Single amino acid substitutions and deletions that alter the G protein coupling properties of the V2 vasopressin receptor identified in yeast by receptor random mutagenesis

To facilitate structure-function relationship studies of the V2 vasopressin receptor, a prototypical G(s)-coupled receptor, we generated V2 receptor-expressing yeast strains (Saccharomyces cerevisiae) that required arginine vasopressin-dependent receptor/G protein coupling for cell growth. V2 receptors heterologously expressed in yeast were unable to productively interact with the endogenous yeast G protein alpha subunit, Gpa1p, or a mutant Gpa1p subunit containing the C-terminal G alpha(q) sequence (Gq5). In contrast, the V2 receptor efficiently coupled to a Gpa1p/G alpha(s) hybrid subunit containing the C-terminal G alpha(s) sequence (Gs5), indicating that the V2 receptor retained proper G protein coupling selectivity in yeast. To gain insight into the molecular basis underlying the selectivity of V2 receptor/G protein interactions, we used receptor saturation random mutagenesis to generate a yeast library expressing mutant V2 receptors containing mutations within the second intracellular loop. A subsequent yeast genetic screen of about 30,000 mutant receptors yielded four mutant receptors that, in contrast to the wild-type receptor, showed substantial coupling to Gq5. Functional analysis of these mutant receptors, followed by more detailed site-directed mutagenesis studies, indicated that single amino acid substitutions at position Met(145) in the central portion of the second intracellular loop of the V2 receptor had pronounced effects on receptor/G protein coupling selectivity. We also observed that deletion of single amino acids N-terminal of Met(145) led to misfolded receptor proteins, whereas single amino acid deletions C-terminal of Met(145) had no effect on V2 receptor function. These findings highlight the usefulness of combining receptor random mutagenesis and yeast expression technology to study mechanisms governing receptor/G protein coupling selectivity and receptor folding.

Binding properties of a selective tritiated vasopressin V2 receptor antagonist, [H]-SR 121463

BACKGROUND

[3H]-SR 121463 is the first radiolabeled selective nonpeptide vasopressin V2 receptor antagonist ligand that has been reported to date. In the present work, we studied the binding properties of [3H]-SR 121463 for renal V2 receptors from animal and human origins.

METHODS

Binding studies were performed with [3H]-SR 121463 in Chinese hamster ovary (CHO) cells transfected with the human V2 receptor and in various kidney preparations expressing the native V2 receptors (rat, rabbit, dog, pig, monkey, and human). Autoradiographies were performed in rat and human kidney sections.

RESULTS

[3H]-SR 121463 binding to CHO cells stably transfected with the cloned human renal V2 receptor was specific, highly stable, time dependent, saturable, and reversible. A single population of high-affinity binding sites was identified (Kd = 0.94 +/- 0.34 nmol/L, Bmax = 9876 +/- 317 fmol/mg protein). Of note, [3H]-SR 121463 revealed a higher number (about 40%) of V2 sites than [3H]-AVP in the same preparation. Displacement of [3H]-SR 121463 binding by reference peptide and nonpeptide vasopressin/oxytocin compounds exhibited a typical AVP V2 profile. [3H]-SR 121463 also displayed a high affinity for native V2 receptors in several kidney preparations from rat, pig, dog, rabbit, bovine, monkey, and human. The autoradiographic experiments using rat and human kidney sections showed intense labeling in the medullopapillary region and lower intensity in the cortex, consistent with a main localization of V2 receptors on collecting tubules.

CONCLUSION

[3H]-SR 121463 is a useful ligand for the specific labeling of animal and human V2 receptors and could be a suitable probe for the search and in situ localization of V2 sites.

Characterization of NPY receptors controlling lipolysis and leptin secretion in human adipocytes

In order to characterize neuropeptide Y (NPY) receptors present in human adipocytes, we used selective ligands together with specific molecular probes able to recognize the different NPY receptor subtypes. RT-PCR experiments revealed the presence of Y(1) receptor transcripts with Y(4) and Y(5) and absence of Y(2) signals. Binding studies, using selective radioiodinated ligands, detected a high number (B(max)=497+/-124 fmol/mg protein) of a high affinity binding site only with [(125)I]peptide YY (PYY) and [(125)I](Leu(31), Pro(34))PYY. These sites exhibited a typical Y(1) profile as indicated by the rank order of affinity of NPY analogs and the high affinity of two selective NPY receptor antagonists, SR120819A and BIBP3226. In [(35)S]GTPgammaS binding experiments, PYY activation was totally inhibited by SR120819A and BIBP3226. Both compounds antagonized, with similar efficiency, the antilipolytic effect exerted by NPY in isolated adipocytes. Finally, PYY and Y(1) ligands enhanced adipocyte leptin secretion, an effect totally prevented by SR120819A. Thus, highly expressed in human adipocytes, the Y(1) receptor sustains the strong antilipolytic effect of NPY and exerts a positive action on leptin secretion.

Vasopressin receptors in human adrenal medulla and pheochromocytoma

The nature of vasopressin (VP) receptors present in normal and tumoral human adrenal was investigated using various experimental approaches. Specific VP-binding sites were detected by autoradiography using [3H]arginine VP as a radioligand in adrenal cortex and medulla. The V1a receptor subtype was expressed in the two parts of the gland, as shown by pharmacological studies and RT-PCR experiments. By contrast, the V1b receptor subtype was only expressed in medullary chromaffin cells. This was confirmed by the characterization of V1b transcripts detected in adrenal medulla tissues. In pheochromocytoma, we also detected functional V1b receptors. These receptors triggered intracellular calcium mobilization from intracellular pools and were involved in catecholamine secretion. Binding experiments performed on pheochromocytoma plasma membrane preparations also revealed V1a vasopressin-binding sites, whose roles and cellular localization have not yet been determined. RT-PCR experiments confirmed these data; 100% and 80% of the five tumors tested exhibited V1a and V1b transcripts, respectively. Perifusion experiments also demonstrated that some pheochromocytomas may secrete large amounts of VP. Our findings imply that VP locally secreted by human adrenal medulla may regulate adrenal function by acting on V1a or V1b receptors. More interestingly, we demonstrate that one pheochromocytoma oversecretes VP. In this particular case, this may contribute to the increase in blood pressure observed.

Binding of the non-peptide vasopressin V1a receptor antagonist SR-49059 in the rat brain: an in vitro and in vivo autoradiographic study

A potent non-peptide vasopressin (AVP) antagonist, SR-49059, displaying high stability and selective affinity for the V1a AVP receptor subtype, has recently been described. The objective of this study was to assess the binding properties and the penetrability of this compound in the rat brain. Both in vitro and in vivo binding autoradiography experiments were performed. In all studies, the liver was used as a reference V1a tissue. In vitro labelling of rat brain sections with [3H]SR-49059 was similar to that previously detected with [3H]AVP, which confirms that the majority of central AVP binding sites are V1a sites similar to peripheral V1a receptors. As expected, intense specific labelling occurred mainly in the lateral septum, the fundus striatum, the hypothalamic stigmoid nucleus and the area postrema-nucleus of the solitary tract complex. In vivo binding autoradiography showed that [3H]SR-49059 injected intravenously did not enter the brain parenchyma. Specific labelling was however clearly detectable in brain regions with permeable hematoencephalic barrier, the choroid plexus and other circumventricular organs expressing V1a receptors, namely the subfornical organ, the pineal gland and the area postrema. The specificity of [3H]SR-49059 binding in the latter structures was confirmed by the fact that labelling was prevented by pretreatment of animals with high doses of nonradioactive SR-49059. In conclusion, our study shows that [3H]SR-49059 is a suitable probe to investigate V1a receptors in the rat brain. We also demonstrate that although this compound is not able to enter the brain tissue from the peripheral circulation, it does bind specifically to regions devoid of blood-brain barrier and known to be involved in autonomic regulations.

V2 receptor antagonism of DDAVP-induced release of hemostasis factors in conscious dogs

The synthetic arginine vasopressin (AVP) analog 1-desamino-8-D-arginine vasopressin (DDAVP) is used in a variety of hemorrhagic disorders. The present experiments were designed to further characterize the mechanism of DDAVP-induced release of hemostasis factors. The [3H]AVP-labeled AVP receptor in canine renomedullary membranes exhibited an AVP V2 profile because the V2 receptor agonist DDAVP displayed similar subnanomolar affinities as the natural hormone AVP, whereas the two selective V1a compounds SR 49059 and d(CH2)5Tyr(Me)AVP as well as the selective V1b agonist D-Pal and oxytocin were much less potent. The rank order of the binding affinities of three V2 receptor antagonists was SR 121463 (a newly described selective V2 receptor antagonist) > OPC 31260 >> d(CH2)5D-Ile2,Ile4AVP. In conscious dogs, DDAVP (0.1-1 microg/kg I.V.) caused a dose-related increase (maximum, 43-52% at 30 min) in plasma levels of factor VIII (FVIII), von Willebrand factor (vWF) and tissue-type plasminogen activator (t-PA), but not in levels of plasminogen activator inhibitor-1. A DDAVP-induced hemostasis factor release was also observed in bilaterally nephrectomized dogs. Pretreatment with SR 121463 inhibited DDAVP-induced (1 microg/kg I.V.) increases in FVIII, vWF and t-PA plasma levels in a dose-dependent manner (ID50 = 14.0 +/- 4.0, 12.4 +/- 3.0 and 16.7 +/- 1.0 microg/kg I.V., respectively). OPC 31260 (300 microg/kg I.V.) revealed a lower activity than SR 121463, and d(CH2)5[D-Ile2,Ile4]AVP (30 microg/kg I.V.) was without effect on the DDAVP response. Pretreatment with SR 49059 (1 mg/kg I.V.) and SR 27417 (a platelet-activating factor receptor antagonist) (1 mg/kg I.V.) had no effect on the DDAVP-induced (1 microg/kg I.V.) increases in FVIII, vWF and t-PA plasma levels. The present results, therefore, strongly suggest that the effect of DDAVP on hemostasis factors occurs via a specific interaction with extrarenal V2 receptors.

Characterization and localization of leptin receptors in the rat kidney

Characterization and localization of leptin binding sites were investigated in rat kidneys using [125I]leptin as a ligand. [125I]Leptin specific binding was found in high amounts in rat renomedullary membranes. This binding was specific, saturable, time-dependent (K(obs) = 0.055 +/- 0.008 min(-1)) and the dissociation of receptor-bound ligand was slowly reversible (K(-1) = 0.048 +/- 0.013 min(-1)). From saturation experiments, a single class of high-affinity binding sites for leptin was identified with an apparent K(d) of 0.57 +/- 0.14 nM and a B(max) of 45 +/- 10 fmol/mg protein. [125I]Leptin binding was inhibited in a dose-dependent manner by cold leptin and was highly selective since not displaceable by a number of other hormones or peptides. Autoradiographic experiments performed on adult rat kidney sections showed the intense presence of [125I]leptin receptors only in specific areas of the renal inner medulla and also consistent labeling associated with vascular structures in the corticomedullary region. The study of the postnatal developmental expression of leptin receptors in the kidney showed very low expression during the early postnatal period (8-21 days). Full expression of leptin sites was achieved at about 30 days and remained stable throughout adulthood (60 days and upwards). Moreover, in vivo administration of leptin (0.5 mg/kg i.p.) induced a significant and rapid diuretic effect in normally hydrated conscious rats. Thus, these data constitute the first characterization and mapping of [125I]leptin specific binding sites in the rat kidney and raise the possibility of a renal control by leptin.

Characterization of SR 121463A, a highly potent and selective, orally active vasopressin V2 receptor antagonist

SR 121463A, a potent and selective, orally active, nonpeptide vasopressin V2 receptor antagonist, has been characterized in several in vitro and in vivo models. This compound displayed highly competitive and selective affinity for V2 receptors in rat, bovine and human kidney (0.6 < or = Ki [nM] < or = 4.1). In this latter preparation, SR 121463A potently antagonized arginine vasopressin (AVP)-stimulated adenylyl cyclase activity (Ki = 0.26+/-0.04 nM) without any intrinsic agonistic effect. In autoradiographic experiments performed in rat kidney sections, SR 121463A displaced [3H]AVP labeling especially in the medullo-papillary region and confirmed that it is a suitable tool for mapping V2 receptors. In comparison, the nonpeptide V2 antagonist, OPC-31260, showed much lower affinity for animal and human renal V2 receptors and lower efficacy to inhibit vasopressin-stimulated adenylyl cyclase (Ki in the 10 nanomolar range). Moreover, OPC-31260 exhibited a poor V2 selectivity profile and can be considered as a V2/V1a ligand. In normally hydrated conscious rats, SR 121463A induced powerful aquaresis after intravenous (0.003-0.3 mg/kg) or oral (0.03-10 mg/kg) administration. The effect was dose-dependent and lasted about 6 hours at the dose of 3 mg/kg p.o. OPC-31260 had a similar aquaretic profile but with markedly lower oral efficacy. The action of SR 121463A was purely aquaretic with no changes in urine Na+ and K+ excretions unlike that of known diuretic agents such as furosemide or hydrochlorothiazide. In addition, no antidiuretic properties have been detected with SR 121463A in vasopressin-deficient Brattleboro rats. Thus, SR 121463A is the most potent and selective, orally active V2 antagonist yet described and could be a powerful tool for exploring V2 receptors and the therapeutical usefulness of V2 blocker aquaretic agents in water-retaining diseases.

Growth inhibitory properties of endothelin-1 in activated human hepatic stellate cells: a cyclic adenosine monophosphate-mediated pathway. Inhibition of both extracellular signal-regulated kinase and c-Jun kinase and upregulation of endothelin B receptors

During chronic liver diseases, hepatic stellate cells (HSC) acquire an activated myofibroblast-like phenotype, proliferate, and synthetize fibrosis components. We have shown that endothelin-1 (ET-1) inhibits the proliferation of activated human HSC via endothelin B (ETB) receptors. We now investigate the transduction pathway involved in the growth inhibitory effect of ET-1 in activated HSC. Endothelin-1 and the ETB receptor agonist, sarafotoxin-S6C, increased synthesis of PGI2 and PGE2, leading to elevation of cAMP. The cyclooxygenase inhibitor ibuprofen and the adenylyl cyclase inhibitor SQ22536 both blunted the growth inhibitory effect of ET-1. Analysis of early steps associated with growth inhibition indicated that: (a) similar to ET-1, forskolin decreased c-jun mRNA induction without affecting c-fos and krox 24 mRNA expression; (b) ET-1, sarafotoxin-S6C, as well as forskolin, reduced activation of both c-Jun kinase and extracellular signal-regulated kinase. Finally, forskolin, PGI2, and PGE2 raised by fivefold the number of ET binding sites after 6 h, and increased the proportion of ETB receptors from 50% in control cells to 80% in treated cells. In conclusion, ET-1 inhibits proliferation of activated HSC via ETB receptors, through a prostaglandin/cAMP pathway that leads to inhibition of both extracellular signal-regulated kinase and c-Jun kinase activities. Upregulation of ETB receptors by prostaglandin/cAMP raises the possibility of a positive feedback loop that would amplify the growth inhibitory response. These results suggest that ET-1 and agents that increase cAMP might be of interest to limit proliferation of activated HSC during chronic liver diseases.

Autoradiographic localization of vasopressin V1a receptors in the rat kidney using [3H]-SR 49059

Localization and characterization of binding sites of the selective non-peptide vasopressin receptor V1a ligand, [3H]-SR 49059, were investigated in the adult rat kidney by quantitative autoradiography using a fast-detecting radioluminographic phosphor-imaging plate system. [3H]-SR 49059, like the other V1a ligands used, showed a total absence of binding in the papilla, discrete and sparse labeling in the cortex and maximal binding in the outer part of the inner medulla. This labeling seemed to be mainly associated with medullary interstitial cells and vascular elements of the vasa recta. Conversely, [3H]-AVP intensely labeled the V2-enriched medulla-papillary portion of the kidney and, to a lesser extent, the cortical structures. [3H]-SR 49059 binding, quantified in the outer part of the inner medulla in rat kidney sections, was time-dependent, reversible, saturable and a single class of high affinity binding sites (Kd = 1.48 +/- 0.16 nM) was identified. The relative potencies of the reference peptide and non-peptide compounds to inhibit [3H]-SR 49059 binding confirm the V1a nature of the site and the stereospecificity of this binding. Thus, [3H]-SR 49059 allows the mapping and characterization of the V1a receptor population present in the rat kidney. The stability and the highly selective affinity of this non-peptide ligand for rat and human V1a receptors make it a suitable probe for the localization of V1a receptors in organs expressing heterogeneous populations of receptors.

Growth inhibitory properties of endothelin-1 in human hepatic myofibroblastic Ito cells. An endothelin B receptor-mediated pathway

Ito cells play a pivotal role in the development of liver fibrosis associated with chronic liver diseases. During this process, Ito cells acquire myofibroblastic features, proliferate, and synthesize fibrosis components. Considering the reported mitogenic properties of endothelin-1 (ET-1), we investigated its effects on the proliferation of human Ito cells in their myofibroblastic phenotype. Both ET receptor A (ETA: 20%) and ET receptor B (ETB: 80%) binding sites were identified, using a selective ETA antagonist, BQ 123, and a selective ETB agonist, sarafotoxin S6C (SRTX-C). ET-1 did not stimulate proliferation of myofibroblastic Ito cells. In contrast, ET-1 inhibited by 60% DNA synthesis and proliferation of cells stimulated with either human serum or platelet-derived growth factor -BB (PDGF-BB). PD 142893, a nonselective ETA/ETB antagonist totally blunted this effect. SRTX-C was as potent as ET-1, while BQ 123 did not affect ET-1-induced growth inhibition. Analysis of the intermediate steps leading to growth-inhibition by ET-1 revealed that activation of mitogen-activated protein kinase by serum or PDGF-BB was decreased by 50% in the presence of SRTX-C. In serum-stimulated cells, SRTX-C reduced c-jun mRNA expression by 50% whereas c-fos or krox 24 mRNA expression were not affected. We conclude that ET-1 binding to ETB receptors causes a potent growth inhibition of human myofibroblastic Ito cells, which suggests that this peptide could play a key role in the negative control of liver fibrogenesis. Our results also point out that, in addition to its well known promitogenic effects, ET-1 may also exert negative control of growth on specific cells.

Effect of SR-49059, a vasopressin V1a antagonist, on human vascular smooth muscle cells

The effects of SR-49059, a new nonpeptide and selective arginine vasopressin (AVP) V1a antagonist, were investigated in binding and functional studies on cultured human aortic vascular smooth muscle cells (VSMC). Characterization of human vascular V1a receptors, using a specific V1a radioiodinated ligand, showed that [125I]-linear AVP antagonist binding to human VSMC membranes was time dependent, reversible, and saturable. A single population of high-affinity binding sites (apparent equilibrium dissociation constant = 15 +/- 6 pM; maximum binding density = 36 +/- 5 fmol/mg protein, i.e., approximately 3,000 sites/cell) with the expected V1a profile was identified. Exposure of these cells to AVP dose-dependently produced cytosolic free [Ca2+] increase [AVP concentration required to obtain a half-maximal response (EC50) = 23 +/- 9 nM] and proliferation (EC50 = 3.2 +/- 0.5 nM). SR-49059 strongly and stereospecifically inhibited [125I]-linear AVP antagonist binding to VSMC V1a receptors [inhibition constant (Ki) = 1.4 +/- 0.3 nM], AVP-evoked Ca2+ increase [concentration of inhibitor required to obtain 50% inhibition of specific binding (IC50) = 0.41 +/- 0.06 nM], and the mitogenic effects induced by 100 nM AVP (IC50 = 0.83 +/- 0.04 nM). OPC-21268, another nonpeptide V1a antagonist, was more than two orders of magnitude less potent than SR-49059 in these models. However, the consistent affinity (Ki = 138 +/- 21 nM) and activity found with OPC-21268 on human VSMC in comparison with the inactivity already observed for other human V1a receptors (liver, platelets, adrenals, and uterus) strongly suggested the existence of human AVP V1a-receptor subtypes.(ABSTRACT TRUNCATED AT 250 WORDS)

Binding of [3H] SR 49059, a potent nonpeptide vasopressin V1a antagonist, to rat and human liver membranes

The new potent and selective nonpeptide vasopressin V1a antagonist, SR 49059, was tritiated and used for the characterization of rat and human liver AVP V1a receptors. Binding of [3H] SR 49059 was time-dependent, reversible and saturable. A single class of high affinity binding sites was identified with Kd values of 0.63 +/- 0.13 and 2.95 +/- 0.64 nM, in rat and human liver membranes, respectively. The maximal binding capacity (Bmax) was about 7 times higher in rat than in human liver preparations. The relative potencies of several AVP/oxytocin agonists or antagonists to inhibit [3H] SR 49059 binding confirmed that this ligand labeled a homogeneous population of sites with the expected AVP V1a profile. Furthermore, [3H] SR 49059 or unlabeled SR 49059 displayed only slight species differences between rat and human V1a receptors, whereas OPC-21268, another nonpeptide V1a antagonist, exhibited a high species-related potency with more than 500 fold higher affinity for rat than for human liver V1a receptors. Thus, [3H] SR 49059 is the first nonpeptide AVP V1a ligand reported having highly specific activity, stability, specificity and affinity. This makes it a suitable probe for labeling AVP V1a receptors in rat and also in human tissues.

Effect of a new, potent, non-peptide V1a vasopressin antagonist, SR 49059, on the binding and the mitogenic activity of vasopressin on Swiss 3T3 cells

The effects of SR 49059, a new non-peptide, selective arginine vasopressin (AVP), V1a antagonist, were investigated both on AVP's receptors and on the mitogenic effects of AVP on Swiss 3T3 fibroblasts. We characterized the AVP V1a receptors on Swiss 3T3 cell membranes using the new highly specific AVP V1a radioiodinated ligand, 125I-linear AVP antagonist. Specific binding of the 125I-linear AVP antagonist was saturable, time-dependent and reversible. A single class of high affinity binding sites was identified with an apparent Kd of 40 +/- 20 pM and a Bmax of 63 +/- 20 fmol/mg protein. 125I-Linear AVP antagonist binding to its receptors was potently inhibited in a concentration-dependent manner by AVP, by the peptide V1a antagonist d(CH2)5Tyr(Me)AVP and by the synthetic V1a antagonist, SR 49059 (IC50 in the nanomolar range) while OPC-21268, another non-peptide compound, was about 100-fold less potent. Both DDAVP, a selective V2 agonist, and oxytocin exhibited low affinity (IC50 > 1 microM) in agreement with the AVP V1a nature of the site identified on Swiss 3T3 cells. In addition, the broad-spectrum antiproliferative agent [Arg6, D-Trp7,9, MePhe8] substance P (6-11), was also able to interact at 3T3 AVP V1a receptors (IC50 = 395 +/- 170 nM). The mitogenic effects of AVP on quiescent Swiss 3T3 cells, assessed through [3H]thymidine incorporation, were selectively, stereospecifically and strongly inhibited by SR 40959 (IC50 = 14 +/- 2 nM) while OPC-21268 was inactive up to 220 nM. SR 49059 was even about six times more efficient than d(CH2)5Tyr(Me)AVP in inhibiting AVP-induced DNA synthesis. Moreover, SR 49059 fully inhibited Swiss 3T3 fibroblast proliferation since it completely blocked AVP-stimulated 3T3 cell growth from the G1/G0 into the S/G2M phase, as evidenced by cell cycle analysis using a cytofluorometer. In summary, SR 49059, through direct interaction at AVP V1a receptors, exerts the most potent antiproliferative effect yet described for any V1a antagonist on Swiss 3T3 cells.

Endothelin action in rat liver. Receptors, free Ca2+ oscillations, and activation of glycogenolysis

High affinity binding sites for endothelin (ET) were identified on rat liver plasma membranes. Binding of 125I-ET-1 with its site was specific, saturable, and time dependent (kobs = 0.019 +/- 0.001 min-1), but dissociation of receptor-bound ligand was minimal. A single class of high affinity binding sites for 125I-ET-1 was identified with an apparent Kd of 32.4 +/- 9.8 pM and a Bmax of 1084 +/- 118 fmol/mg protein. ET-3 and big-ET-1 (1-38) (human) inhibited 125I-ET-1 binding with IC50 values of 1.85 +/- 1.03 nM and 43 +/- 6 nM, respectively. Aequorin measurements of cytosolic free Ca2+ in single, isolated rat hepatocytes showed that ET-1 at subnanomolar concentrations induced a series of repetitive, sustained Ca2+ transients. ET-1 had no effect on cAMP production. Finally, ET-1 caused a rapid and sustained stimulation of glycogenolysis in rat hepatocytes. A 1.8-fold maximal increase in glycogen phosphorylase alpha was observed at 1 pM ET-1, with an EC50 of 0.03 pM. Stimulation of the enzyme was specific for ET-1 since the order of potency of related peptides was similar to that in binding experiments (ET-1 greater than ET-3 greater than big ET-1). These data constitute the first demonstration of the presence of ET-1 binding sites in liver which is associated with a rise in cytosolic free Ca2+ and a potent glycogenolytic effect. We conclude that ET-1 behaves as a typical Ca2+ mobilizing hormone in liver.