Despite these slight differences in design, our conclusions are in full agreement with and extend the observations of Monauni et al. Together, these findings caution for the extrapolation of findings in rodents to human physiology and metabolism.
In summary, in this study we did not find any effect of GlcN infusion for up to 300 min on insulin-stimulated glucose uptake at the whole body level or at the forearm level. These results indicate that the hexosamine pathway does not play an important role in humans, at least not over the short term. Whether longer administration of GlcN or a higher dose would have an effect remains to be determined. The reasons for discordance between the animal and human data are not clear. Further human studies concerning the activity of the hexosamine pathway are warranted.
|
Footnotes |
|---|
Received July 24, 2000.
Revised November 8, 2000.
Revised January 23, 2001.
Accepted January 30, 2001.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B A Biggee, C M Blinn, M Nuite, J E Silbert, and T E McAlindon Effects of oral glucosamine sulphate on serum glucose and insulin during an oral glucose tolerance test of subjects with osteoarthritis Ann Rheum Dis, February 1, 2007; 66(2): 260 - 262. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Muniyappa, R. J. Karne, G. Hall, S. K. Crandon, J. A. Bronstein, M. R. Ver, G. L. Hortin, and M. J. Quon Oral Glucosamine for 6 Weeks at Standard Doses Does Not Cause or Worsen Insulin Resistance or Endothelial Dysfunction in Lean or Obese Subjects Diabetes, November 1, 2006; 55(11): 3142 - 3150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L Stumpf and S.-W. Lin Effect of Glucosamine on Glucose Control Ann. Pharmacother., April 1, 2006; 40(4): 694 - 698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Buse Hexosamines, insulin resistance, and the complications of diabetes: current status Am J Physiol Endocrinol Metab, January 1, 2006; 290(1): E1 - E8. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Scroggie, A. Albright, and M. D. Harris The Effect of Glucosamine-Chondroitin Supplementation on Glycosylated Hemoglobin Levels in Patients With Type 2 Diabetes Mellitus: A Placebo-Controlled, Double-blinded, Randomized Clinical Trial Arch Intern Med, July 14, 2003; 163(13): 1587 - 1590. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Yu, S. M. Boies, and J. M. Olefsky The Effect of Oral Glucosamine Sulfate on Insulin Sensitivity in Human Subjects Diabetes Care, June 1, 2003; 26(6): 1941 - 1942. [Full Text] [PDF]
|
|
|
|
 |
|
 M.-J. J. Pouwels, P. N. Span, C. J. Tack, A. J. Olthaar, C. G. J. Sweep, B. G. van Engelen, J. G. de Jong, J. A. Lutterman, and A. R. Hermus Muscle Uridine Diphosphate-Hexosamines Do Not Decrease Despite Correction of Hyperglycemia-Induced Insulin Resistance in Type 2 Diabetes J. Clin. Endocrinol. Metab., November 1, 2002; 87(11): 5179 - 5184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Pavelka, J. Gatterova, M. Olejarova, S. Machacek, G. Giacovelli, and L. C. Rovati Glucosamine Sulfate Use and Delay of Progression of Knee Osteoarthritis: A 3-Year, Randomized, Placebo-Controlled, Double-blind Study Arch Intern Med, October 14, 2002; 162(18): 2113 - 2123. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, GlcN150 (n = 6); 
, GlcN300
(n = 6).