| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Medicine M (Endocrinology and Diabetes) (T.Ø., K.B.D., O.S.), University Hospital of Aarhus, and Department of Clinical Pharmacology (O.S.), University of Aarhus, DK-8000 Aarhus, Denmark; Novo Nordisk A/S (M.G., R.D.C., M.K.T.), 2880 Bagsvaerd, Denmark; and Division of Endocrinology (J.D.V., R.A.R.), Mayo Clinic and Foundation, Rochester, Minnesota 55905
Address all correspondence and requests for reprints to: Torben Østergård, M.D., Department of Medicine M (Endocrinology and Diabetes), University Hospital of Aarhus, AKH Nørrebrogade 4244, DK-8000 Aarhus C, Denmark. E-mail: oest{at}dadlnet.dk.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
In the ß-cell, SUs cause insulin secretion by inhibiting ATP-sensitive potassium (KATP) channels. Closure of the channels leads to a membrane depolarization, which in turn results in calcium influx and, eventually, stimulation of calcium-related insulin exocytosis (3). However, KATP channels are expressed extensively in endocrine cells, including the pituitary cells (10). In vitro experiments in the latter cells have demonstrated that exposure to the SU compound glipizide leads to release of GH in a manner analogous to the mechanism by which SUs stimulate insulin secretion in the pancreatic ß-cells (11). Clearly, if these findings are applicable to humans, they may well have important clinical implications. Even minimal elevations of GH promote insulin resistance in humans (12, 13), and there is substantial evidence that GH is involved in the pathogenesis of the microvascular complications of diabetes (14, 15). Finally, it cannot be excluded that chronic hypersomatotrophism may exhaust the ß-cells (16).
The metiglinide compound, repaglinide, is one of the novel non-SU insulin secretagogues that also causes insulin secretion by interacting with the KATP channels. In contrast to SUs, repaglinide has been shown to have direct actions only on the SU receptor (SUR) and to have distinct binding sites to the receptor (17), whereas SUs also interfere with the secretory machinery at a level distal to the SUR by directly stimulating exocytosis (18). Furthermore, in contrast to the SUs, repaglinide has recently been shown not to stimulate GH secretion in rat pituitary cells (19).
Thus, based on the previous in vitro experiments (10, 11, 17, 19), we explored whether SUs and repaglinide induced GH secretion in humans by either 1) augmenting the stimulating effect of GHRH or 2) antagonizing the suppressive effect of somatostatin on GH secretion. This was done by measuring GH concentrations after administration of a SU compound (glibenclamide), repaglinide, or placebo, after a GHRH bolus and during somatostatin infusion. Furthermore, because diabetic subjects, in addition to insulin deficiency, are also characterized by glucagon excess (20) and SUR are present in
-cells (21), we also sought to examine the effect of the two compounds (glibenclamide and repaglinide) on circulating glucagon levels. C-peptide-negative type 1 diabetic patients were chosen to preclude secretion of endogenous insulin during the experiments.
| Subjects and Methods |
|---|
|
|
|---|
Ten subjects (six men and four women) diagnosed with type 1 diabetes, aged 38 ± 11 yr, with BMI of 23.4 ± 2.3 kg/m2 and diabetes duration of 19.5 ± 9.8 yr, (mean ± SD), and without any significant endogenous insulin production [mean fasting serum C-peptide, 81 pmol/liter (range, 17183 pmol/liter)] were included. The amount of glycosylated hemoglobin in blood at entrance was 7.7 ± 0.9% (mean ± SD). None were known to have proliferative retinopathy, overt nephropathy or neuropathy, atherosclerotic vascular diseases, or clinical relevant abnormalities in GH secretion. Other medications apart from insulin (simvastatin, n = 1; perindopril, n = 1; citalopram, n = 1) were withdrawn 24 h before the experiments. The subjects were recruited from the outpatient clinic at the Medical Department M, University Hospital of Aarhus, Denmark. All gave written consent to participate in the study, which was approved by the local ethical committee of the county of Aarhus, Denmark. The study complies with the guidelines proposed in The Declaration of Helsinki and the guidelines for Good Clinical Practice.
Experimental protocol
To evaluate the impact of insulin secretagogues on GH secretion the subjects were examined in a single-blinded, placebo-controlled crossover study. On a screening visit, a medical history and physical examination were performed, as well as standard laboratory measurements to confirm good general health apart from diabetes. Each eligible subject was studied on three occasions separated from each other by 610 wk. On the day before the study day, subjects administered only short-acting insulin (last dose administered at
1800 h).
Subjects were admitted to the clinical research unit at 2200 h on the evening before each study day, and an iv catheter was inserted into an arterialized vein for blood sampling and another catheter into the contralateral antecubital vein for infusions. To obtain and maintain near-normoglycemia throughout the night, and to avoid hypoglycemia, an iv insulin infusion (insulin Actrapid, Novo Nordisk A/S; 50 U of insulin Actrapid in 1 liter of isotonic sodium chloride) was started after admittance and adjusted regularly according to measurements of plasma glucose every 3060 min. The experimental protocol is outlined in Fig. 1
. At 0800 h (0 min), a single dose of either 10.5 mg glibenclamide (Daonil, Aventis Pharma A/S, Hørsholm, Denmark), 8 mg repaglinide (Novonorm, Novo Nordisk A/S), or placebo in randomized order was administered per os together with 200 ml tap water. Doses were chosen to be approximately a half-maximal dose of the recommended daily dose in a clinical setting (recommended maximal daily dose of glibenclamide vs. repaglinide is 20 mg vs. 16 mg) (22). Coincident with administration of the study medicine, the variable insulin infusion was shifted to a fixed insulin infusion at a rate of 0.25 mU·kg1·min1 and continued to the end of the experiments. Blood glucose concentration was monitored every 510 min and exogenous glucose (200 g/liter) infused when appropriate to ensure that plasma glucose was always above 4.5 mmol/liter throughout the study. An iv bolus of GHRH (Somatrel, 0.1 µg/kg; Ferring Pharmaceuticals, Copenhagen, Denmark) was given at time 60 min, and an infusion of somatostatin (7 ng·kg1·min1; Ferring) was commenced at 240 min and continued until 480 min. A second iv bolus of GHRH (0.1 µg/kg) was given at time 300 min. The insulin infusion was discontinued at 480 min, and the usual sc insulin regimen of the patient was resumed. A standard meal was then served, and the patient was discharged after completion of the meal.
|
Analytical methods
Plasma glucose was measured in duplicate immediately after sampling (Beckman Instruments, Palo Alto, CA). Serum GH was determined by a double monoclonal immunofluorometric assay (DELFIA, Wallac Inc., Turku, Finland). Serum C-peptide and insulin were measured with an immunoassay (Dako Diagnostics Ltd., Cambridgeshire, UK). Plasma glucagon was measured by an in-house RIA, and serum NEFA was determined by a colorimetric method using a commercial kit (Wako Pure Chemical Industries, Neuss, Germany). Glycosylated hemoglobin was determined by high-pressure liquid chromatography (normal range, 4.86.4%).
Data analysis
Comparisons were made between the three study days of maximal concentration of serum GH (Cmax) after GHRH administration (0150 min and 300390 min) and integrated GH concentrations [areas under the curve (AUCs)] during the following time windows according to the experimental protocol: after GHRH bolus administration (60150 min), in the post-GHRH-effect window (180240 min), during somatostatin infusion (240300 min), after GHRH bolus during somatostatin infusion (300390 min), and finally in the post-GHRH-effect window during somatostatin (420480 min) (Fig. 1
).
Plasma glucagon concentrations were compared by mean concentrations and incremental AUCs in the intervals t = 0240 (without somatostatin) and t = >240480 min (with somatostatin), respectively.
Deconvolution analysis. Serum GH time series were analyzed by deconvolution analysis to quantify secretory burst mass (23). Deconvolution was performed with a previously validated iterative multiparameter technique given the following assumptions: 1) the hormone is secreted in a finite number of bursts superimposed on a basal time-invariant secretory rate with 2) an individual amplitude, 3) a common half-duration, and 4) a biexponential disappearance rate (24).
Statistical analysis
Data are given as means ± SEM. Comparisons of data were done by two-way ANOVA, followed by Bonferroni post hoc analysis where applicable. AUC was calculated using the trapezoidal method. GH Cmax, AUCs, and secretion rates were not parametrically distributed, thus a logarithmical transformation was done to obtain normality before further statistical processing. Glucagon AUCs were compared by the Friedman test for several related nonparametrically distributed samples, and post hoc analysis done by Wilcoxons pair-wise comparisons. All statistical analyses were performed using SPSS for Windows version 11.0 (SPSS, Chicago, IL), and P < 0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
During the constant insulin infusion rate, plasma glucose was kept within a narrow range near euglycemia: 5.8 ± 0.03 vs. 6.0 ± 0.04 vs. 5.8 ± 0.02 mmol/liter (glibenclamide, repaglinide, and placebo days, respectively); P = not significant (NS). Average circulating insulin concentrations (122 ± 3 vs. 120 ± 4 vs. 120 ± 3 pmol/liter; P = NS), and NEFAs (0.31 ± 0.02 vs. 0.31 ± 0.02 vs. 0.33 ± 0.02 mmol/liter; P = NS) were comparable in the three study conditions (Fig. 2
).
|
Cmax post GHRH, and AUCs of serum GH in the intervals as described above, were similar after glibenclamide, repaglinide, and placebo, except in the interval 420480 min, in which the AUC of GH after repaglinide was significantly lower than the AUC of placebo (P < 0.05) (Table 1
and Fig. 3
). There were no differences between AUCs of GH after glibenclamide and repaglinide or after glibenclamide and placebo.
|
|
Glucagon
Baseline plasma glucagon (mean of glucagon levels at t = -15 and 0 min) were similar on all three study days (Table 2
). After both glibenclamide and repaglinide, average plasma glucagon concentrations were higher than after placebo in the period before infusion of somatostatin (0240 min) (Table 2
). Likewise, incremental AUCs of plasma glucagon were increased after glibenclamide and repaglinide, although not statistically significant after repaglinide (Table 2
and Fig. 4
). During infusion of somatostatin (>240480 min), plasma glucagon concentrations were suppressed equally in all three conditions as shown by incremental AUCs (Table 2
).
|
|
| Discussion |
|---|
|
|
|---|
In humans, GH secretion is influenced by a variety of metabolic stimuli, e.g. NEFAs, insulin, and glucose. NEFAs inhibits GH release (25, 26), and there is evidence that hyperinsulinemia per se reduces circulating GH concentrations (27, 28). Finally, in nondiabetic subjects, hyperglycemia exerts a profound suppressive effect on GHRH-mediated GH secretion (29, 30, 31, 32). In this study, insulin levels were comparable because type 1 diabetic subjects, without any insulin production of their own, were studied. Insulin levels were determined exclusively by an individually calculated infusion rate (by weight). NEFAs were also comparable and therefore minimizing the probability of any confounding effect on GH secretion. Fluctuations in glucose concentrations during the experiments could potentially disturb the interpretation of GH levels in the three study days. However, the effect of hyperglycemia on GH secretion in type 1 diabetic subjects differs substantially from what is seen in a nondiabetic population. Even severe hyperglycemia has only a weak suppressive effect on GH concentrations in type 1 diabetic subjects (31, 33). Thus, subtle variations in glucose concentrations within subjects in this study are most likely to be of only insignificant importance when analyzing GH secretion patterns.
As shown, in this study overall GH release was not affected by either glibenclamide or repaglinide. Only in the last 60 min of the protocol, intended to capture the final effects of somatostatin on GH release, was a lower GH response observed after repaglinide compared with placebo. Thus, these data suggest that repaglinide may be capable of enforcing the inhibitory effect of somatostatin on GH secretion, although the mechanisms behind this are unknown.
In contrast to the present study, in vitro experiments have shown that SUs augment GH release in pituitary somatotrophs (10, 11, 34), whereas this is not the case with repaglinide (19). Both SUs and repaglinide cause closure of KATP channels on cell membranes, inducing depolarization and ultimately leading to exocytosis (17, 19). In contrast to repaglinide, SUs can cause exocytosis in voltage-clamped cells, indicating that SUs interact with the secretory machinery at a more distal level as well (17). The mechanism has been found to be dependent on protein kinase C and does not involve intracellular increases in Ca2+ concentration (18). This mechanism is not shared by repaglinide, which instead has been found to act specifically on the SUR causing closure of KATP channels (17, 19). This feature of repaglinide has been suggested to account for the discrepancy in the in vitro observations that repaglinide, in contrast to SUs, does not stimulate GH release (19).
The discrepancy between the in vitro (rodents) (19) and in vivo (present study) observations may relate to several mechanisms. In contrast to in vitro experiments, human studies will explore the summarized effect of numerous interrelated pathways, and only the net result of these will be measurable by simple techniques as used in the present study. For example, agents that mediate closure of KATP channels will probably interact with somatostatin-producing cells, which have also been shown to contain KATP channels (35) and to respond to stimulation with SUs (36, 37, 38). The effect is stimulation of somatostatin exocytosis, as has been shown in pancreatic
-cells (39). In the pituitary gland, this scenario would most likely cause a decrease in GH release, and the net effect might thus be an unaltered GH secretion. Finally, one cannot exclude species differences.
After administration of either glibenclamide or repaglinide, glucagon release was enhanced in these type 1 diabetic subjects. The interpretation of this observation should be done with a few reservations. As the average diabetes duration in our study subjects was quite long, the glucagon response to the two compounds of course may have been altered. In addition, the GH and glucagon levels theoretically may have been affected by previous hypoglycemia although none were reported by the patients. The observation of enhanced glucagon release is in contrast to what is seen in subjects with preserved insulin secretory capacity, in whom insulin secretagogues will cause insulin secretion, which in turn, through intraislet control, leads to a diminished glucagon release from the pancreatic
-cells (40, 41). Accordingly, in subjects with preserved insulin secretory capacity, it has been reported that during treatment with SUs, the counterregulatory response to hypoglycemia by glucagon is diminished (42, 43). The present finding of increased glucagon levels in the absence of insulin secretory capacity supports previous reports of SUs being direct stimulants of
-cells both in vitro (36, 44) and in vivo (45). The mechanism by which SUs cause glucagon release has recently been elucidated by the demonstration of KATP channels in pancreatic
-cells (21). SUs seem to generate glucagon secretion in
-cells in a manner comparable to the mechanism whereby SUs cause insulin secretion in ß-cells, i.e. by closure of KATP channels resulting in Ca2+-dependent exocytosis. Moreover, in ß-cells, SUs are characterized by an ability to stimulate exocytosis at a level distal to the SUR (17), as also has been proposed in
-cells (34). This is in contrast to what has been shown for repaglinide in ß-cells (17), in pituitary cells (19), and in
-cells (46). The clinical relevance of this has not yet been explored. Theoretically, KATP channel inhibitors could have an unfavorable effect in type 2 diabetic subjects with a very low ß-cell reserve, namely to increase glucagon levels and, consequently, possibly increase hepatic glucose production. On the other hand, in clinical practice these patients have often long been insulin treated. Finally, it is important to emphasize that we used type 1 diabetic individuals to completely avoid interference from endogenous insulin release, meaning that extrapolation to the type 2 diabetic population should be done with some caution.
In summary, this experiment demonstrates that neither glibenclamide nor repaglinide have any significant acute effect on GH secretion in humans. Additionally, this study shows that KATP channel inhibitors seem to stimulate glucagon secretion in people with severe intraislet insulin deficiency, i.e. C-peptide-negative type 1 diabetic subjects. The clinical relevance of the latter for type 2 diabetic individuals is probably limited, although a contributory role to the failure of oral insulin secretagogues cannot be completely disregarded.
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: AUC, Area under curve; Cmax, maximal concentration; NEFA, nonesterified fatty acid; NS, not significant; SU, sulfonylurea; SUR, sulfonylurea receptor.
Received June 23, 2003.
Accepted September 23, 2003.
| References |
|---|
|
|
|---|
2-cells by insulin/glucose and amino acids. Biochim Biophys Acta 1310:212216[Medline]
This article has been cited by other articles:
![]() |
S. Porksen, L. B. Nielsen, A. Kaas, M. Kocova, F. Chiarelli, C. Orskov, J. J. Holst, K. B. Ploug, P. Hougaard, L. Hansen, et al. Meal-Stimulated Glucagon Release Is Associated with Postprandial Blood Glucose Level and Does Not Interfere with Glycemic Control in Children and Adolescents with New-Onset Type 1 Diabetes J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 2910 - 2916. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Takahashi, H. Ishihara, A. Tamura, S. Yamaguchi, T. Yamada, D. Takei, H. Katagiri, H. Endou, and Y. Oka Cell type-specific activation of metabolism reveals that {beta}-cell secretion suppresses glucagon release from {alpha}-cells in rat pancreatic islets Am J Physiol Endocrinol Metab, February 1, 2006; 290(2): E308 - E316. [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 |