Endothelin receptors: what's new and what do we need to know?

Abstract

Receptors are at the heart of how a molecule transmits a signal to a cell. Two receptor classes for endothelin (ET) are recognized, the ET_A and ET_B receptors. Intriguing questions have arisen in the field of ET receptor pharmacology, physiology, and function. For example, a host of pharmacological studies support the interaction of the ET_A and ET_B receptor in tissues (veins, arteries, bronchus, arterioles, esophagus), but yet few have been able to demonstrate direct ET_A/ET_B receptor interaction. Have we modeled this interaction wrong? Do we have a truly selective ET_A receptor agonist such that we could selectively stimulate this important receptor? What can we learn from the recent phylogenic studies of the ET receptor family? Have we adequately addressed the number of biological molecules with which ET can interact to exert a biological effect? Recent mass spectrometry studies in our laboratory suggest that ET-1 interacts with other hereto unrecognized proteins. Biased ligands (ligands at the same receptor that elicit distinct signaling responses) have been discovered for other receptors. Do these exist for ET receptors and can we take advantage of this possibility in drug design? These and other questions will be posed in this minireview on topics on ET receptors.

in 1988, the endothelial cell-derived peptide endothelin (ET) was discovered (121). In the 21 years since its discovery, ET-1 has been implicated in a wide number of physiological systems. Implication here means that synthetic machinery for ET peptides, ET peptides themselves, and ET receptors have been localized to tissues in these systems, as well as demonstration that ET-1 causes a physiological function within that system. An incomplete list includes the cardiovascular (11, 12, 16, 98), renal (35, 58, 87), central nervous (103), gastrointestinal (45), pituitary (61), adrenal (24, 43, 95, 96), nasal (83), peripheral nervous (20, 33, 82), immune (76), hepatic (70), genitourinary (15), and endocrine systems (75).

Given ET's enormously diverse presence in physiological systems, it comes as no surprise that ET-1 has been targeted as a pathophysiological culprit in multiple diseases and/or conditions. An incomplete list of these diseases includes hypertension [pregnancy, pulmonary, portopulmonary, and essential systemic (11, 31, 65, 77, 86, 113)], congestive heart failure (81), cancer (8, 37, 38, 85), diabetes (50, 55, 60), glaucoma (17), pain (41, 56), sexual dysfunction (112), fibrosis (108), renal failure (4, 63, 93), inflammation (4), and cerebral vasospasm (6).

If the involvement of ET-1 in disease is as pervasive as the many diseases listed above suggests, then development of compounds that block the effects of ET is imperative. This requires knowledge as to how ET exerts its biological activity. The classical view is that ET peptides stimulate their biological and pathophysiological effects through plasma membrane-bound ET receptors, the ET_A and ET_B receptors. This review is focused on a few current ideas/issues regarding ET receptor pharmacology and functioning in mammalian systems. Development of compounds for individual receptors as it pertains to therapy will not be discussed. There are several brilliant reviews that discuss, in far more detail than will be done here, ET peptides, classic knowledge of ET receptors (pharmacology, physiology and function), and models for study of ET in disease, and the reader is directed to some of them here (9, 22, 23, 26, 73, 88, 90, 91, 100, 104, 109, 110, 116).

CURRENTLY CLASSIFIED ET RECEPTORS AND THEIR PHARMACOLOGY

The International Union of Pharmacology (IUPHAR) officially defines receptor classification. This was done for ET receptors in a first report in 1994 (74) and updated in 2002 (21). At this time, only two classes are recognized as ET receptors: the ET_A receptor and the ET_B receptor (http://www.iuphar.org/).

ET_A Receptor

Originally cloned in 1990 (5), this protein is a 7α helical transmembrane domain molecule that is a Class A member of G protein-coupled receptors, possessing an extracellular amino terminus and intracellular carboxy terminus. Antagonists selective for this subtype of ET receptor include ZD4054, atrasentan, darusentan, macitentan, ambrisentan, and sitaxsentan. ET_A receptors are recognized by their relative insensitivity to stimulation by the agonist ET-3, whereas ET-2 and ET-1 have similar affinities at the ET_A receptor. Selective agonists of this receptor have been poorly defined, although the ET peptide ET-1[1–31] has been used as an ET_A receptor agonist (95). While it has some utility, this peptide falls short of the ideal agonist because ET-1[1–31] can be subjected to proteolytic degradation, thereby modifying the molecule and its pharmacological properties (30). It is unclear what contributions the different enzymatic products of ET-1[1–31] metabolism make to ET_A receptor stimulation, such that this peptide is limited in its usefulness. The ET_A receptor has been localized to virtually every physiological system, including cardiovascular (heart, blood, blood vessels), respiratory (lung, trachea), central nervous, sensory nervous, immune (neutrophils, macrophages), genitourinary (kidney, prostate, ovary), gastrointestinal (intestines, stomach, liver), and endocrine systems (pancreas).

ET_B Receptor

The ET_B receptor was cloned shortly after the ET_A receptor. Similarly to the ET_A receptor, it belongs to class A of G protein-coupled heptahelical receptors with an external amino terminus and internal carboxy terminus and binding sites intrinsic to the heptahelical portions of the receptor (99). Similar to the ET_A receptor, the endogenous agonist ET-1 has a high affinity for the ET_B receptor. The pharmacology of the ET_B receptor is somewhat richer than that of the ET_A receptor in that multiple agonists of the ET_B receptor are recognized, including sarafotoxin 6c (S6c) and IRL1620. Selective antagonists of the ET_B receptor include BQ788, A192621, RES7011, and IRL2500.

However, a number of real differences between the ET_A and ET_B receptor exist. First, unlike the ET_A receptor, all three endogenous ET peptides (ET-1, ET-2, and ET-3) have similar affinity for the ET_B receptor. Second, variants in the ET_B receptor gene exist in the human population to a seemingly greater extent than for the ET_A receptor (80, 107). In a similar vein, it has been suggested that ET_B1 and ET_B2 receptors exist, here discriminated by their geographical location (B1 = endothelial cell, B2 = smooth muscle in a blood vessel). The IUPHAR and ET receptor committee have not formally adopted this nomenclature in the absence of data that support the existence of two different ET_B receptor genes or proteins that are distinct pharmacological entities. This is thus a nomenclature that would seem wise to not use. Third, the ET_B receptor is not expressed as extensively as the ET_A receptor, or at least the data supporting ET_B receptor expression are not as readily available. The ET_B receptor has been localized to cardiovascular tissues, the pulmonary system, neurons, bone, pancreas, and kidney.

Thus we know, at the present, that two bona fide ET receptor subtypes exist. What are some of the outstanding questions regarding these two receptors and their stimulation by ET-1?

OUTSTANDING QUESTIONS IN THE FIELD

In this section, topic areas in which progress could provide our field with better tools, or issues with which the field of ET (and, in particular, the author's laboratory) is wrestling will be addressed. It should be noted that these topics are the opinion of the author of the present review, and thus are arguable.

Identification of an ET_A Receptor Agonist

For pharmacologists, it is well established that receptors are best characterized when a full armamentarium of tools is available to define that receptor. This includes positive agonists (substances that bind to and stimulate a biological response through the receptor) and antagonists (substances that bind to and prevent a biological response through the receptor). Collectively, these pharmacological tools can be used to provide a “fingerprint” for a receptor. The ET_B receptor has S6c as an agonist, but the ET_A receptor has no known parallel tool. The limitation of ET-1[1–31] as an ET_A receptor agonist was described above, and this is not an ideal situation. Without an ET_A receptor selective agonist, one lacks the ability to activate the ET_A receptor in the absence of other ET receptors. Such a tool would enable the study of receptor trafficking, desensitization, resensitization, and signal transduction elements directly utilized by the ET_A receptor. To be sure, ET-1 itself can do this, but there is always the caveat that ET-1 will stimulate any ET_B receptors present. In artificial systems, one can test the proof-of-concept studies of how ET stimulation modifies ET_A receptor movement in the absence of an ET_B receptor, but this only approaches an endogenous physiological situation. For example, most blood vessels (arteries and veins) possess both smooth muscular ET_A and ET_B receptors. It would be ideal to compare and contrast how the ET_A receptor moves and signals independently in these two different blood vessel types that serve significantly different purposes in the body. Using ET-1 is not good enough. An agonist (ideally a small molecule) that activated the ET_A receptor selectivity (>50-fold difference from ET_B) would be a beneficial tool. Recently, several studies have dissected structural determinants for selective recognition of ET_A and ET_B receptor ligands and developed imaging proteins for the ET_A receptor (7, 44, 64). With such information, we can hope to develop better tools for studying the ET_A receptor.

Biased Ligands for ET Receptors?

Receptor signal transduction has classically been considered in a sequential fashion. The agonist (ET-1 in our case) activates a heptahelical receptor, which activates a G protein, which activates an effector that produces or modifies production of a second messenger such that a biological message is sent. This scenario suggests that a message is black/white (it is there or not), and once the receptor is stimulated, the message is sent. In the late 1990s, the idea of differential signaling was introduced (18); some of the original work was described in the serotonin receptor field. This idea stems from findings that different agonists could elicit different signal transduction events through the same receptor (for review see Refs. 53, 54, 114, 115). The mechanism by which this occurs has best been illustrated by the ANG II (AT₁) receptor. This receptor couples to both G proteins (G protein dependent) and β-arrestin (G protein independent; a src/Erk MAPK pathway) to elicit biological effects. When ANG II is an agonist, both pathways (G protein-dependent and β-arrestin-dependent) are stimulated. However, the substituted ANG II peptide Sar¹, Ile⁴, Ile⁸-ANG II (SII) activates only the β-arrestin pathways (115). SII has thus been called a biased agonist, in that its activation of the AT₁ receptor is biased towards one particular signal transduction pathway.

Endothelin activates src and the Erk MAPK pathways in multiple cell types; and more recently, the ET_A receptor has been linked to β-arrestin in signaling to β-catenin in ovarian cancer cells (94). Thus, it is probable that, like the AT₁ receptor, ET receptors can function in a biased manner. Which ligands that we use possess this capability? Would the design of ligands that are biased be beneficial experimental or therapeutic tools? Does the context of the receptor modify whether an agonist is biased? Makita et al. (68) demonstrated that ET-3 showed significant affinity for the ET_A receptor in superior cervical ganglion. Is this an issue of truly different ET receptor subtypes or ET receptor behaving with different signaling/pharmacology bias? This field is open for discovery.

ET_A/ET_B Receptor Heterodimerization

Heptahelical receptors like the ET receptors have classically been described as functional monomers which, when stimulated, activate G proteins and subsequent effectors. Evidence accrued over the last decade suggests that heptahelical receptors function as physical homodimers (e.g., 2 GABA receptors together) or heterodimers (e.g., angiotensin AT₁ and bradykinin BK₂) (14, 25, 34, 39, 72, 79, 119). Heptahelical receptors can dimerize through venus fly trap elements of the NH₂ terminus, the heptahelical domains, covalent cysteine bonds in the NH₂ termini, and twinning motifs in the COOH terminus (39, 66). The pharmacology and function of a multisubunit receptor complex (e.g., heterodimer, homodimer) can be different from that of the individual monomers, with changes in agonist potency, affinity, and antagonist-sensitivity reported. One member of the dimer pair may serve as a chaperone or modify the function of the other partner just as the receptor activity modifying proteins do for the B family of G protein-coupled receptors (105). Physical interaction of receptors occurs not only in the plasma membrane but also in the endoplasmic reticulum. If heterodimerization occurs physiologically, it could have a profound impact on pharmacological therapy of disease (117, 123). Dimerization is, however, controversial for class A/rhodopsin-like heptahelical receptors, which include the ET receptors, because they irregularly include classical motifs for interaction. Sequences of the rat and human ET_A and ET_B receptor contain NH₂ terminus cysteine residues that could form covalent bonds and incomplete twinning motifs in the COOH terminus. Additionally, exactly what constitutes functional dimerization is arguable (39). The partners in heterodimers that can be coimmunoprecipited with one another typically have significant physical interaction, such as covalent bonds through cysteine residues in the NH₂ terminus. This suggests these receptors live and operate together as a functional unit for extended periods of time. Do two partners that come together for a brief and noncovalent protein-protein interaction (a “kiss-and-run” interaction) also constitute a heterodimer? This should be so if that interaction, however brief, modifies the pharmacology and/or function of one of the partners. The question can then be asked: is this type of kiss-and-run association simply par for the course in signal transduction, a process clearly dependent on noncovalent and brief interactions?

Most studies investigating receptor heterodimerization use artificial systems in which receptors are transfected and are expressed at supraphysiological levels. Heterodimer pairs studied in such systems include the ANG II AT₁ receptor with bradykinin B₂ receptor (106), AT₂ with B₂ receptor (1), thromboxane α- and β-receptor (120), dopamine D₁ and D₂ receptors (27), α_2C- and β₂-adrenergic receptors (92), α_1B- and α_1D-adrenergic receptors (40), and vasopressin V_1a and V₂ receptors (111). This list is far from complete. One exception to transfected cell experiments is AT₁/B₂ receptor heterodimerization in cultured renal mesangial cells from the spontaneously hypertensive rat but not from normotensive controls, with heterodimerization contributing to the hyperresponsiveness of spontaneously hypertensive rat cells to ANG II (2).

We have revealed an interaction of the ET_A and ET_B receptor in venous but not arterial tissues. A similar interaction of ET receptors has been described in a number of tissues, being first described in the vasculature in 1995 (3, 32, 46, 49, 67, 69, 78, 89, 102). At the base of this interaction is the failure of ET_A receptor antagonists to shift venous contraction to ET-1, unless ET_B receptor function is ablated. S6c/ET_B receptor desensitization renders veins sensitive to ET_A receptor blockade. An explanation for the failure of the ET_A receptor antagonist to shift ET-1-induced venous contraction may be that activated ET_B receptors physically/pharmacologically uncouple the ET_A receptor through heterodimerization. Exposure to high concentrations of S6c removes the actions of the ET_B receptor. This allows ET_A receptors to participate in contraction in a pharmacologically definable manner. Heterologously expressed human ET_A and ET_B receptors can form heterodimers (19, 28, 29, 36). More recently, ET_A and ET_B receptor heterodimerization has been suggested to occur in pulmonary arteries, but this is the only example in a physiological and pathological situation (101, 102). The question foremost in this field is under what conditions heterodimer formation contributes to physiological function. We have been unable to demonstrate coimmunoprecipitation of the ET_A and ET_B receptor in veins, having tried in at least a dozen different immunoprecipitation approaches (different kits, buffers, urea gels, etc.). Interaction of the ET_A and ET_B receptor is the most logical reason for our pharmacological results, but we’ve begun to look at other possible explanations given our lack of ability to explain the unusual pharmacology described above by ET_A/ET_B receptor heterodimerization. ET receptors may interact with other non-ET receptors to form a new pharmacophore. This was recently described for the ET_B receptor, which heterodimerize with the dopamine D₃ receptor in a cell culture of renal proximal tubule cells (122). This represents one possible explanation for the pharmacological results observed. Another question that deserves thought is whether ET-1 itself is a bivalent ligand, being capable of stimulating both the ET_A and ET_B receptor with different portions of the ET-1 molecule. This idea has been proposed by Hirada et al. (42). This group took a slightly different tactic, concluding that the ET_B receptor could not independently recognize ET-1 without the ET_A receptor and suggesting that ET-1 itself is a bivalent ligand. Another explanation includes alternative targets for ET itself, and these are next described.

Other Targets for ET-1?

Dual angiotensin ET receptor.

The dual angiotensin ET receptor (DEAR) was characterized in 1998 by Ruiz-Opazo et al. (97). Structural analysis predicted a receptor polypeptide that, unlike the native ET and AT receptors, possesses only one transmembrane region, an extracellular amino terminus and intracellular carboxy terminus. This receptor possesses both ANG II- and ET-1-specific binding sites, as well as the ability to mobilize calcium. Expression of this receptor mRNA was found in the brain and the heart, suggesting it participates in cardiovascular control. More recently, DEAR has been implicated in the female Dahl salt-sensitive model of hypertension (51). The interaction of ET-1 and ANG II is supported by the basic knowledge that ANG II can cause ET-1 release, and both modify the function of NADPH oxidase (62). The pharmacology and physiology of DEAR has not been developed. Could the DEAR receptor be a biological target of endogenous ET peptides?

The G protein-coupled peptide receptor 37 gene.

In 1997, GPR37, a gene located on chromosome 7 encoding a putative G protein-coupled peptide receptor, was cloned in human frontal brain (71). This has been called the ET receptor B-like protein receptor 1 because of its high degree of homology with the human ET_B receptor. The most recent scientific research suggests this protein more likely functions as a substrate for Parkin, a polypeptide with similarities to ubiquitin protein ligase E3. This protein is now called the Parkin-associated ET receptor-like receptor (Pael-R; 48). Research of this protein as a bona fide ET receptor has not progressed.

ET_C.

The ET_C receptor was proposed early in ET receptor history, being cloned in Xenopus laevis (52, 74). A similar receptor from Xenopus was cloned by Kumar et al. (59). Recent phylogenetic reports suggest that the ET_C receptor gene and protein was lost in placental and marsupial mammals and thus is not relevant to human physiology and pharmacology (13, 47).

Other targets of ET-1.

A tenet of pharmacology is that an agonist interacts with a receptor to exert a biological effect. This is certainly true, and pharmacology has taken advantage of this interaction to develop pharmaceutical therapies that have, without a doubt, benefited human health. However, it is becoming increasingly apparent that agonists (hormones, neurotransmitters, peptides) may exert their biological activity in decidedly different ways. For example, the small molecule serotonin (5-HT) can activate multiple plasma membrane receptors, but it can also be used to covalently modify proteins (serotonylation) and change protein function (118).

Could ET-1 Interact with Other Biological Molecules to Exert its Effects?

This is a long-term question that we are just beginning to address with experiments such as the following one. We used an ET-1 molecule biotinylated on the lysine 9 position (Genscript). This compound retains its biological activity as validated by its ability to cause arterial contraction (not shown). We incubated 100 μg of rat aortic homogenate (not whole tissue) with biotinylated ET-1 (∼1 μM final concentration) for 3 h at 4°C. The assumption is that ET-1 will find and, as it does with ET receptors, avidly bind to interacting proteins. Streptavidin-coated magnetic beads were added to the samples, and the samples tumbled overnight in a nonstringent buffer (PBS) at 4°C. Samples were washed three times and then the beads were collected, heated at 60°C for 10 min, and a sample (beads + eluted proteins) loaded onto a 10% polyacrylamide gel. The gel was run under standard conditions and taken through ProteoSilver staining. Figure 1 shows one example of such an experiment with no modifications. ET-1 clearly interacted with many proteins in the homogenate. These are in the process of being identified by tandem mass spectrometry. While these experiments must be performed in a whole tissue/whole cell, it suggests that we might need to take a broader look as to what ET interacts with in biological systems.

Fig. 1.

Gel of proteins brought down by streptavidin-coated magnetic beads from homogenates of normal Sprague-Dawley male rats incubated without (−) or with (+) biotinylated endothelin-1 (ET-1). Right, molecular weight markers.

PERSPECTIVES: WHAT DO WE NEED TO DO?

ET receptors have a rich, focused, but a still relatively new history compared with families like the adrenergic receptors. There is thus much we can learn and discover. I would hope that in the future we strive to

• identify a truly selective ET_A receptor agonist;

• determine whether and how the concept of biased agonism applies to ET receptors;

• determine mechanisms to show with confidence that ET_A and ET_B receptors heterodimerize in a physiological situation;

• identify other biological targets for ET-1;

• interrogate the current tools we have to study ET receptor pharmacology [Definitions of the tools we have used have been narrow (a compound is an agonist or antagonist), and one could speculate we have overlooked subtle differences in compounds such as bivalency, functional selectivity, and so on.];

• answer the question of why ET receptors exist in the nucleus of the cell and modify calcium (10, 57, 84) (This fascinating finding is still at the fringes of ET pharmacology and deserves attention not only because of its relevance to ET-1 but because of the germane idea that classically known plasma membrane receptors may modify nuclear biochemistry.); and

• reflect on recently published phylogenic studies (13, 47) (These studies described several gene losses in the evolution of the ET system, including that of the ET_C receptor from placental and marsupial mammals. It has also been suggested that the ET core system is a vertebrate-specific innovation. These reports may be revealing as to other functions of ET-1.).

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grant PO1-HL-70687.