Multihormonal regulation of pituitary melanotrophs

Development of the hypothalamo-hypophyseal system in amphibians with special reference to metamorphosis

In this review article, topics of the embryonic origin of the adenohypophysis and hypothalamus and the development of the hypothalamo-hypophyseal system for the completion of metamorphosis in amphibians are included. The primordium of the adenohypophysis as well as the primordium of the hypothalamus in amphibians is of neural origin as shown in other vertebrates, and both are closely associated with each other at the earliest stage of development. Metamorphosis progresses via the interaction of thyroid hormone and adrenal corticosteroids, of which secretion is enhanced by thyrotropin and corticotropin, respectively. However, unlike in mammals, the hypothalamic releasing factor for thyrotropin is not thyrotropin-releasing hormone (TRH), but corticotropin-releasing factor (CRF) and the major releasing factor for corticotropin is arginine vasotocin (AVT). Prolactin, the release of which is profoundly enhanced by TRH at the metamorphic climax, is another pituitary hormone involved in metamorphosis. Prolactin has a dual role: modulation of the metamorphic speed and the development of organs for adult life. The secretory activities of the pituitary cells containing the three above-mentioned pituitary hormones are elevated toward the metamorphic climax in parallel with the activities of the CRF, AVT, and TRH neurons.

Some aspects of the hypothalamic and pituitary development, metamorphosis, and reproductive behavior as studied in amphibians

In this review article, information about the development of the hypothalamo-hypophyseal axis, endocrine control of metamorphosis, and hormonal and pheromonal involvements in reproductive behavior in some amphibian species is assembled from the works conducted mainly by our research group. The hypothalamic and pituitary development was studied using Bufo embryos and larvae. The primordium of the epithelial hypophysis originates at the anterior neural ridge and migrates underneath the brain to form a Rathke's pouch-like structure. The hypothalamo-hypophyseal axis develops under the influence of thyroid hormone (TH). For the morphological and functional development of the median eminence, which is a key structure in the transport of regulatory hormones to the pituitary, contact of the adenohypophysis with the undeveloped median eminence is necessary. For the development of proopiomelanocortin-producing cells, contact of the pituitary primordium with the infundibulum is required. The significance of avascularization in terms of the function of the intermediate lobe of the pituitary was evidenced with transgenic Xenopus frogs expressing a vascular endothelial growth factor in melanotropes. Metamorphosis progresses via the interaction of TH, adrenal corticosteroids, and prolactin (PRL). We emphasize that PRL has a dual role: modulation of the speed of metamorphic changes and functional development of organs for adult life. A brief description about a novel type of PRL (1B) that was detected was made. A possible reason why the main hypothalamic factor that stimulates the release of thyrotropin is not thyrotropin-releasing hormone, but corticotropin-releasing factor is considered in light of the fact that amphibians are poikilotherms. As regards the reproductive behavior in amphibians, studies were focused on the courtship behavior of the newt, Cynops pyrrhogaster. Male newts exhibit a unique courtship behavior toward sexually developed conspecific females. Hormonal interactions eliciting this behavior and hormonal control of the courtship pheromone secretion are discussed on the basis of our experimental results.

Angiogenesis in the intermediate lobe of the pituitary gland alters its structure and function

The pars distalis (PD) and the pars intermedia (PI) have the same embryonic origin, but their morphological and functional characteristics diverge during development. The PD is highly vascularized, whereas the highly innervated PI is essentially non-vascularized. Based on our previous finding that vascular endothelial growth factor-A (VEGF-A) is involved in vascularization of the rat PD, attempt was made to generate transgenic Xenopus expressing VEGF-A specifically in the melanotrope cells of the PI as a model system for studying the significance of vascularization or avascularization for the functional differentiation of the pituitary. The PI of the transgenic frogs, examined after metamorphosis, were distinctly vascularized but poorly innervated. The experimentally induced vascularization in the PI resulted in a marked increase in tissue volume and a decrease in the expression of both alpha-melanophore-stimulating hormone (α-MSH) and prohormone convertase 2, a cleavage enzyme essential for generating α-MSH. The transgenic animals had low plasma α-MSH concentrations and displayed incomplete adaptation to a black background. To our knowledge, this is the first report indicating that experimentally induced angiogenesis in the PI may bring about functional as well as structural alterations in this tissue.

Analysis of the melanotrope cell neuroendocrine interface in two amphibian species, Rana ridibunda and Xenopus laevis: a celebration of 35 years of collaborative research

This review gives an overview of the functioning of the hypothalamo-hypophyseal neuroendocrine interface in the pituitary neurointermediate lobe, as it relates to melanotrope cell function in two amphibian species, Rana ridibunda and Xenopus laevis. It primarily but not exclusively concerns the work of two collaborating laboratories, the Laboratory for Molecular and Cellular Neuroendocrinology (University of Rouen, France) and the Department of Cellular Animal Physiology (Radboud University Nijmegen, The Netherlands). In the course of this review it will become apparent that Rana and Xenopus have, for the most part, developed the same or similar strategies to regulate the release of α-melanophore-stimulating hormone (α-MSH). The review concludes by highlighting the molecular and cellular mechanisms utilized by thyrotropin-releasing hormone (TRH) to activate Rana melanotrope cells and the function of autocrine brain-derived neurotrophic factor (BDNF) in the regulation of Xenopus melanotrope cell function.

TRH acts as a multifunctional hypophysiotropic factor in vertebrates

Thyrotropin-releasing hormone (TRH) is the first hypothalamic hypophysiotropic neuropeptide whose sequence has been chemically characterized. The primary structure of TRH (pGlu-His-Pro-NH(2)) has been fully conserved across the vertebrate phylum. TRH is generated from a large precursor protein that contains multiple repeats of the TRH progenitor tetrapeptide Gln-His-Pro-Gly. In all tetrapods, TRH-expressing neurons located in the hypothalamus project towards the external zone of the median eminence while in teleosts they directly innervate the pars distalis of the pituitary. In addition, in frogs and teleosts, a bundle of TRH-containing fibers terminate in the neurointermediate lobe of the pituitary gland. Although TRH was originally named for its ability to trigger the release of thyroid-stimulating hormone (TSH) in mammals, it later became apparent that it exerts multiple, species-dependent hypophysiotropic activities. Thus, in fish TRH stimulates growth hormone (GH) and prolactin (PRL) release but does not affect TSH secretion. In amphibians, TRH is a marginal stimulator of TSH release in adult frogs, not in tadpoles, and a major releasing factor for GH and PRL. In birds, TRH triggers TSH and GH secretion. In mammals, TRH stimulates TSH, GH and PRL release. In fish and amphibians, TRH is also a very potent stimulator of alpha-melanocyte-stimulating hormone release. Because the intermediate lobe of the pituitary of amphibians is composed by a single type of hormone-producing cells, the melanotrope cells, it is a suitable model in which to investigate the mechanism of action of TRH at the cellular and molecular level. The occurrence of large amounts of TRH in the frog skin and high concentrations of TRH in frog plasma suggests that, in amphibians, skin-derived TRH may exert hypophysiotropic functions.

Identification of novel genes involved in the plasticity of pituitary melanotropes in amphibians

Melanotrope cells from the amphibian intermediate lobe are composed of two subpopulations that exhibit opposite secretory behavior: hypersecretory and hormone-storage hyposecretory melanotropes. Isolation of these subpopulations allowed a comparison of their gene expression profiles by differential display, leading to the identification of a number of genes differentially expressed in hypersecretory or hyposecretory melanotropes. Among them, we chose two (preferentially expressed in hyposecretory cells) of unknown function but structurally related to proteins involved in the secretory process: Rab18 and KIAA0555. We demonstrate that, upon activation of the regulated secretory pathway, Rab18 associates with secretory granules, inhibits their mobilization, and, consequently, reduces the secretory capacity of neuroendocrine cells. The other gene, KIAA0555, was predicted by in silico analysis to encode a protein with a long coiled-coil domain, a structural feature also shared by different proteins related to intracellular membrane traffic (i.e., golgins), and a hydrophobic C-terminal domain that could function as a transmembrane domain. A database search unveiled the existence of a KIAA0555 paralogue, KIAA4091, displaying a long coiled-coil region highly similar to that of KIAA0555 and an identical C-terminal transmembrane domain. Both KIAA0555 and KIAA4091 were found to be predominantly expressed in tissues containing cells with regulated secretory pathway, that is, endocrine and neural tissues. Moreover, when exogenously expressed in HEK293 cells, both proteins showed a yuxtanuclear distribution, which partially overlaps with that of a Golgi complex marker, thus suggesting a possible role of these two proteins in the control of the secretory process.

Melanotrope cells as a model to understand the (patho)physiological regulation of hormone secretion

Regulation of hormone secretion is a complex process that comprises the sequential participation of numerous subcellular mechanisms. Hormone secretion is dictated by extracellular stimuli that are transduced intracellularly into activation/deactivation of different mechanisms, such as hormone expression, processing and exocytosis, which will ultimately determine the precise availability of hormone to be secreted. Malfunction in any of these steps may result in deficient or excessive hormone release and the subsequent appearance of endocrine disorders. Given the complexity of this system, it is difficult to find appropriate cellular models wherein to investigate the multiple components of the secretory process in a physiologically relevant, experimentally manipulable setting. In this review, we present recent evidence on the use of the intermediate lobe (IL) of the pituitary as a powerful tool to understand different aspects of the regulated secretory pathway. IL is composed of a single endocrine cell type, alpha-melanocyte stimulating hormone (alpha-MSH)-producing melanotropes, a fact that greatly facilitates its study. Furthermore, melanotropes can be separated using classic cell separation techniques into two cell subtypes showing opposite morphophysiological phenotypes of hypo- and hypersecretory cells. Comparison of their gene expression fingerprints has unveiled the existence of certain genes preferentially expressed in each melanotrope subtype. Because of their direct participation in the secretory pathway, we postulate that characterization of these gene products in an endocrine cell type may represent novel and useful markers for reliably determining the general secretory status in an endocrine gland, as well as a valuable new tool to further investigate this complex process.

The regulation of alpha-MSH release by GABA is mediated by a chloride-dependent [Ca2+]c increase in frog melanotrope cells

In frog melanotrope cells, gamma-aminobutyric acid (GABA) induces a biphasic effect, i.e. a transient stimulation followed by a more sustained inhibition of alpha-MSH release, and both phases of the GABA effect are mediated by GABAA receptors. We have previously shown that the stimulatory phase evoked by GABAA receptor agonists can be accounted for by calcium entry. In the present study, we have investigated the involvement of the chloride flux on GABA-induced [Ca2+]c increase and alpha-MSH release. We show that GABA evokes a concentration-dependent [Ca2+]c rise through specific activation of the GABAA receptor. The GABA-induced [Ca2+]c increase results from opening of voltage-activated L- and N-type calcium channels, and sodium channels. Variations of the extracellular Cl- concentration revealed that GABA-induced [Ca2+]c rise and alpha-MSH release both depend on the Cl- flux direction and driving force. These observations suggest for the first time that GABA-gated Cl- efflux provokes an increase in [Ca2+]c increase that is responsible for hormone secretion.

Localization of the mRNA encoding prolyl endopeptidase in the rat brain and pituitary

Prolyl endopeptidase (EC 3.4.21.26, PEP), a serine protease that hydrolyzes peptides at the carboxyl side of proline residues, is involved in the breakdown of several proline-containing neuropeptides and, thus, may contribute to the regulation of behavioral activities. In this study, the distribution of PEP mRNA was investigated in the central nervous system and pituitary of rat by means of quantitative reverse transcriptase-polymerase chain reaction analysis and in situ hybridization histochemistry. High densities of PEP transcripts were found in cerebellar Purkinje and granule cells, within most hypothalamic nuclei, in pyramidal neurons of the Ammon's horn, in granule cells of the dentate gyrus, and within the basolateral complex of the amygdala. Moderate levels of PEP mRNA were observed in layers 3-5 of the cerebral cortex, the anterior thalamic group, the septal region, the substantia nigra, the magnocellular neurons of the red nucleus, and the motor nuclei of the cranial nerves. Low concentrations of PEP mRNA were detected in the deep mesencephalic nuclei, the reticular formation, the pretectum, and the tectum. A high density of PEP mRNA was found in the intermediate and the anterior lobes of the pituitary, while the neural lobe was devoid of labeling. In several brain regions, the distribution pattern of PEP mRNA overlapped that of various neuropeptide receptors, suggesting that PEP is actually involved in the inactivation of regulatory neuropeptides.

Comparison of long-term perifused pars intermedia of Anolis carolinensis, Rana pipiens and Hyla chrysoscelis: their responses to dopamine

Pituitary pars nervosa-pars intermedia of Anolis carolinensis, Rana pipiens and Hyla crysoscelis were perifused with synthetic medium 199 for up to 35 h. The pre- and post-perifused tissues were examined by electron microscopy. No neuronal endings were found in Anolis tissue, but both Rana and Hyla had occasional synaptic end bulbs, which remained visible in the post-perifused tissue, although the synaptic vesicles appeared to cluster in the center of the end bulbs. Exposure to dopamine HCl from 10(-8) to 10(-5) M had little effect on Anolis pituitary but inhibited Rana and Hyla pituitaries from releasing skin-darkening substances. The skin-darkening substances, presumably derivatives of the proopiomelanocortin molecule, were assayed on Anolis skin. No dose-dependent responses to dopamine were seen at the concentrations used. We saw the possibility of a short-loop feedback.

Neurotensin modulates the amplitude and frequency of voltage-activated Ca2+ currents in frog pituitary melanotrophs: implication of the inositol triphosphate/protein kinase C pathway

Many excitatory neurotransmitters and neuropeptides regulate the activity of neuronal and endocrine cells by modulating voltage-operated Ca2+ channels. Paradoxically, however, excitatory neuromediators that provoke mobilization of intracellular calcium from inositol trisphosphate (IP3)-sensitive stores usually inhibit voltage-gated Ca2+ currents. We have recently demonstrated that neurotensin (NT) stimulates the electrical and secretory activities of frog pituitary melanotrophs, and increases intracellular calcium concentration in these cells. In the present study, we have investigated the effects of NT on Ca2+ currents in cultured frog melanotrophs by using the perforated patch-clamp technique. Frog neurotensin (f NT) reduced the amplitude and facilitated the inactivation of both L- and N-type Ca2+ currents. Application of the membrane-permeant Ca2+ chelator BAPTA-AM, the sarcoendoplasmic reticulum Ca2+-ATPase inhibitor thapsigargin, or the IP3 receptor antagonist 2-APB suppressed the reduction of Ca2+ currents induced by f NT. Incubation of melanotrophs with the diacylglycerol analogue PMA, which causes desensitization of protein kinase C (PKC), or with the PKC inhibitors chelerythrine and calphostin C, reduced the inhibitory effect of f NT. The NT-induced action potential waveforms, applied as voltage-clamp commands, decreased the amplitude of Ca2+ currents, and enhanced Ca2+ influx by increasing the Ca2+ spike frequency. Altogether, these data indicate that the inhibitory effect of f NT on Ca2+ currents results from activation of the IP3/PKC pathway. The observation that NT controls Ca2+ signalling through both amplitude and frequency modulations of Ca2+ currents suggests that NT might induce spacial and temporal changes of intracellular Ca2+ concentration leading to stimulation of exocytosis.

Neuropeptide Y inhibits spontaneous alpha-melanocyte-stimulating hormone (alpha-MSH) release via a Y(5) receptor and suppresses thyrotropin-releasing hormone-induced alpha-MSH secretion via a Y(1) receptor in frog melanotrope cells

In amphibians, the secretion of alpha-MSH by melanotrope cells is stimulated by TRH and inhibited by NPY. We have previously shown that NPY abrogates the stimulatory effect of TRH on alpha-MSH secretion. The aim of the present study was to characterize the receptor subtypes mediating the action of NPY and to investigate the intracellular mechanisms involved in the inhibitory effect of NPY on basal and TRH-induced alpha-MSH secretion. Y(1) and Y(5) receptor mRNAs were detected by RT-PCR and visualized by in situ hybridization histochemistry in the intermediate lobe of the pituitary. Various NPY analogs inhibited in a dose-dependent manner the spontaneous secretion of alpha-MSH from perifused frog neurointermediate lobes with the following order of potency porcine peptide YY (pPYY) > frog NPY (fNPY) > porcine NPY (pNPY)-2-36) > pNPY-(13-36) > [D-Trp(32)]pNPY > [Leu(31),Pro(34)]pNPY. The stimulatory effect of TRH (10(-8)6 M) on alpha-MSH release was inhibited by fNPY, pPYY, and [Leu(31),Pro(34)]pNPY, but not by pNPY-(13-36) and [D-Trp(32)]pNPY. These data indicate that the inhibitory effect of fNPY on spontaneous alpha-MSH release is preferentially mediated through Y(5) receptors, whereas the suppression of TRH-induced alpha-MSH secretion by fNPY probably involves Y(1) receptors. Pretreatment of neurointermediate lobes with pertussis toxin (PTX; 1 microg/ml; 12 h) did not abolish the inhibitory effect of fNPY on cAMP formation and spontaneous alpha-MSH release, but restored the stimulatory effect of TRH on alpha-MSH secretion, indicating that the adenylyl cyclase pathway is not involved in the action of fNPY on TRH-evoked alpha-MSH secretion. In the majority of melanotrope cells, TRH induces a sustained and biphasic increase in cytosolic Ca(2+) concentration. Preincubation of cultured cells with fNPY (10(-7) M) or omega-conotoxin GVIA (10(-7) M) suppressed the plateau phase of the Ca(2+) response induced by TRH. However, although fNPY abrogated TRH-evoked alpha-MSH secretion, omega-conotoxin did not, showing dissociation between the cytosolic Ca(2+) concentration increase and the secretory response. Collectively, these data indicate that in frog melanotrope cells NPY inhibits spontaneous alpha-MSH release and cAMP formation through activation of a Y(5) receptor coupled to PTX- insensitive G protein, whereas NPY suppresses the stimulatory effect of TRH on alpha-MSH secretion through a Y(1) receptor coupled to a PTX-sensitive G protein-coupled receptor.

Potassium-induced secretion of melanocyte-stimulating hormone from the melanotrophs of the neurointermediate lobe of the lizard Anolis carolinensis

Secretion of melanocyte-stimulating hormone (MSH) from the melanotrophs of the neurointermediate lobe (NIL) of the lizard Anolis carolinensis was studied to investigate the role of membrane potential and extracellular calcium ions in the control of secretion in this species. MSH secretion was monitored from perifused NILs which had been in organ culture for 7-14 days prior to experiment to allow the nerve terminals present in the tissue to degenerate. Elevation of the K(+) concentration in the perifusate induced a marked increase in MSH secretion. Perifusion of the cultured NILs with a nominally Ca-free solution did not reduce basal MSH secretion but blocked K-induced secretion. Moreover, nimodipine, an antagonist of voltage-gated Ca(2+) channels, inhibited K-induced secretion, whereas BAY K 8644, a Ca(2+) channel agonist, potentiated it; but neither drug affected basal secretion. High ¿K(+) perifusate also stimulated MSH secretion from freshly excised (acute) NILs. Furthermore, in these preparations nominally Ca-free solution reduced basal secretion by 40% in addition to blocking K-induced secretion. As in the cultured NILs, nimodipine blocked and BAY K 8644 potentiated K-induced secretion in the acute NILs while not affecting basal secretion. The results from the cultured NILs are consistent with the hypothesis that stimulated MSH secretion from the melanotrophs of the anole is dependent upon Ca(2+) influx through voltage-gated calcium channels. The qualitatively similar results obtained from the acute NILs in response to high ¿K(+), nimodipine, and BAY K 8644 suggest that, for the most part, what is being observed are the direct effects of these substances on the melanotrophs. Basal secretion of MSH in cultured NILs is significantly less than that in the acute preparations. The calcium-sensitive fraction of basal secretion present in acute NILs may require the presence of nerve terminals.

The pituitary-skin connection in amphibians. Reciprocal regulation of melanotrope cells and dermal melanocytes

In amphibians, alpha-MSH secreted by the pars intermedia of the pituitary plays a pivotal role in the process of skin color adaptation. Reciprocally, the skin of amphibians contains a number of regulatory peptides, some of which have been found to regulate the activity of pituitary melanotrope cells. In particular, the skin of certain species of amphibians harbours considerable amounts of thyrotropin-releasing hormone, a highly potent stimulator of alpha-MSH release. Recently, we have isolated and sequenced from the skin of the frog Phyllomedusa bicolor--a novel peptide named skin peptide tyrosine tyrosine (SPYY), which exhibits 94% similarity with PYY from the frog Rana ridibunda. For concentrations ranging from 5 x 10(-10) to 10(-7) M, SPYY induces a dose-related inhibition of alpha-MSH secretion. At a dose of 10(-7) M, SPYY totally abolished alpha-MSH release. These data strongly suggest the existence of a regulatory loop between the pars intermedia of the pituitary and the skin in amphibians.

Differential effects of dopamine on two frog melanotrope cell subpopulations

The frog intermediate lobe consists of a single endocrine cell type, the melanotrope cells, which are under the tonic inhibitory control of dopamine. Separation of dispersed pars intermedia cells in a Percoll density gradient has revealed the existence of two melanotrope cell subpopulations, referred to as high-density (HD) and low-density (LD) cells. The aim of the present study was to investigate the effects of dopamine on each of these melanotrope cell subsets. Increasing doses of dopamine, ranging from 10(-9)-10(-6) M, inhibited the release of alpha-melanocyte-stimulating hormone (alpha-MSH) in LD (but not in HD) melanotrope cells. In addition, dopamine provoked a significant reduction of the rate of acetylation of alpha-MSH in LD cells but not in HD cells. Similarly, dopamine significantly decreased the accumulation of POMC messenger RNA in LD cells, whereas it did not affect POMC gene expression in the HD melanotrope subset. On the other hand, microfluorimetric studies revealed that dopamine induced a significant reduction of KCl-stimulated cytosolic free calcium concentration in both LD and HD cells. The present study provides additional evidence for functional heterogeneity of melanotrope cells in the frog pars intermedia. Because dopamine plays a pivotal role in the regulation of alpha-MSH secretion, these data suggest the involvement of cell heterogeneity in the physiological process of background color adaptation in amphibians.

Characterization of the cDNA encoding the prohormone convertase PC2 and localization of the mRNA in the brain of the frog Rana ridibunda

A number of precursors for neuropeptides have recently been cloned in amphibians, but little is known concerning the endoproteases responsible for the processing of these precursors. Here we report on the molecular cloning of the cDNA encoding the proprotein convertase PC2 and the distribution of the corresponding mRNA in the European green frog Rana ridibunda. The full cDNA structure (2125 bp) was obtained from the analysis of the PCR products combined with the sequence from a clone isolated from a frog pituitary cDNA library. The deduced amino acid sequence revealed that frog PC2 comprises 636 amino acid residues including a 22-residue signal peptide. RT-PCR analysis showed that PC2 is expressed not only in the brain and pituitary but also in various peripheral organs including the pancreas, stomach, intestine, liver, kidney and testis. In situ hybridization histochemistry revealed that, in the central nervous system, PC2 mRNA is widely distributed, the highest concentrations being found in the pallium, the anterior preoptic area, the hypothalamus and the medial amygdala. High levels of PC2 mRNA were also detected in the intermediate lobe of the pituitary. The overall distribution of PC2 mRNA in the frog brain is consistent with its involvement in the processing of a number of neuropeptide and hormone precursors.

Isolation, primary structure, and effects on alpha-melanocyte-stimulating hormone release of frog neurotensin

Neurotensin (NT) was isolated in pure form from the small intestine of the European green frog, Rana ridibunda, and its primary structure was established as pGlu-Ala-His-Ile-Ser-Lys-Ala-Arg-Arg-Pro-Tyr-Ile-Leu. This sequence contains five amino acid substitutions (Leu2-->Ala, Tyr3-->His, Glu4-->Ile, Asn5-->Ser, and Pro7-->Ala) compared with human NT. A peptide with identical chromatographic properties was identified in an extract of frog brain. Synthetic frog NT produced a concentration-dependent increase in alphaMSH release from perifused frog pars intermedia cells, with an ED50 of 5 x 10(-9) M. A maximum response (276.3 +/- 45.5% above basal release) was produced by a 10(-8) M concentration. Repeated administration of NT to melanotrope cells revealed the occurrence of a rapid and pronounced desensitization mechanism. The data are consistent with a possible role for the peptide as a hypophysiotropic factor in amphibians.

Inhibitory effect of adenosine on electrical activity of frog melanotrophs mediated through A1 purinergic receptors

1. The effects of adenosine were studied in cultured frog melanotrophs by the patch-clamp technique. 2. In cell-attached experiments, most cells responded to adenosine (50 microM) by a reversible inhibition of action current discharges without any apparent desensitization. 3. In whole-cell experiments, adenosine provoked a hyperpolarization accompanied by a depression of spontaneous action potentials and a decrease in membrane resistance. When adenosine was repeatedly applied, tachyphylaxis was observed. Addition of GTP (100 microM) in the intracellular solution augmented the percentage of cells hyperpolarized by adenosine, and the duration and amplitude of the hyperpolarization, and prevented the tachyphylaxis. 4. Pretreatment with pertussis toxin (1 microgram ml-1) blocked adenosine-induced inhibition. 5. In cells dialysed with the non-hydrolysable GTP analogue GTP gamma S (100 microM), adenosine caused a sustained, strong hyperpolarization and an irreversible inhibition of spikes. 6. The effect of adenosine was mimicked by the A1 receptor agonist R-PIA (R-N6-phenylisopropyl-adenosine; 50 microM) and blocked by the A1 receptor antagonist CPDPX (8-cyclopentyl-1,3-dipropylxanthine, 50 microM). The A2 receptor antagonist CGS15943 (9-chloro-2-(2-furanyl)-5,6-dihydro-1,2,4-triazolo[1,5-c] quinazoline-5-imine; 50 microM) did not affect the adenosine-induced response. 7. The results suggest that, in frog melanotrophs, adenosine exerts a direct hyperpolarizing effect accompanied by blockage of spontaneous action potentials. The effect of adenosine is mediated through A1 receptors coupled to a Gi/o protein.