Down-regulation of imidazoline sites in rabbit kidney

Cystathionine-γ lyase-derived hydrogen sulfide mediates the cardiovascular protective effects of moxonidine in diabetic rats

Blunted cystathionine-γ lyase (CSE) activity (reduced endogenous H2S-level) is implicated in hypertension and myocardial dysfunction in diabetes. Here, we tested the hypothesis that CSE derived H2S mediates the cardiovascular protection conferred by the imidazoline I1 receptor agonist moxonidine in a diabetic rat model. We utilized streptozotocin (STZ; 55mg/kg i.p) to induce diabetes in male Wistar rats. Four weeks later, STZ-treated rats received vehicle, moxonidine (2 or 6mg/kg; gavage), CSE inhibitor DL-propargylglycine, (37.5mg/kg i.p) or DL-propargylglycine with moxonidine (6mg/kg) for 3 weeks. Moxonidine improved the glycemic state, and reversed myocardial hypertrophy, hypertension and baroreflex dysfunction in STZ-treated rats. Ex vivo studies revealed that STZ caused reductions in CSE expression/activity, H2S and nitric oxide (NO) levels and serum adiponectin and elevations in myocardial imidazoline I1 receptor expression, p38 and extracellular signal-regulated kinase, ERK1/2, phosphorylation and lipid peroxidation (expressed as malondialdehyde). Moxonidine reversed these biochemical responses, and suppressed the expression of death associated protein kinase-3. Finally, pharmacologic CSE inhibition (DL-propargylglycine) abrogated the favorable cardiovascular, glycemic and biochemical responses elicited by moxonidine. These findings present the first evidence for a mechanistic role for CSE derived H2S in the glycemic control and in the favorable cardiovascular effects conferred by imidazoline I1 receptor activation (moxonidine) in a diabetic rat model.

Imidazoline receptors in the heart: a novel target and a novel mechanism of action that involves atrial natriuretic peptides

Chronic stimulation of sympathetic nervous activity contributes to the development and maintenance of hypertension, leading to left ventricular hypertrophy (LVH), arrhythmias and cardiac death. Moxonidine, an imidazoline antihypertensive compound that preferentially activates imidazoline receptors in brainstem rostroventrolateral medulla, suppresses sympathetic activation and reverses LVH. We have identified imidazoline receptors in the heart atria and ventricles, and shown that atrial I1-receptors are up-regulated in spontaneously hypertensive rats (SHR), and ventricular I1-receptors are up-regulated in hamster and human heart failure. Furthermore, cardiac I1-receptor binding decreased after chronic in vivo exposure to moxonidine. These studies implied that cardiac I1-receptors are involved in cardiovascular regulation. The presence of I1-receptors in the heart, the primary site of production of natriuretic peptides, atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP), cardiac hormones implicated in blood pressure control and cardioprotection, led us to propose that ANP may be involved in the actions of moxonidine. In fact, acute iv administration of moxonidine (50 to 150 microg/rat) dose-dependently decreased blood pressure, stimulated diuresis and natriuresis and increased plasma ANP and its second messenger, cGMP. Chronic SHR treatment with moxonidine (0, 60 and 120 microg kg(-1) h(-1), sc for 4 weeks) dose-dependently decreased blood pressure, resulted in reversal of LVH and decreased ventricular interleukin 1beta concentration after 4 weeks of treatment. These effects were associated with a further increase in already elevated ANP and BNP synthesis and release (after 1 week), and normalization by 4 weeks. In conclusion, cardiac imidazoline receptors and natriuretic peptides may be involved in the acute and chronic effects of moxonidine.

Chronic treatment with rilmenidine in spontaneously hypertensive rats: differences between two schedules of administration

Rilmenidine is one of the lead compounds of the second generation of centrally acting antihypertensive drugs. In the first part of this study, 2 routes of administration of chronic treatment (1 month) with rilmenidine were compared. In conscious and pentobarbital-anesthetized spontaneously hypertensive rats (SHR), rilmenidine was delivered intraperitoneally either 250 microg/kg b.i.d. or 500 microg/kg/d infusion by means of minipumps. The possibility of rilmenidine-induced desensitization of central (brain cortex) and/or peripheral (kidney) alpha2-adrenoreceptors was studied in saturation experiments with the classic alpha2-adrenergic antagonist [H]rauwolscine. In the second part of this study, the cardiovascular and cardiac antihypertrophic effects of the most efficient procedure were investigated. The discontinuous administration of the drug was more effective than infusion. In rats treated with rilmenidine b.i.d., mean blood pressure was reduced by nearly 15% when no reduction occurred in SHRs treated with minipumps. With the first schedule of administration, plasma concentration of the drug reached a maximum of approximately 30 ng/ml when it was only 12 ng/ml with the continuous infusion of the same dose. Anesthesia with pentobarbital potentiated the antihypertensive effect of rilmenidine in rats treated discontinuously and unmasked an antihypertensive action in rats receiving the drug with minipumps. In saturation binding experiments, no significant changes in adrenergic receptors were observed in kidney membrane preparations. In contrast, in brain cortical membranes a reduction by about 50% of the Bmax of [H]rauwolscine value was observed in rats treated discontinuously with rilmenidine. In contrast, a 400% increase of the Bmax was observed in the brain of rats treated with minipumps. Over the one-month period of the second study, the discontinuous treatment with the 500 microg/kg/d dose of rilmenidine was still able to reduce blood pressure, at least at the peak concentration time, but did not induce any significant reduction of the ventricular mass. In conclusion, rilmenidine has only weak antihypertensive effects in conscious SHRs, even at doses higher than those that are active in rabbits and humans. As a consequence, it lacks significant cardiac antihypertrophic effects in this species. Pharmacokinetic data show that the rapid plasma withdrawal of this drug may explain this particular feature in rats.

Imidazoline Receptors but Not α2-Adrenoceptors Are Regulated in Spontaneously Hypertensive Rat Heart by Chronic Moxonidine Treatment

We have recently identified imidazoline I(1)-receptors in the heart. In the present study, we tested regulation of cardiac I(1)-receptors versus alpha(2) -adrenoceptors in response to hypertension and to chronic exposure to agonist. Spontaneously hypertensive rats (SHR, 12-14 weeks old) received moxonidine (10, 60, and 120 microg/kg/h s.c.) for 1 and 4 weeks. Autoradiographic binding of (125)I-paraiodoclonidine (0.5 nM, 1 h, 22 degrees C) and inhibition of binding with epinephrine (10(-10)-10(-5) M) demonstrated the presence of alpha(2)-adrenoceptors in heart atria and ventricles. Immunoblotting and reverse transcription-polymerase chain reaction identified alpha(2A)-alpha(2B)-, and alpha(2C), and -adrenoceptor proteins and mRNA, respectively. However, compared with normotensive controls, cardiac alpha(2) -adrenoceptor kinetic parameters, receptor proteins, and mRNAs were not altered in SHR with or without moxonidine treatment. In contrast, autoradiography showed that up-regulated atrial I(1)-receptors in SHR are dose-dependently normalized by 1 week, with no additional effect after 4 weeks of treatment. Moxonidine (120 microg/kg/h) decreased B(max) in right (40.0 +/- 2.9-7.0 +/- 0.6 fmol/unit area; p < 0.01) and left (27.7 +/- 2.8-7.1 +/- 0.4 fmol/unit area; p < 0.01) atria, and decreased the 85- and 29-kDa imidazoline receptor protein bands, in right atria, to 51.8 +/- 3.0% (p < 0.01) and 82.7 +/- 5.2% (p < 0.03) of vehicle-treated SHR, respectively. Moxonidine-associated percentage of decrease in B(max) only correlated with the 85-kDa protein (R(2) = 0.57; p < 0.006), suggesting that this protein may represent I(2)-receptors. The weak but significant correlation between the two imidazoline receptor proteins (R(2) = 0.28; p < 0.03) implies that they arise from the same gene. In conclusion, the heart possesses I(1)-receptors and alpha(2)-adrenoceptors, but only I(1)-receptors are responsive to hypertension and to chronic in vivo treatment with a selective I(1)-receptor agonist.

Importance of imidazoline-preferring receptors in the cardiovascular actions of chronically administered moxonidine, rilmenidine and clonidine in conscious rabbits

OBJECTIVE

To determine the involvement of central imidazoline receptors in the cardiovascular actions of the chronically administered antihypertensive agents moxonidine, rilmenidine and clonidine.

DESIGN AND METHODS

In 21 rabbits with implanted fourth-ventricular catheters, we investigated the central effects of three cumulative doses of an I(1)-imidazoline/alpha(2)-adrenoceptor antagonist, efaroxan, and of an alpha(2)-adrenoceptor antagonist, 2-methoxyidazoxan (2-MI), on the changes in blood pressure and heart rate (HR) elicited by chronic subcutaneous administration of moxonidine, rilmenidine and clonidine, after 1 and 3 weeks of treatment. A low, medium and high dose of 2-MI was matched to three doses of efaroxan, such that each produced equal reversal of the hypotension induced by fourth-ventricular alpha-methyldopa and hence produced a similar degree of alpha(2)-adrenoceptor blockade.

RESULTS

Clonidine and moxonidine, at doses of 1 mg/kg per day, and rilmenidine at 5 mg/kg per day, produced sustained reductions in mean arterial pressure of 13 +/- 3, 15 +/- 2 and 13 +/- 2 mmHg, respectively over the 3-week treatment period, but did not alter HR. Central administration of efaroxan on day 9 and day 23 of treatment produced a greater increase in blood pressure than did 2-MI with all three antihypertensive agents. Blood pressure reached levels that were significantly above the original control values. By contrast, the alpha(2)-adrenoceptor antagonist 2-MI only induced a rebound blood pressure effect in clonidine- and to a lesser extent in rilmenidine-treated rabbits. Both efaroxan and 2-MI produced a similar degree of tachycardia in moxonidine-, rilmenidine- and clonidine-treated animals.(2)

CONCLUSIONS

The greater effect of efaroxan compared to the alpha(2)-adrenoceptor antagonist 2-MI suggests that the hypotension induced by chronic subcutaneous administration of moxonidine, rilmenidine and clonidine is mediated predominantly via an action on central imidazoline receptors. Furthermore, all agents showed a propensity to produce rebound hypertension with imidazoline receptor blockade. However, only clonidine showed a rebound phenomenon when challenged by acute central alpha(2)-adrenoceptor blockade

The I1-imidazoline receptor and its cellular signaling pathways

Two primary questions are addressed. First, do I1-imidazoline binding sites fulfill all the essential criteria for identification as a true receptor? Second, what are the cellular signaling pathways coupled to this novel receptor? I1-imidazoline binding sites show specificity in binding assays, linkage to physiologic functions, appropriate anatomic, and cellular and subcellular localization. Most important, binding affinities correlate with functional drug responses. I1-imidazoline binding sites meet several additional criteria identified with functional receptors: they show physiologic regulation and endogenous ligands and, most crucially, are coupled to cellular signaling events. A series of studies have identified cellular events triggered by I1-imidazoline receptor occupancy. This receptor is not coupled to conventional pathways downstream of heterotrimeric G-proteins, such as activation or inhibition of adenylyl or guanylyl cyclases, stimulation of inositol phospholipid hydrolysis, or induction of rapid calcium fluxes. The I1-imidazoline receptor is coupled to choline phospholipid hydrolysis, leading to the generation of diacylglyceride, arachidonic acid, and eicosanoids. Additional cellular responses include inhibition of Na+/H+ exchange and induction of genes for catecholamine synthetic enzymes. The signaling pathways linked to the I1-imidazoline receptor are similar to those of the interleukin family, implying that I1-receptors may belong to the family of neurocytokine receptors.

Sigma receptor ligands and imidazoline secretagogues mediate their insulin secretory effects by activating distinct receptor systems in isolated islets

The effects of two potent sigma receptor agonists (+)-3-PPP ((R)-3-(3-hydroxyphenyl)-N-(1-propyl)piperidine) and DTG (N,N'-di-(o-tolyl)guanidine) on the insulin secretory responses in rat islets of Langerhans were investigated. Both sigma receptor ligands were able to potentiate the insulin secretory response of islets incubated at 6 mM glucose, in a dose-dependent manner and were also able to reverse the effects of diazoxide on insulin release. When islets were treated with efaroxan, a well-characterised imidazoline insulin secretagogue, and either (+)-3-PPP or DTG together, there was an unexpected and profound absence of stimulation of insulin release as compared to when islets were incubated with each compound alone. Experiments performed with islets where there was desensitization of DTG/sigma receptor or efaroxan/imidazoline binding site mediated responses suggest that at least two distinct receptor systems appear to be involved. The complex interactions of these two classes of drug require further investigation.

Differential effects of rANF and chronic guanabenz to presynaptic and postsynaptic alpha 2-adrenoceptor-mediated cardiovascular response in pithed rats

1. In normal rats, intracerebroventricular (i.c.v.) guanabenz induced a decrease in blood pressure (BP) and heart rate (HR), and this hypotension or bradycardia was not changed by rANF pretreatment (3 micrograms i.c.v.). 2. In pithed rats, intravenous (i.v.) guanabenz, an alpha 2-adrenoceptor agonist, produced an increase in mean blood pressure (MBP) in a dose-dependent manner. The pressor response by guanabenz was attenuated by infusion of rANF. This attenuation was additive when incubated in combination with yohimbine. 3. In pithed rats, the pressor response due to the increase of sympathetic outflow with electrical stimulation was partially blocked by rANF infusion or chronic guanabenz treatment. This reduction was not augmented by chronic guanabenz plus rANF treatment. 4. The inhibitory action of guanabenz in tachycardia evoked by electrical stimulation at the C7-T1 site was attenuated by rANF, but not by chronic treatment with guanabenz.

The imidazoline receptor in control of blood pressure by clonidine and allied drugs

Clonidine, moxonidine, and rilmenidine are centrally acting antihypertensive agents that lower arterial pressure by inhibiting the tonic activity of sympathoexcitatory neurons in the rostral ventrolateral medulla. Competing hypotheses have been put forward by different investigators to explain the sympathoinhibition evoked by "imidazoline drugs": either via central actions at alpha 2-adrenergic receptors or novel I1-imidazoline receptors. These different perspectives are presented in the accompanying reviews.

Effect of rilmenidine on arterial pressure and urinary output in the spontaneously hypertensive rat

Rilmenidine is an antihypertensive agent acting at the imidazoline receptor that may have both central effects in the ventral lateral medulla and direct effects on the kidney to alter Na+ excretion. The present experiments examined whether rilmenidine induces a leftward shift or change in the slope of the pressure-natriuresis curve in the spontaneously hypertensive rat (SHR). A single oral gavage dose indicated that 3 and 10 mg/kg rilmenidine significantly lowers arterial pressure at 4-12 h after administration by oral gavage. The effect of rilmenidine on pressure-natriuresis was studied using twice daily doses of 1 and 3 mg/kg for control and treated SHR drinking tap water or 1% NaCl for 3 days. Na+ excretion was measured over 24 h, and mean arterial pressure was measured 6-8 h after the morning dose of rilmenidine. The results indicate that 1 mg/kg had no effect, while the pressure-natriuresis relationship for the rats receiving the 3 mg/kg dose was shifted to the left and was not significantly different from the vertical slope of the untreated SHR. This experiment also suggested that rilmenidine may attenuate the salt preference of the rats. This was confirmed in an additional series of experiments in which the rats had access to both tap water and 1% NaCl. Thus, rilmenidine shifts the pressure-natriuresis relationship to the left and reduces salt preference in SHR.

A series of novel imidazoline I2-receptor selective Schiff bases of 1-(benzylidenamino)-3,3-dimethylguanidine

A series of 12 Schiff bases of 1-(benyzylidenamino)-3,3-dimethylguanidine were evaluated for their affinities at imidazoline I2-receptors and alpha 2-adrenoceptors in the guinea pig cerebral cortex by using a 2-curve assay with [3H]idazoxan as ligand and the simultaneous computer analysis of (-)adrenaline masked and non-masked competition curves. All of the dimethylguanidines were selective for I2-receptors with the selectivity ranging 6-110-fold, and the most selective being WAP8 (2-bromobenzylidenamino-3,3-dimethylguanidine) closely followed by FLA151 (2-chlorobenzylidenamino-3, 3-dimethylguanidine). Some of the dimethylguanidines were also tested for affinity at bovine kidney cortex I2-receptors and guinea pig alpha 1-adrenoceptors. These substances were found to have similar affinities for bovine kidney I2-receptors as for guinea pig cerebral cortex I2-receptors, but their affinities for the alpha 1-adrenoceptors were considerably lower.

Current status of putative imidazoline (I1) receptors and renal mechanisms in relation to their antihypertensive therapeutic potential

1. A 'second generation' of centrally acting antihypertensive agents has recently been developed. Unlike the 'first generation' of these agents (e.g. alpha-methyldopa, clonidine, guanabenz), which act predominantly by an agonist action at a alpha 2-adrenoceptors, these agents (e.g. rilmenidine, moxonidine) are believed to exert their antihypertensive effects chiefly by an interaction at putative imidazoline (I) receptors of the I1-type, and so have a reduced profile of alpha 2-adrenoceptor-mediated side effects. There is also evidence from studies in experimental animals that activation of I1-receptors mediates a natriuretic effect. This review evaluates the evidence that they mediate renal effects different from those of alpha 2-adrenoceptors that could contribute to their long-term efficacy. 2. Data from binding studies suggest that I1-binding sites are heterogeneous. There is conflicting evidence concerning whether any of these binding sites are truly receptors. Indeed, the best evidence for the existence of I1-receptors comes from in vivo experiments indicating that imidazoline compounds act at non-adrenoceptor receptive sites in the central nervous system to reduce sympathetic drive and blood pressure. 3. There are a wide range of potential sites and mechanisms through which centrally acting antihypertensive agents can affect renal function, including actions mediated within the central nervous system, heart, systemic circulation and within the kidneys themselves. 'First generation' centrally acting antihypertensive agents cause diuresis and natriuresis in rats, while in dogs and humans a diuresis is often seen with variable effects on sodium excretion. 4. Evidence from studies in anaesthetized rats indicates that rilmenidine and moxonidine can promote sodium excretion by interacting with both central nervous system and renal putative I1-receptors. This does not appear to necessarily be the case in other species. At this time there are few or no published data from clinical studies to suggest that 'second generation' centrally acting antihypertensive agents affect salt and water balance differently from 'first generation' agents.

Dynamic properties of I2-receptors

Evidence is presented that drug binding to I2-sites in the guinea pig kidney and liver is subjected to dynamic homeotropic allosteric regulation. Drugs have been found that are capable of inducing both positive and negative cooperativity on both kidney and liver I2-sites when studied in competition with [3H]idazoxan. Moreover the kinetics of [3H]idazoxan association with guinea pig kidney I2-sites was complex and occurred with three different rate constants. However, on dissociation of [3H]idazoxan from its binding sites in the kidney a substantial number of the slower sites had been converted to faster sites, thus supporting the idea of a dynamic interconversion of I2 sites. The guinea pig liver and kidney I2-sites appear to be similar in their pharmacological properties. They also bear some resemblance to the I2-sites of the smooth muscle of the guinea pig ileum, but they appear to be more distant to the guinea pig cerebral cortex I2-sites. Nevertheless, data indicate that considerable overall heterogeneity exists between the guinea pig kidney and liver I2-sites and the guinea pig ileum and cerebral cortex I2-sites.

Why imidazoline receptor modulator in the treatment of hypertension?

The influence of the sympathetic nervous system on blood pressure control was impressively demonstrated in 1940 by bilateral excision of sympathetic nerve fibers. Thereafter, the first generation of drugs lowering blood pressure by central modulation of the sympathetic outflow through alpha 2-adrenoceptor for stimulation, such as alpha-methyldopa, guanabenz, clonidine, and guanfacine, were marketed. However, these compounds were often tolerated poorly, because they caused orthostatic hypotension, sedation, tachycardia or bradycardia, dry mouth, and reduced cardiac output. The mode of action of the second generation centrally acting antihypertensive drugs moxonidine and rilmenidine is different from that of the first generation compounds (e.g., clonidine). Contrary to clonidine, the newer drugs bind more selectively to I1-imidazoline receptors rather than to alpha 2-adrenoceptors where first-generation drugs act. The high affinity and selectivity of these two drugs for this recently discovered new receptor class make it possible to discriminate between I1-imidazoline receptor-mediated blood pressure lowering, on the one hand, and alpha 2-adrenoceptor-mediated side effects, on the other. Discrimination of the two effects was substantiated either by studies using moxonidine alone or in interaction experiments with I1-imidazoline receptor or alpha 2-adrenoceptor antagonists. The high selectivity of moxonidine at the I1-imidazoline receptor allows discrimination between alpha 2-adrenoceptors and I1-imidazoline receptors and is reflected in man by the relatively low incidence of adverse drug events during moxonidine treatment. Concentration of endazoline, a specific mediator of I1-imidazoline receptors, is elevated in some patients with essential hypertension. Modulation of I1-imidazoline receptors by moxonidine could be interpreted as antagonism with regard to the endogenous agonistic effect of the endogenous "transmitter" endazoline. On the other hand, moxonidine acted directly as an agonist at the putative I1-imidazoline receptor. Therefore, to clear the ground, characterization as well as physiological function of the mediator for imidazoline receptors seems essential. The therapeutic relevance of using drugs selective for I1-imidazoline receptors for blood pressure reduction in hypertensive patients is substantiated by the finding that in human rostral ventrolateral medulla (RVLM), which is essential in central blood pressure regulation, the relation between alpha 2-adrenoceptors and I1-imidazoline receptors is about one to ten (1:10). Reduction of a long-lasting sympathetic overdrive may avoid the deteriorating effects on the heart and peripheral circulation. These recent findings give a rational explanation for the very low incidence of sedation and the absence of respiratory depression, orthostatic hypotension, and rebound hypertension that banned the former central acting antihypertensive drugs from first-line treatment despite the advantages of central mediated blood pressure control.

Characterization of binding sites for [3H]-idazoxan, [3H]-P-aminoclonidine and [3H]-rauwolscine in the kidney of the dog

1. We characterized the binding of [3H]-rauwolscine, [3H]-p-aminoclonidine and [3H]-idazoxan in a dog kidney membrane preparation. Our aim was to determine the pharmacological nature of the alpha 2-adrenoceptor- and imidazoline-preferring binding sites in this organ. 2. [3H]-Rauwolscine bound to an apparent single site with an affinity (KD) of 2.2 nmol/L and a maximum density (Bmax) of 58.5 fmol/mg protein, when 10 mumol/L idazoxan defined non-specific binding. However displacement studies demonstrated that a number of compounds, including prazosin, inhibited [3H]-rauwolscine binding in a complex manner consistent with displacement from two distinct binding sites. The majority (69%) of the [3H]-rauwolscine binding sites had a relatively low affinity for prazosin (KI = 398 nmol/L), while the remainder had a relatively high affinity for prazosin (KI = 7.9 nmol/L). 3. [3H]-p-Aminoclonidine bound to an apparent single site (KD = 5.2 nmol/L; Bmax = 72.4 fmol/mg protein), when 10 mumol/L phentolamine defined non-specific binding. When 1 mumol/L of the potent and selective alpha 2-adrenoceptor antagonist 2-methoxyidazoxan was included in the incubate, no specific binding was detected. We therefore conclude that under the conditions of this experiment [3H]-p-aminoclonidine binds only to alpha 2-adrenoceptors in the dog kidney. 4. [3H]-Idazoxan bound to two sites, with a higher (KD = 0.95 nmol/L; Bmax = 43.9 fmol/mg protein) and lower (KD = 9.1 nmol/L; Bmax = 93.8 fmol/mg protein) affinity, respectively, when 1 mmol/L phentolamine defined non-specific binding. When 10 mumol/L GTP gamma S was included in the incubate, the low affinity site was unaffected but the maximum binding at the higher affinity site was reduced by 79%. 2-Methoxyidazoxan displaced [3H]-idazoxan in a monophasic manner and with low potency (IC50 = 11.5 mumol/L). Yohimbine, efaroxan, clonidine, rilmenidine, guanabenz and idazoxan itself displaced [3H]-idazoxan in a complex manner; the slope of the displacement curves being less than unity. 5. We conclude that the dog kidney contains a heterogeneous population of alpha 2-adrenoceptors that can be labelled either with [3H]-rauwolscine or [3H]-p-aminoclonidine. The dog kidney also contains a heterogeneous population of non-adrenoceptor imidazoline-preferring binding sites of the I2-subtype, that can be labelled with [3H]-idazoxan. The binding site for which [3H]-idazoxan has the highest affinity appears to be coupled to a guanine nucleotide binding regulatory protein.

The imidazoline site involved in control of insulin secretion: characteristics that distinguish it from I1- and I2-sites

1. The nature of the binding site mediating the insulin secretagogue activity of certain imidazoline compounds remains unclear and the pharmacology of the I1- and I2-imidazoline sites, described in many tissues, does not correlate with the observed responses to imidazolines in islets. In the present paper, we describe further results which support the concept that the islet imidazoline site may represent a novel subtype of imidazoline receptor. 2. Culture of rat isolated islets in the presence of imidazoline secretagogues (either efaroxan or phentolamine) resulted in loss of responsiveness on subsequent re-exposure to these agents. However, culture of islets with either idazoxan or UK14,304 (imidazoline ligands that do not stimulate insulin secretion) did not lead to any loss of response when the islets were subsequently exposed to efaroxan. By contrast, islets cultured with UK14,304 (a potent alpha 2-adrenoceptor agonist), displayed loss of sensitivity to noradrenaline, consistent with down-regulation of alpha 2-adrenoceptors. 3. In order to characterize the imidazoline site further, radioligand binding studies were performed in membranes from RINm5F insulinoma cells using [3H]-RX821002, an imidazoline insulin secretagogue that does not interact significantly with imidazoline sites in other tissues. [3H]-RX821002 labelled alpha 2-adrenoceptors with high affinity (2.01 +/- 0.7 nM) but also labelled a second, non-adrenoceptor site with much lower affinity. 4. Under conditions of alpha 2-adrenoceptor blockade (in the presence of adrenaline), efaroxan displaced [3H]-RX821002 binding to the low affinity site, in a dose-dependent manner. Competition studies employing additional imidazoline compounds of varying secretagogue activity revealed that the pharmacological profile of the low affinity site correlates well with that observed in secretion experiments.5. The results obtained from the down-regulation experiments with isolated islets and from the radioligand binding studies suggest that the low affinity [3H]-RX821002 binding site may represent the functional receptor responsible for the secretagogue activity of imidazoline compounds in the endocrine pancreas and that it has a pharmacological profile distinct from those of I,- and 12-sites.