Abstract
The total and regional peripheral resistance and capacitance of the vascular system is regulated by the sympathetic nervous system, which influences the vasculature mainly through changes in the release of catecholamines from both the sympathetic nerve terminals and the adrenal medulla. The knowledge of the targets for noradrenaline and adrenaline, the main endogenous catecholamines mediating that influence, has recently been greatly expanded. From two types of adrenoceptors (α and β), we have now nine subtypes (α1A, α1B, α1D, α2A/D, α2B, α2A/D, β1, β2, and β3) and two other candidates (α1L and β4), which may be conformational states of α1A and β1-adrenoceptors, respectively. The vascular endothelium is now known to be more than a pure anatomical entity, which smoothly contacts the blood and forms a passive barrier against plasma lipids. Instead, the endothelium is an important organ possessing at least five different adrenoceptor subtypes (α2A/D, α2C, β1, β2, and β3), which either directly or through the release of nitric oxide actively participate in the regulation of the vascular tone. The availability of transgenic models has resulted in a stepwise progression toward the identification of the role of each adrenoceptor subtype in the regulation of blood pressure and fine-tuning of blood supply to the different organs: α2A/D-adrenoceptors are involved in the central control of blood pressure; α1-(primarily) and α2B-adrenoceptors (secondarily) contribute to the peripheral regulation of vascular tone; and α2A/D- and α2C-adrenoceptors modulate transmitter release. The increased knowledge on the involvement of vascular adrenoceptors in many diseases like Raynaud's, scleroderma, several neurological degenerative diseases (familial amyloidotic polyneuropathy, Parkinson disease, multiple-system atrophy), some kinds of hypertension, etc., will contribute to new and better therapeutic approaches.
I. Introduction
“The nerves controlling the blood-vessels that supplied his face functioned so well that the skin, robbed of all its blood, went quite cold, the nose looked peaked, and the hollows beneath the young eyes were lead-couloured as any corpse's. And theSympathicus caused his heart, Hans Castorp's heart, to thump, in such a way that it was impossible to breathe except in gasps; and shivers ran over him, due to the functioning of the sebaceous glands, which, with the hair follicles, erected themselves”.—Thomas Mann, 1924
The operation of the sympathetic nervous system, especially of its cardiovascular branches, is nowhere in literature described better than in this passage from Thomas Mann's Magic Mountain, that great novel on pre-1914 Europe that the author places in a sanatorium at Davos in the Swiss mountains. Vasoconstriction, tachycardia, and contraction of the musculi arrectorum pilorum are Hans Castorp's autonomic responses when he first addresses his beloved Claudia Chaucat on Walpurgis-Night to borrow a pencil from her. This review probes the mechanisms that noradrenaline, the classical transmitter substance of the sympathetic vasoconstrictor fibers, uses to make blood vessels constrict; probes, in other words, the events that occurred in Hans Castorp when he borrowed the pencil.
Directly or indirectly, the blood vessels are the source of many and serious diseases that affect millions of people. In many respects, vascular physiology and pharmacology have changed dramatically over the last years. The discovery by Furchgott and Zawadzski in 1980 of endothelium-derived relaxing factor (EDRF2) revolutionized our knowledge and placed the endothelium in the center of the physiology and pathophysiology of the vascular tree; the cloning of many receptors brought about a true “Renaissance” in receptor pharmacology (Kenakin, 1997); and the possibility to “knock out” specific genes in experimental animals represents a new and important tool for a detailed study of the adrenoceptors, including those of the vascular system.
The present review aims at updating adrenoceptors in blood vessels, particularly on a functional point of view. Occasionally, some information is derived from nonvascular tissues; however, emphasis is placed on results obtained in blood vessels. Some reviews covering part of the present theme were published in the last few years (Insel, 1996;Strosberg, 1997; Summers et al., 1997; Docherty, 1998; Miller, 1998;Brodde and Michel, 1999; Bünemann et al., 1999; Freissmuth et al., 1999; Guimarães, 1999; Hein, 1999; Zhong and Minemann, 1999;Garcia-Sáinz et al., 2000; Gauthier et al., 2000; Hein, 2000;Kable et al., 2000).
II. Subclassification of Adrenoceptors
The adrenoceptors are the cell membrane sites through which noradrenaline and adrenaline act as important neurotransmitters and hormones in the periphery and in the central nervous system. The adrenoceptors are targets for many therapeutically important drugs, including those for some cardiovascular diseases, asthma, prostatic hypertrophy, nasal congestion, obesity, and pain.
The first step leading to the discovery of the adrenoceptors was made in the cardiovascular system—the observation by Dale (1905) that the pressor effect of adrenaline was reversed by ergotoxine into a depressor effect. An explanation for this phenomenon was not apparent until 43 years later! In 1948, Ahlquist noted two patterns in the relative ability of several sympathomimetic agonists to cause pharmacological responses in a series of organs and proposed the division of adrenoceptors into two types, α and β. This was subsequently confirmed by the identification of selective antagonists for these two sites: phentolamine and ergotamine for α-adrenoceptors; dichloroisoprenaline (Powell and Slater, 1958) and propranolol (Black et al., 1964) for β-adrenoceptors. Nineteen years later, it was shown that certain agonists and antagonists could distinguish β-adrenoceptor-mediated responses among tissues such as cardiac muscle and bronchial smooth muscle, implying the existence of subtypes of β-adrenoceptors (β1 in cardiac muscle and β2 in the bronchi) (Furchgott 1967, 1972; Lands et al., 1967a,b). Later on, the existence and differential tissue localization of α1 and α2 subtypes of α-adrenoceptors were discovered and characterized. The existence of subclasses of α-adrenoceptors has become evident from the results obtained by Starke and coworkers, who showed that pre- and postjunctional α-adrenoceptors differ with respect to the relative potencies of some agonists: low concentrations of clonidine and oxymetazoline selectively activate the prejunctional α-adrenoceptors, whereas phenylephrine and methoxamine selectively activate the postjunctional α-adrenoceptors (Starke, 1972; Starke et al., 1974, 1975b). Similarly, the relative potency of antagonists supported this differentiation: phenoxybenzamine was about 30 times more potent in blocking postjunctional than prejunctional α-adrenoceptors (Dubocovich and Langer, 1974) and yohimbine preferentially blocked prejunctional α-adrenoceptors (Starke et al., 1975a). Langer (1974) suggested that α-adrenoceptors mediating responses of effector organs should be referred to as α1 and those mediating a reduction of the transmitter release during nerve stimulation as α2. Later, it was found that α-adrenoceptors pharmacologically very similar to the prejunctional α2-adrenoceptors are also found postjunctionally. Consequently, the nomenclature of α1- and α2-adrenoceptors, depending exclusively on the relative potencies of certain α-agonists and antagonists, was accepted (Berthelsen and Pettinger, 1977). In the late 1980s, the development of more selective drugs and the use of molecular cloning technology showed that there are more adrenoceptor subtypes than previously suspected. Nine different subtypes have now been cloned and pharmacologically characterized (Alexander and Peters, 1999).
A. α1-Adrenoceptors
α1-Adrenoceptors were first divided into two subtypes, α1A and α1B, based on the differential affinity of the receptors for 5-methyl urapidil (5-MU), WB-4101 (Morrow and Creese, 1986; Gross et al., 1988; Hanft and Gross, 1989; Boer et al., 1989) and the irreversible antagonist chloroethylclonidine (Han et al., 1987). α1A-Adrenoceptors showed high affinity for 5-MU and WB-4101 and were insensitive to chloroethylclonidine, and α1B-adrenoceptors were sensitive to CEC and had low affinity for 5-MU and WB-4101. At the present time, a consensus has been reached, such that the subdivision of α1-adrenoceptors into three subtypes is generally accepted: α1A (formerly α1c; Schwinn et al., 1990), α1B (Cotecchia et al., 1988), and α1D (formerly α1a/d;Lomasney et al., 1991; Perez et al., 1991; Bylund et al., 1994; Ford et al., 1994). In humans, α1A-, α1B-, and α1D-adrenoceptors are encoded by distinct genes located on chromosomes 8, 5, and 20, respectively (Hieble et al., 1995;Michel et al., 1995). Furthermore, human α1A-adrenoceptor heterogeneity comes from the existence of multiple variants that differ in length and sequence of their C-terminal domains (Hirasawa et al., 1995). Additional truncated α1A-adrenoceptor proteins have been reported (Chang et al., 1998). More importantly, no pharmacological or signaling differences were observed on expression of these different splice variants. According to Lattion et al. (1994), they may exhibit differential susceptibility to desensitization. A fourth α1-adrenoceptor, the so-called α1L-adrenoceptor, has been postulated (Holck et al., 1983; Flavahan and Vanhoutte, 1986a; Muramatsu et al., 1990), based exclusively on pharmacological criteria (e.g., relatively low affinity for prazosin and other antagonists such as RS-17053). This α1L-adrenoceptor seems to mediate constriction of human (Ford et al., 1996) and rabbit (Van der Graaf et al., 1997; Kava et al., 1998) lower urinary tract, guinea pig aorta (Muramatsu et al., 1990), and rat small mesenteric arteries (Stam et al., 1999). However, this hypothetical additional subtype resisted identification by biochemical and/or molecular techniques so far. Recent studies indicate that the α1L-adrenoceptor may not be derived from a distinct gene, but represents a particular, energetically favorable, conformational state of the α1A-adrenoceptor (Ford et al., 1998). Why these two pharmacological phenotypes occur requires further investigation (Ford et al., 1997, 1998).
It is well known that α1-adrenoceptors are mainly coupled to Gq/11-protein to stimulate phospholipase C activity and that this enzyme promotes the hydrolysis of phosphatidylinositol bisphosphate producing inositol trisphosphate and diacylglycerol. These molecules act as second messengers mediating intracellular Ca2+ release from nonmitochondrial pools and activating protein kinase C, respectively (for reviews, seeHein and Kobilka, 1995; Zhong and Minneman, 1999;Garcı́a-Sáinz et al., 2000). The three cloned α1-adrenoceptor subtypes have different efficiencies in activating phospholipase C. According to Theroux et al. (1996), the ranking order of coupling efficiency (increase in inositol triphosphate formation and intracellular Ca2+) after agonist occupation of recombinant α1-adrenoceptors expressed in human embryonic kidney 293 cells was: α1A > α1B > α1D. All three α1-adrenoceptor subtypes can couple to phospholipase C through protein Gαq/11, only α1A- and α1B-subtypes couple to protein Gα14, and only the α1B-subtype couples to protein Gα16 (Wu et al., 1992). Other studies support that native α1B-adrenoceptors (but not α1A- or α1D-adrenoceptors) can also couple to protein Gαo in rat aorta (Gurdal et al., 1997) suggesting a functional role for this coupling. Other signaling pathways have also been shown to be activated by α1-adrenoceptors: Ca2+influx, arachidonic acid release, phospholipase D activation, and activation of mitogen-activated protein kinase (for a review, see Zhong and Minneman, 1999). Currently, no close relationship can be established between specific subtypes and signaling mechanisms.
α1-Adrenoceptor subtypes are differentially regulated. Although the maximal down-regulation after a prolonged exposure to phenylephrine was similar for α1A- and α1B-adrenoceptors, the threshold concentration of phenylephrine for significant reduction was 100-fold higher for α1A- than for α1B-adrenoceptors. In contrast, phenylephrine up-regulated α1D-adrenoceptors in a time- and concentration-dependent manner (Yang et al., 1999).
B. α2-Adrenoceptors
It is now clear that there are three subtypes of α2-adrenoceptors: α2A/D, α2B, and α2C. This subdivision, although primarily based on radioligand binding data, was preceded by results obtained in functional studies and confirmed by molecular cloning. For the α2B- and α2C-adrenoceptors, the pharmacological characteristics are consistent across mammalian species; however, the α2A-adrenoceptor cloned from human and porcine tissue differs slightly in its amino acid composition from the homologous receptor cloned from the rat, mouse, or guinea pig in having a serine residue rather than a cysteine, at the position corresponding to Cys201. To the three different genes, four pharmacological subtypes correspond since the Ser201 receptor possesses pharmacological properties different from the Cys201 receptor, and the two have been distinguished as α2A(e.g., humans) and as α2D (e.g., rodents) (Bylund et al., 1992; Starke et al., 1995; Trendelenburg et al., 1996;Paiva et al., 1997; Guimarães et al., 1998). These two orthologs will be simply referred to as α2A/D, unless some distinction between them has to be made. In humans, the genes coding for α2A-, α2B-, and α2C-adrenoceptors are localized in chromosomes 10, 2, and 4, respectively (Regan et al., 1988; Lomasney et al., 1990; Weinshank et al., 1990).
Pharmacologically it is well known that the different α-adrenoceptor antagonists possess different potency/affinity for the different α2-adrenoceptor subtypes: prazosin for example, has relatively high affinity for α2B- and α2C-adrenoceptors and very low affinity for α2A- and α2D-adrenoceptors (Latifpour et al., 1982;Nahorski et al., 1985; Bylund et al., 1988); yohimbine and rauwolscine are more potent than phentolamine and idazoxan on α2A-adrenoceptors, whereas reversed relative potencies are observed for α2D-adrenoceptors (Starke, 1981; Ennis, 1985; Lattimer and Rhodes, 1985; Alabaster et al., 1986; Limberger et al., 1989). The comparison of the functional potency of several antagonists with their affinity to all subtypes, as determined either in radioligand assays in native tissues possessing only one subtype or in cells transfected with recombinant α2-adrenoceptors, shows full agreement. So, this functional approach has been extensively used to characterize α2-autoreceptor subtypes in the different tissues (Hieble et al., 1996). Systematic studies recently undertaken to characterize prejunctional α2-adrenoceptor subtypes in different species confirmed that receptors with α2A properties occur in some species and receptors with α2D properties occur in others (Bylund et al., 1994; Starke et al., 1995; Trendelenburg et al., 1996; Paiva et al., 1997; Guimarães et al., 1998) (Table1).
However, some rare discrepancies to this postulate have been reported: in the rat vena cava (Molderings and Göthert, 1993) and rat atria (Connaughton and Docherty, 1990), where the prejunctional receptors were classified as α2B, and in the human kidney cortex (Trendelenburg et al., 1994) and human right atrium (Rump et al., 1995), where they appeared to belong to the α2C-subtype. However, a reinvestigation of these unexpected subclassifications showed that the prejunctional receptors in rat vena cava and atria and in guinea pig urethra were α2D, and those of human kidney were α2A. Thus, in contrast to previous suggestions, all these receptors conform to the rule that α2-autoreceptors belong, at least predominantly, to the genetic α2A/D-subtype (Trendelenburg et al., 1997).
Although the vast majority of tissues express more than one subtype, there are rare tissues expressing only one subtype: α2A in human platelets (Bylund et al., 1988), α2B in the rat neonatal lung (Bylund et al., 1988), α2C in opossum cells (Murphy and Bylund, 1988), and α2D in the rat submaxillary gland (Michel et al., 1989).
α2-Adrenoceptors are predominantly coupled to the inhibitory heterotrimeric GTP-binding protein inhibiting the activity of adenylyl cyclase (Cotecchia et al., 1990; Wise et al., 1997), inhibiting the opening of voltage-gated Ca2+ channels (Cotecchia et al., 1990) and activating K+ channels (Surprenant et al., 1992). The α2-adrenoceptors may also couple to other intracellular pathways involving Na+/H+ exchange and the activation of phospholipase A2, C, and D (Limbird, 1988; Cotecchia et al., 1990; MacNulty et al., 1992; Kukkonen et al., 1998). In neurons, α2-adrenoceptors inhibit N-, P-, and Q-type voltage-gated Ca2+channels (Waterman, 1997; Delmas et al., 1999; Jeong and Ikeda, 2000).
Like the α1-adrenoceptors, the three α2-adrenoceptor subtypes are regulated differentially. Human α2C-adrenoceptors do not appear to down-regulate following exposure to agonists (Eason and Liggett, 1992; Kurose and Lefkowitz, 1994); α2A/D- and α2B-adrenoceptors down-regulate apparently due to an increase in the rate of receptor disappearance (Heck and Bylund, 1998).
C. β-Adrenoceptors
Three distinct β-adrenoceptor subtypes have been cloned so far: β1, β2, and β3 (Bylund et al., 1994). These subtypes are encoded by three different genes located on human chromosomes 10 (β1), 5 (β2), and 8 (β3). The human β3-adrenoceptor has 49 and 51% overall homology at the amino acid level with human β2- and β1-adrenoceptors, respectively (Emorine et al., 1989; Granneman and Lahners, 1994). Other species homologs of the human β3-adrenoceptor have also been cloned (for a review, see Strosberg, 1997). β1 and β2-Adrenoceptors are well known pharmacologically since the classical papers by Lands et al. (1967a,b). They mediate cardiovascular responses to noradrenaline released from sympathomimetic nerve terminals and to circulating adrenaline. They are stimulated or blocked by many compounds that are used to treat important and common diseases, such as hypertension, cardiac arrhythmias, and ischemic heart disease.
The existence of a third β-adrenoceptor subtype (β3-adrenoceptor), which was previously shown to mediate lipolysis in rat adipocytes (Harms et al., 1974; Arch et al., 1984; Wilson et al., 1984; Bojanic et al., 1985; Emorine et al., 1989), was also found in blood vessels where it mediates vasodilation (Cohen et al., 1984; Molenaar et al., 1988; Rohrer et al., 1999). β3-Adrenoceptors are not blocked by propranolol, and other conventional β-adrenoceptor antagonists are activated by β3-adrenoceptor selective agonists like BRL 37344 and CL 316243 (for reviews, see Manara et al., 1995;Strosberg, 1997; Summers et al., 1997; Fischer et al., 1998) and are blocked by β3-adrenoceptor antagonists like SR-59230, which has been described as β3-adrenoceptor selective in rat brown adipocytes (Nisoli et al., 1996), rat colonic motility assays (Manara et al., 1996), and human colonic circular smooth muscle relaxation activity assays (De Ponti et al., 1996). More recently, Candelore et al. (1999) did not confirm the selectivity of SR-59230 for human β3-adrenoceptors, but described two compounds, namely L-748328 and L-748337 that display greater than 90-fold selectivity for human β3- versus β1-adrenoceptors, and 20- and 45-fold selectivity versus human β2-adrenoceptors, respectively. The pharmacology of β3-adrenoceptors is clearly distinct from that of β1- and β2-adrenoceptors; however, one has to bear in mind that there are differences between rodents, where β3-adrenoceptors were studied initially, and humans, and this contributes to some confusion in the subclassification of β-adrenoceptors (Wilson et al., 1996; Arch, 1998). Furthermore, there are also differences depending on the methodological approach used. For example, the potency of catecholamines at the human β3-adrenoceptor was found to be 1 to 2 orders of magnitude higher when determined in an intact cell cAMP accumulation assay than in a membrane-based adenylyl cyclase activation assay (Wilson et al., 1996).
On the basis of many pharmacological and molecular studies, the existence of a fourth β-adrenoceptor subtype was postulated (for reviews, see Arch and Kaumann, 1993; Barnes, 1995; Strosberg and Pietri-Rouxel, 1996; Kaumann, 1997; Strosberg, 1997; Summers et al., 1997; Galitzky et al., 1998; Strosberg et al., 1998; Brodde and Michel, 1999). These receptors would include the receptor in rat soleus muscle, which mediates glucose uptake (Roberts et al., 1993) and the receptor in human and rat heart, which mediates positive chronotropism and inotropism (Kaumann and Molenaar, 1996, 1997; Kaumann et al., 1998; Oostendorp and Kaumann, 2000) (putative β4-adrenoceptor). A receptor cloned from turkey (βt-adrenoceptor) has no mammalian counterpart (Chen et al., 1994). In mouse brown adipose tissue (±)-CGP-12177, a partial agonist at β3-adrenoceptors, which is also antagonist at β1- and β2-adrenoceptors, evoked a full metabolic response that was of a similar magnitude in wild-type and β3-adrenoceptor knockout mice; however, the metabolic response to CL-316243 was abolished (Preitner et al., 1998). This unexpected result supports the view that a new β-adrenoceptor, distinct from β1-, β2-, and β3-adrenoceptor and referred to as putative β4-adrenoceptor, is present in brown adipose tissue and can mediate a maximal lipolytic stimulation (Preitner et al., 1998). A similar occurrence was reported for the heart. In β3-adrenoceptor knockout mice, CGP-12177A increased the force and rate of atrial contractions, and these effects were not antagonized by propranolol, but were antagonized by bupranolol (Kaumann et al., 1998). Furthermore, the binding of (−)-[3H]CGP-12177A was similar in ventricular membranes from hearts of wild-type and β3-adrenoceptor knockout mice; this provides evidence that the cardiac putative β4-adrenoceptor is distinct from the β3-adrenoceptor (Kaumann et al., 1998). More recently, evidence was obtained that this putative fourth β-adrenoceptor subtype is a particular state of β1-adrenoceptor (see Section III.C.1.c.).
All β-adrenoceptor subtypes signal by coupling to the stimulatory G-protein Gα s leading to activation of adenylyl cyclase and accumulation of the second messenger cAMP (Dixon et al., 1986; Frielle et al., 1987; Emorine et al., 1989). However, some recent studies indicate that, under certain circumstances, β-adrenoceptors, and particularly the β3-adrenoceptor, can couple to Gi as well as to Gs (Asano et al., 1984; Chaudry et al., 1994; Xiao et al., 1995; Gauthier et al., 1996).
Intracellular events following β-adrenoceptor activation are also linked to ion transport. It is well known, for example, that protein kinase A activated by cAMP phosphorylates L-type Ca2+ channels, facilitating Ca2+ entry, and producing the positive inotropic effect in atria and ventricles, increased heart rate in the sino-auricular node, and accelerated the conduction in the atrio-ventricular node. In addition to mechanisms that indirectly lead to alterations in ion transport, β-adrenoceptor activation is more directly linked to ion channels: β-adrenoceptor stimulation is able to activate L-type Ca2+ channels via Gα s (Brown, 1990); in airway smooth muscle, β-adrenoceptor activation opens Ca2+-dependent K+ channels and charybdotoxin—a specific inhibitor of the high conductance Ca2+-activated K+channel—antagonizes the relaxant effects of β-adrenoceptor agonists (Miura et al., 1992; Jones et al., 1993).
Multiple mechanisms control the signaling and density of G-protein-coupled receptors. The termination of G-protein-coupled receptor signals involves binding of proteins to the receptor. This process is initiated by serine-threonine phosphorylation of agonist-occupied receptors, both by members of the G-protein-coupled receptor kinase family and by second-messenger-activated protein kinases such as protein kinase A and protein kinase C. Receptor phosphorylation by G-protein-coupled receptor kinase is followed by binding of proteins termed arrestins, which bind to the phosphorylated receptor and sterically inhibit further G-protein activation (Luttrell et al., 1999). Desensitized receptor-arrestin complexes undergo arrestin-dependent targeting for sequestration through clathrin-coated pits (Goodman et al., 1996; Luttrell et al., 1999). Sequestrated receptors are ultimately either dephosphorylated and recycled to the cell surface or targeted for degradation (Luttrell et al., 1999).
In addition, many other G-protein-coupled receptors are sequestrated from the cell membrane and become inaccessible to their ligands. Both receptor/G-protein uncoupling and receptor sequestration may involve the participation of arrestins or other proteins. A model for receptor regulation has been developed on the basis of data from studies of the β-adrenoceptors. However, according to recent reports, other G-protein-coupled receptors, like muscarinic receptors in the cardiovascular system, may be regulated by mechanisms other than those that regulate the β-adrenoceptors (for a review, see Bünemann et al., 1999).
III. Postjunctional Adrenoceptors in Vascular Smooth Muscle
Because vascular smooth muscles possess both α- and β-adrenoceptors, the net response to agonists that like adrenaline stimulate both types of receptors depends on the relative importance of each population. For example, while in the dog saphenous vein, in vitro adrenaline causes contraction, which is enhanced by β-adrenoceptor blockade (Guimarães, 1975); in the rabbit facial vein, adrenaline causes relaxation, which is enhanced by α-adrenoceptor blockade (Pegram et al., 1976). On the other hand, the contractile response of the saphenous vein to adrenaline is converted into a relaxation when an α-adrenoceptor antagonist is present (Guimarães and Paiva, 1981a), and the relaxation caused by adrenaline in the rabbit facial vein is converted into a contraction when a β-adrenoceptor antagonist is present (Pegram et al., 1976). Thus, while in the dog saphenous vein, the α-adrenoceptor-mediated influence dominates, in the rabbit facial vein the dominating influence is exerted by β-adrenoceptor.
In the vast majority of vascular tissues, α-adrenoceptor-mediated effects predominate, such that to demonstrate in vitro β-adrenoceptor-mediated responses using adrenaline as agonist, both α-adrenoceptor blockade and active tone of the tissue must be present. When a pure or almost pure β-adrenoceptor agonist like isoprenaline is used, the only requirement to obtain β-adrenoceptor-mediated responses is the presence of tone. The threshold for α-adrenoceptor-mediated effects in large arteries and veins is between 1 and 10 nM noradrenaline (Guimarães, 1975;Bevan, 1977). The levels of noradrenaline and adrenaline in human arterial plasma at rest are about 2 and 0.5 nM, respectively (Engleman and Portnoy, 1970; DeQuattro and Chan, 1972). In the dog, the level of noradrenaline is similar. Thus, at rest, most vessels are scarcely influenced by circulating catecholamines. However, in the rat mesentery, precapillary sphincters have a threshold response to adrenaline and noradrenaline of 0.1 to 1 nM (Altura, 1971), and rat plasma adrenaline and noradrenaline levels average 2.5 and 3 nM, respectively (Donoso and Barontini, 1986). Although in vivo sensitivity cannot be directly related to plasma catecholamine levels, these data suggest that precapillary sphincters may be affected by circulating catecholamines even under resting conditions, in contrast to other vessels. In humans, during exercise, plasma noradrenaline and adrenaline may reach levels 30 times higher than those at rest, which may have a profound effect on vessels.
A. α1-Adrenoceptors
It is important to underline that many of the advances made in the last years in the field of receptors in general and on vascular adrenoceptors in particular were due to the possibility to generate knockout mice. However, one should not forget that the lack of a given receptor from conception may be compensated by adequate adjustments, whereas its functional elimination by an antagonist is not acutely compensated (Rohrer and Kobilka, 1998). This is something one must bear in mind when results obtained in wild-type mice are compared with results obtained in knockout mice. It is dangerous to assume that knockout animals differ from the wild-type by no more than the absence of one receptor subtype.
1. In Vitro.
In most mammalian species, contraction of vascular smooth muscle is predominantly mediated via α1-adrenoceptors. Although the existence of both α1- and α2-adrenoceptors has been shown by functional studies in vivo, it has been difficult to demonstrate functional postjunctional α2-adrenoceptors in most arteries in vitro (De Mey and Vanhoutte, 1981; McGrath, 1982;Timmermans and van Zwieten, 1982; Polónia et al., 1985;Guimarães, 1986; Aboud et al., 1993; Burt et al., 1995, 1998). In isolated canine aorta and canine femoral, mesenteric, jejunal, renal, and splenic arteries, contractile responses were exclusively α1-adrenoceptor-mediated (Polónia et al., 1985; Shi et al., 1989; Daniel et al., 1999). In the arteries of other mammalian species, α1-adrenoceptors also predominate: in rat aorta (Han et al., 1990; Aboud et al., 1993); in rat carotid, mesenteric, renal, and tail arteries (Han et al., 1990;Villalobos-Molina and Ibarra, 1996); and in human arteries (Flavahan et al., 1987a).
In veins, particularly in cutaneous veins, at the postjunctional level, α1- and α2-adrenoceptors both contribute to vasoconstriction (Flavahan and Vanhoutte, 1986a; Guimarães et al., 1987). In dog and human saphenous veins, α2-adrenoceptors are the predominant receptors mediating contraction (Müller-Schweinitzer, 1984; Guimarães and Nunes, 1990; Docherty, 1998).
The question of which α1-adrenoceptor subtype is involved in vasoconstrictive responses to sympathomimetic agonists is not easy to answer. Vascular smooth muscle tissues express mixtures of α1-adrenoceptor subtypes (Miller et al., 1996) and in most cases responses to α1-adrenoceptor agonists are probably due to activation of more than one subtype (Van der Graaf et al., 1996a; Zhong and Minneman, 1999). mRNA for the α1A-adrenoceptor is expressed at very high levels in peripheral arteries, around 90% of the total α1-adrenoceptors message pool (Guarino et al., 1996), but in most cases, there is lack of correlation between protein expression of one adrenoceptor subtype and the function this receptor mediates (Hrometz et al., 1999; Ohmi et al., 1999). The rat is a good case to exemplify this pharmacological problem (Aboud et al., 1993; Kong et al., 1994; Saussy et al., 1996; Piascik et al., 1997; Stam et al., 1999). Despite the fact that the mRNA for all three cloned α1-adrenoceptor subtypes has been found in the rat mesenteric artery, as well as the aorta and pulmonary artery (Xu et al., 1997), the contraction in response to phenylephrine in these three vessels is primarily α1D-adrenoceptor-mediated, α1B-adrenoceptor being secondarily involved (Hussain and Marshall, 2000). A similar lack of correlation was demonstrated in a study involving several arteries of the rat: although in terms of level of mRNA expression for α1-adrenoceptor subtypes, the ranking order was α1A- > α1B- > α1D, only α1B-adrenoceptors played a functional role in mesenteric resistance artery, whereas α1D-adrenoceptors were implicated in mediating the contraction of the aorta and femoral, iliac, and superior mesenteric arteries (Piascik et al., 1997). Similarly, in the rabbit, all α1-adrenoceptor subtypes coexist in the aorta and in the mesenteric, renal, and iliac arteries. However, although the renal and iliac arteries contract predominantly via the activation of α1D-adrenoceptors in response to noradrenaline and secondarily via activation of α1A- and α1B-adrenoceptors, the aorta contracts via the activation of α1A- and α1B-adrenoceptors (Satoh et al., 1998, 1999). According to functional results, it seems that in the rat the α1A- and α1D-adrenoceptor subtypes regulate the larger vessels, whereas the α1B-adrenoceptors control the small resistance vessels (Leech and Faber, 1996; Piascik et al., 1997; Gisbert et al., 2000). In the dog mesenteric artery, α1-adrenoceptors are predominantly of the α1A-subtype (Daniel et al., 1999).
Table 2 summarizes the α1-adrenoceptors subtypes primarily responsible for the contractile responses of the main arteries from species mostly currently used in research: α1A- and α1D-subtypes are those mainly involved in the contractions evoked by α1-adrenoceptor agonists.
α1-Adrenoceptors are also involved in the regulation of vascular smooth muscle growth. Findings by some authors suggest that prolonged stimulation of chloroethylclonidine-sensitive, possibly α1B-adrenoceptors, induce hypertrophy of arterial smooth muscle cells, whereas stimulation of α1A-adrenoceptors attenuates this growth response (Chen et al., 1995; Siwik and Brown, 1996).
2. In Vivo.
There is also longstanding evidence that multiple α1-adrenoceptor subtypes are involved in the regulation of peripheral vascular function in vivo (McGrath, 1982;Minneman, 1988; Bylund et al., 1995b). However, the individual contribution of each of the α1-adrenoceptor subtypes has not been established. Of the three known α1-adrenoceptor subtypes, α1A- and α1D-adrenoceptors have most often been implicated in the regulation of vascular smooth muscle tone (see Table2). There are discrepancies between results obtained in vitro and in vivo involving α1-adrenoceptors. Although in vitro studies in rats had indicated a predominant role of the α1D-adrenoceptor in the vascular contractions caused by α1-adrenoceptor agonists (Piascik et al., 1995; Hussain and Marshall, 2000), surprisingly experiments in α1B-knockout mice show that the maximal contractile response of aortic rings to phenylephrine was reduced by 40% and the mean arterial blood pressure response to phenylephrine was decreased by 45%, showing that the α1B-adrenoceptor is important for blood pressure and the contractile response of the aorta evoked by α1-adrenoceptor agonists (Cavalli et al., 1997). In the pithed rat, the systemic blood pressure is tonically regulated by the interaction of peripheral sympathetic nerves with vascular α1A-adrenoceptors (Vargas et al., 1994), although vascular α1D-adrenoceptors have a role in the pressor response to phenylephrine (Zhou and Vargas, 1996). Also, in the pithed rat, it was shown that the selective α1D-adrenoceptor antagonist BMY-7378 not only antagonized the pressor effect of phenylephrine, but also was more potent in young prehypertensive spontaneously hypertensive rats (SHRs) than in young WKY rats. The presence of α1D-adrenoceptors in the resistance vasculature of prehypertensive and hypertensive rats may indicate that they are involved in the development/maintenance of hypertension (Villalobos-Molina et al., 1999). Thus, it may be concluded that, in rats in vivo, the pressor response to phenylephrine is mediated by vascular α1A- and α1D-adrenoceptors (Vargas et al., 1994; Guarino et al., 1996; Zhou and Vargas, 1996).
In human vasculature, as in that of other mammals, α1-adrenoceptors play a crucial role in the regulation of vascular tone. In healthy volunteers, Schäfers et al. (1997, 1999) showed that, whereas 2 mg of doxazosin (a selective α1-adrenoceptor antagonist) nearly completely antagonized the blood pressure increasing effect of i.v. administered noradrenaline (10 to 160 ng/kg · min), 15 mg of yohimbine (a selective α2-adrenoceptor antagonist) only slightly attenuated noradrenaline effect. With regard to this finding, one should bear in mind that the administration of exogenous noradrenaline does not necessarily result in identical concentrations in the biophase of the postjunctional α1- and α2-adrenoceptors; there may develop a certain ratio biophase α1/biophase α2. Moreover, this ratio may be different for noradrenaline released from sympathetic nerves (see Distribution of vascular adrenoceptors). The available information regarding α1-adrenoceptor subtypes mediating vasoconstriction in humans is still very scarce.
In conclusion to the role played by each α1-adrenoceptor subtype in the maintenance of vascular tone and in vascular responses to α1-adrenoceptor ligands, one can say that there is a lack of correlation between two sets of results disturbing their interpretation. First, the lack of correlation between protein expression of a given adrenoceptor and the functional role this adrenoceptor plays; second, the lack of correlation between the results obtained in vitro (Table 2) and in vivo. Despite that, according to the vast majority of the authors, it seems that in the rat α1A-adrenoceptors have a prominent role in the regulation of blood pressure, although α1B- and α1D-adrenoceptors are also functionally present and participate in the responses to exogenous agonists (Piascik et al., 1990; Schwietert et al., 1992; Vargas et al., 1994; Guarino et al., 1996; Zhou and Vargas, 1996).
3. α1-Adrenoceptor Antagonists in the Symptomatic Treatment of Prostatic Hypertrophy.
Clinical interest in this target comes from the fact that selective α1A-adrenoceptor antagonists may have significant therapeutic advantages over nonsubtype selective α1-adrenoceptor antagonists in the treatment of benign prostatic hypertrophy. Which is the basis for the hypothetical differential effect of α1A-adrenoceptor antagonists at vascular tissue and prostate? Are α1-adrenoceptors of vascular- and prostatic smooth muscle different? Several studies have shown that the α1A-adrenoceptor subtype accounts for the majority of α1-adrenoceptor mRNAs and expressed protein in human prostatic smooth muscle and mediates contraction in this tissue (Price et al., 1993; Faure et al., 1994; Lepor et al., 1995; Michel et al., 1996; Schwinn and Kwatra, 1998). However, recent experiments carried out in rat mesenteric arteries (a tissue the α1-adrenoceptors of which, like those of the prostate, have low affinity for prazosin and RS-17053) (Ford et al., 1996), showed that the affinity of prazosin and RS-17053 was not altered by changing the experimental conditions (lowering temperature, inducing tone via KCl or U-46619—a derivative of prostaglandin F2α), calling again our attention to the problem of the putative α1L-adrenoceptors (Yousif et al., 1998; Stam et al., 1999). On the other hand, which is the α1-adrenoceptor subtype that mediates contractile vascular responses in humans? The few reports on α1-adrenoceptors in resistance arteries failed to show that a particular α1-adrenoceptor subtype is of primary importance in the sympathetic control of these vessels. Probably, as animal studies have suggested, each vessel possesses mixtures of α1-adrenoceptor subtypes, and responses to α1-adrenoceptor agonists are due to stimulation of more than one subtype (Michel et al., 1998b;Ruffolo and Hieble, 1999; Zhong and Minneman, 1999; Argyle and McGrath, 2000). In a very recent study, it was shown that the receptor subtype mediating the constriction of canine resistance vessels is an α1A-/α1L adrenoceptor (Argyle and McGrath, 2000), which is the same that has been proposed as mediating the adrenergic responses in prostate (McGrath et al., 1996). Thus, the relative selectivity of α1A-adrenoceptor antagonists, if there is any, may not depend on differences between subtypes, but rather on differences between local functional expressions of the receptors. In single human prostatic smooth muscle cells, MacKenzie et al. (2000)showed that the affinity of a prazosin analog for native human α1A-adrenoceptors was higher than for human cloned α1A-adrenoceptors expressed in cell cultures. This suggests that a tissue-specific affinity state of the same receptor genotype exists, and this could be a potential differentiator of drug action (MacKenzie et al., 2000).
Halotano et al. (1994) reported a slightly lower potency for 5-MU and WB-4101 in the human iliac artery, compared with the human urethra suggesting a therapeutic benefit in prostatic symptoms without causing the vascular side effects associated with α1-adrenoceptor blockade. However, the degree of selectivity of the different compounds until now available to treat benign prostatic hypertrophy (doxazosin, alfusosin, terazosin) is not enough to eliminate cardiovascular side effects, such as dizziness, orthostatic hypotension, asthenia, and occasionally syncope (Michel et al., 1998a; Chapple and Chess-Williams, 1999; Pulito et al., 2000). The moderately selective α1A-adrenoceptor antagonist tamsulosin has been introduced for this purpose (Foglar et al., 1995). When directly comparing equieffective dosis of terazosin (a selective α1-adrenoceptor antagonist) with tamsulosin in patients with prostatic hyperplasia, Lee and Lee (1997)observed that tamsulosin caused significantly fewer side effects; however, Schäfers et al. (1998) less enthusiastically concluded that further experimental and clinical work was required to unequivocally demonstrate this advantage of selective α1A-adrenoceptor antagonists. Very recently, the selectivity of tamsulosin, doxazosin, and alfuzosin was determined by comparing their effects on the human prostate and human mesenteric arteries in vitro. It was observed that tamsulosin exhibited a 10-fold selectivity for the prostate over the artery, a degree of selectivity that was compatible with its claimed clinical benefit (Davis et al., 2000). A possible explanation for the clinical advantages of tamsulosin was given by Hein et al. (2001), who showed that the α1-antagonist with the least vascular effects in humans in vivo also was the drug with the least inverse agonism in vitro (tamsulosin).
Some new aryl piperazine compounds were recently synthesized, which in binding experiments to recombinant human α1-adrenoceptors showed high α1A-adrenoceptor subtype selectivity (Pulito et al., 2000). Furthermore, some of them were more potent in inhibiting noradrenaline-evoked contraction of rat prostate tissue than those of rat aorta tissue: RWJ-38063 and RWJ-69736 were 319- and 100-fold more potent in their effects on prostate tissue than aorta tissue. In anesthetized dogs, both compounds suppressed the intraurethral pressure response to phenylephrine to a greater extent than the mean arterial pressure response (Pulito et al., 2000). Other new compounds like RO-70–0004 and RS-100329 (Williams et al., 1999) and some aryldihydropyrimidinones (Barrow et al., 2000) also show an approximately 100-fold selectivity for α1A- versus α1B- and α1D-adrenoceptor subtypes.
α1-Adrenoceptor agonists have been used clinically in the treatment of stress incontinence, acting to increase urethral tone by contracting urethra smooth muscle. Efforts are also being made to identify agents of this kind, selective enough to act on the urethra without causing increases in blood pressure (Ruffolo and Hieble, 1999).
B. α2-Adrenoceptors
1. In Vitro.
At the postjunctional level, α2-adrenoceptors were not found in vitro in the vast majority of the arterial vessels (Table 1). No constrictor activity of α2-adrenoceptor agonists is present in large arteries; when it appears, it is generally restricted to small arteries/arterioles (Docherty and Starke, 1981; Polónia et al., 1985; Aboud et al., 1993; Leech and Faber, 1996; Daniel et al., 1999). Using rabbit polyclonal antibodies for the α2-adrenoceptor subtypes, it was observed that α2A/D- and α2C-adrenoceptors are present in the smooth muscle of mouse tail arteries, the expression of α2C-adrenoceptors being smaller in distal arteries than in proximal arteries (Chotani et al., 2000). In contrast to the difficulty in demonstrating postjunctional α2-adrenoceptors in arteries in vitro, they are consistently found in many isolated veins of different species (De Mey and Vanhoutte, 1981; Constantine et al., 1982; Shoji et al., 1983;Guimarães et al., 1987). This is why the characterization of α2-adrenoceptor subtypes involved in vascular responses to sympathomimetic agonists is being made in veins (or in vivo). The α2A/D-subtype is the predominant one in almost all the veins until now studied: α2A/D in dog saphenous vein (Hicks et al., 1991; MacLennan et al., 1997); α2A/D in rabbit skeletal muscle venules (although predominantly α1D) (Leech and Faber, 1996); and α2A/D (most probably) in the porcine palmar lateral vein (Blaylock and Wilson, 1995). In good agreement with the premise that the α2A- and α2D-adrenoceptors represent species orthologs (Bylund et al., 1995a)—α2A occurring in humans, dogs, pigs, and rabbits and α2Doccurring in rats, mice, and cows—it was observed that postjunctional α2-adrenoceptors of the canine mesenteric vein are predominantly α2A, whereas those of the rat femoral vein are predominantly α2D (Paiva et al., 1999). In human saphenous vein, correlation of α2-adrenoceptors antagonist potency with binding affinity suggests the contribution of the α2C-subtype (Gavin et al., 1997).
2. In Vivo.
α2-Adrenoceptors are essential components of the neural complex system regulating cardiovascular function (Ruffolo et al., 1991) (see Section IV.). When clonidine-like α2-adrenoceptor agonists are intra-arterially administered to wild-type mice, they cause an initial brief pressor effect that is gradually reversed to hypotension at the same time as the animal experiences a severe bradycardia (Link et al., 1996;MacMillan et al., 1996). This is a typical cardiovascular response to intravenous administration of an α2-adrenoceptor agonist also in humans and other species (Hoefke and Kobinger, 1966; Kallio et al., 1989; Bloor et al., 1992). Three main factors are now known to participate in this biphasic response: activation of α2-adrenoceptors on vascular smooth muscle cells that is responsible for the initial and transient hypertensive phase; activation of α2-adrenoceptors in the brainstem leading to a reduction in sympathetic tone with a resultant decrease in blood pressure and heart rate (this hypotensive effect has been the rationale for the use of clonidine in the treatment of hypertension); and a third factor, namely the stimulation of prejunctional α2-adrenoceptors located on sympathetic terminals innervating the vascular smooth muscle cells, an effect that augments the hypotensive effect due to stimulation of central α2-adrenoceptors (Ruffolo et al., 1993;Urban et al., 1995). Bearing in mind that all the three subtypes of α2-adrenoceptors (α2A/D, α2B, and α2C) are present in the vascular tree (Hicks et al., 1991; Gavin et al., 1997; MacLennan et al., 1997; Paiva et al., 1999), which is the involvement of each subtype, if any, in these responses?
The lack of subtype-selective ligands and the cross-reactivity of α2-adrenoceptor agonists with imidazoline receptors made it impossible, until the recent development of knockout animals, to assess the involvement of the individual α2-adrenoceptor subtypes not only in the hypotensive response to α2-adrenoceptor agonists, but also in the general physio-pharmacology of the cardiovascular system.
3. Blood Pressure Regulation in α2-Adrenoceptor-Deficient Mice.
Five mouse strains with genetic alterations of α2-adrenoceptor expression have been generated; they offer new pathways to further identify the role of each subtype (Hein, 1999; Kable et al., 2000): deletion of α2A/D- (α2A/D-knockout), α2B- (α2B-knockout), or α2C-gene (α2C-knockout) (Link et al., 1995, 1996; Altman et al., 1999). More recently, mouse strains lacking α2A/D-, α2B-, or α2C-subtypes were crossed to generate double knockout mice. However, the only viable animals were those lacking both α2A/D- and α2C-adrenoceptors (α2A/DC-knockout) (Hein et al., 1999). Mice have been also developed with a point mutation of the α2A/D-gene (α2A/D-D79N). The D79N mutation causes a replacement of the aspartate with an asparagine residue at position 79, in the second transmembrane domain of the α2A/D-adrenoceptor. In cultured cell lines, the α2A/D-D79N mutant receptor failed to activate K+ currents, but exhibited normal inhibition of voltage-gated calcium channels and cAMP production (Surprenant et al., 1992). It was expected that the expression of this mutation in the intact animal would provide insight into the signal transduction mechanisms mediating the effect of α2A-adrenoceptor stimulation. However, α2A/D-D79N mice showed about an 80% reduction in α2-adrenoceptor binding, as determined by radioligand studies in the brain (MacMillan et al., 1996). Thus, the α2A/D-D79N receptor expressed in vivo showed different characteristics, compared with its expression in vitro, thus behaving as a functional knockout.
In these D79N α2A/D-adrenoceptor mice, the hypotensive response to intra-arterial infusion of α2-adrenoceptor agonists was almost absent, while the initial hypertensive response remained unchanged. This alteration in the cardiovascular response demonstrates that α2A/D-adrenoceptors mediates the brainstem hypotensive response not only to endogenous catecholamines, but also to imidazoline-based α2-adrenoceptor agonists (MacMillan et al., 1996). Accordingly, any hypothetical important role of the so-called imidazoline receptors as mediators of this effect can be ruled out.
As an alternative approach, knockout mice deficient in α2A/D-, α2B-, or α2C-adrenoceptors were developed (Link et al., 1996). All these different strains of knockout animals are viable, fertile, and develop normally (Hein et al., 1998).
In α2B-knockout mice, the initial pressor response to α2-adrenoceptor agonists was abolished and the hypotensive effect occurred immediately and was significantly greater than that observed for control animals showing that the initial hypertensive phase was due to activation of α2B-adrenoceptors on vascular smooth muscle and that the vasoconstrictory α2B-adrenoceptors in the peripheral vasculature counteract the therapeutic hypotensive action of α2-adrenoceptor agonists (Link et al., 1996; Hein et al., 1998). Thus, if α2A/D-adrenoceptor agonists are developed selective enough to avoid α2B-adrenoceptor stimulation, their antihypertensive effect should be enhanced. The bradycardia evoked by α2-adrenoceptor agonists in α2B-knockout mice was not changed.
Mice with a deletion of the α2C-adrenoceptor gene showed no differences from wild-type mice in their hypertensive, hypotensive, and bradycardic responses to α2-adrenoceptor agonists (Link et al., 1996;Hein et al., 1998). Results obtained in α2A/D-knockout mice confirmed and extended those obtained with α2A/D-D79N mice, demonstrating that most of the classical effects ascribed to α2-adrenoceptor agonists are mediated by α2A/D-adrenoceptors. Both α2A/D-D79N and α2A/D-knockout mice failed to become hypotensive in response to exogenous α2-adrenoceptor agonists (MacMillan et al., 1996; Altman et al., 1999). However, whereas the heart rate of α2A/D-D79N mice was not significantly different from their controls—indicating that prejunctional regulation of catecholamine release was preserved—α2A/D-knockout mice had: 1) basal resting heart rates more than 180 beats/min greater than their control littermates (α2A/D-knockout 581 ± 21/min versus wild-type 395 ± 21/min) (Hein et al., 1998); 2) a significant depletion of tissue noradrenaline stores, which can be ascribed to an enhanced release of noradrenaline from sympathetic nerves; and 3) higher plasma noradrenaline levels, as indicated by a 25% reduction in the density of β-adrenoceptors (Altman et al., 1999). All these changes can be explained by a higher basal level of sympathetic tone resulting from the loss of α2A/D-adrenoceptor-mediated inhibition of the vasomotor center.
It is somewhat surprising that resting blood pressure was unaffected in α2A/D-knockout mice. Several factors can be advanced to explain the lack of an hypertensive response to the disappearance of α2A/D-adrenoceptors: first of all, the main vasoconstrictory influence exerted by the sympathetic nervous system is mediated through α1-adrenoceptors (Rohrer et al., 1998a); and some other changes like the redistribution of blood to individual vascular beds (which differ with respect to their α2-adrenoceptor population) (MacMillan et al., 1996) and some compensatory adjustments from the renin-angiotensin and nitric oxide systems may also contribute to keep the resting blood pressure at roughly its normal value (Cavalli et al., 1997; Altman et al., 1999). It is also difficult to interpret the role played by tachycardia; it will depend on the amount of venous blood delivered to the heart. Interestingly, a study in which the effect of antisense to α2A/D-adrenoceptors was checked in the rat, the antisense sequence when given intrathecally, caused an increase of the systolic blood pressure (Nunes, 1995), indicating that in this animal species central α2-adrenoceptors regulating blood pressure belong to the α2A/D-subtype. Moreover, this increase in blood pressure suggests that the centrally mediated hypotensive effect of α2-agonists may be more important in the rat than in the mouse.
C. β-Adrenoceptors
1. In Vitro
a. β1- and β2-Adrenoceptors.
According to Lands and coworkers (1967a,b), the β-adrenoceptors in the peripheral vessels were classified as β2. Later studies using more selective agonists and antagonists showed that relaxation of vascular smooth muscle cells resulted from activation of either β1- or β2-adrenoceptor subtypes, and that the involvement of each subtype depended on the vascular bed and the species (O'Donnell and Wanstall, 1984;Guimarães et al., 1993; Shen et al., 1994, 1996; Begonha et al., 1995). β2-Adrenoceptors represent the predominant subtype in most vascular smooth muscles, although β1-adrenoceptors may contribute to vasodilation (for a review, see Osswald and Guimarães, 1983). In a few vessels, β1-adrenoceptors appear to predominate, e.g., coronary arteries (O'Donnell and Wanstall, 1985;Begonha et al., 1995) and cerebral arteries (Edvinsson and Owman, 1974). In contrast to the heart where a maximal increase in force of contraction is obtained by stimulation of β1-adrenoceptors only and activation of β2-adrenoceptors causes no more than a submaximal effect (Kaumann et al., 1989; Motomura et al., 1990), in the vessels (at least in the veins), the maximum relaxation evoked by β2-adrenoceptors is larger than that evoked by β1-adrenoceptor stimulation (Guimarães and Paiva, 1981c) (Fig. 1). The maximal β-adrenoceptor-mediated relaxation varies from vascular bed to vascular bed (see Section VI.C.) and crucially depends on the level of the tone of the tissue (Guimarães, 1975; Begonha et al., 1995) (Table 3). The maximal β-adrenoceptor-mediated relaxation of the canine veins vary markedly from vein to vein (Table 4). After precontraction with methoxamine, isoprenaline antagonized more than 80% of the precontraction in cephalic, external jugular, azygos, renal, femoral, saphenous, pulmonary, splenic, and the superior part of the inferior vena cava; whereas in the portal, mesenteric, and inferior segments of vena cava, the maximal relaxation antagonized less than 25% of the previous contraction (Furuta et al., 1986).
b. β3-Adrenoceptors.
The participation of a third β-adrenoceptor subtype in β-adrenoceptor-mediated vasodilatation was suggested by results obtained in several studies. Pindolol, a nonselective β-adrenoceptor antagonist with significant agonist activity, caused relaxation of canine isolated perfused mesenteric vessels (Clark and Bertholet, 1983) and rat aorta precontracted with KCl (Doggrell, 1990). In both instances, the vasorelaxant effect of pindolol was not significantly antagonized by propranolol, thus suggesting the presence of a β-adrenoceptor subtype different from the conventional β1- and β2-adrenoceptors, and the effect of isoprenaline was ascribed not only to activation of β2- and β1-adrenoceptors, but also to that of an additional adrenoceptor. Similar propranolol-resistant components to isoprenaline-induced relaxations have been observed in rat carotid artery (Oriowo, 1994; MacDonald et al., 1999), rat mesenteric artery (Sooch and Marshall, 1995), rat aorta (Gray and Marshall, 1992; Oriowo, 1995; Sooch and Marshall, 1997), rat pulmonary artery (Sooch and Marshall, 1996; Dumas et al., 1998), and canine pulmonary artery (Tamaoki et al., 1998; Tagaya et al., 1999). The involvement of β3-adrenoceptors in isoprenaline-induced relaxation of vascular smooth muscle was demonstrated by the use of preferential β3-adrenoceptor agonists and antagonists. In the rat carotid artery, the selective β3-adrenoceptor agonist BRL-37344 and the selective β2-adrenoceptor agonist, salbutamol, were not antagonized by propranolol (100 nM), and pretreatment of the artery segments with BRL-37344 did not desensitize the tissue to the relaxant effect of isoprenaline and salbutamol; it is noteworthy to point out that the pD2 for salbutamol was 5.0, a value that is not consistent with the activation of β2-adrenoceptors (Oriowo, 1994 ). In the same tissue, MacDonald et al. (1999) confirmed the presence of β3-adrenoceptor by the relaxant effects of two selective β3-adrenoceptor agonists, BRL-37344 and ZD-2079. A β3-adrenoceptor-mediated vasorelaxation was also observed in the canine pulmonary artery, an effect that was exerted through a cAMP-dependent pathway (Tagaya et al., 1999).
The presence of β3-adrenoceptors has been reported also in veins. In the rat portal vein, activation of β3-adrenoceptors stimulates L-type Ca2+ channels through a Gαs-induced stimulation of the cyclic AMP/protein kinase A pathway and the subsequent phosphorylation of the channels (Viard et al., 2000).
c. Putative β4-Adrenoceptors.
Although there is now convincing evidence supporting the functional presence of β3-adrenoceptors in vascular tissue, various observations cannot be fully explained by the existence of β1-, β2-, and β3-adrenoceptors only. For example, the propranolol-resistant component of the effect of isoprenaline in rat aorta is not antagonized by either the selective β3-adrenoceptor antagonist SR-59230 (Brawley et al., 2000a) or cyanopindolol, a β3-adrenoceptor antagonist in the guinea pig ileum (Blue et al., 1989; Oriowo, 1994;MacDonald et al., 1999). Similarly, in rat aorta, the selective β3-adrenoceptor agonist CGP-12177 was resistant to the blockade by SR-59230 (Brawley et al., 2000a). Also in agreement with these data, cyanopindolol did not inhibit the relaxant effect of isoprenaline in rat aorta, although it did antagonize the effect of isoprenaline at β3-adrenoceptors in the distal colon and fundic strips (Oriowo, 1994). Furthermore, the order of potency for unconventional partial adrenoceptor agonists in rat aorta was contrary to that obtained at β3-adrenoceptors in other tissues (Oriowo, 1994; Brawley et al., 2000a). According to these data, the propranolol-resistant component of the effect of isoprenaline was suggested to be mediated by the putative β4-adrenoceptor. However, the selective β3-adrenoceptor agonist BRL-37344, which does not activate this putative β4-adrenoceptor (Malinowska and Schlicker, 1996; Galitzky et al., 1998), caused concentration-dependent relaxation in the rat aorta (Brawley et al., 2000a). Also in disagreement with the β4-adrenoceptor hypothesis was that cyanopindolol showed higher potency than CGP-12177 in rat aorta, whereas it had been shown to have lower potency at the putative β4-adrenoceptor (Malinowska and Schlicker, 1996). The identity of the receptor mediating the β3-adrenoceptor–independent effects of CGP-12177 was clarified in a recent study in which it was shown that activation of adenylyl cyclase by CGP-12177 in β3-adrenoceptor-knockout mice is mediated by β1-adrenoceptors (Konkar et al., 2000a). The same authors showed that activation of β1-adrenoceptors by CGP-12177 or LY-362884 (a second aryloxypropanolamine) is significantly more resistant to inhibition by β-adrenoceptor antagonists, compared with activation by catecholamines and suggests that catecholamines and aryloxypropanolamines interact with two distinct active conformational states of the β1-adrenoceptor (Konkar et al., 2000b): one state that is responsive to catecholamines and is inhibited with high affinity by CGP-12177 and LY-362884, and a novel state that is activated by aryloxypropanolamines but is resistant to inhibition by classical β-adrenoceptor antagonists (Konkar et al., 2000b). Results leading to a similar conclusion were obtained in a rat model of cardiac failure: desensitization to isoprenaline and CGP-12177 after myocardial infarct and resensitization (after pertussis toxin treatment) occurs in parallel, suggesting that the β1- and the putative β4-adrenoceptor use the same pathway. Furthermore, antagonist affinity studies confirmed that drugs acting at β1-adrenoceptors also interact with putative β4-adrenoceptors with approximately 100 times lower affinity, suggesting that CGP-12177 causes its cardiac effects by interacting with a low affinity state of the β1-adrenoceptor (Kompa and Summers, 2000). Recently Kaumann et al. (2001) showed that, in double β1-/β2-adrenoceptor knockout mice CGP-12177 did not at all affect force of contraction or heart rate, indicating an obligatory role of β1-adrenoceptor for effects evoked by stimulation of the putative β4-adrenoceptor. Accordingly, it is quite likely that there exist no fourth β-adrenoceptor, but the effects of CGP-12177 are due to an atypical interaction of this compound with the β1-adrenoceptor.
2. In Vivo.
β-Adrenoceptor-mediated vasodilation is thought to play an important physiological role in the regulation of vascular tone. Stimulation of peripheral β-adrenoceptors leads to relaxation of the vascular smooth muscle, thereby controlling the peripheral vascular resistance and consequently the distribution of blood to the different organs. During exercise, for example, activation of β-adrenoceptors contributes to the increased blood flow to skeletal muscle.
Although much has been learned about the role of the individual β-adrenoceptor subtype using classical pharmacological approaches in vitro, experiments in awake and unrestrained animals are of crucial importance to determine the real influence of factors coming into play, since reflex pathways can be obscured in anesthetized animals and are absent in the in vitro preparations. Experiments carried out in vivo in wild-type animals of several species had already indicated that the activation of β2- and β1-adrenoceptors led to relaxation of vascular smooth muscle of both arteries and veins (Taira et al., 1977; Vatner et al., 1985). However, it is now widely accepted that β-adrenoceptors other than the classical β1- and β2-adrenoceptors are also involved in the relaxation of the vasculature.
β3-Adrenoceptors were recently shown to play a role in regulating peripheral vasodilatation, although this role was highly species-dependent (Shen et al., 1996) (Fig.2).
In conscious dogs, the selective β3-adrenoceptor agonist BRL-37344 caused a long-lasting vasodilation even in the presence of propranolol, whereas isoprenaline was totally ineffective when given in a dose that was equieffective prior to propranolol. This vasodilation occurred primarily in the skin and fat, and persisted in the presence of a complete blockade of all known neural or hormonal pathways, indicating that probably it was due to activation of β3-adrenoceptors (Tavernier et al., 1992;Berlan et al., 1994; Shen et al., 1994). Similarly, BRL-26830, another selective β3-adrenoceptor agonist caused a marked increase in blood flow to brown adipose tissue in the anesthetized rat (Takahashi et al., 1992). Although these results support the view that the effect of BRL-26830 is β3-adrenoceptor-mediated, they do not provide an unequivocal demonstration. The increase in blood flow may well be secondary to an augmented metabolic process (Shen and Claus, 1993), since BRL-37344 causes marked increases in the plasma levels of free fatty acids and insulin. In the rat, the selective β3-adrenoceptor agonist CL-316243 induced a marked increase in both islet blood flow and plasma insulin level, and these increases were abolished by bupranolol, a β1,β2,β3-adrenoceptor antagonist but not by nadolol—a β1,β2-adrenoceptor antagonist, indicating that β3-adrenoceptors caused a vasodilation of microvessels in the islets of Langerhans (Atef et al., 1996).
Experiments carried out in rats, dogs, and monkeys showed a graded decrease in responsiveness to CL-316243 from the dog to the rat to the monkey, and suggest that β3-adrenoceptor agonists do not evoke cardiovascular effects in primates (Shen et al., 1996). Furthermore, it was also shown that while BRL-37344 was selective enough to discriminate between β3-adrenoceptors and the other two β-adrenoceptor subtypes (β1 and β2) in the dog (until now no data has been reported on the existence of putative β4-adrenoceptors in the dog), only CL-316243 was able to provide a similar discrimination in the rat (Shen et al., 1994, 1996). This species-dependent selectivity of β3-adrenoceptor agonists (Shen et al., 1996) is one possible explanation for the lack of effect of β3-adrenoceptor agonists in primates. Another obvious explanation is that there are few or no β3-adrenoceptors in primates. The presence of functional β3-adrenoceptors in primates either in vessels or in the adipocytes is still a debatable question. Although some authors did not find evidence supporting this hypothesis (Zaagsma and Nahorski, 1990; Langin et al., 1991; Rosenbaum et al., 1993), many others found evidence supporting their existence in humans (Arner, 1995; Clément et al., 1995; Lönnqvist et al., 1995; Lowell and Flier, 1995; Walston et al., 1995) (see Section IX.).
Most recently, the compound L-750355 was identified as a potent and selective β3-adrenoceptor agonist in cloned human and rhesus monkey β3-adrenoceptors expressed in Chinese hamster ovary cells (Forrest et al., 2000). Furthermore, it was shown that L-750355 stimulates lypolisis in isolated human and rhesus adipocytes in vitro and causes tachycardia (propranolol-sensitive) and hyperglycerolemia (propranolol-resistant) in anesthetized rhesus monkeys (Forrest et al., 2000).
In summary, stimulation of β3-adrenoceptors by the selective β3-adrenoceptor agonists BRL-37344 and CL-316243 in mice, rats, and dogs causes long-lasting reductions in both blood pressure and total peripheral resistance, indicating that β3-adrenoceptors are present in the vasculature (Tavernier et al., 1992; Shen et al., 1994, 1996;Rohrer et al., 1999). In contrast, in conscious monkeys and baboons, neither drug caused significant cardiovascular effects (Shen et al., 1996). Furthermore, CL-326243 appears to be a more selective β3-adrenoceptor agonist than BRL-37344.
The recent progress in the development of knockout mice made it possible to selectively disrupt the gene for each of the β-adrenoceptor subtypes (β1, β2, and β3) (Susulic et al., 1995; Rohrer et al., 1996; Revelli et al., 1997; Chruscinski et al., 1999); or for both β1- and β2-adrenoceptors (Rohrer et al., 1999). Animals lacking β1- or β2- or β3-adrenoceptors had normal prenatal development, appeared grossly normal, were fertile, and showed normal resting cardiovascular parameters (Chruscinski et al., 1999). One should bear in mind, as already pointed out, that the lack of a given receptor from the conception may lead to compensatory changes.
In β1-adrenoceptor-knockout mice, basal cardiovascular indices were unchanged and the capacity to respond to stresses like exercise was normal (Rohrer et al., 1996, 1998a), although the high variability in blood pressure in conscious unrestrained mice may have obscured a slight trend toward lower values in β1-adrenoceptor-knockout animals (Desai et al., 1997; Rohrer et al., 1998b). The hypotensive response to isoprenaline was not significantly different from that in wild-type mice; although the percent increase in heart rate after isoprenaline was significantly smaller in β1-adrenoceptor-knockout than in wild-type animals (Rohrer et al., 1998b), and the nonselective β-adrenoceptor antagonist propranolol caused a modest pressor response in both wild-type and β1-adrenoceptor-knockout mice (Rohrer et al., 1998b). In β2-adrenoceptor-knockout mice, the resting cardiovascular parameters (heart rate and blood pressure) appeared completely unaltered. The major effects of β2-adrenoceptor gene deletion were observed only during exercise. Apparently, β2-adrenoceptor-knockout mice tolerate the workload better than wild-type controls. However, they were hypertensive during exercise, suggesting an imbalance between the vasoconstrictive and vasorelaxant effects of endogenous catecholamines (Chruscinski et al., 1999). In β2-adrenoceptor-knockout mice, the hypotensive response to the nonselective β-adrenoceptor agonist isoprenaline is significantly attenuated, confirming that β2-adrenoceptors play an important role in vascular relaxation and indicating that part of the hypotensive response to isoprenaline depends on the activation of other β-adrenoceptor subtypes.
In general terms, in double knockout mice lacking β1- and β2-adrenoceptors, changes in basal physiological cardiovascular functions are virtually nonexistent. However, functional deficits in vascular reactivity are revealed when β-adrenoceptors are stimulated by β-adrenoceptor agonists or exercise. In double β1- and β2-adrenoceptor-knockout mice, the hypotensive response to the selective β3-adrenoceptor agonist CL-316243 is markedly enhanced (Fig.3) (Rohrer et al., 1999). This enhancement may be ascribed to a vascular β3-adrenoceptor up-regulation, since it was demonstrated that β1-adrenoceptors are up-regulated in the adipose tissue of β3-adrenoceptor-knockout mice (Susulic et al., 1995). Conversely, in the ileum, β3-adrenoceptor-knockout mice, β1-adrenoceptors functionally compensate for the lack of β3-adrenoceptors (Hutchinson et al., 2001). Deficiencies in β-adrenoceptor signaling in double β1- and β2-adrenoceptor-knockout mice can be compensated for by increases in the density and/or signaling efficiency of other β-adrenoceptor subtypes (Rohrer et al., 1999).
The involvement of β3-adrenoceptors in vascular relaxation is more and more widely accepted. According to Liggett et al. (1993), β3-adrenoceptors, unlike β1- and β2-adrenoceptors, lack regulatory phosphorylation sites for G-protein receptor kinases, a characteristic that increases the resistance to agonist-evoked desensitization. Thus, under conditions of a persistent overstimulation of the sympathetic nervous system, β1- and β2-adrenoceptors are desensitized and β3-adrenoceptors may represent a functional alternative (Rohrer et al., 1999). The fact that double knockout β1- and β2-adrenoceptor animals are supersensitive to β3-adrenoceptor agonists is consistent with this tentative explanation.
Bearing in mind all these data, particularly those from knockout animals, it seems that β1-adrenoceptors are those that predominantly regulate cardiac contractility and rate of heart, β2-adrenoceptors are those that predominantly mediate the vasodilation evoked by sympathomimetic agonists, and β3-adrenoceptors those that predominantly control lipolysis in adipose tissue. This is an oversimplified conclusion, which, however, may be valid for the majority of species.
IV. Prejunctional Adrenoceptors
A. α2-Adrenoceptors
At the prejunctional level, α2-adrenoceptors have been found in vitro in every vascular tissue (arteries and veins) until now studied where they mediate a negative modulation of the release of noradrenaline (Starke, 1987; Langer, 1997). Experiments in isolated organs support the view that the α2A/D-subtype is the principal prejunctional α2-adrenoceptor; some studies indicate that, in certain tissues, including the heart and some vessels, another α2-subtype might also be involved in the regulation of the release of noradrenaline (Oriowo et al., 1991; Limberger et al., 1992; Guimarães et al., 1997;Trendelenburg et al., 1997; Docherty, 1998; Ho et al., 1998). Results obtained in knockout mice clearly confirmed this hypothesis. In the vas deferens, for example, it was observed that maximal inhibition by α2-adrenoceptor agonists of the electrically evoked contractions was reduced to 50% in mice lacking α2A/D, whereas the effect of these agonists in α2C-adrenoceptor-deficient mice was unchanged. Furthermore, in mice lacking both the α2A/D- and the α2C-adrenoceptors, the prejunctional effect of α2-adrenoceptor agonists was abolished, indicating that the residual 50% response to α2-adrenoceptor agonists in α2A/D-knockout animals was due to prejunctional α2C-adrenoceptors (Hein et al., 1998). Also in the heart atria and in the brain cortex, deletion of α2A/D-adrenoceptors reduced, but did not abolish, the inhibitory effect of the α2-selective agonist UK-14304, indicating that a second α2-autoreceptor operates in both sympathetic and central adrenergic neurons. However, the loss inE max of α2-adrenoceptor agonists was smaller in the heart than in the brain, supporting the view that, in the peripheral tissues, prejunctional non-α2A/D-adrenoceptors are functionally more important (Hein et al., 1999; Trendelenburg et al., 1999). Again, in α2A/DC-knockout mice, the concentration-dependent inhibition of noradrenaline release caused by α2-adrenoceptor agonists in the atria was abolished, indicating that the α2C-adrenoceptor is the second type involved in regulating noradrenaline release (Hein et al., 1999). Thus, the hypothesis that more than one α2-adrenoceptor subtype might be present at the prejunctional level in the same tissue was confirmed in knockout animals.
According to Link et al. (1992), noradrenaline has a higher affinity for α2C- than for α2A/D-adrenoceptors. In mouse atria, Hein et al. (1999) confirmed that noradrenaline was more potent on α2C- than on α2A/D-adrenoceptors (EC50= 16 nM, 20 nM, and 156 nM in wild-type, α2A/D-knockout, and α2C-knockout animals, respectively) and showed that α2C-adrenoceptors inhibit transmitter release at low levels of sympathetic tone and that α2A/D-adrenoceptors are required to control release at higher levels of sympathetic tone.
Unfortunately there are no studies in isolated vascular tissues from knockout animals. However, the calculation of the correlation between the pK i values of several antagonists at some canine and human vessels and their pK d values at prototypical α2A, α2B, α2C, and α2Dradioligand binding sites (Altman et al., 1999) shows that, as in other peripheral tissues, the main modulatory role is played by α2A/D-adrenoceptors and suggests that more than one α2-adrenoceptor subtype participate in the feedback inhibition of transmitter release (Table5). This calculation still suggests that α2B or α2C or both may be involved in this modulatory influence. Additionally, it can be concluded that the existence of a second type of prejunctional receptor in vascular tissues of α2A/D-knockout mice does not result from a compensatory mechanism (Altman et al., 1999), since these results were obtained in wild-type animals.
There is no doubt that in vessels the control mechanism of noradrenaline release is highly active at physiological frequencies of electrical stimulation. Evidence supporting the possibility of a tonic physiological role of a negative feedback control of noradrenaline release appeared before the discovery of prejunctional α-adrenoceptors. The first report pointing in this direction was that by Malik and Muscholl (1969), who showed that noradrenaline, in doses that did not alter basal resistance, slightly reduced the response of the perfused mesenteric artery to sympathetic nerve stimulation. This was confirmed by the observation, in the rabbit pulmonary artery, that the inhibitory influence of a fixed concentration of an exogenous agonist was not as effective at high as at low frequencies of stimulation (Starke and Endo, 1975). Indeed, at the higher frequencies, the negative influence of α-adrenoceptor agonists is less because the prejunctional α-adrenoceptors are already more intensely activated by endogenous noradrenaline (Vizi et al., 1973; McCulloch et al., 1975).
Numerous pharmacological findings obtained by many authors in several species, including humans, support the view that a negative feedback mechanism mediated by prejunctional α2-adrenoceptor operates under physiological conditions of noradrenergic neurotransmission (for reviews, see Starke, 1977, 1987; Langer, 1981, 1997). However, this hypothesis was contested by Kalsner (1982) and Kalsner and Westfall (1990). The latest observations in knockout animals clearly show that prejunctional α2-adrenoceptors are really autoreceptors that accomplish a physiological function.
B. β-Adrenoceptors
The release of noradrenaline from sympathetic varicosities is modulated by a large number of prejunctional auto- and heteroreceptors. Facilitatory β-adrenoceptors have been shown in various sympathetically innervated tissues of many species both in vitro (Adler-Graschinsky and Langer, 1975; Majewski, 1983; Misu and Kubo, 1986; Encabo et al., 1996) and in vivo (Boudreau et al., 1993; Tarizzo et al., 1994; Vila and Badia, 1995). Those receptors, which have been demonstrated to be of the β2-subtype (Dahlöf et al., 1978; Guimarães et al., 1978; Göthert and Hentrich, 1985; Coopes et al., 1993), may be activated by circulating adrenaline (Stjärne and Brundin, 1975) and/or by adrenaline taken up by and released from sympathetic varicosities as a cotransmitter (Majewski et al., 1981a,b; Misu et al., 1989; Tarizzo and Dahlöf, 1989). Very recently, the existence of prejunctional β2-adrenoceptors located on nerve terminals that release nitric oxide (NO: NOergic nerve terminals) of the porcine basilar arteries was proposed by Lee et al. (2000). According to these authors, noradrenaline released from sympathetic nerves upon application of nicotine, acts on prejunctional β2-adrenoceptors of NOergic nerve terminals to release NO, resulting in vasodilatation. In humans, there is also evidence for facilitatory β-adrenoceptors on noradrenergic nerve endings, both in vitro (Stjärne and Brundin, 1975; Stevens et al., 1982) and in vivo (Brown and Macquin, 1981; Vincent et al., 1982;Blankestijn et al., 1988). However, whereas prejunctional α2-adrenoceptors (of the α2A/D- and α2C-subtype) play an important physiological role in the feedback inhibition of neurotransmitter release (Starke, 1987; MacMillan et al., 1996; Altman et al., 1999; Trendelenburg et al., 1999), the physiological role of the facilitatory prejunctional β-adrenoceptors remains controversial (Abrahamsen and Nedergaard, 1989, 1990, 1991; Floras, 1992;Apparsundaram and Eikenburg, 1995; Coopes et al., 1995). A hypothetical pathological implication of the facilitatory prejunctional β2-adrenoceptors originated whenGuimarães et al. (1978) first showed that, in canine saphenous vein, propranolol was able to reduce the electrically evoked tritium overflow after preloading of the venous tissue with3H-adrenaline, but not when it had been preloaded with 3H-noradrenaline. Confirmatory results were obtained in different in vitro vascular peparations (Majewski et al., 1981a; Misu et al., 1989; Rump et al., 1992) and also in vivo studies: in the anesthetized rabbit (Majewski et al., 1982); in the anesthetized dog (Boudreau et al., 1993); in the pithed rat (Vila and Badia, 1995); and in the demedulated rat (Coopes et al., 1993, 1995). On the basis of these results, the hypothesis was put forward that high levels of adrenaline within the synaptic cleft could either directly stimulate prejunctional β2-adrenoceptors or be taken up by sympathetic nerves, and then be coreleased and activate prejunctional β2-adrenoceptors. In both cases, stimulation of this positive feedback loop would lead to an increased release of noradrenaline, which might represent an early step in the development of hypertension (Majewski et al., 1981b; Misu et al., 1990; for a review, see Floras, 1992). Depletion of plasma adrenaline by surgical adrenal demedullation attenuates the development of hypertension in 4-week-old SHRs (Borkowski and Quinn, 1985), and this effect is antagonized by depot implants of adrenaline. Moreover, the prohypertensive effect of adrenaline in demedullated SHRs was abolished by concomitant treatment with β2-adrenoceptor antagonists (Borkowski and Quinn, 1985). The development of hypertension was attenuated only in SHRs demedullated at 6 weeks of age or younger, indicating that a period of critical sensitivity of prejunctional β2-adrenoceptors to the facilitatory effect of catecholamines may exist (Borkowski, 1991). In humans, episodes of sympathoadrenal activation repeatedly causing increases in plasma adrenaline concentration might, by direct (as a hormone), indirect (as a cotransmitter), or both ways, initiate or facilitate the development of primary hypertension (Brown and Macquin, 1981; Blankestijn et al., 1988). Although attractive and based on a large amount of experimental work, this theory has not received conclusive support so far. Floras et al. (1988) reported that 30 min after a local infusion of adrenaline into the forearm of volunteers, there was a facilitation of neurogenic vasoconstriction that was due to a delayed facilitation of noradrenaline release caused by the previous infusion of adrenaline. Stein et al. (1997), under identical conditions, did not observe any delayed facilitatory effect on noradrenaline spillover in the forearm of normotensive or borderline hypertensive subjects. However, these authors observed an increase in systemic noradrenaline spillover 30 min after the infusion of adrenaline, suggesting that there is a delayed facilitatory response to adrenaline in specific organs, the identification of which would be of importance in elucidating the role of this mechanism in the pathogenesis of hypertension. In a comparative study involving patients with longstanding essential hypertension and normotensive control subjects, Chang et al. (1994) evaluated the kinetics of noradrenaline by measuring in the forearm the appearance rate of noradrenaline in plasma and the spillover of noradrenaline into plasma before and after the infusion of adrenaline and found no differences between the two groups. Similarly, in 40 healthy volunteers, loading of sympathetic nerve terminals of the human forearm with adrenaline did not augment subsequent neurogenic vasoconstriction or noradrenaline release in response to sympathetic stimulation (Goldstein et al., 1999). Using a different methodological approach, Thompson et al. (1998) also obtained negative results: isometric handgrip contraction evoked similar responses in total and cardiac noradrenaline spillovers, and in muscle sympathetic activity before and after an infusion of adrenaline. The influence of muscle contraction and blood flow on noradrenaline and adrenaline spillover was also studied in the in situ canine gracilis muscle, and spillover of both amines was observed (Lavoie et al., 2000 ). Since adrenaline is not synthesized locally, adrenaline spillover means that it was taken up from the circulation, stored in the vesicles, and then re-released with noradrenaline (Lavoie et al., 2000).
Although convincing evidence for the positive feedback loop hypothesis is still lacking, it may be inadvisable to regard it as disproven. As suggested by Folkow (1982), the development of hypertension by this mechanism may be restricted to a specific subset of individuals genetically predisposed to high blood pressure. As pointed out byFloras (1992), there are no prospective evaluations of the predictive value of augmented plasma adrenaline concentrations in childhood and adolescence as an indicator for later hypertension development; it is possible that, as shown for SHRs (Borkowski 1991), a period of critical sensitivity of prejunctional β2-adrenoceptors to catecholamines exists during which the pathological development starts. If this were true for humans and if it were known into which age this critical period fell, an adequate treatment might avoid the progress toward an established hypertension.
V. Endothelial Adrenoceptors
A. α2-Adrenoceptors
It is now widely accepted that vascular endothelium plays an important role in the function of the cardiovascular system (Moncada et al., 1991a; Vaz-da-Silva et al., 1996; Busse et al., 1998). Functional evidence suggesting that α2-adrenoceptors play a role in the physiology of the vasculature was first reported by Cocks and Angus (1983), who showed that, in isolated coronary, renal, and mesenteric arteries, noradrenaline and clonidine caused relaxation that was inhibited by selective α2-adrenoceptor antagonists and was eliminated by removal of the endothelium. Similar results were obtained in several isolated arteries (Egleme et al., 1984; Angus et al., 1986) and veins (Miller and Vanhoutte, 1985). Soon after the discovery of the EDRF (Furchgott and Zawadzki, 1980)—later on identified with NO—(Ignarro et al., 1987; Palmer et al., 1987), it was demonstrated that activation of α2-adrenoceptors on endothelial cells stimulates the release of NO, an action that would tend to attenuate vasoconstriction produced by activation of postjunctional vascular α1-adrenoceptors (Angus et al., 1986; Vanhoutte and Miller, 1989; Richard et al., 1990). Furthermore, it was suggested that endothelial α2-adrenoceptors mediate release of EDRF in coronary microvessels (Angus et al., 1986). Thus, it appears that α2-adrenoceptor agonists do indeed have the capability of modulating vascular responsiveness via stimulation of the release of NO in both large arteries and microcirculation. Furthermore, it was reported that noradrenaline-induced release of nitric oxide is enhanced in mineralocorticoid hypertension (Bockman et al., 1992), indicating that endothelial α2-adrenoceptors may play an important role in the regulation of vascular tone not only in physiological, but also in pathological conditions.
Which α2-adrenoceptor subtype is responsible for this modulatory influence? The first study aiming at characterizing the α2-adrenoceptor subtypes present on vascular endothelium was carried out in pig coronary arteries and showed that the endothelium of this vessel possesses both α2A/D- and α2C-adrenoceptors, the latter predominating (77% of α2C versus 23% of α2A/D). However, despite the prominent presence of α2C-adrenoceptors, the α2A/D-adrenoceptor subtype is the one mediating endothelium-dependent relaxation (Bockman et al., 1993). Interestingly, it was shown that in the rat mesenteric artery the α2-adrenoceptor that is coupled to endothelium-dependent NO-mediated relaxation belongs to the α2A/D-subtype appearing in its α2D-version (Bockman et al., 1996). Also in the endothelium of different species, the α2A/D-adrenoceptors serve the same function (Bylund et al., 1995a). Contrary to what was expected, cAMP is not involved in the signal transduction pathway for α2A/D-adrenoceptor-mediated NO formation (Bockman et al., 1996).
B. β-Adrenoceptors
It is now widely accepted that β-adrenoceptors exist on endothelial cells and contribute to the regulation of vasomotor tone. The role these receptors play, the mechanisms by which this role is played and the β-adrenoceptor subtypes that are involved are still debatable questions. Some of the first studies on the existence of endothelial β-adrenoceptors and some others carried out later on did not find evidence supporting their existence. Removal of the endothelium or inhibitors of NO synthase were found to have no influence on isoprenaline-evoked relaxations in rat aorta (Konishi and Su, 1983), canine coronary arteries (Cohen et al., 1983, 1984; White et al., 1986), rat carotid artery (Oriowo, 1994), or human internal mammary artery (Molenaar et al., 1988). In contrast, many other authors reported that removal of endothelium reduces the relaxations caused by β-adrenoceptor agonists in several isolated vessels from different species, including humans (Grace et al., 1988; Kamata et al., 1989;Dainty et al., 1990 ; Gray and Marshall, 1992; Delpy et al., 1996; Toyoshima et al., 1998; Ferro et al., 1999; Trochu et al., 1999;Brawley et al., 2000a,b; Vanhoutte, 2000). Surprisingly, in the same preparation (the thoracic aorta), different authors found opposite results. This discrepancy may be ascribed to one or more of the following factors: 1) the agent used to precontract the vessel was either noradrenaline (which also activates β-adrenoceptors) or phenylephrine (which lacks affinity for β-adrenoceptors) (Guimarães, 1975); 2) the level of the precontraction at which the β-adrenoceptor-mediated relaxation might not be the same and the magnitude of the relaxant effect is critically dependent on the extent of the pre-existing tone (Guimarães, 1975); and 3) the β3-adrenoceptor agonist used may be more or less active on β1- and β2-adrenoceptors.
According to Eckly et al. (1994), the reduction in response to isoprenaline after pretreatment with l-NAME or endothelial removal can be explained by the fact that the precontraction of the vessel is greater than in control tissues due to the disappearance of NO production under resting conditions. This enhancement of the precontraction would counteract the relaxation (Guimarães, 1975). However, it has been shown that endothelial removal does not consistently increase the preconstriction caused by noradrenaline or phenylephrine (Delpy et al., 1996; Brawley et al., 2000b). Many other kinds of evidence also support the view that β-adrenoceptors are present in endothelial cells and mediate relaxing responses in which NO is involved. First of all, the presence of β-adrenoceptors was confirmed by radioligand binding studies in cultured bovine aortic endothelial cells (Steinberg et al., 1984), by autoradiography in human cardiac endocardium (Buxton et al., 1987), in the endothelium of internal mammary artery and saphenous vein (Molenaar et al., 1988), and by biochemical data obtained in cultured human umbilical vein endothelial cells (Ferro et al., 1999). Furthermore, in vivo studies in cat hindlimb (Gardiner et al., 1991), canine coronary artery (Parent et al., 1993), and newborn pig pial arteries (Rebich et al., 1995) support a role of vascular endothelium in β-adrenoceptor-mediated relaxation. In humans, it was also found that forearm blood flow increases by infusion of either isoprenaline or salbutamol into the brachial artery, and coinfusion of the nitric oxide synthase inhibitorl-NMMA blocks this response to either drug (Dawes et al., 1997). Additionally, it had been shown that relaxant responses of the rat aorta to isoprenaline are inhibited by methylene blue and hemoglobin (Grace et al., 1988), indicating that the endothelium-dependent NO/cGMP system may be activated by stimulation of β-adrenoceptors (Grace et al., 1988; Gray and Marshall, 1992; Iranami et al., 1996).
The second point concerns the role played by endothelial β-adrenoceptors and the mechanism through which they induce their effects. Pretreatment with l-NAME and endothelium removal exert a similar inhibitory influence on isoprenaline-evoked relaxation, and the combination of the two procedures has no additional effect, compared with either treatment alone (Gray and Marshall, 1992; Ferro et al., 1999). On the basis of this evidence, these authors concluded that β-adrenoceptor-mediated vasorelaxation is totally endothelium-dependent: isoprenaline-evoked relaxation is due to the elevation of cyclic AMP caused by β2-adrenoceptor stimulation, and this elevation activates the l-arginine/NO system and gives rise to vasorelaxation (via cGMP formation) (Gray and Marshall, 1992; Ferro et al., 1999). However, other authors found little or no effect of endothelium removal on isoprenaline-evoked relaxations (Konishi and Su, 1983; Moncada et al., 1991b; Eckly et al., 1994; Satake et al., 1996). The hypothesis that incomplete removal of endothelium might account for some remaining relaxation to isoprenaline (Gray and Marshall, 1992) can be discarded at least in some cases in which part of the relaxation to isoprenaline remained, although removal of endothelium had abolished acetylcholine-evoked relaxation (Brawley et al., 2000b). In these cases, treatment with l-NAME of the endothelium-free preparations caused no further effect on the isoprenaline-evoked relaxations (Brawley et al., 2000b). Thus, it appears that the relaxant effect of isoprenaline involves two components: one endothelium-dependent and another endothelium-independent (Brawley et al., 2000b). The endothelium-dependent component is triggered by β-adrenoceptor activation and leads to the promotion of NO production/release (Gray and Marshall, 1992; Ferro et al., 1999;Brawley et al., 2000b), or to some kind of enhancement of smooth muscle β-adrenoceptor-mediated relaxation by basal release of NO (Grace et al., 1988; Delpy et al., 1996).
A third point refers to the endothelial β-adrenoceptor subtype(s) involved in isoprenaline-evoked relaxation of the vascular smooth muscle. It is now recognized that β-adrenoceptors located in the endothelium play an important role in the relaxant response to isoprenaline, since the nonselective β1-and β2-adrenoceptor antagonist propranolol antagonized this relaxant effect (Oriowo, 1995; Sooch and Marshall, 1996; Brawley et al., 2000a,b). However, recent studies carried out in humans—either in umbilical veins in vitro (Ferro et al., 1999) or in the forearm in vivo (Dawes et al., 1997)—showed that vasorelaxation to isoprenaline was abolished by the selective β2-adrenoceptor antagonist ICI-118551 and remained unchanged in the presence of the β1-adrenoceptor antagonist CGP-20712, indicating that as in the vascular smooth muscle cells (Lands et al., 1967a,b), the endothelial β-adrenoceptors are totally or at least predominantly of the β2-subtype (Dawes et al., 1997; Ferro et al., 1999). Furthermore, it was observed that, afterl-NAME treatment or removal of endothelium, relaxant responses to isoprenaline were still unaffected by propranolol, suggesting that they were mediated by β3- and/or the low-affinity state of β1-adrenoceptors, formerly proposed as putative β4-adrenoceptors (Brawley et al., 1998).
Additionally, it was shown that relaxation of rat thoracic aorta was also caused by selective β3-adrenoceptor agonists like CGP-12177 (Mohell and Dicker, 1989), cyanopindolol (Engel et al., 1981), ZD-2079 (Grant et al., 1994), ZM-215001 (Tesfamariam and Allen, 1994), and SR-58611 (Trochu et al., 1999), further supporting the presence of β3-adrenoceptors (Brawley et al., 2000a,b).
Endothelial removal or pretreatment with l-NAME significantly reduced the relaxation caused by isoprenaline or SR-58611 (Trochu et al., 1999), but had less effect on the relaxation caused by another selective β3-adrenoceptor agonist BRL-37344 (MacDonald et al., 1999) or CGP-12177 (Brawley et al., 2000a,b). Furthermore, sodium nitroprusside enhanced isoprenaline effects as previously reported (Maurice and Haslam, 1990), but had little or no effect on the response to CGP-12177 (Brawley et al., 2000a,b) or on the relaxation to isoprenaline in the presence of propranolol. After l-NAME had reduced responses to both isoprenaline and CGP-12177, sodium nitroprusside restored them, but the contribution of NO to the atypical β-adrenoceptor-mediated response was less than to β1- and β2-adrenoceptor-mediated relaxation (Brawley et al., 2000a,b). Both of these procedures indicate that exogenously applied NO interacts with the β1- and β2-adrenoceptor signaling pathway to a greater extent than with the non-β1-/β2-adrenoceptor pathway. However, the non-β1-/β2-adrenoceptor-mediated component of the response to isoprenaline appears to be partially endothelium-dependent, since l-NAME or endothelium removal attenuated isoprenaline relaxation in the presence of propranolol (Shafiei and Mahmoudian, 1999; Trochu et al., 1999; Brawley et al., 2000b).
All of these findings show that the endothelium/NO pathway modulates β1- and β2-adrenoceptor-mediated responses in rat aorta to a greater extent than non-β1-/β2-adrenoceptor-mediated responses (Brawley et al., 1998; MacDonald et al., 1999) and indicate that non-β1-/β2-adrenoceptors are present in the endothelium of some mammalian arteries.
Which non-β1-/β2-adrenoceptor? This is still a difficult question to answer. First of all, in the rat pulmonary vessels, several β3-adrenoceptor agonists (SR-58611, SR-59119, and SR-59104) caused relaxant effects. However, only the effect of SR-59104 was antagonized by the selective β3-adrenoceptor antagonist SR-59230 (Dumas et al., 1998). In the rat thoracic aorta, β3-adrenoceptors are mainly located on endothelial cells, and act in conjunction with β1- and β2-adrenoceptors to mediate relaxation through activation of an NO synthase pathway and subsequent increase in cyclic GMP levels (Trochu et al., 1999).
VI. Distribution of Vascular Adrenoceptors
A. Localization in Relation to Sympathetic Nerve Terminals
In vascular tissue, α- and β-adrenoceptors are not situated close to each other (Guimarães et al., 1981a,b; Guimarães, 1982), such that when there is a change in the concentration of circulating adrenaline coming from the blood or of noradrenaline coming from the sympathetic nerve terminals, it does not affect α- and β-adrenoceptors equally; there are two different biophases for sympathomimetic agonists: one for α-adrenoceptors around the nerve terminals, where the concentration of the agonist available for α-effect is mainly governed by uptake into these terminals; and one for β-adrenoceptors in the neighborhood of catechol-O-methyl transferase whose activity is the main factor determining the concentration of the agonist available for the β-effect (for a review, see Guimarães, 1982; Guimarães et al., 1982). This will contribute to the fact that the different vessels have different sensitivities to sympathomimetic amines, such that some of them may be under the control of circulating catecholamines, whereas others are not.
In 1980, Yamaguchi and Kopin showed that, in the pithed rat, the pressor response to sympathetic nerve stimulation is the result of activation of α1-adrenoceptors, whereas the pressor effects of the exogenous catecholamines are medited by α2-adrenoceptors. This conclusion was consistent with the suggestion that postjunctional vascular α2-adrenoceptors might be located at extrajunctional sites (Langer et al., 1981). This differential location of α1- and α2-adrenoceptors in relation to the nerve terminals was confirmed in conscious rabbits, where it was shown that pretreatment with 6-hydroxydopamine augmented the pressor response to α1-adrenoceptor agonists without changing the responses to α2-adrenoceptor agonists (Hamilton and Reid, 1981). All these results are consistent with the hypothesis that α1-adrenoceptors are located in the vicinity of sympathetic nerve terminals, strategically situated to be activated by noradrenaline coming out from the nerves, whereas α2-adrenoceptors are situated extrajunctionally and may be activated preferentially by circulating catecholamines, particularly adrenaline. However, in the isolated rabbit portal vein, the selective α1-adrenoceptor antagonist prazosin failed to antagonize the contractions evoked by electrical stimulation of the vessel, whereas the selective α2-adrenoceptor antagonist rauwolscine did antagonize these contractions (Docherty and Starke, 1981). In agreement with this report, it was shown that, in the canine saphenous vein, the inhibition of neuronal uptake by cocaine enhanced the contractile response to noradrenaline more in the presence of prazosin (α2-adrenoceptor-mediated response) than in that of yohimbine (α1-adrenoceptor-mediated response) (10- versus 6-fold; Guimarães et al., 1983). Furthermore, yohimbine was more potent than prazosin in antagonizing the effect of noradrenaline released from nerve terminals either by electrical stimulation or tyramine (Guimarães et al., 1983; Cooke et al., 1984; Flavahan et al., 1984; Pereira et al., 1991). More recently, it was reported that, in the canine saphenous vein in vitro, the contractile response evoked by electrical stimulation is mediated by three receptors: α1- and α2-adrenoceptors and P2X-receptors; the α1-adrenoceptor and the P2X-receptor-mediated contractions develop immediately after starting the stimulation and reach the maximum very quickly, whereas the α2-adrenoceptor-mediated contraction develops slowly, although reaching a maximum of similar magnitude. The purinergic component was smaller than the other two. Cocaine, which did not change the purinergic response, enhanced both adrenoceptor-mediated components but enhanced more markedly the α2-adrenoceptor-mediated- than the α1-adrenoceptor-mediated component (Hiraoka et al., 2000).
In surgically denervated canine saphenous veins, Flavahan et al. (1987b) showed that α2-adrenoceptor-mediated responses to noradrenaline are augmented, whereas α1-adrenoceptor-mediated responses are not. All these results indicate that, in contrast to the arteries, in these veins α2-adrenoceptors are situated closer to the sympathetic nerve terminals than α1-adrenoceptors.
A similar differential location was also encountered among β-adrenoceptors. β1-Adrenoceptors that are very responsive to noradrenaline are “innervated” and mediate responses to sympathetic nerve activity, whereas β2-adrenoceptors that are insensitive to noradrenaline are functionally “noninnervated” and function as hormone receptors for adrenaline from the adrenal medulla (Russel and Moran, 1980; Bryan et al., 1981). In electrically driven rabbit papillary muscles, it was shown that both α1- and β-adrenoceptors are located near or within the synaptic clefts of the sympathetic nerve endings (Dybvik et al., 1999). However, in the severely failing human heart, whereas α1-adrenoceptors are apparently located close to the nerve endings, the down-regulated β-adrenoceptors are situated outside the range of the neuronal influence, a fact that may have functional implications (Skomedal et al., 1998). As far as the veins are concerned, it was shown that, in normal strips of canine saphenous vein precontracted by prostaglandin F2α in the presence of phentolamine, there was no relaxant response to either electrical stimulation or tyramine. However, in strips preloaded with adrenaline and precontracted by prostaglandin F2α in the presence of phentolamine, electrical stimulation or tyramine caused frequency- or dose-dependent relaxations up to a maximum of 53.6% and 49%, respectively, of the steady-state precontraction (Guimarães and Paiva, 1981b). These results seem to indicate that, in this venous tissue, β2-adrenoceptors are located relatively close to the nerve terminals, whereas β1-adrenoceptors are not innervated (or are not abundant enough), since adrenaline coming from the nerves can efficiently cause relaxation whereas noradrenaline does not.
Apparently, the adrenoceptors under sympathetic control vary from vascular bed to vascular bed, and the vascular tone results from the simultaneous activation of receptors that are differentially influenced in the different vascular areas by the transmitters coming out from the sympathetic nerve endings. The receptors, the activation of which more importantly contribute to the basal vascular tone, are those that are “innervated”: the α1-adrenoceptors in the arterial vessels, the α1- and β1-adrenoceptors in the heart, and α2- and β2-adrenoceptors in the veins.
B. Distribution Upstream and Downstream
In this section, we deal with those smooth muscle receptors that are functionally involved in the responses to activation of adrenoceptors.
Noradrenaline contracts the vascular smooth muscle of most major arteries by activating postjunctional α1-adrenoceptors (see Section III.A.1.). However, in several blood vessels, α2-adrenoceptors also contribute to the vasoconstriction caused by noradrenaline, particularly in cutaneous arteries and veins (Polónia et al., 1985; Flavahan and Vanhoutte, 1986a). Most interesting is that the contribution of α1- and α2-adrenoceptors to the vasoconstriction caused by noradrenaline changes along the length of a single vessel (Bevan et al., 1980). In the arteries of the limbs, the participation of α2-adrenoceptors for the α-adrenoceptor-mediated vasoconstriction caused by noradrenaline increases from the proximal to the distal parts of these vessels. A comparison between human proximal (dorsalis pedis and arcuate arteries of the foot and superficial palmar arch of the hand) and distal arteries (digital arteries of the foot and hand) showed an increased prominence of α2-adrenoceptors on distal, compared with proximal, arteries (Flavahan et al., 1987a). The reduction in α2-adrenoceptor responsiveness from distal to proximal arteries continues in more proximal blood vessels such that α2-adrenoceptor-mediated responses are not present in larger arteries of the limbs (Thom et al., 1985). In the mouse tail artery, the contractile response resulting from α2-adrenoceptor activation is also greater at the distal than at the proximal level, whereas the opposite pattern was observed for α1-adrenoceptors (Chotani et al., 2000).
This pattern of increasing α2-adrenoceptor responsiveness from proximal to distal arteries is apparently not observed in the cerebral circulation. Bevan et al. (1987) showed that, in the cerebral arteries, the α-adrenoceptor responsiveness becomes progressively less important with successive branching of arteries. Small branches seem to have no functional α-adrenoceptors (Bevan et al., 1987). This may explain the observation that, in mice lacking the α2A/D-adrenoceptors, the pressor response to intra-arterial injection of α2-adrenoceptor agonists (UK-14304 or dexmedetomidine) was blunted when the injection was given into the femoral artery, but not when the injection was given into the carotid artery (MacMillan et al., 1996).
In contrast to what was observed in arteries, it was shown that on isolated strips of canine saphenous and cephalic veins the maximum contractile effect of α2-adrenoceptor activation was markedly smaller at the distal than at the proximal level; however, there was no change in the potencies of the selective α2-adrenoceptor agonists UK-14304 or BHT 920 along the length of the vessels (Guimarães and Nunes, 1990), indicating that the density of α2-adrenoceptors is higher at the proximal than at the distal level. This was confirmed in perfusion experiments with distal segments of canine saphenous vein where it was observed that they responded very poorly or did not respond at all to α2-adrenoceptor agonists applied to either the intima or the adventitia (Nunes et al., 1991). Since α2-adrenoceptors are very sensitive to changes in temperature and are abundant in cutaneous blood vessels (of the limbs in humans and dogs and of the tail in the rat) (seeSection VIII.), they appear to be important for thermoregulation, by constricting on cooling and dilating when exposed to a warmer environment. It is not surprising that the distribution of α2-adrenoceptors in cerebral and cutaneous arteries is different, since the functional role of arteries in these two beds is not comparable.
Large coronary vessels possess both α- and β-adrenoceptors, whereas small vessels of the coronary circulation possess only β-adrenoceptors (Bohr, 1967).
β-Adrenoceptors are also not uniformly distributed along the length of a single vessel. The thoracic aorta of the rabbit shows considerable β-adrenoceptor-mediated activity, whereas in the abdominal aorta β-adrenoceptor-mediated activity is practically nonexistent (Bevan et al., 1980). In the dog saphenous vein, the maximal relaxation to isoprenaline was much larger in the distal than in the proximal vein, whereas the effectiveness of forskolin did not vary, irrespective of the tone and the segment of vein used (Guimarães et al., 1993). Furthermore, the effectiveness of dobutamine increased from the proximal to the distal part, whereas that of the selective β2-adrenoceptor agonist terbutaline decreased. This indicates that the effectiveness of β-adrenoceptor activation and the contribution of β1-adrenoceptors to the relaxation increase from the proximal to the distal part of the canine saphenous vein (Guimarães et al., 1993).
It is not easy to propose an explanation for this differential distribution of β-adrenoceptors, since the functional consequences of their activation by any of the endogenous ligands (adrenaline and noradrenaline) is always masked by the predominant effect resulting from simultaneous activation of α-adrenoceptors.
C. Distribution in Particular Vascular Beds
There is also a regional variation in the distribution of vascular adrenoceptors. For many years, it has been accepted, at least in humans, that splanchnic and skeletal muscle vascular beds dilate to adrenaline because β-adrenoceptors predominate in their vessels, whereas adrenaline consistently reduces renal and skin blood flow, because in renal and skin vessels α-adrenoceptors are predominant (for reviews, see Innes and Nickerson, 1970; Hoffman and Lefkowitz, 1995).
The cerebral and coronary arteries are of particular importance in the whole of the vascular system, because of the vital functions of the organs they supply. The cerebral circulation of many species has an abundant and dense sympathetic innervation. However, the response of the cerebral vasculature to sympathetic nerve activity is comparatively small (Bevan et al., 1980; Toda, 1983). In humans, the influence of sympathetic innervation on the tone of cerebral vasculature is weak and reflects not only a low density of innervation, but also a reduced number of α-adrenoceptors (Bevan et al., 1998a). Furthermore, the sympathetic neurogenic control of cerebral arteries decreases with decreasing diameter of the vessel, such that the human pial arteries pratically do not contract in response to nerve stimulation (VanRipper and Bevan, 1991; Bevan et al., 1998a). Whereas the maximum vasoconstriction to noradrenaline in the middle meningeal artery reaches 34% of the maximum to KCl, in the pial artery it reaches only about 10% of the maximum. The cerebral arteries of the rat and pig do not contain functional α-adrenoceptors (Bevan et al., 1987). There is also little evidence for a significant β-adrenoceptor population in cerebral arteries (Bevan et al., 1998a). However, this relative lack of postjunctional adrenoceptors does not necessarily mean a lack of influence of the sympathetic nerves on the cerebral circulation. Some influence may be exerted through a cross-talk between sympathetic nerves and other neuronal systems. The nicotine-induced relaxation in the porcine basilar artery appears to result from the activation of nicotinic receptors on the presynaptic adrenergic nerve terminals; this activation causes release of noradrenaline that activates β1-adrenoceptors located on NOergic nerves and promotes the release of NO (Toda et al., 1995; Zhang et al., 1998; Lee et al., 2000). Another indirect effect mediated by adenoceptors is that observed in segments of rabbit middle cerebral arteries, where the activation of endothelial α2-adrenoceptor causes a reduction in endothelin-1 production and promotes vascular relaxation (Thorin et al., 1997).
In the coronary circulation, the relative amount of α- and β-adrenoceptors and the relative functional role they play also does not fit into the general pattern of the vascular beds. In the pig, the small coronary vessels exhibit little or no α-adrenoceptor-mediated activity, and the large coronary artery contains α1-adrenoceptors, mainly of the α1A-subtype, but the functional importance of their vasoconstrictive effect is unclear (Yan et al., 1998). Also in the coronary artery of the dog, the functional role of α-adrenoceptors varies between undetectable and of little expression (Begonha et al., 1995). In vessels with spontaneous tone, isoprenaline causes concentration-dependent relaxations, whereas noradrenaline and adrenaline cause either contraction (of small magnitude) or relaxation. However, after the tone had been elevated by phenylephrine, both adrenaline and noradrenaline cause concentration-dependent relaxations with a maximum effect that sometimes did not fully antagonize the previous tone (Ross, 1976; Guimarães et al., 1993; Begonha et al., 1995). This is not true for isoprenaline, which, at any level of tone, causes a relaxation that totally antagonizes the previous contraction (Table 3). In contrast to the mesenteric, splenic, and pulmonary arteries, where β2-adrenoceptors predominate, in the coronary arteries β1-adrenoceptors are largely predominant if not exclusive (Begonha et al., 1995). In fact, whereas in the systemic arteries adrenaline was much more potent than noradrenaline as an agonist and the selective β2-adrenoceptor antagonists ICI-118551 was much more potent than atenolol at antagonizing the responses to isoprenaline; in the coronary arteries, noradrenaline was more potent than adrenaline; and ICI-118551 and atenolol were equipotent as antagonists of isoprenaline (Begonha et al., 1995). Also in the dog, it was shown that the magnitude of β-adrenoceptor-mediated responses of epicardial coronary arteries is inversely related to the size of the vessel (Krauss et al., 1992). The difference was independent of α-adrenoceptors, endothelium, and second messenger processing, suggesting a mechanism based on β-adrenoceptor density (Krauss et al., 1992). As far as the splanchnic vascular bed is concerned, experiments carried out on isolated mesenteric and splenic arteries showed, at least in the dog, that α-adrenoceptor-mediated effects always predominate over β-adrenoceptor-mediated responses when adrenaline is used as agonist (Guimarães and Paiva, 1981a; Begonha et al., 1995). It is possible that β-adrenoceptors are associated only with arterioles and precapillary sphincters, which regulate the peripheral resistance observed in vivo and that are not available for studies in vitro. However, in experiments in which the hindlimb of the dog was perfused, adrenaline, despite reaching arterioles and precapillary sphincters, caused concentration-dependent increases of the perfusion pressure, showing that α-adrenoceptor-mediated vasopressor effect predominates also in this vascular bed (Teixeira, 1977). An interesting peculiarity was shown in the acral regions of the cutaneous circulation, where the vascular tone is primarily controlled by humoral mechanisms mediated at postjunctional α2-adrenoceptors; α1-adrenoceptors that mediate neuronally evoked constriction in the cutaneous vasculature contribute little to the sympathetic regulation of this bed (Willette et al., 1991).
VII. Influence of Maturation and Aging
Maturation and aging are associated with many alterations in vascular adrenergic mechanisms. From birth to adulthood (maturation) and from adulthood to old age (aging or senescence), important changes occur in animal models as in humans at the receptor level, neurotransmitter process, and catecholamine inactivation.
In general terms, one can accept that maturation is associated with an increase, whereas aging is associated with a reduction in the adrenergic influence on the physiological processes.
A. On α-Adrenoceptors
It is well documented that the responsiveness of vascular smooth muscle to α1-adrenoceptor activation is present at birth (Guimarães et al., 1994) and that it changes with age, although in the majority of functional studies no important alterations in responses to noradrenaline had been demonstrated either during maturation or aging (for a review, see Docherty, 1990). In the dog mesenteric artery and rat aorta, small reductions in the responsiveness to sympathomimetic amines were reported during maturation (McAdams and Waterfall, 1986; Toda and Shimizu, 1987), whereas a decrease in α-adrenoceptor-mediated functions with aging was observed in the rat tail artery (Fouda and Atkinson, 1986) and in the rat aorta (Hyland et al., 1987; Wanstall and O'Donnell, 1989). According to Satoh et al. (1995), the potency of noradrenaline in the rat aorta increased with age from 3 to 10 weeks, but decreased from 10 to 40 weeks. In the pig coronary arteries, the endothelium-dependent relaxation to noradrenaline via the α2-adrenoceptors decreases with aging (Murohara et al., 1991).
It has been suggested that the age-related changes in α1-adrenoceptor-mediated vasoconstrictor responses in isolated blood vessels might result from changes in the expression of the α1-adrenoceptor subtypes; accordingly, functional, radioligand binding, and molecular biology studies using rat aortic tissue have shown that with age the expression of the α1A subtype is increased, that of the α1B subtype is decreased, and that of the α1D-subtype does not change (Gurdal et al., 1995a,b). However, in the pithed rat, it was shown that the selective α1D-adrenoceptor antagonist BMY-7378 displaced the dose-response curve to phenylephrine in young prehypertensive SHRs, but had no effect in young WKY rats; whereas in adult WKY rats, BMY-7378 caused a greater shift in the concentration-response curve to phenylephrine than in younger animals (Villalobos-Molina et al., 1999). The presence of α1D-adrenoceptors in the resistance vasculature of the prehypertensive and hypertensive rats may indicate that α1D-adrenoceptors are involved in vascular hyperreactivity (Villalobos-Molina et al., 1999). This apparent contradiction may well be due to the fact that the aorta and the resistance vessels are functionally totally different. The results obtained by Xu et al. (1997) confirm that aging changes heterogeneously the expression of α1-adrenoceptor subtypes. These authors determined the changes in mRNA levels of α1-adrenoceptor subtypes during maturation and aging in aortae and in renal, pulmonary, and mesenteric arteries isolated from 3-, 12-, and 24-month-old rats. They observed that, in the aorta, α1A-, α1B-, and α1D-adrenoceptors declined with aging, whereas in the renal artery there was a decrease in mRNA for the α1B-adrenoceptor in aged rats. However, in mesenteric and pulmonary arteries, there were no changes in mRNA levels for any of the subtypes. The results obtained on the aggregatory responses of human platelets in radioligand binding studies also show no important differences with maturation and aging in the affinity of ligands for the binding site (Buckley et al., 1986: Davis and Silski, 1987). Vascular contractile responsiveness seems to increase with aging, and this supersensitivity may be related to the pronounced increase in the maximal pressor effect of α1-adrenoceptor stimulation observed in the adult pithed rat (Ibarra et al., 1997). Because α1D-adrenoceptors represent the predominant subtype that mediates contraction in the aorta, carotid, and mesenteric arteries of SHRs (Villalobos-Molina and Ibarra, 1996), it may be that α1D-adrenoceptors play some role in the pathogenesis/maintenance of hypertension (Ibarra et al., 1998).
The α2-adrenoceptor-mediated negative modulation of noradrenaline release is fully developed at birth (Guimarães et al., 1991, 1994). However, while at the postjunctional level phenylephrine is equipotent in adults and neonates, indicating that postjunctional α1-adrenoceptors do not change during maturation, UK-14304 is about 4 times more potent at inhibiting noradrenaline release evoked by electrical stimulation in adults than in neonates (Guimarães et al., 1991). A similar difference in potency of UK-14304 at inhibiting noradrenaline release in adults and neonates was also observed in mouse atria (A.U. Trendelenburg and K. Starke, personal communication). One possible explanation for this difference is that the fractional release of noradrenaline is much higher in neonates than in adults (Guimarães et al., 1991; Moura et al., 1993). Hence, the concentration of noradrenaline in the biophase during electrical stimulation is higher in neonates than in adults; consequently, the inhibitory effect of any given concentration of UK-14304 is smaller, and its IC50 is higher in neonates than in adults (Starke, 1972; Fuder et al., 1983). Based on the temporal and regional pattern of α2-adrenoceptor mRNA expression in rat brain, it has been suggested that the perinatal increase in receptor density may serve specific roles in development, including neuronal migration, maturation of neurons, and mediation of sensory functions (Winzer-Serhan and Leslie, 1997; Winzer-Serhan et al., 1997a,b). According to Happe et al. (1999), α2-adrenoceptors are functionally coupled to G-protein throughout postnatal development and, therefore, are able to mediate signal transduction upon stimulation by noradrenaline and adrenaline. In the neonatal rat lung, there is a pure and dense population of α2B-adrenoceptors, which is assumed to be the only one existing in this tissue. However, the number of these receptors falls to undetectable levels in adults (Latifpour and Bylund, 1983). In mice, functionally important prejunctional α2-adrenoceptors exist in atria and vas deferens already at the age of 1 day, which are mainly α2A/D. With maturation, the α2A/D-adrenoceptors increase their functional influence. However, the development of prejunctional α2C-adrenoceptors is much more impressive. They are almost absent at birth, become influencial after birth, and reach maximum activity in adult life. In atria from adult α2A/D-adrenoceptor knockout mice, there is as much autoinhibition as in adult wild-type atria (A. U. Trendelenburg and K. Starke, personal communication).
B. On β-Adrenoceptors
In the canine saphenous vein, pre- and postjunctional β-adrenoceptor-mediated effects are lacking at birth. However, the responses to forskolin, a direct-acting stimulant that bypasses the need for β-adrenoceptors on their linkage to stimulatory G-protein subunits, are already present at birth; this shows that the lack of responses to isoprenaline is linked to either a lack or some kind of immaturity of the receptors or G-protein (Guimarães et al., 1994). Furthermore, it was shown that β2-adrenoceptor-mediated effects and the increase in the adrenaline content of the adrenal gland have a parallel time course (Paiva et al., 1994). Thus, both the prejunctional and the postjunctional β2-adrenoceptor-mediated effects increase with increasing age (until adulthood), as does the adrenaline content of the adrenal gland, such that at 2 weeks the β2-adrenoceptor-mediated maximum effect is about 50% of that of the adult; and at 1 month, it is fully developed (Paiva et al., 1994). The relationship between the content of adrenaline of the adrenal medulla and the development of β2-adrenoceptor-mediated responses was analyzed also in the rat, a species in which β2-adrenoceptor-mediated responses develop earlier than in the dog, such that at birth these responses are already fully expressed. Interestingly, whereas the adrenaline content of the canine adrenal medulla at birth is about 3% that of the adult, in the rat it is about 50%. This suggests a link between adrenaline and the maturation of β2-adrenoceptor-mediated effects, indicating that either adrenaline triggers the expression of β2-adrenoceptor-mediated effects or that the expression of adrenaline formation and β2-adrenoceptor-mediated effects are simultaneously evoked by the same event (Moura et al., 1997). Similarly, in the mouse, a species in which adrenaline represents about 60% to 70% of total catecholamines in the adrenal medulla of 1-day-old animals, prejunctional β2-adrenoceptors fully operate at that age (A.U. Trendelenburg and K. Starke, personal communication). This hypothesis is in good agreement with the report that in early neonatal life isoprenaline, instead of producing desensitization of responses, enhances expression or efficiency of β-adrenoceptor signaling (Giannuzzi et al., 1995). Recently, it was demonstrated that agonist treatment in the neonate causes an enhancement of coupling rather than an uncoupling of receptors from G-proteins (Zeiders et al., 1999), and the reversal from enhancement of coupling to uncoupling occurs (in cardiac cells) between postnatal days 11 and 14 (Zeiders et al., 2000).
β-Adrenoceptor-mediated relaxation was compared in the pulmonary vein of the fetal (145 ± 2 days of gestation) and newborn lamb. Isoprenaline caused greater relaxation in newborn than in fetal lambs. Also, in humans, the sympathetic nerves play a more important role in the regulation of cerebrovascular tone in the infant than in the adult (Bevan et al., 1998b). Biochemical studies showed that isoprenaline and forskolin evoked a greater increase in cAMP content and in adenyl cyclase activity of pulmonary veins in the newborn than in the fetal lamb. These results show that β-adrenoceptor-mediated relaxation of the pulmonary veins increases with maturation (Gao et al., 1998). However, according to Conlon et al. (1995), there is no change in myocardial ventricle β-adrenoceptor G-protein coupling capacity or adenylate activation with aging beyond maturity. These authors showed that aging between 6 and 26 months in male Wistar rats is not accompanied by changes in myocardial β-adrenoceptor signal transduction and capacity for formation of the high-affinity β-adrenoceptor G-protein coupled complex with the agonist. It was also found that an age-related impairment of myocardial β-adrenoceptor up-regulation occurs with aging (Conlon et al., 1995).
This β-adrenoceptor-mediated relaxing capacity, which increases during the first weeks of life, then declines as the age increases. The loss of vasodilator response to isoprenaline in the rat aorta has been reported at different ages ranging from 3 to 22 months (for a review, see Docherty, 1990). There is not only a decrease in the maximum relaxation to isoprenaline with aging, which has been reported for the rabbit aorta, rat pulmonary artery, rat mesenteric artery, human saphenous vein, canine mesenteric artery, but also an increase in the EC50 of isoprenaline: in the aorta of 5- and 20-week-old rats preconstricted with phenylephrine, the pD2 values for isoprenaline were 7.97 and 6.57, respectively (Borkowski et al., 1992), indicating a marked reduction in the potency of this β-adrenoceptor agonist. According to Dohi et al. (1995), with increasing age, maximum β-adrenoceptor-mediated relaxation decreases in most arteries, but not in veins. Also SHRs exhibit an age-related loss in vasodilator β-adrenoceptor responsiveness. However, the maximum relaxation to sodium nitrite or to sodium nitroprusside is not reduced (O'Donnell and Wanstall, 1986;Küng and Lüscher, 1995).
Because most studies show no change with age in the number of β1- or β2-adrenoceptor-binding sites of the human lymphocytes and rat heart and because cAMP production in response to forskolin and dibutyril cyclic adenosine monophosphate is also reduced by aging in the rat myocardium and human lymphocytes, it seems likely that the change is not at the receptor level but in the coupling to the adenylate cyclase via G-proteins. In healthy volunteers of different ages, isoprenaline-induced increases in heart rate were significantly greater in young than in old ones (Brodde et al., 1998). However, β-adrenoceptor numbers and subtype distribution were unchanged as determined in patients undergoing open heart surgery. The decrease in β-adrenoceptor-mediated efficiency is due to a reduced activity of the catalytic unit of the adenylyl cyclase (Brodde and Pönicke, 1998).
Prolonged or repeated exposure to β-agonists in adults results in a compensatory desensitization that reduces responsiveness (for a review, see Summers et al., 1997). In older animals, the predominant effect is heterologous desensitization mediated at the level of the G-protein. During development, however, responses in most systems increase with age and with the maturation of neuronal inputs (Giannuzzi et al., 1995). Instead of producing desensitization of responses, agonist exposure promotes receptor signaling by enhancing expression and/or catalytic efficiency of adenylyl cyclase. These developmental differences are likely to be important in the maintenance of tissue responsiveness during the period in which innervation develops (Guimarães et al., 1994; Giannuzzi et al., 1995; Moura et al., 1997).
Regarding the age-related involvement of the endothelium in β-adrenoceptor-mediated responses, it has been shown that aging reduced endothelium-dependent relaxations to acetylcholine and isoprenaline in aortas from both normotensive and SHR rats (Arribas et al., 1994; Küng and Lüscher, 1995; Satake et al., 1995; van der Zypp et al., 2000). However, the responses to the NO-donor nitroprusside sodium, which directly activates the soluble guanylyl cyclase and formation of cGMP, were very similar in adult and old rats of either strain. This indicates that the impairment of the response to acetylcholine and isoprenaline is due to functional changes of the endothelium, rather than the vascular smooth muscle (Küng and Lüscher, 1995; van der Zypp et al., 2000). Furthermore, it was observed that, in endothelium-intact aortas, the nitric oxide synthase inhibitor L-NMMA attenuated the isoprenaline-induced relaxation to a similar extent in both age groups, suggesting that although NO was involved in the response to isoprenaline, it cannot have been responsible for the age-related difference (van der Zypp et al., 2000). It was also observed that the age-dependent reduction in isoprenaline-mediated relaxation in aorta was greater in K+ than in phenylephrine-constricted aortas (Borkowski et al., 1992; Chapman et al., 1999), supporting the view that the signaling pathway involved in isoprenaline-induced relaxation switches toward an increased role of K+ channels in older rats. Thus, the signaling pathways involved in β-adrenoceptor-mediated responses are multifactorial. They include a NO-dependent pathway, that does not depend on age; an endothelium-independent pathway involving cAMP, which appears to decline with age; and a third factor apparently endothelium-dependent that involves tetraethylammonium-sensitive K+channels and increases with age (Satake et al., 1996, 1997; van der Zypp et al., 2000).
VIII. Influence of Temperature on Vascular Adrenoceptor-Mediated Responses
It is common knowledge that cold makes the skin pale and heat makes the skin red. In intact organisms, exposure to cold causes cutaneous veins to constrict, whereas deeper veins dilate thus transferring venous blood from the superficial to the deep circulation to reduce heat loss (Vanhoutte, 1980). In isolated veins of the dog contracted with exogenous noradrenaline or sympathetic nerve stimulation, cooling enhances (saphenous) or reduces (femoral) the contractile responses (Vanhoutte and Lorenz, 1970). Furthermore, it was reported that cooling enhances the contractile response to the selective α2-adrenoceptor agonist UK-14304 (Flavahan and Vanhoutte, 1986b; Nunes and Guimarães, 1993), does not change that to the selective α1-adrenoceptor agonist phenylephrine (Flavahan and Vanhoutte, 1986b; Nunes and Guimarães, 1993), and markedly reduces the response to chloroethylclonidine (Nunes and Guimarães, 1993). Similarly, cooling enhanced the α2-adrenoceptor-mediated contractile effect evoked by electrical stimulation on the human saphenous vein (Harker et al., 1994). Flavahan and Vanhoutte (1986b) explained the different sensitivity of cutaneous and deep blood vessels to cooling on the basis of a different α1-adrenoceptor reserve: in the saphenous vein, an enhanced α2-adrenoceptor-mediated effect is added to a nonreduced α1-adrenoceptor-mediated response (because in this vein there is a large α1-adrenoceptor reserve that buffers the α1-adrenoceptor-mediated response from the inhibitory influence of cooling); in the femoral vein, the α2-adrenoceptor-mediated effect is so inefficient that its enhancement does not compensate for the markedly reduced α1-adrenoceptor-mediated responses (which is depressed because there is no α1-adrenoceptor reserve). The same authors had previously shown in the saphenous vein that under normal conditions, cooling to 24°C did not affect the responses to phenylephrine, whereas it did reduce markedly this response after partial irreversible blockade of α1-adrenoceptors with phenoxybenzamine (Flavahan and Vanhoutte, 1986b). However, the enhancement of the α2-adrenoceptor-mediated responses remained to be explained.
In the deer digital arteries, a different reactivity was found in winter and summer: in the cold winter, they were either insensitive or had a reduced sensitivity to the vasodilator action of histamine, compared with arteries collected in summer (Callingham et al., 1998;Milton et al., 1999).
Very recently, it was shown that, in the mouse tail artery, at 37°C, vasoconstriction to the α2-adrenoceptor agonist UK-14304 was antagonized by the selective α2A/D-adrenoceptor antagonist BRL-44408, but was not antagonized by the α2B- and α2C-adrenoceptor antagonist ACR-239 or the preferential α2C-adrenoceptor antagonist MK-912. However, at 28°C, the enhanced vasoconstrictor response to UK-14304 was inhibited by low concentrations of the preferential α2C-adrenoceptor antagonist MK-912, whereas ACR-239 was ineffective and the selective α2A/D-adrenoceptor antagonist BRL-44408 showed an inhibitory effect that was not different from that observed at 37°C. These results indicate that, at 28°C, α2C-adrenoceptors contribute to α2-adrenoceptor-mediated vasoconstriction and probably are responsible for the supersensitivity to α2-adrenoceptor agonists caused by cold (Chotani et al., 2000). Interestingly, this is not a phenomenon exclusively occurring with vascular smooth muscle of superficial vessels, since hypothermia enhanced α2-adrenoceptor-mediated responses in rat vas deferens in such a way that the lack of any response to UK-14304 at 37°C was converted to evident contractions at 20°C (Gonçalves et al., 1989).
IX. Vascular Adrenoceptors in Some Diseases
Vascular adrenoceptors may be affected in many diseases, sometimes as a consequence of alterations suffered by the vessels and sometimes by participating themselves in the genesis of diseases or by being their primary cause.
For many diseases, there are no animal models. Even in well studied animal models, such as the SHRs, relevance to human “essential hypertension” is unknown. As a consequence, one often has to rely on observations with patients. Additionally, in vivo experiments in animals (e.g., measurements of blood pressure) often involve multifactorial systems, the analysis of which is far more complex than in the in vitro experiments. For instance, the determination of maximum responses of the blood pressure to pressor agents is often impossible; hence, it is difficult or impossible to provide a full and satisfactory description of an “enhanced pressor response” observed in this or that disease: is this phenomenon due to a parallel shift of the dose-response curve to the left, to an increase of the maximum response, or to both? Is this phenomenon generated by a change in the structure of the blood vessels by a change in the mechanisms that inactivate the test agonist, by a change in G-proteins or second messenger mechanisms, or in change of the adrenoceptors? Given these unavoidable limitations of attempts to delineate the role played by adrenoceptors in various diseases, it is not surprising that hard facts are rare. However, such studies have provided valuable hints.
In 1929, Lewis postulated that Raynaud's disease resulted from a “local fault” of the blood vessel wall. This fault could be an anomalous regulation of α2-adrenoceptors. It appears now clear that α2-adrenoceptors play a role in the development of Raynaud's disease (Freedman et al., 1995;Chotani et al., 2000). Cold augments constriction to α2-adrenoceptor activation without affecting the responses to α1-adrenoceptor stimulation or to any other vasoconstrictor agent (Flavahan and Vanhoutte, 1986a;Chotani et al., 2000). Nonselective α2-adrenoceptor antagonists abolish cold-induced vasospastic crises in patients with primary Raynaud's disease (Freedman et al., 1995). However, according to many authors, there is no abnormal reactivity of α2-adrenoceptors in subjects with primary Raynaud's disease (Lindblad et al., 1989; Coffman and Cohen, 1990;Freedman et al., 1993). Therefore, the local fault in Raynaud's crisis may represent a consequence of a cold-induced functional expression of α2C-adrenoceptors that appear to be silent at normal temperature (Chotani et al., 2000). Thus, inhibition of α2C-adrenoceptors may provide a highly selective therapeutic measure for this disease.
Vasospasm and ischemic organ injury are functional changes that play an important role in the pathogenesis of scleroderma (Kahaleh, 1990). These functional changes were attributed to a failure of vascular endothelium in releasing nitric oxide (Freedman et al., 1999). However, in arterioles isolated from uninvolved skin of patients with scleroderma, the constrictor responses to the selective α2-adrenoceptor agonist UK-14304 were increased, whereas those to KCl or the selective α1-adrenoceptor agonist phenylephrine were similar to controls. This selective increase in the reactivity of α2-adrenoceptors was not altered by removing the endothelium, indicating that the enhanced constrictor effect was not due to changes in endothelial dilator activity, but to an enhancement of the α2-adrenoceptor-mediated responses of the vascular smooth muscle cells (Flavahan et al., 2000).
In several neurological degenerative and genetic disorders, there are also important changes in α-adrenoceptor-mediated responses. In the majority of these situations, the pathological process involves primarily the sympathetic postganglionic neurones, leading to a progressive denervation of some organs, including the blood vessels. The most prominent cardiovascular symptom in all these conditions is orthostatic hypotension, which is a common complaint due to a sympathetic neurocirculatory failure and sometimes forces the patients to be bedridden, even if they are still able to work. In the familial amyloidotic polyneuropathy, an autosomal dominant disorder with an estimated prevalence of about 1/1000 in the population of the most affected areas in the northwest of Portugal (Andrade, 1952; Carvalho et al., 1997), there is a progressive degeneration of the sympathetic nerves leading to complete denervation. Concomitantly, there is a marked supersensitivity to the vasoconstrictor action of noradrenaline (Falcão-de-Freitas, 1996; Carvalho et al., 1997), probably due to an up-regulation of α-adrenoceptors. It is not yet known whether all α-adrenoceptor subtypes are equally involved or some particular subtype(s) is predominantly implicated. A similar enhancement of α-adrenoceptor-mediated responses occurs in patients with congenital dopamine-hydroxylase deficiency (Man in't Veld et al., 1987). In this disorder, there is no conversion of dopamine to noradrenaline causing a lack of the transmitter at postganglionic sympathetic neurones (Man in't Veld et al., 1987; Rea et al., 1990). Recently, it was shown that many patients with Parkinson's disease have evidence of peripheral sympathetic denervation causing a deficient release of noradrenaline in the heart and blood vessels with a consequent α-adrenoceptor up-regulation (Magalhaes et al., 1995;Netten et al., 1995; Goldstein et al., 2000). In diabetic polyneuropathy, there is also an autonomic dysfunction leading to a progressive sympathetic denervation and to a more or less marked increase in α-adrenoceptor-mediated responses (for review, seeWatkins, 1998). There are other degenerative neurological diseases, like the multiple system atrophy (Shy Drager syndrome), in which the main pathological changes leading to the neurocirculatory failure with severe orthostatic hypotension occur in the central nervous system without marked alterations of the peripheral sympathetic nerves (Zoukos et al., 1999; Goldstein et al., 2000). Very recently, it was shown that overexpression of the α1B-adrenoceptor causes apoptotic neurodegeneration with a corresponding multiple system atrophy (Zuscik et al., 2000). The resulting symptoms (impaired hindlimb function and seizures) could be rescued with the α1-adrenoceptor antagonist terazosin, indicating that α1-adrenoceptors participated directly in the pathology. These findings suggest a link between α1B-adrenoceptor function and the etiology of Shy-Drager syndrome (Zuscik et al., 2000).
Sometimes, β-adrenoceptors are also involved in these neurological degenerative diseases. Both in multiple sclerosis and multiple system atrophy, there is sympathetic neurocirculatory failure with supersensitivity to α-adrenoceptor agonists. However, whereas in multiple sclerosis there is also an up-regulation of the β-adrenoceptors expressed on peripheral blood mononuclear cells, in multiple system atrophy there is not (Zoukos et al., 1999).
According to many authors, α-adrenoceptors may be involved in the pathogenesis/maintenance of some kinds of hypertension. Not only sensitivity to salt is a common trait in patients with essential hypertension, but there is also experimental evidence suggesting that salt loading causes hypertension via a mechanism involving α2-adrenoceptors. A recent comparison of the effect of subtotal nephrectomy and salt loading in α2B-adrenoceptor knockout mice, in α2C-adrenoceptor knockout mice, and in wild-type mice showed that only the α2B-adrenoceptor knockout mice have no significant increase in blood pressure (Makaritsis et al., 1999). Both the wild-type and α2C-adrenoceptor knockout mice had significant blood pressure increases, indicating that α2B-adrenoceptors are relevant for the development of this kind of hypertension. On the other hand, some data draw attention to the possible role played by changes in α1-adrenoceptors in the development/maintenance of hypertensive states. In the resistance vasculature of young prehypertensive and hypertensive rats, a high density of α1D-adrenoceptors was found (Stassen et al., 1997; Ibarra et al., 1998; Xu et al., 1998). Furthermore, in endothelium-denuded tail artery and aorta, the maximum contractile response to phenylephrine and chloroethylclonidine was higher in SHR than in WKY rats (Villalobos-Molina et al., 1999; Ibarra et al., 2000), suggesting that α1D-adrenoceptors may be involved in vascular supersensitivity leading (or in some way being linked) to the hypertensive state. It is also possible that changes at postadrenoceptor level, such as altered levels of G-proteins may be related to the higher reactivity to agonists in hypertension (Li et al., 1994; Kanagy and Webb, 1996). These results indicate that future studies with α1D-adrenoceptor knockout animals may be helpful to explain the role (if any) of these alterations in genesis/maintenance of hypertension.
Trp64Arg mutation of β3-adrenoceptor has been suggested to confer susceptibility to essential hypertension (Morris et al., 1994); this thesis was contested by Fujisawa et al. (1997) and confirmed by Tonolo et al. (1999). These authors concluded that the Trp64Arg polymorphism of the β3-adrenoceptor gene is associated more often with high blood pressure than with normal blood pressure.
A naturally occurring variation found in about 8% of Europeans and North Americans actually restores in humans the arginine residue present in animals (Strosberg, 1997). This variation was found to be associated with 1) an increased capacity of obese French patients to gain weight (Clément et al., 1995); 2) an early onset of noninsulin-dependent diabetes mellitus in obese Pima Indians (Walston et al., 1995); and 3) an early onset of noninsulin-dependent diabetes mellitus and clinical features of the insulin resistance syndrome in Finns (Widén et al., 1995). Although these alterations are related to β3-adrenoceptors in adipocytes, they may be particularly important since it is believed that in terms of the risks of cardiovascular disorders, visceral obesity is the most dangerous form of regional fat accumulation, the form of obesity that is more directly linked to β3-adrenoceptor activity (Arner, 1995).
X. Conclusions
New possibilities are now offered by molecular biology (knockout animals, genetically altered receptors, measurements of mRNA, etc.) that will help in clarifying receptor function. However, data obtained in experiments carried out in knockout animals must be carefully interpreted, keeping in mind the multiple ways to compensate for the lack of this (or these) adrenoceptor subtype(s).
Regarding the subclassification of adrenoceptors, two controversial points are now on the way to being solved: the existence of a fourth α1-adrenoceptor subtype (the α1L-adrenoceptor) and a fourth β-adrenoceptor (the β4-adenoceptor). It is now accepted that these hypothetical subtypes correspond to low-affinity states of the α1A- and the β1-adrenoceptors, respectively.
The pharmacology of human α1-adrenoceptors often differs from that of the corresponding α1-adrenoceptor subtypes of experimental animals. Then, the identification of α1-adrenoceptor subtypes present in human vasculature may be useful for the discovery of new selective compounds effective in the treatment of prostatic hypertrophy, pulmonary hypertension, and coronary insufficiency.
A fascinating point is that, in the vast majority of the organs, the adrenoceptors expressed there do not correspond to the functional roles they play. The potential role of some adrenoceptor subtypes apparently unimportant under normal conditions, should be kept in mind and carefully taken into consideration. It has been described that some sodium channels (II and III α-isoforms), which are functionally important during the earlier stages of life, loose their important roles in adult life and reappear functionally active under some pathological conditions (Aronica et al., 2001). This fact should be linked to the interesting fact that expression and function of a given adrenoceptor subtype changes their role during a lifetime. For example the α2B-adrenoceptor, which is densely represented during intrauterine development, disapears after birth, whereas the α1D-adrenoceptors that have no important role under physiological conditions may become important in some hypertensive states. This dynamic balance can also be exemplified by the fact that α2C-adrenoceptors, which at 37°C are nonfunctional, become functionally predominant at lower temperatures.
Prejunctional α2A/D-adrenoceptors are now well established to be primarily responsible for the regulation of the release of noradrenaline under physiological conditions. However, also, α2C-adrenoceptors play a minor role in this regulatory mechanism.
The existence of α- and β-adrenoceptors in the endothelium and the importance of the endothelial system in the physiology and pathophysiology of the vascular system has to be considered; however, it is intriguing that the activation of α- and β-adrenoceptors lead to the same effect: an increase in NO formation/release.
From a therapeutic standpoint (and as far as β-adrenoceptors are concerned), there are many instances where β-adrenoceptor-subtype selective stimulation (asthma, atrioventricular block, obesity) or block (hypertension, coronary insufficiency) is desired. Therefore, a still more detailed knowledge of subtype-specific functions is necessary as drugs, which are more selective, are required.
Acknowledgments
This work was supported by Project PRAXIS/P/SAU/14294/1998. We are thankful to Professors U. Trendelenburg (Tübingen, Germany) and K. Starke (University of Freiburg, Germany) for thoughtful comments and valuable suggestions concerning the manuscript.
Footnotes
-
↵1 Address for correspondence: Dr. Serafim Guimarães, Institute of Pharmacology and Therapeutics, Faculty of Medicine, Alameda Hernani Monteiro, 4200-319, Porto, Portugal. E-mail:sguimara{at}med.up.pt
Abbreviations
- EDRF
- endothelium-derived relaxing factor
- 5-MU
- 5-methyl urapidil
- WB-4101
- 2-(2′,6′-dimethoxyphenoxyethyl)-aminomethyl-1,4-benzodioxane hydrochloride
- BRL-37344
- 1-(3-chlorophenyl)-2-[2-(4-(carboxymethoxy)phenyl)-1-methyl-ethylamino]ethanol
- CL-316243
- disodium (R,R)-5-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl] amino] propyl]-1,3-benzodioxole-2,2-dicarboxylate
- SR-59230
- 3-(2-ethylphenoxy)-1-[(1S)-1,2,3,4-tetrahydronaphtalen-1-yl-amino]-2S-2 propanol oxalate
- L-748328
- (S)-N-[4-[2-[[3-[3-(aminosulfonyl)phenoxy]-2-hydroxypropyl] amino] ethyl] phenyl] benzenesulfonamide
- L-748337
- (S)-N-[4 -[2-[[3-[3-(acetamidomethyl)phenoxy]-2-hydroxypropyl] amino] ethyl] phenyl]benzenesulfonamide
- cAMP
- cyclic adenosine monophosphate
- CGP-12177
- (−)4-(3-t-butylamino-2-hydroxypropoxy)-benzimidazol-2-one
- BMY-7378
- (8-[2-[4-(2-methoxyphenyl)-1-piperazinyl]-ethyl]-8-azaspiro[4,5] decane-7,9-dionedihydrochloride)
- SHR
- spontaneously hypertensive rat
- WKY
- Wistar-Kyoto rat
- RS-17053
- (N-[2-(2-cyclopropylmethoxyphenoxy)ethyl]-5-chloro-α,α-dimethyl-1H-indole-3-ethanamine hydrochloride)
- U-46619
- 9,11-dideoxy-11α,9α-epoxy-methano prostaglandin F2α
- RWJ-38063
- N-(2-{4-[2-(methylethoxy)phenyl] piperazinyl}ethyl)-2-(2-oxopiperidyl)acetamide
- RWJ-69736
- N-(3-{4-[2-(methylethoxy)phenyl]piperazinyl} propyl)-2-(2-oxopiperidyl)acetamide
- RO-70–0004
- 3-(3-{4-[fluoro-2-(2,2,2-trifluoroethoxy)-phenyl]-piperazin-1-yl}-propyl)-5-methyl-1H-pyrimidine-2,4-dione
- RS-100329
- 3-(3-{4 -[ 2,2,2-trifluoroethoxy)-phenyl]-piperazin-1-yl}-propyl)-5-methyl-1H-pyrimidine-2,4-dione
- ZD-2079
- ((R)-N-(2-[4-(carboxymethyl)phenoxy)] ethyl-N-(β-hydroxyphenethyl)ammonium chloride
- LY-362884
- 6-{4-[2-({2-hydroxy-3-[(2-oxo-2,3-dihydro-1H-benzimidazol-4-yl)oxy] propyl} amino)-2-methylpropyl]phenoxy}}nicotinamide
- BRL-26830
- (R,R)(±)-methyl-4-{2-[(2-hydroxy-2-phenyethyl)amino] propyl}benzoate
- L-750355
- (S)-N-[4-[2-[[3-[(2-amino-5-pyridinyl)oxy]-2-hydroxypropyl] amino]-ethyl]-phenyl]-4-isopropylbenzenesulfonamide
- UK-14304
- 5-bromo-6-(imidazoline-2-ylamino)quinoxaline
- NO
- nitric oxide
- l-NAME
- Nω-nitro-l-arginine methyl ester
- l-NMMA
- NG-monomethyl-l-arginine
- cGMP
- cyclic guanosine monophosphate
- ICI-118551
- ((±)-1-[2,3-(dihydro-7-methyl-1H-inden-4-yl)oxy]-3-[(1-methylethyl)amino]-2-butanol
- CGP-20712
- 2-hydroxy-5(2-(2-hydroxy-3–4((1-methyl-4 -trifluoromethyl)1H-imidazole-2-yl)-phenoxy)propyl)amino)ethoxy)-bezamide monomethane sulfonate
- ZM-215001
- (S)-4-(2-hydroxy-3-phenoxypropylaminoethoxy)-N-(2-methoxyethyl)-phenoxyacetic acid
- SR-58611
- (RS)-[(25)-ethoxycarbonyl-methoxy-1,2,3,4-tetrahydronaphth-2-yl]-2-(chlorophenyl)-2 hydroethanamine hydrochloride
- SR-59119
- N-[(7-methoxy-1,2,3,4-tetrahydronaphtalen-(2R)-2-yl)methyl]-(2R)-2-hydroxy-2(3-chlorophenyl)ethanamine hydrochloride
- SR-59104
- N-[(6-hydroxy-1,2,3,4-tetrahydronaphtalen-(2R)-2yl)methyl]-(2R)-2-hydroxy-2-(3-chorophenyl)ethanamine hydro chloride
- BRL-44408
- 2-[(4,5-dihydro-1H-imidazol-2-yl)methyl]-2,3-dihydro-1-methyl-1H-isoindole
- ACR-239
- 2-[2,4-(2-methoxyphenyl)piperazin-1-yl] ethyl-4,4-dimethyl-1,3-(2H,4H)-isoquinolindione
- MK-912
- 1′,3′-dimethylspiro(1,3,4,5′,6′,7,12b)-octahydro-2H-benzo[b]furo[2,3-a]quinazoline)-2,4′-pyrimidin-2′one
- The American Society for Pharmacology and Experimental Therapeutics
References
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵