Interactions between imidazoline compounds and sulphonylureas in the regulation of insulin secretion

Mechanisms of antihyperglycaemic action of efaroxan in mice: time for reappraisal of α2A-adrenergic antagonism in the treatment of type 2 diabetes?

AIMS/HYPOTHESIS

Inspired by recent speculation about the potential utility of α(2A)-antagonism in the treatment of type 2 diabetes, the study examined the contribution of α(2)-antagonism vs other mechanisms to the antihyperglycaemic activity of the imidazoline (±)-efaroxan.

METHODS

Effects of the racemate and its pure enantiomers on isolated pancreatic islets and beta cells in vitro, as well as on hyperglycaemia in vivo, were investigated in a comparative manner in mice.

RESULTS

In isolated perifused islets, the two enantiomers of efaroxan were equally potent in counteracting inhibition of insulin release by the ATP-dependent K(+) (K(ATP)) channel-opener diazoxide but (+)-efaroxan, the presumptive carrier of α(2)-antagonistic activity, was by far superior in counteracting inhibition of insulin release by the α(2)-agonist UK14,304. In vivo, (+)-efaroxan improved oral glucose tolerance at 100-fold lower doses than (-)-efaroxan and, in parallel with observations made in vitro, was more effective in counteracting UK14,304-induced than diazoxide-induced hyperglycaemia. The antihyperglycaemic activity of much higher doses of (-)-efaroxan was associated with an opposing pattern (i.e. with stronger counteraction of diazoxide-induced than UK14,304-induced hyperglycaemia), which implicates a different mechanism of action.

CONCLUSIONS/INTERPRETATION

The antihyperglycaemic potency of (±)-efaroxan in mice is almost entirely due to α(2)-antagonism, but high doses can also lower blood glucose via another mechanism. Our findings call for reappraisal of the possible clinical utility of α(2A)-antagonistic compounds in recently identified subpopulations of patients in which a congenitally higher level of α(2A)-adrenergic activation contributes to the development and pathophysiology of type 2 diabetes.

α2-adrenoceptor regulation of blood glucose homeostasis

The α(2A)-adrenoceptor has been identified as an important regulator of blood glucose homeostasis. α(2A)-Adrenoceptors on pancreatic β-cells inhibit insulin secretion, and α(2A)-adrenoceptors on sympathetic nerves and on adrenomedullary chromaffin cells limit sympathoadrenal output. Recently, human α(2A)-adrenoceptor gene polymorphisms that influence α(2A)-adrenoceptor expression and function have been described. Increased α(2A)-adrenoceptor expression has been associated with impaired glucose-stimulated insulin secretion, elevated fasting blood glucose levels and an increased risk of type 2 diabetes. Accordingly, administration of α(2)-adrenoceptor agonists generally increases blood glucose levels, in spite of the ensuing sympatholysis that would be expected to lower blood glucose as a result of diminished α(1)- and β-adrenoceptor activation. α(2)-Adrenoceptor antagonists increase insulin secretion and reduce blood glucose levels by inhibiting tonically active α(2A)-adrenoceptors on pancreatic β-cells, but may also enhance sympathoadrenal output. In addition, α(2)-adrenoceptor antagonists potentiate the insulinotropic effect of sulphonylurea drugs, pointing to a potentially serious adverse drug interaction when the two classes of drugs are combined. The α(2)-adrenoceptor antagonist atipamezole is widely used in veterinary medicine, and sulphonylureas are prescribed for the treatment of type 2 diabetes in cats and dogs. Even if no dedicated α(2)-adrenoceptor antagonists are in clinical use in humans, some antipsychotic and antidepressant drugs are relatively potent α(2)-adrenoceptor antagonists. In the treatment of type 2 diabetes, α(2)-adrenoceptor agonists could possibly protect against sulphonylurea-induced hypoglycaemia, and α(2)-adrenoceptor antagonist drugs could improve insulin secretion. The potential usefulness of such drugs may vary between individuals, depending on α(2A)-adrenoceptor genetics, sympathetic tone and concomitant pathological conditions, such as cardiovascular disease and obesity.

alpha2A-adrenoceptor antagonism increases insulin secretion and synergistically augments the insulinotropic effect of glibenclamide in mice

BACKGROUND AND PURPOSE

The imidazoline-type alpha2-adrenoceptor antagonists (+/-)-efaroxan and phentolamine increase insulin secretion and reduce blood glucose levels. It is not known whether they act by antagonizing pancreatic beta-cell alpha2-adrenoceptors or by alpha2-adrenoceptor-independent mechanisms. Many imidazolines inhibit the pancreatic beta-cell KATP channel, which is the molecular target of sulphonylurea drugs used in the treatment of type II diabetes. To investigate the mechanisms of action of (+/-)-efaroxan and phentolamine, alpha2A-adrenoceptor knockout (alpha2A-KO) mice were used.

EXPERIMENTAL APPROACH

Effects of (+/-)-efaroxan, 5 mg kg(-1), and phentolamine, 1 mg kg(-1), on blood glucose and insulin levels were compared with those of the non-imidazoline alpha2-adrenoceptor antagonist [8aR,12aS,13aS]-5,8,8a,9,10,11,12,12a,13,13a-decahydro-3-methoxy-12-(ethylsulphonyl)-6H-isoquino[2,1-g][1,6]naphthyridine (RS79948-197), 1 mg kg(-1), and the sulphonylurea glibenclamide, in alpha2A-KO and control (wild type (WT)) mice.

KEY RESULTS

In fed WT mice, (+/-)-efaroxan, phentolamine and RS79948-197 reduced blood glucose and increased insulin levels. Fasting abolished these effects. In fed alpha2A-KO mice, (+/-)-efaroxan, phentolamine and RS79948-197 did not alter blood glucose or insulin levels, and in fasted alpha2A-KO mice, blood glucose levels were increased. Glibenclamide, at a dose only moderately efficacious in WT mice (5 mg kg(-1)), caused severe hyperinsulinaemia and hypoglycaemia in alpha2A-KO mice. This was mimicked in WT mice by co-administration of RS79948-197 with glibenclamide.

CONCLUSIONS AND IMPLICATIONS

These results suggest that (+/-)-efaroxan and phentolamine increase insulin secretion by inhibition of beta-cell alpha2A-adrenoceptors, and demonstrate a critical role for alpha2A-adrenoceptors in limiting sulphonylurea-induced hyperinsulinaemia and hypoglycaemia.

Glucose concentration-dependent potentiation of insulin secretion by a new chemical entity, KCP256

The insulinotropic activity of KCP256 [(R)-8-benzyl-2-cyclopentyl-7, 8-dihydro-4-propyl-1H-imidazo[2,1-i]purin-5(4H)-one hydrochloride] was examined using MIN6 cells (a pancreatic beta-cell line) and pancreatic islets isolated from rats. Unlike sulfonylurea anti-diabetic drugs, KCP256 dose-dependently (0.1-10 microM) enhanced insulin secretion from MIN6 cells and its insulinotropic effect was exerted only at high concentrations of glucose (8.3-22 mM) but not at low concentrations of glucose (3.3-5.5 mM). Furthermore, the action mechanism of KCP256 was different because, unlike sulfonylurea drugs, KCP256 did not displace the binding of [3H]glibenclamide, and did not inhibit the 86Rb+ efflux nor K(ATP) channel activity. In isolated islets, KCP256 also enhanced insulin secretion in a dose- and a glucose-concentration-dependent manner. Plasma levels of insulin after glucose challenge in KCP256-administrated rats were higher than those in vehicle-administrated animals, indicating that KCP256 can enhance insulin secretion in vivo. Since the insulinotropic activity of KCP256 only occurs at high concentrations of glucose, this novel drug may exhibit a decreased risk of drug-induced hypoglycemia compared with sulfonylurea drugs when treating patients with diabetes.

Imidazoline receptors: new targets for antihyperglycaemic drugs

In both animal models of Type 2 diabetes and in man, it has been evident for many years that certain imidazoline drugs can stimulate insulin secretion and improve glycaemia. This suggests that they may be useful new reagents for use in the management of Type 2 diabetes. However, despite their promise, no imidazoline compound has yet come into clinical use as an effective therapeutic agent in diabetes. This should not be taken as evidence of a flaw in the basic hypothesis, but derives, in part, from continuing ignorance about the molecular characteristics of imidazoline binding proteins, and the precise structure-activity relationships of their ligands. In this review, the mode of action of antihyperglycaemic imidazoline compounds is considered, and the possibility discussed that these agents may interact with a unique subtype of imidazoline binding site associated with ATP-sensitive potassium channels. The functional consequences of this interaction are summarised together with evidence that the binding site may actually lie within the channel complex. Additional data implicating the participation of alpha2-adrenoceptors in some actions of imidazolines are evaluated, and examples of relevant drugs having encouraging therapeutic profiles are highlighted. The possibility that some anti-diabetic imidazoline reagents may exert extra-pancreatic effects is also considered. Overall, the article aims to highlight important developments within the field but also draws attention to those areas where controversy remains.

The role of I(1)-imidazoline and alpha(2)-adrenergic receptors in the modulation of glucose metabolism in the spontaneously hypertensive obese rat model of metabolic syndrome X

We examined glucose metabolism after I1-imidazoline (I1R) and alpha2-adrenergic receptor (alpha2AR) activation in an animal model of metabolic syndrome X. Fasted spontaneously hypertensive obese rats (SHROB) were given the I1R/alpha2AR agonists moxonidine and rilmenidine or the alpha2AR agonist guanabenz. Because of the dual specificity of moxonidine, its actions were split into adrenergic and nonadrenergic components by using selective antagonists: rauwolscine (alpha2AR) efaroxan (I1R/alpha2AR), or 2-endo-amino-3-exo-isopropylbicyclo[2.2.1.]heptane (AGN 192403) (I1R). Hyperglycemia induced by moxonidine, rilmenidine, and guanabenz resulted from inhibition of insulin secretion. Similar responses were observed after oral dosing and in lean littermates. Glucagon was reduced by the I1R agonists (moxonidine, 32 +/- 5%; rilmenidine, 24 +/- 7%) but elevated by guanabenz (71 +/- 32%). The hyperglycemic and hypoinsulinemic responses to moxonidine were blocked by rauwolscine. In contrast, rauwolscine potentiated the reduction in glucagon (39 +/- 6%). AGN 193402 blocked the glucagon response without affecting hyperglycemia and hypoinsulinemia. Efaroxan blocked all responses to moxonidine. When SHROB rats were treated with moxonidine 15 min before an oral glucose tolerance test, the glucose area under the curve (AUC) was increased. Antagonizing the alpha2AR component of moxonidine's action with rauwolscine improved glucose AUC 3-fold and facilitated the insulin secretory response and reduced glucagon secretion. Testing fasting glucose and insulin during 3 weeks of oral moxonidine revealed early hyperglycemia that later faded, and a progressive drop in fasting insulin. The acute hyperglycemia and hypoinsulinemia elicited by moxonidine and rilmenidine was mediated by alpha2AR, whereas I1R may reduce glucagon and increase insulin, particularly after a glucose load.

Evidence that the novel imidazoline compound FT005 is an alpha(2)-adrenoceptor agonist

1: The aim of this study was to determine whether the hyperglycaemic action of the novel imidazoline compound FT005 could be mediated by activation of alpha(2)-adrenoceptors, using a variety of in vivo and in vitro methods including radioligand binding. 2: FT005 produced a dose-dependent increase in blood glucose levels of CBA/Ca mice (0.125-25 mg kg(-1), i.p.). The time course of this hyperglycaemic effect matched that of adrenaline (1 mg kg(-1)) more closely than glucagon (1 mg kg(-1)) or the K(ATP) channel opener diazoxide (25 mg kg(-1)). The hyperglycaemic effect of FT005 (1 mg kg(-1)) was significantly reduced by the alpha(2)-adrenoceptor antagonist rauwolscine (0.5 mg kg(-1)). 3: FT005 produced a significant reduction in plasma insulin levels of mice 30 min after administration. The hyperglycaemic effect of FT005 (25 mg kg(-1)), although still present, was significantly less in fasted mice in which insulin levels are lower, suggesting that a reduction of insulin secretion contributes to the action of FT005. 4: When studied in the mouse isolated vas deferens preparation, FT005 produced a complete inhibition of neurogenic contractions, which was blocked by rauwolscine. This is consistent with activation of pre-synaptic alpha(2)-adrenoceptors. 5: In radioligand binding studies FT005 completely displaced the alpha(2)-adrenoceptor antagonist [(3)H]-RX821002 from mouse whole brain homogenates. The displacement was best described by a two-site model of interaction comprising high and low affinity components. 6: The results indicate that FT005 is an agonist at alpha(2)-adrenoceptors. A reduction in insulin secretion contributes to the hyperglycaemic action of FT005, although an additional mechanism can not be excluded.

The potential antidiabetic activity of some alpha-2 adrenoceptor antagonists

The effect of alpha-2 adrenoceptor antagonists, yohimbine and efaroxan, on the plasma glucose and insulin levels was studied in non-diabetic control, type-I (insulin-dependent) and type-II (non-insulin-dependent) diabetic rats. Pretreatment with either yohimbine or efaroxan potentiated glucose-induced insulin release in non-diabetic control rats and produced an improvement of the oral glucose tolerance and potentiated glucose-induced insulin release in type-II but not in type-I diabetic rats. Treatment with either yohimbine or efaroxan reduced the plasma glucose level and increased the plasma insulin level of non-diabetic control and type-II diabetic rats but not of type-I diabetic rats. Effects of efaroxan were more marked. Pretreatment of non-diabetic control and type-II diabetic rats with either yohimbine or efaroxan inhibited clonidine-induced hyperglycaemia and suppressed or reversed clonidine-induced hypoinsulinaemia. Also, pretreatment of these animals with either yohimbine or efaroxan enhanced the hypoglycaemic and insulinotropic effects of glibenclamide. The combination of glibenclamide and efaroxan led to a synergistic increase in insulin secretion, while that of glibenclamide and yohimbine led to an additive increase. The hyperglycaemic effect of diazoxide in non-diabetic control and type-II diabetic rats was inhibited by pretreatment with either yohimbine or efaroxan. The hypoinsulinaemic effect of diazoxide in these animals was antagonized and reversed by pretreatment with yohimbine and efaroxan, respectively. In type-I diabetic rats, there was no change in the plasma glucose and insulin levels induced by the treatment of animals with each of clonidine or diazoxide alone or in combination with either yohimbine or efaroxan. Glibenclamide produced a slight decrease in the plasma glucose level of type-I diabetic rats, at the end of the 120 min period of investigation but there was no change in the plasma insulin level. Pretreatment of these animals with either yohimbine or efaroxan produced no change in glibenclamide effects. Additionally, bath application of efaroxan or glibenclamide inhibited the relaxant effects of different concentrations of diazoxide on the isolated norepinephrine-contracted aortic strips, while the application of yohimbine produced insignificant changes. The combination of glibenclamide and efaroxan led to complete inhibition of the relaxant effects of different concentrations of diazoxide, while that of glibenclamide and yohimbine did not produce such an effect. It is concluded that yohimbine, via blockade of postsynaptic alpha-2 adrenoceptors, and efaroxan, via blockade of postsynaptic alpha-2 adrenoceptors and adenosine triphosphate-sensitive potassium channels in the pancreatic beta-cell membrane, produce insulinotropic and subsequent hypoglycaemic effects.

Multiple effector pathways regulate the insulin secretory response to the imidazoline RX871024 in isolated rat pancreatic islets

When isolated rat islets were cultured for 18 h prior to use, the putative imidazoline binding site ligand, RX871024 caused a dose-dependent increase in insulin secretion at both 6 mM and 20 mM glucose. By contrast, a second ligand, efaroxan, was ineffective at 20 mM glucose whereas it did stimulate insulin secretion in response to 6 mM glucose. Exposure of islets to RX871024 (50 microM) for 18 h, resulted in loss of responsiveness to this reagent upon subsequent re-exposure. However, islets that were unresponsive to RX871024 still responded normally to efaroxan. The imidazoline antagonist, KU14R, blocked the insulin secretory response to efaroxan, but failed to prevent the stimulatory response to RX871024. By contrast with its effects in cultured islets, RX871024 inhibited glucose-induced insulin release from freshly isolated islets. Efaroxan did not inhibit insulin secretion under any conditions studied. In freshly isolated islets, the effects of RX871024 on insulin secretion could be converted from inhibitory to stimulatory, by starvation of the animals. Inhibition of insulin secretion by RX871024 in freshly isolated islets was prevented by the cyclo-oxygenase inhibitors indomethacin or flurbiprofen. Consistent with this, RX871024 caused a marked increase in islet PGE2 formation. Efaroxan did not alter islet PGE2 levels. The results suggest that RX871024 exerts multiple effects in the pancreatic beta-cell and that its effects on insulin secretion cannot be ascribed only to interaction with a putative imidazoline binding site.

Imidazolines and the pancreatic B-cell. Actions and binding sites

Stimulation of insulin secretion by imidazoline compounds displays variable characteristics. Phentolamine (10-100 microM) increased secretion of perifused mouse islets at nonstimulatory glucose concentrations (5 mM) and even in the absence of glucose. Idazoxan (20-100 microM) elicited a moderate increase in insulin secretion, which required the presence of a stimulatory glucose concentration (10 mM). Phentolamine is therefore a stimulator of secretion in its own right, whereas idazoxan may be termed an enhancer of secretion. Both compounds inhibited the activity of ATP-dependent K+ channels in inside-out patches from B-cells; however, idazoxan achieved only an incomplete block. Both compounds depolarized the B-cell plasma membrane to an extent that permitted the opening of voltage-dependent Ca2+ channels (-40 to -30 mV). An increase in cytoplasmic Ca2+ concentration was induced by phentolamine and much less so by idazoxan. Activation of protein kinase C, a possible mechanism to amplify Ca(2+)-induced secretion, could not be verified for phentolamine. It thus appears that stimulation of insulin secretion by phentolamine is due to its blocking effect on KATP channels, which may be the correlate of non-adrenergic imidazoline binding sites which were characterized in insulin-secreting HIT cells. Whether incomplete closure of KATP channels by idazoxan or additional effects are responsible for the requirement of high glucose to stimulate secretion remains to be clarified.

Novel targets and techniques in imidazoline receptor research

Imidazoline binding sites are now generally accepted as being receptors. Despite this acceptance, the molecular structure and signal transduction mechanisms of these receptors are still poorly understood. The I1-imidazoline binding site (I1-receptor) is localized to the plasma membrane, but it is not clear if this represents a conventional receptor. It is also not clear if there are multiple forms of the I1-receptor. The signal transduction mechanisms of I1-receptors are similarly unclear, but much progress has been made. Evidence clearly indicates that ligands with high affinity for I1-receptors stimulate a novel signal transduction pathway, phosphatidylcholine-selective phospholipase C, in the rat adrenal medullary tumor cell line PC-12. However, this may not be the case in all cell types as microphysiometry, a novel technique for determining cellular activation, could not detect receptor activation in cultured bovine adrenal medullary cells exposed to a number of imidazolines considered to be agonists at the I1-receptor. This suggests that there is no I1-receptor-mediated stimulation of phosphatidylcholine-specific phospholipase C in these cells. By contrast, nicotine-stimulated increases in ion entry were blocked by clonidine. Ion channels have been suggested as another possible I1-imidazoline "receptor" family and may represent the low affinity I1-receptor. I1-Receptor ligands can be shown to bind to, or block, the following members of the ligand-gated ion channel super family, the 5HT3, K+ATP, NMDA, and nicotinic acetylcholine receptors. The site of action appears to be the phencyclidine binding site in these channels, but other possibilities cannot be excluded. Molecular modeling suggests that I1-receptor-selective ligands share a common three-dimensional structure with phencyclidine, providing a basis for these actions. This suggests that a phencyclidine-binding site motif may represent a novel site of action for I1-receptor ligands and that searches for receptors based on this motif may reveal novel imidazoline "receptors."