Brief Review. Adenosine Receptors and Signaling in the Kidney. William S. Spielman and Lois J. Arend

Transcription

1 Brief Review 117 Adenosine Receptors and Signaling in the Kidney William S. Spielman and Lois J. Arend It is now generally accepted that adenosine is capable of regulating a wide range of physiological functions. Nowhere is the diversity of this action better illustrated than in the kidney. When adenosine binds to plasma membrane receptors on a variety of cell types in the kidney, it stimulates functional responses that span the entire spectrum of renal physiology, including alterations in hemodynamics, hormone and neurotransmitter release, and tubular reabsorption. These responses to adenosine appear to represent a means by which the organ and its constituent cell types can regulate their metabolic demand such that it is maintained at an appropriate level for the prevailing metabolic supply. Extracellular adenosine, produced from the hydrolysis of adenosine 5'-monophosphate and stimulated by increased substrate availability and enzyme induction, acts on at least two types of cell surface receptors to stimulate or inhibit the production of cyclic adenosine-3',5'-monophosphate and also acts in some renal cells to stimulate the production of inositol phosphates and elevation of cytosolic calcium concentration. To understand when and why this complicated system becomes activated, how it interacts with other known extracellular effector systems, and ultimately how to manipulate the system to therapeutic advantage by selective agonists or antagonists, requires a detailed knowledge of renal adenosine receptors and their signaling mechanisms. The following discussion attempts to highlight our knowledge in this area, to present a modified hypothesis for adenosine as a feedback regulator of renal function, and to identify some important questions regarding the specific cellular mechanisms of adenosine in renal cell types. (Hypertension 1991;17: ) The kidney, like many organs, is endowed with intrinsic mechanisms that allow it to self-regulate many of its functions. Two well-known examples are the relatively constant glomerular filtration rate (GFR) and renal blood flow over a wide range of arterial pressure (i.e., autoregulation) and the ability to release renin in response to changes in renal perfusion pressure independent of neural or humoral signals. It was the search to understand the renal mechanism responsible for this intrinsic regulation that prompted investigators to turn their interest toward adenosine. Previous reviews 1-5 have dealt specifically with a proposed role for adenosine in the control of GFR and renin release. Although this may have prompted much of the original interest in intrarenal adenosine, recent investigations have revealed adenosine to have a much broader regulatory role. The present discussion focuses From the Departments of Physiology and Biochemistry, Michigan State University, East Lansing, Mich. Work from the authors' laboratory was supported by grants HL and DK from the US Public Health Service and grants from the American Heart Association, Michigan Affiliate, and the National Kidney Foundation of Michigan. Address for correspondence: Dr. William S. Spielman, Department of Biochemistry, 502 Biochemistry Building, Michigan State University, East Lansing, MI principally on data obtained subsequent to those earlier reviews and attempts to assemble what knowledge is available on the specific adenosine receptor subtypes and signaling pathways in renal cells. Adenosine Metabolism: Sources and Sinks Adenosine exists in the intracellular space mainly in its various phosphorylated forms, that is, adenosine 5'-monophosphate (5'AMP), adenosine 5'-diphosphate (ADP), and adenosine 5'-triphosphate (ATP). Adenosine can be thought of as a cellular power station in which energy is stored and from which energy can be obtained through phosphorylation and dephosphorylation. Based on data from a wide variety of nonrenal tissues, several enzymes maintain the intracellular concentration of free adenosine in a range well below 1 /im. These include the deamination of adenosine to inosine by adenosine deaminase, the phosphorylation of adenosine to 5'AMP by adenosine kinase, and coupling of adenosine to S-adenosylhomocysteine (SAH) by SAH hydrolase (Figure 1). Although not well investigated in the kidney, it has been reported for other tissues that the majority of basal adenosine production during normoxia derives from the action of SAH hydrolase. 6 However, under circumstances of

2 118 Hypertension Vol 17, No 2, February 1991 S -odenosylhomocysteine DIPYRIOAMOLE NBTG INOSINE* ADENOSINE i ADENOSINE ATP ADP * 5 1 AMP lidcm Metabolism of Adenosine FIGURE 1. Schematic diagram showing adenosine sources and sinks. Model of the intracellular disposition of adenosine and its transport out of the cell. See text for details. NBTG, nitrobenzylthioguanosine; ATP; adenosine 5'-triphosphate; ADP, adenosine 5'-diphosphate; 5'-AMP, adenosine 5'- monophosphate. enhanced oxygen demand or reduced supply (i.e.^ hypoxia), increased amounts of adenosine are formed 7 almost exclusively by the action of 5'-nucleotidase through increased availability of substrate (i.e., 5'AMP) and by disinhibition of the enzyme. 6 Subcellular distribution studies in the heart have demonstrated that 5'-nucleotidase is present in the cytosol and tissue homogenate. 8-9 The membrane-bound form of the enzyme is an ectoenzyme. Dawson et al 10 recently reinvestigated the location of the ecto-5'-nucleotidase in the kidney using improvements in the methodology for histochemical and immunohistochemical studies. Their findings, schematically summarized in Figure 2, demonstrate ecto-5'-nucleotidase activity in the proximal tubule brush border and intercalated cells of the connecting tubule and collecting duct. Ecto-5'-nucleotidase activity was also observed in cortical peritubular and perivascular connective tissue but was absent in the medullary interstitium. In the heart, experiments using either chemical or immunological inhibitors of the ecto-5'-nucleotidase were without effect on hypoxia-induced adenosine release, indicating that most of the extracellular adenosine is from an intracellular source Additional investigation is needed to determine if the formation of extracellular adenosine in the kidney is indeed restricted to these specific sites or if, as in the heart, other forms of the 5'-nucleotidase (i.e., soluble) are also contributory. 13 Regardless, the highly specific localization of the renal ecto-5'-nucleotidase raises FIGURE 2. Schematic representation summarizing the localization of ecto-5'-nucleotidase in rat kidney; the enzyme has been revealed in the shaded areas. Dense stippling, cortical interstitial tissue in the labryinth; light stippling, interstitial tissue in the medullary rays; dashed line delimits the medullary ray from the labyrinth; hatching, brush border in the proximal tubule; black dots, intercalated cells in connecting tubules and collecting duct; different shading intensities indicate corresponding differences in staining for the enzyme. C, cortex; OM, outer medulla; OS, outer stripe; IS, inner stripe; IM, inner medulla. Reproduced with permission." interesting questions about its function and begs further investigation. Once formed within the cells, adenosine can pass the cell membrane by facilitated diffusion, which is a concentration-dependent process This process is blocked by dipyridamole and related compounds. In the extracellular space, adenosine can activate specific membrane-bound receptors (see below). Extracellular adenosine is inactivated mainly by uptake via the nucleoside carrier (see Figure 1) with subsequent deamination to inosine and phosphorylation to 5'AMP as well as through extracellular deamination by adenosine deaminase. Free adenosine that exits the cell and enters the blood is also rapidly inactivated by these processes in the erythrocyte. For this reason it is normally unlikely that adenosine would ever reach sufficient concentration in blood to act as a blood-borne mediator transferring information from one organ to another. However, during situations in which adenosine uptake is pharmacologically

3 Spielman and Arend Adenosine and Renal Function 119 inhibited with nucleoside transport inhibitors (e.g., dipyridamole, papaverine), plasma levels of adenosine may be markedly elevated, 16 affecting function in multiple organs. Recently, proximal tubule brush border membranes have been demonstrated to transport adenosine via a sodium-dependent, concentrative process. 17 -' 9 Although the function of this transporter is not yet clear, it is likely involved in the reabsorption of plasma adenosine filtered at the glomerulus. Adenosine Feedback Hypothesis Because of the relation of adenosine formation to oxygen supply and demand and its property of relaxing vascular smooth muscle, it has been postulated that adenosine acts as a negative feedback regulator to return the oxygen supply/demand ratio toward normal. This hypothesis was first suggested by Berne 20 in the United States and Gerlach et al in Germany for the local regulation of coronary blood flow by adenosine acting as a vasodilator. This feedback hypothesis was extended to explain various types of active hyperemia in other tissues including exercising skeletal muscle and postprandial hyperemia of the gastrointestinal tract. 23 Adenosine was often disregarded as a hemodynamic regulator in the kidney because of the following general concept: a vasodilator, produced in association with the increased metabolism of active transport, would lead to an increase in GFR and, in turn, further stimulate active reabsorption with the result of additional generation of metabolic vasodilator and thus little hope of maintaining constancy of blood flow and GFR. 24 What is general knowledge today, but not then widely appreciated, is that among the vast array of physiological actions of adenosine, 25 one of the most curious, in contrast to its ability to relax most vascular smooth muscle, is renal vasoconstriction. (Adenosine-induced renal vasoconstriction was first reported by Drury and Szent-Gyorgyi in ) xh e vasoconstrictive response and fall in GFR in response to increased energy use results in a curtailment of metabolism through the reduction of solute delivered to the tubular epithelium. This seemingly paradoxical vasoconstrictive action of adenosine and the resultant energy-sparing effect on tubular transport by reducing GFR has led various investigators to propose that endogenous adenosine production mediates changes in vascular resistance that maintain a constant GFR This hypothesis was extended to the single nephron level by OBwald et al, 28 who suggested that adenosine is the mediator of the preglomerular vasoconstriction in tubuloglomerular feedback (see below). Recent findings that adenosine acts directly on the renal tubular epithelium to inhibit hormone-stimulated cyclic AMP production and, therefore, transport (see below for description) has led to modification of the original hypothesis. Figure 3 attempts to assemble a paradigm that incorporates the wellknown action of adenosine, presumably produced by CONSTRICTION FIGURE 3. Schematic diagram showing adenosine feedback hypothesis for the kidney. Adenosine, produced by transporting epithelium, reduces glomerular filtration rate (GFR) by preglomerular and postglomerular mechanisms, thereby decreasing solute delivery to the nephron. Adenosine reduces the metabolic demand of the epithelium by inhibiting hormonestimulated cyclic adenosine 5'-monophosphate both at the epithelium and by reducing neurotransmitter release. PTH, parathyroid hormone; VP, vasopressin; ATP, adenosine 5'- triphosphate. transporting epithelium, to reduce GFR by vasoconstriction at the afferent arteriole and vasodilation at the efferent arteriole, thereby regulating the "supply of delivered solute" to the nephron. In addition, the action of adenosine to inhibit hormone-stimulated cyclic AMP in various segments of the nephron, both directly on the cell and indirectly through the inhibition of neurotransmitter release, serves to reduce the "metabolic demand" of the tubular cells. Therefore, the hemodynamic and tubular actions of adenosine work together to adjust the metabolic supply and demand toward a level of transport activity appropriate for the oxygen and substrate availability of the tissue. (Any attempt to formulate a working hypothesis for renal adenosine acting to couple metabolic supply and demand, based primarily.on the activation of inhibitory A r receptors, must also take into account the directionally opposite effects of activating the A 2 -receptor. However, because of the higher affinity of the A,-receptor population to ligand [i.e.,

4 120 Hypertension Vol 17, No 2, February 1991 ADENOSttC -ADENO3NE FIGURE 4. Model for adenosine receptors and their interaction with adenytyl cyclase. See text for details. GTP, guanosine 5'-triphosphate; ATP, adenosine 5'-triphosphate; camp, cyclic adenosine 5'-monophosphate. activation of receptors at lower concentrations of ligand], the presence of A 2 -receptors does not preclude a feedback hypothesis from operating, assuming that adenosine fails to reach concentrations that would activate the A 2 -receptor.) Classification of Adenosine Receptors Adenosine receptors are members of a large class of hormone receptors that, like the visual pigment rhodopsin, are coupled to their intracellular effector systems via guanine nucleotide binding regulatory proteins. The primary basis for the classification of adenosine receptors comes from work in nonrenal tissue, which has been extensively reviewed in previous articles Basically, it was observed that in some tissues, adenosine results in the stimulation of cyclic AMP production and in others results in the inhibition of cyclic AMP. These observations have led to the conclusion that these directionally opposite responses were mediated by different receptors. 31 This was soon followed by the observation that analogues of adenosine, substituted at the N 6 position, for which i?-iv 6 -phenylisopropyladenosine (R- PIA) is the prototype, had greater activity for the inhibitory adenosine receptor (designated A,), whereas 5'-substituted analogues, for which 5'-Nethylcarboxamidoadenosine (NECA) is the prototype, had greater potency for the stimulatory adenosine receptor (designated A 2 ) (Figure 4). From that time forward, our knowledge of the adenosine receptor, its subtypes, and its coupling to second messengers has, to a great extent, followed the development of specific agonist and antagonist ligands. Figure 4 shows the current paradigm for the receptor classification and signaling pathways for adenosine based on studies from a wide variety of cells. A high-affinity receptor (A,) acts to inhibit the activity of adenylyl cyclase, and a low-affinity A 2 receptor acts to stimulate adenylyl cyclase. Both A! and A 2 receptor populations are coupled to adenylyl cyclase through the guanine nucleotide binding proteins G and G,, respectively. In addition, there is an adenosine binding site on the catalytic subunit of adenylyl cyclase itself, referred to as the P-site, which inhibits cyclase activity. The functional significance of the P-site is not clear, although the very low affinity for adenosine makes it an unlikely candidate as the normal ligand. Recent work has demonstrated that 2'-deoxy-3'AMP and 3'AMP, metabolites of nucleic acids, have sufficient affinity and concentration to be candidates as the endogenous ligands. 32 In general, adenosine receptors appear to exhibit no distinguishing properties, other than their ability to recognize adenosine, that set them apart from the large family of receptors that modify adenylyl cyclase activity. Although both A, and A 2 receptors are often depicted, as the cartoon in Figure 4 shows, coexisting on the same cell, the general dogma is that a given population of cells has either A, or A 2 receptors but not both. 33 As described below, a clear exception appears to be cells in the kidney. The adenosine A, receptor from brain has been photoaffinity labeled and followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE), revealing a specifically labeled band with an apparent molecular weight of 35,000 Da Total deglycosylation by enzymatic 36 or chemical 37 methods gives a core protein of 32,000 Da. Recently, the A, receptor from rat brain has been purified to homogeneity by affinity chromatography. 38 This step should greatly facilitate the molecular characterization of the A, receptor, including sequencing of the receptor protein and cloning of the gene. Development of Selective Ligands Since the definition of adenosine receptor subclasses, 3139 considerable progress has been made in the development of selective adenosine agonist and antagonist ligands, with an eye toward their potential use as therapeutic neuropharmaceuticals. This effort has followed the recognition that adenosine may be an important modulator of transmitter release in the central nervous system and is likely the system through which common chemical stimuli such as caffeine and theophylline act. This effort in ligand development has contributed important tools for the investigation of adenosine's action outside the nervous system. For example, these analogues, of which there are now many, have in addition to relative specificity for the receptor subtypes the advantage that for the most part they are less effectively transported by the nucleoside carrier and are poor substrates for the enzymes that deaminate and phosphorylate adenosine. (2-Chloroadenosine is a substrate for the nucleoside transporter and R-PIA enters some cells, probably by the carrier. 140 ) The resistance of these analogues to enzymatic degradation by adenosine deaminase has provided an important means of managing the endogenous adenosine environment in studies using isolated cells and membranes. Whole cells have been reported to release adenosine into the medium 41 as do membrane preparations. 42 This endogenous adenosine occupies the available adenosine receptors and thereby precludes

5 Spielman and Arend Adenosine and Renal Function 121 odenosine DPCPX FIGURE 5. Schematic diagrams of structures of adenosine agonists and antagonists. NECA, 5' -N-ethylcarbaxamidoadenosine; CHA, N 6 -cyclohexyladenosine; R-PIA, R-N 6 -phenylisopropyladenosine; DPCPX, 8-cyclopentyl-l,3-dipropylxanthine. the observation of functional responses and specific receptor binding to added adenosine analogues. The inclusion of adenosine deaminase to the incubation medium provides a convenient means of eliminating the contaminating endogenous adenosine. The nonmetabolized adenosine analogues can then be added to the incubation medium, providing a precisely controlled concentration of agonist at an unoccupied receptor. So far the outcome of the search for specific receptor ligands, which has been extensively reviewed elsewhere, 2943 has been to yield a number of A r selective agonist and antagonist compounds. A,- selective agonist ligands are all A r6 -substituted analogues, with increased A, selectivity conferred by cyclic substitutions at position 6 (Figure 5). On the other hand, only very recently have compounds with appreciable A 2 selectivity 44 become available. The synthesis of antagonist ligands, which are almost all xanthine derivatives, was based on the observations that A! selectivity of xanthines can be enhanced by alkyl substitutions in positions 1 and 3 and by ring substitutions in position 8 (Figure 5). l,3-diethyl-8-phenylxanthine (DPX) was the first of such compounds with moderate affinity and A, selectivity. A similar approach was used for the synthesis of a xanthine amine congener (XAC). 45 An 8-cyclopentyl substitution leads to 8-cyclopentyl-l,3-dipropylxanthine (DPCPX). This compound combines a high affinity for A! receptors with a 1,000-fold selectivity. Radioligand Binding Studies The use of radiolabeled agonist and antagonist ligands for binding studies has confirmed and extended our knowledge of adenosine receptor classification based on adenylyl cyclase studies. Although not as yet used extensively for investigation of renal adenosine receptors, A, receptors have been identified in rabbit renal glomeruli using the radiolabeled agonist (-)[' 25 I]A f6 -(4-hydroxyphenylisopropyl)-adenosine (['^IJHPIA). 46 A[ receptors have been visualized in both human and guinea pig glomeruli by the specific localization of the A, ligand [ 3 H]JV 6 -cyclohexyladenosine ([ 3 H]CHA), using autoradiography; however, binding was not localized to a specific glomerular cell type. 47 Radiolabeled ligand binding has also been used to identify the presence of A 2 receptors in human renal papillae. 48 Autoradiographic localization of the A, agonist [ 3 H]CHA was reported in the inner and outer medulla of the guinea pig kidney. 47 Studies in the rat kidney using [ 12? I]HPIA and [ 3 H]DPCPX demonstrated specific binding in crude membranes from the inner stripe of outer medulla and the inner papilla, as well as isolated medullary thick ascending limb (mtal) tubules, suggesting the presence of A, receptors in the thick ascending limb and papillary collecting duct In addition to the simple and direct demonstration of adenosine receptors by specific binding, these radioligands have allowed the study of the functional regulation of receptors. For example (Figure 6), R-PIA competition for [ 3 H]DPCPX binding sites demonstrates two affinity states of the A[ receptor for agonist ligands in membranes prepared from a cell line (28A) derived from the rabbit cortical collecting tubule. The addition of guanosine-5'triphosphate (GTP) to the preparation leads to a decrease in the number of high affinity binding sites and a resultant shift of the displacement curve to the left (because of the effect of GTP to uncouple the inhibitory guanine nucleotide binding protein [Gj] from the receptor, thereby decreasing affinity) as is the case for numerous G protein-coupled receptors Antagonist ligand binding, on the other hand, appears to recognize only a single affinity state and is unaffected 33 or actually increased by guanine nucleotides, making them especially useful in studies of whole cells where GTP levels are high. Physiological Actions of Renal Adenosine Receptor Activation With the increased recognition of a wide array of renal cellular actions and the continuing development of relatively specific adenosine receptor agonist and antagonist ligands, investigators have undertaken the

6 122 Hypertension Vol 17, No 2, February «RCT-2BA WHBRAMES o c c ou o CD X a a. a BO GTP \ \ \ \ H L ^ K -O.5nU Kj-10.5nU K -18nU \ FIGURE 6. Line graph showing competition binding curve for the A t receptor and the effect of guanosine 5'-triphosphate (GTP). In the absence of GTP, two binding sites are observed, with a KJor the high affinity site (H) of 0.5 nm and a K.,for the low affinity site (L) of 10.5 nm. Addition of 100 fim GTP results in a shift to a single low affinity binding site with a K, of 18 nm. DPCPX, 8-cycbpentyl-l,3-dipmpylxanthine; R-PIA, R-N 6 -phenylisopmpyladenosine. 0 o "' 10"' CONCENTRATION OF R-PIA (M) 10' task of assigning the different renal actions of adenosine to the known adenosine receptor types, as previously identified in other tissues, by comparison of relative agonist and antagonist potencies (Table 1). Hemodynamic Actions The hemodynamic actions of adenosine are varied and demonstrate an interesting interaction with angiotensin (See previous reviews for a detailed description. 2-5 ) Briefly, the infusion of adenosine into the renal artery of dogs results in a prompt fall in blood flow, which rapidly wanes, with renal blood flow returning to a value below, at, or above the control level. Despite the return of bloodflowtoward preinfusion levels during the continued infusion of adenosine, GFR generally remains depressed until TABLE 1. Renal Actions of Adenosine Effect ' Adenosine receptor Hemodynamic ( GFR) Vasoconstriction (preglomerular) A! Vasodilation (postglomerular) A 2 Hormonallneurotrawmitter Renin release Inhibition A, Stimulation. A 2 Erythropoietin Inhibition A] Stimulation A 2 Adrenergic transmission Inhibition (presynaptic) A[ Tubular Collecting tubule t Hydraulic conductivity A 2 Thick ascending limb i Sodium chloride transport A] GFR, glomerular filtration rate; I, decrease; t > increase. the infusion is terminated. 55 The vascular response of the kidney is vasoconstriction of the outer cortex accompanied by vasodilation of the deep cortex. 56 Studies directed at the mechanism of the adenosineinduced fall in GFR demonstrate a fall in glomerular hydrostatic pressure resulting from a preglomerular vasoconstriction and a more slowly developing postglomerular vasodilation, findings in keeping with the biphasic action on bloodflowand the sustained effect on GFR. 55 More recent investigations into the subtype of adenosine receptor responsible for these hemodynamic effects, taking advantage of the nonmetabolized adenosine receptor agonists, convincingly indicate that the vasoconstrictive response is mediated by activation of an A, receptor, whereas the renal vasodilation develops from stimulation of the lower affinity A 2 receptor No evidence has been gathered to demonstrate an effect of adenosine on hydraulic conductivity or surface area of the glomerular membrane. However, adenosine and adenosine analogues have been reported to stimulate an increase in cyclic AMP production by isolated glomeruli This observation of adenosine-induced cyclic AMP production in glomerular cells, presumably via activation of A 2 receptors, is consistent with the action of other effectors of the glomerular filtration coefficient (K { ) on cyclic AMP, raising the possibility that the decrease in GFR during adenosine infusion is also the result of a decrease in K f. It has long been recognized that an increase in the perfusion rate of fluid in the distal nephron results in a vasoconstrictive response of the afferent arteriole such that the glomerular filtration rate of that nephron is diminished. It has been hypothesized that this phenomenon, termed tubuloglomerular feedback, serves as a mechanism whereby each nephron has the ability to limit large increases in tubular fluid and solute delivery and the consequent alterations in

7 tubular and excretory function It is proposed that changes in perfusion of the thick ascending limb are detected by a specialized group of epithelial cells, collectively termed the macula densa, which in turn signals the afferent arteriole to constrict. Although a detailed summary of the extensive research directed at determining the mediator of this phenomenon is beyond the scope of this discussion, it is sufficient to state that it remains unclear how information is transmitted from the macula densa to the adjacent afferent arteriole. An interesting hypothesis relevant to the present discussion is that with an increased delivery of solute to the cells of the macula densa, an increased reabsorptive sodium chloride transport ensues that is associated with stimulation of ATP hydrolysis and adenosine formation and release from the epithelial cells, resulting in afferent arteriolar vasoconstriction. 2-5 This hypothesis suggests that the macula densa might then serve as a "transport sensor" and uses adenosine as a paracrine signal to reduce delivered solute, and thereby transport, toward normal. 2-5 Important evidence was recently published demonstrating that the highly specific A, antagonist DPCPX was effective in the inhibition of tubuloglomerular feedback when administered either into the tubule lumen or into the peritubular capillary circulation. 64 Interestingly, it was recently reported that intraluminal administration of cyclopentyladenosine (CPA) produced an enhanced rubulogjomerular feedback response, raising the possibility that adenosine A, receptors on the luminal side of the tubule epithelium play a role in the transmission of the feedback signal. 65 (A similar observation was also made by Schnermann, 66 although he argues that the highly lipophilic adenosine analogues may have diffused out of the tubule lumen and acted directly on the vascular smooth muscle.) Spielman and Arend Adenosine and Renal Function 123 Renin Secretion Adenosine has long been known to inhibit the release of renin and is the specific topic of a recent review by Churchill and Churchill. 6 Studies both in vivo 67 and in vitro 668 " 72 have provided evidence that this action is likely a direct effect of adenosine on the renin-secreting cells rather than mediated by adenosine-induced alterations in hemodynamics, tubular reabsorption, or neurotransmitter release. At the same time that adenosine was proposed as the mediator of the vasoconstriction associated with increased perfusion of the macula densa (tubuloglomerular feedback), it was also proposed as a mediator of the inhibition of renin release by the macula densa. 2 Evidence in support of a role for adenosine in control of renin release has come principally from studies that have evaluated the action of exogenously administered adenosine and evidence in support of a role for endogenously produced adenosine in the control of renin release is largely circumstantial. A more direct assessment of the actions of intrarenal adenosine is to elevate endogenous levels of adenosine through pharmacological, physiological, or pathophysiological perturbations while monitoring renal function and then using adenosine receptor antagonists. Evidence that endogenous adenosine production can suppress renin secretion was obtained in studies in which the administration of maleic acid, a substance that alters ATP formation by the kidney and produces a generalized tubular transport defect resembling the Fanconi syndrome, results in the elevation of renal adenosine production as determined by measurements of adenosine in renal venous blood and urine and results in a marked suppression of renin release. 73 This decrease in renin release was largely reversed by the administration of theophylline. 73 Additional evidence for an inhibitory action of endogenous adenosine on renin release has been obtained in studies in which administration of l,3-dipropyl-8-(p-sulfophenyl)xanthine (DPSPX), a relatively nonselective but more potent adenosine receptor antagonist than theophylline, to sodium-depleted and normal rats resulted in an increased rate of renin release. 74 Evidence for mediation of macula densa control of renin release by adenosine was reported by Itoh et al, 75 who demonstrated that in the microdissected glomerulus, renin release was lower when the macula densa was left intact as compared with when the macula densa segment was removed. Furthermore, this difference was abolished with the administration of theophylline, suggesting that adenosine produced by the macula densa segment was responsible for the inhibition of renin release. Finally, data obtained using a single glomerulus with an attached, perfused macula densa segment demonstrated that the inhibition of renin release by elevated sodium chloride was inhibited by the A, adenosine antagonist DPCPX. 76 Collectively, these findings are supportive of a role for endogenous adenosine in regulating both basal renin release and the renin response to sodium restriction and that these effects may be mediated by a macula densa mechanism. Although most theories concerning the role of adenosine in renin release stress its inhibitory action, Churchill and Churchill 6 point out in their review that the recent use of adenosine receptor agonist ligands has revealed both an inhibitory and stimulatory action of adenosine mediated by selective activation of the A, and A 2 receptor populations, respectively. In general, studies using either isolated perfused rat kidneys or rat renal cortical slices have demonstrated that renin release can either be inhibited or stimulated by activation of adenosine receptors depending on the dose of agonist. The rank order of potency for the adenosine agonists on the renin response indicates that inhibition of renin release is by activation of the A, receptor, whereas stimulation is the result of activation of A 2 receptors (Figure 7). In addition, it has been reported that pretreatment of rats with pertussis toxin blocked the actions of CHA to inhibit renin release in kidney slices. 77 This observation suggests that the renin inhibitory effect of adenosine at the A) receptor is

8 124 Hypertension Vol 17, No 2, February 1991 OJMM Micromotor FlGURE 7. Line graph showing renin secretory rate of renal cortical slices versus the logarithm of the concentrations of N 6 -cyclohexyladenosine (CHA), 2-chloroadenosine (2- CADO), and 5' -N-ethylcarboxamidoadenosine (NECA). Reproduced with permission. n coupled to Gi. However, because a number of G proteins are now known to act as pertussis toxin substrates and are inactivated by ADP ribosylation, this observation alone cannot be taken as evidence that the A, receptor-mediated inhibition is the result of the suppression of adenylyl cyclase activity. As is discussed in detail below, adenosine A, receptor activation has recently been shown to stimulate turnover of membrane inositol phospholipids with the resultant stimulation of inositol trisphosphate (IP 3 ) production and mobilization of cytosolic free Ca 2+ via a pertussis toxin-sensitive G protein in renal epithelial cells. This observation raises the possibility of a similar mechanism existing in juxtaglomerular cells. The hypothesis that adenosine acts to inhibit renin release via elevation of cytosolic Ca 2+ has been discussed previously 678 and is particularly appealing in light of the vasoconstriction due to activation of A, receptors in afferent arteriolar vascular smooth muscle cells, the cells from which the renin-secreting cells apparently derive. Recently, it was demonstrated in isolated juxtaglomerular cells that activation of A, receptors is associated with an elevation of cyclic guanosine-3',5'- monophosphate (cyclic GMP) but not with changes in either cytosolic calcium or cyclic AMP, suggesting the involvement of yet another second messenger system for adenosine. 79 This interesting observation fails to adequately explain why adenosine activation of the A, receptor results in vasoconstriction at the afferent arteriole, a result directionally opposite to that expected with elevations of cyclic GMP. Furthermore, the previously mentioned pertussis toxin sensitivity of the response is difficult to resolve with elevation of cyclic GMP since there are no reports of G protein coupling to guanylate cyclase. Erythropoietin Secretion Erythropoietin is secreted from an unknown site in the renal cortex and is important in stimulating the production of red blood cells in the bone marrow. It was recently reported that radioiron incorporation into red blood cells of exhypoxic polycythemic mice, indicative of erythropoietin production, was inhibited by A, receptor activation and stimulated by A 2 receptor activation. 80 In a preliminary report by the same workers, activation of A, receptors inhibited and A 2 receptor activation stimulated erythropoietin production by a renal carcinoma cell culture system. 81 These intriguing findings raise the possibility that adenosine, produced from the degradation of ATP due to limited oxygen availability, is involved in the regulation of erythropoietin production. Neurotransmitter Release Adenosine acts to inhibit the release of neurotransmitter from postganglionic sympathetic neurons. 82 In the kidney, adenosine acts at a prejunctional A, receptor that reduces the release of norepinephrine. 83 " 85 Curiously, the postjunctional effects of adenosine increase the sensitivity of the kidney to transmitter. For example, the effects of nerve stimulation in the presence of elevated adenosine levels are enhanced despite diminished transmitter release. 83 Although it has been suggested that the inhibition of neurotransmitter release by adenosine could be the initiating factor for many of the other observed renal actions of adenosine, it appears unlikely that this is the case since adenosine-induced inhibition of renin release and changes in hemodynamics and excretion occur in the absence of functional renal nerves. 2-6 Although it is generally agreed that neurotransmitter release is inhibited by activation of A, adenosine receptors, the postreceptor mechanism remains unclear. Neither increases nor decreases in cyclic AMP appear to account for the presynaptic effects of adenosine, and various hypotheses for the inhibition of transmitter release have been suggested To date, no studies have directly determined the effect of adenosine analogues on postsynaptic second messengers in renal sympathetic nerves. Excretory and Tubular Epithelial Effects of Adenosine Elevation of intrarenal adenosine by exogenous administration or by pharmacological manipulation of endogenous adenosine has been reported to result in a fall of urine flow and solute excretion. 17 ' 55 Despite the fact that adenosine has been reported to change the fractional excretion rate of sodium, the concomitant changes in GFR and renal blood flow have generally made it difficult to determine in the intact kidney if adenosine has direct actions to affect tubular reabsorption. However, in a recent study in which adenosine infusions had no effect on either GFR or renal bloodflow,it was observed that urinary volume and sodium excretion were increased, 88-89

9 Spielman and Arend Adenosine and Renal Function ANALOG NECA CHA PIA FIGURE 8. Line graph showing effect of adenosine agonists on cyclic adenosine 5'-monophosphate production in renal cortical collecting tubule (RCCT) cells. See text for details. NECA, 5'-N-ethylcarboxamidoadenosine; CHA, N 6 -cyclohexyladenosine; PIA, R-N 6 -phenylisopropyladenosine. Reproduced with permission o 9 (-log M) leading the authors to suggest a direct tubular effect for adenosine. A direct action of adenosine and adenosine analogues has been reported in a wide variety of nonrenal epithelia and was recently summarized. 2 In the isolated, perfused rabbit collecting tubule, NECA has been shown to stimulate hydraulic conductivity. 90 Anand-Srivastava et al 91 demonstrated that NECA increases adenylyl cyclase activity in isolated segments of dog medullary thick ascending limb and collecting tubule. Furthermore, recent preliminary data indicate that adenosine A[ receptor activation results in a decrease in transtubular voltage in isolated cortical thick ascending limb, suggesting a direct action on sodium transport. 92 Using recently developed immunoselection techniques and the subsequent development of a clonally expanded cell line, we have investigated the presence of adenosine receptors and signaling mechanisms in tubular cells from the rabbit renal cortical collecting tubule (RCCT) and the rabbit thick ascending limb (RTAL). 93 " 95 Figure 8 illustrates the concentrationresponse curves of three adenosine receptor agonists on cyclic AMP production by primary cultures of RCCT cells. Cyclic AMP production was stimulated by micromolar concentrations of each analogue with a rank order of potency of NECA>R-PIA>CHA, indicating activation of adenylyl cyclase by an A 2 receptor. At nanomolar concentrations, the receptor agonists produced an inhibition of basal cyclic AMP production. Furthermore, stimulation of cyclic AMP in RCCT cells by isoproterenol or vasopressin was inhibited by simultaneous addition of CHA, an A, receptor agonist. This inhibitory action of CHA on hormone-stimulated cyclic AMP production was inhibited by the highly specific A, receptor antagonist DPCPX or by prior treatment of the cells with pertussis toxin (Figure 9), indicating that activation of A, adenosine receptors in RCCT cells regulates adenylyl cyclase activity. Similar evidence for both A, and A 2 receptors coupled via G proteins to adenylyl cyclase was also reported for the RCCT-28A cell, a recently established cell line derived from the rabbit cortical collecting tubule 96 and for cultured cells of the thick ascending limb. 95 A notable difference of the renal epithelial adenosine receptors and their coupling to adenylyl cyclase when compared with adenosine receptors described in brain, fat cells, and platelets is that in the renal epithelial cells studied, both stimulatory (A 2 ) and inhibitory (A,) receptors appear to exist in the same cell type, contrary to the current perception that cells may demonstrate one but not both receptor subtypes. 33 (Both adenosine A, and A 2 receptor subtypes have also been reported in the elasmobranch rectal gland, a tissue of homogeneous transporting epithelia, with similarities to the thick ascending limb of the loop of Henle. 2 ) Although the functional significance of A! and A 2 receptor populations on collecting tubule and thick limb cells is not as yet entirely clear, the ability of adenosine to regulate basal and hormone-stimulated cyclic AMP production via activation of two distinct receptor populations makes it a potentially important regulator of hormonally controlled transport. Several observations in nonrenal tissues suggest that adenosine may alter cell function independently of changes in cyclic AMP production We have recently reported in cortical collecting tubule cells, 94 in the 2&A cell line, 96 and in cultured thick ascending limb cells 95 that adenosine analogues result in the mobilization of cytosolic free calcium and the increased turnover of inositol phosphates. The adenosine analogueinduced increase in cytosolic calcium and inositol phosphate is inhibited by the specific A, receptor

10 126 Hypertension Vol 17, No 2, February 1991 CONTROL PERTUSSIS TOXIN FIGURE 9. Bar graph showing effect of pretreatment with pertussis toxin (1 fjg/mlfor 12 hours) on the modulation by VP-cyclohexyladenosine (CHA) (50 nm) of 1 fim arginine vasopressin (AVP) -stimulated cyclic adenosine 5'- monophosphate (camp) production in renal cortical collecting tubule (RCCT) cells. Reproduced with permission. 93 AVP CHA/AVP antagonist DPCPX and by prior treatment of the cells with pertussis toxin. Although the blockade of the response by a specific A, adenosine receptor antagonist and the observation that the G protein is a pertussis toxin substrate are consistent with a phospholipase C response coupled to an Aj receptor, the lack of significantly different potencies among various adenosine receptor agonists (Figure 10) raises the possibility of a previously unidentified adenosine receptor subtype similar to the A! receptor [ADENOSINE ANALOG] FIGURE 10.. Line graph showing effect of adenosine analogues on cytosolic free calcium in renal cortical collecting tubule (RCCT) cells. NECA, 5'-^i-ethylcarboxamidoadenosine; CHA, f^-cyclohexyladenosine; PIA, R-N 6 -phenylisopropyladenosine. Reproduced with permission. 94 Figure 11 illustrates models to explain the coupling of adenosine receptors in tubular epithelial cells to multiple effector systems. There may be different subtypes of adenosine A, receptor, designated A la and A 1/h that cannot be differentiated by the currently available agonist and antagonist ligands, or alternatively, differ not in their ligand recognition domains but in their coupling domains. In this scheme (Figure 11A), A lo couples via Gi to inhibit adenylyl cyclase, and A ip couples via an as yet unidentified G protein to phospholipase C. This situation would be analogous to the adrenergic receptor system in which a t receptor couples to phospholipase C and a 2 couples to the inhibition of adenylyl cyclase. The second model (Figure 11B) suggests a single A, receptor type, at least as distinguished by pharmacological criteria, that has divergent effector systems by coupling to more than one G protein. Precedents for this are the angiotensin, prostaglandin E, and P 2 - purinergic receptors, which in some systems attenuate adenylyl cyclase and provoke IP 3 production through a single receptor population The observation that both of the adenosine-mediated responses are blocked by pertussis toxin does not indicate that a single subtype of nucleotide binding protein is involved, since a family of G proteins that act as pertussis toxin substrates, ranging in size from kda, has been identified. 101 The third model (Figure 11C) depicts one subtype of Aj receptor coupled to a single type of G protein, which in turn interacts with both adenylyl cyclase and phospholipase C. At this time, but without definitive evidence to the contrary, this scheme seems the least likely. In this model, activation of any of several receptor populations that couple to Gj would invariably result in both the inhibition of adenylyl cyclase and stimulation of phospholipase C, a situation with-

11 Spielman and Arend Adenosine and Renal Function 127 FIGURE 11. Models of coupling of adenosine receptors to several effectors. Panel A: Subtypes of adenosine A j receptor, each coupled to different G proteins. Panel B: One A, receptor type, coupling to different G proteins. Panel C: One A, receptor type, coupling to one G protein. PLC, phospholipase C; AC, adenylyl cyclase; PI-P 2, phosphatidylinositol-4,5 bisphosphate; IP 3, inositol-1,4,5 trisphosphate; DAG, diacylglycerol; ATP, adenosine 5'-triphosphate; camp, cyclic adenosine 5''-monophosphate. out precedent and representing a somewhat inflexible system of control. At this time, it is difficult to speculate which of these models might be correct, although without any clear evidence for an additional subtype of A, receptor, we hypothesize that a single population of A, receptors in the collecting tubule and thick ascending limb couples to two separate messenger systems (Figure 11B). An unequivocal test of the hypothesis that one receptor mediates both responses would be insertion of the cloned adenosine A! receptor into cells lacking this receptor with demonstration that adenosine elicits both responses in these cells. As for the functional significance of adenosine receptor-activated phospholipase C stimulation, relatively little is known. Recent preliminary data indicate that adenosine A, receptor activation of phospholipase C in the 28A cell results in the opening of a 300 ps chloride channel on the apical membrane, via stimulation of diacylglycerol and activation of protein kinase C. 102 This response is inhibited by the adenosine A, receptor antagonist DPCPX or treatment of the cells with pertussis toxin. Future Considerations The presence of both stimulatory and inhibitory adenosine receptors capable of regulating several aspects of renal function illustrates the "dual-control" nature of adenosine as a regulator of kidney cell function. This dual-control regulation provides for an interesting model with which to examine renal cellular regulation and provides the impetus for the development of highly selective adenosine agonist and antagonist ligands with which to therapeutically manipulate renal function. However, it must be kept in mind that the affinity of the endogenous ligand adenosine for the inhibitory A, receptor is generally felt to be 100-1,000-fold higher than for the stimulatory A 2 receptor. Therefore, if endogenous adenosine participates in the control of renal function, its action at the inhibitory A[ receptor will most likely dominate. In the nephron, Aj adenosine receptors linked to the inhibition of adenylyl cyclase may also evoke the mobilization of cytosolic calcium via the turnover of inositol phosphates through the activation of phospholipase C coupled with a pertussis toxin-sensitive G protein. Additional molecular information is needed to determine whether a single receptor population can effect both the inhibition of adenyh/1 cyclase and the acceleration of inositol polyphosphate production, or whether particular subpopulations are linked to different effector systems. Although G proteins link receptor occupancy to changes in both inhibition of cyclase and the increase in inositol phosphate production, the identity of the G proteins and the mechanisms involved in vivo are not clear. Finally, further work is required to establish 1) which of the possible signaling events induced by occupancy of receptors coupled to adenylyl cyclase and to phospholipase C are causal in mediating a physiological event, 2) which are permissive, and 3) which are without functional consequence. It also remains to be determined if the coupling of adenosine receptors to both adenylyl cyclase and phospholipase C is unique for renal epithelia or represents a more general pattern of adenosine signal transduction. References 1. Spielman WS, Thompson CI: A proposed role for adenosine in the regulation of renal hemodynamics and renin release. Am J Physiol 1982;242:F423-F Spielman WS, Arend LJ, Forrest JN: The renal and epithelial actions of adenosine, in Gerlach E, Becker BF (eds): Topics and Perspectives in Adenosine Research. Berlin/Heidelberg, Springer-Verlag, 1987, pp OBwald H: The role of adenosine in the regulation of glomerular filtration rate and renin secretion. Trends Pharmacol Sci 1984^: OBwald H: Adenosine and renal function, in Berne RM, Rail TW, Rubio R (eds): Regulatory Function of Adenosine. Boston, Martinus Nijhoff Publishing, 1983, pp Churchill PC, Churchill MC: Effects of adenosine on renin secretion. ISI Atlas of Science: Pharmacology, 1988, pp

12 128 Hypertension Vol 17, No 2, February Lloyd HGE, Schrader J: The importance of the transmethylation pathway for adenosine metabolism in the heart, in Gerlach E, Becker BF (eds): Topics and Perspectives in Adenosine Research. Berlin/Heidelberg, Springer-Vcrlag, 1987, pp Miller WL, Thomas RA, Berne RM, Rubio R: Adenosine production in the ischemic kidney. Ore Res 1978;43:39O Lowenstein JM, Yu M-K, Narto Y: Regulation of adenosine metabolism by 5'-nucleotidase, in Berne RM, Rail TW, Rubio R (eds): Regulatory Function of Adenosine. Boston, Martinus Nijhoff Publishing, 1983, pp Schrader J: Metabolism of adenosine and sites of production in the heart, in Berne RM, Rail TW, Rubio R (eds): Regulatory Function of Adenosine. Boston, Martinus Nijhoff Publishing, 1983, pp Dawson TP, Gandhi R, LeHir M, Kaissling B: Ecto- 5'nucleotidase: Localization in rat kidney by light microscopic histochemical and immunohistochemical methods. / Histochem Cytochem 1988;37: Collinson AR, Peuhkurinen KJ, Lowenstein JM: Regulation and function of 5'-nucleotidase, in Gerlach E, Becker BF (eds): Topics and Perspectives in Adenosine Research. Berlin/ Heidelberg, Springer-Verlag, 1987, pp Schutz W, Schrader J, Gerlach E: Different sites of adenosine formation in the heart. Am J Physiol 1981;240: H963-H Newby AC, Worku Y, Meghji P: Critical evaluation of the role of ecto- and cytosolic 5'nucleotidase in adenosine formation, in Gerlach E, Becker BF (eds): Topics and Perspectives in Adenosine Research. Berlin/Heidelberg, Springer- Verlag, 1987, pp Plagemann PGW, Wohlhueter RM: Permeation of nucleosides, nucleic acid bases, and nucleotides in animal cells. Curr Top Membr Transp 1980;14: Pearson JD, Gordon JL: Nucleotide metabolism by endothe- Ymm.Annu Rev Physiol 1985;47: Arend LJ, Thompson CI, Spielman WS: Dipyridamole decreases glomerular filtration in the sodium-depleted dog. Ore Res 1985^6: LeHir M, Dubach UC: Sodium gradient-energized concentrative transport of adenosine in renal brush border vesicles. PflugersArch 1984;401: LeHir M, Dubach UC: Concentrative transport of purine nucleosides in brush border vesicles of the rat kidney (abstract). EurJ Clin Invest 1985;15: LeHir M, Dubach UC: Uphill transport of pyrimidine nucleosides in renal brush border vesicles. PflugersArch 1985;404: Berne RM: Cardiac nucleotides in hypoxia: Possible role in regulation of coronary blood flow. Am J Physiol 1963;204: Gerlach E, Deuticke B, Dreisbach RH: Der Nucleotide- Abbau in Herzmuskel bei Sauerstoffmangel und seine mogliche Bedeutung fur die Coronardurchblutung (abstract). Naturwissenschaften 1963;50: Gerlach E, Deuticke B, Dreisbach RH, Rosarius CW: Zum Verhalten von Nucleotiden und ihren dephosphorylieten Abbauprodukten in der Niere bei Ischaemie und kurzzeitiger postischaemischer Wiederdurchblutung. Pflugers Arch 1963; 278: Haddy FJ, Scott JB: Metabolicalry linked vasoactive chemicals in local regulation of blood flow. Physiol Rev 1968;48: Thurau L: Renal hemodynamics. Am J Med 1964;36: Berne RM: Adenosine: An important physiological regulator. News Physiol Sci 1986;1: Drury A, Szent-Gy6rgi A: The physiological activity of adenine compounds with special reference to their action upon the mammalian heart. / Physiol (Lond) 1929;68: Tagawa H, Vander AJ: Effects of adenosine compounds on renal function and renin secretion in dogs. Ore Res 1970;26: OBwald H, Hermes HH, Nabakowski G: Role of adenosine in signal transmission of tubuloglomerular feedback (abstract). Kidney Int 1982;22(suppl):S Lohse MJ, Klotz K-N, Schwabe U, Cristalli G, Vittori S, Grifantini M: Pharmacology and biochemistry of adenosine receptors, in Melchiorre C, Giannella M (eds): Recent Advances in Receptor Chemistry. Amsterdam, Elsevier Science Publishing Co, Inc, 1988, pp Stiles GL: Adenosine receptors: Structure, function and regulation. Trends Pharmacol Sci 1986;7: Van Calker D, Muller M, Hamprecht B: Adenosine inhibits the accumulation of cyclic AMP in cultured brain cells. Nature 1978;276: Johnson RA, Yeung SMH, Stubner D, Bushfield M, Schoshani I: Cation and structural requirements for P-site mediated inhibition of adenylate cyclase. Mol Pharmacol 1989;35: Jacobson KA: Chemical approaches to the definition of adenosine receptors, in Cooper DMF, Londos C (eds): Adenosine Receptors. New York, Alan R Liss, Inc, 1988, pp Klotz K-N, Cristalli G, Grifantini M, Vittori S, Lohse MJ: Photoaffinity labeling of A r adenosine receptors. / Biol Chem 1985;260: Stiles GL, Daly DT, Olson RA: The A, adenosine receptor: Identification of the binding subunit by photoaffinity crosslinking. / Biol Chem 1986;260: Stiles GL: Photoaffinity crosslinked A, adenosine receptor binding subunits. J Biol Chem 1986,261: Klotz K-N, Lohse MJ: The glycoprotein nature of A, receptor. Biochem Biophys Res Commun 1986; 140: Nakata H: Purification of Aj adenosine receptor from rat brain membranes. / Biol Chem 1989;264: Londos C, Cooper DMF, Wolff J: Subclasses of external adenosine receptors. Proc Natl Acad Sci USA 1980;77: Clanachan AS, Heaton TP, Parkinson FE: Drug interactions with nucleoside transport systems, in Gerlach E, Becker BF (eds): Topics and Perspectives in Adenosine Research. Berlin/ Heidelberg, Springer-Verlag, 1987, pp Londos C: Receptor G, and Gi complexes in rat adipocyte plasma membranes: Regulators of camp-related and campindependent processes, in Cooper DMF, Londos C (eds): Adenosine Receptors. New York, Alan R Liss, Inc, 1988, pp Bruns RF: Adenosine receptor assays (abstract), in Cooper DMF, Londos C (eds): Adenosine Receptors. New York, Alan R Liss, Inc, 1988, pp Bruns RF, Lu GH, Pugsley TA: Characterization of the A 2 adenosine receptor labeled by 'H-NECA in rat striatal membranes. Mol Pharmacol 1986;29: Williams M: Adenosine receptors as drug targets: Fulfilling the promise? in Jacobson K, Daly J, Mangianello V (eds): Purines in Cell Signalling: Targets for New Drugs. New York, Springer-Verlag, 1989, pp Jacobson KA, Ukena D, Kirk KL, Daly JW: [ 3 H]Xanthine amine congener of l,3-dipropyl-8-phenylxanthine: An antagonist radioligand for adenosine receptors. Proc Natl Acad Sci USA 1986;83: Freissmuth M, Hausleithner V, Tuisl E, Nannoff C, Schuetz W: Glomeruli and microvessels of the rabbit kidney contain both A! and A2 adenosine receptors. Naunyn Schmiedebergs Arch Pharmacol 1987;335: Palacios JM, Fastbom J, Wiederhold K-H, Probst A: Visualization of adenosine A! receptors in the human and the guinea-pig kidney. Eur J Pharmacol 1987;138: Woodcock EA, Leung E, Johnson CI: Adenosine receptors in papilla of human kidneys. Clin Sci 1986;70: Brines ML, Forrest JN: Autoradiographic localization of A, adenosine receptors to tubules in the red medulla and papilla of the rat kidney (abstract). Kidney Int 1987;33: Weber RG, Brines ML, Hebert SC, Forrest JN: Demonstration of Ai adenosine receptors on rat medullary thick ascend-

13 Spielman and Arend Adenosine and Renal Function 129 ing limb tubules by radioligand binding (abstract). Kid Int 1990;37: Goodman RR, Cooper MJ, Garvish, Snyder SH: Guanine nucleotide and cation regulation of the binding of [ 3 H]cyclohexyladenosine and phjdiethylphenylxanthine to adenosine A[ receptor in brain membranes. Mol Pharmacol 1982;21: Rodbell M: The role of hormone receptors and GTPregulatory proteins in membrane transduction. Nature 1980; 284: Spielman WS, OBwald H: Characterization of the postocclusive response of renal blood flow in the cat. Am J Physiol 1978;235:F286-F Spielman WS, OBwald H: Blockade of postocclusive renal vasoconstriction by an angiotensin II antagonist: Evidence for an angiotensin-adenosine interaction. Am J Physiol 1979;237: F463-F OBwald H, Spielman WS, Knox FG: Mechanism of adenosine-mediated decreases in glomerular filtration rate in dogs. Ore Res 1978;43: Spielman WS, Britton SL, Fiksen-Olsen MJ: Effect of adenosine on the distribution of renal blood flow in dogs. Ore Res 1980;46: Murray RD, Churchill PC: The effects of adenosine receptor agonists in the isolated, perfused rat kidney. Am J Physiol 1984;247:H343-H Murray RD, Churchill PC: The concentration-dependency of the renal vascular and renin secretory responses to adenosine receptor agonists. J Pharmacol Exp Ther 1985;232: Rossi NF, Churchill PC, Amore B: Mechanism of adenosine receptor induced renal vasoconstriction in the rat. Am J Physiol 1988;255:H885-H Rossi NF, Churchill PC, Jacobson KA, Leahy AE: Further characterization of the renovascular effects of N 6 -cyclohexyladenosine in the isolated perfused rat kidney. J Pharmacol Exp Ther 1987;240: Abboud H, Dousa TP: Action of adenosine on cyclic 3',5'- nucleotides in glomeruli. Am J Physiol 1983;244:F633-F Schnermann J, Briggs JP: The function of the juxtaglomerular apparatus: Local control of glomerular hemodynamics, in Seldin DW, Giebisch G (eds): The Kidney: Physiology and Pathophysiology, Volume 1. New York, Raven Press, Publishers, 1985, pp Briggs JP, Schnermann J: The tubuloglomerular feedback mechanism, in Laragh JH, Brenner BM (eds): Hypertension. New York, Raven Press, Publishers, 1990, pp Schnermann J, Weihprect H, Briggs J: Inhibition of tubuloglomerular feedback during adenosine 1-receptor blockade. Am J Physiol 1990;258:F553-F Franco M, Bell PD, Navar LG: Effect of adenosine A, analogue on tubuloglomerular feedback mechanism. Am J Physiol 1989;257:F231-F Schnermann J: Effect of adenosine analogues on tubuloglomerular feedback responses. Am J Physiol 1989;255: F33-F Arend LJ, Haramati A, Thompson CI, Spielman WS: Adenosine-induced decrease in renin release: Dissociation from hemodynamic effects. Am J Physiol 1984;247:F447-F Barchowsky A, Data JL, Whorton AR: Inhibition of renin release by analogues of adenosine in rabbit renal cortical slices. Hypertension 1987;9: Churchill PC, Bidani AK: Renal effects of selective adenosine receptor agonists. Am J Physiol 1987;252:F299-F Churchill PC, Jacobson KA, Churchill MC: XAC, a functionalized congener of 1,3-dialkylxanthine, antagonizes A[ adenosine receptor-mediated inhibition of renin secretion in vitro. Arch Int Pharmacodyn Ther 1987;290: Churchill PC, Rossi NF, Churchill MC: Renin secretory effects of N 6 -cyclohexyladenosine: Effects of dietary sodium. Am J Physiol 1987;252:F872-F Churchill PC, Churchill MC: A[ and A 2 adenosine receptor activation inhibits and stimulates renin secretion of rat renal cortical slices. J Pharmacol Exp Ther 1985;232: Arend LJ, Thompson CI, Brandt MA, Spielman WS: Elevation of intrarenal adenosine by maleic acid decreases GFR and renin release. Kidney Int 1986;30: Kuan CI, Wells JN, Jackson EKj Endogenous adenosine restrains renin release during sodium restriction. J Pharmacol Exp Ther 1989;249: Itoh S, Carretero OA, Murray RD: Possible role of adenosine in the macula densa mechanism of renin release in rabbits. J CUn Invest 1985;76: Weihprecht H, Lorenz JN, Schnermann J, Briggs JP: Effect of adenosine, receptor blockade on renin release from the rabbit isolated perfused juxtaglomerular apparatus. / CUn Invest 1990;85: Rossi NF, Churchill PC, Churchill MC: Pertussis toxin reverses adenosine receptor-mediated inhibition of renin secretion in rat renal cortical slices. Life Sci 1987;40: Churchill PC: Second messengers in renin secretion. Am J Physiol 1985;249:F175-F Kurtz A, Bruna RD, Pfeilschifter J, Bauer C: Role of cgmp as second messenger of adenosine in the inhibition of renin release. Kidney Int 1988;33: Ueno M, Brooking J, Beckman B, Fisher JW: A, and A 2 adenosine receptor regulation of erythropoietin production. Life Sci 1988;43: Ueno M, Brooking J, Beckman B, Fisher JW: Adenosine receptor regulation of erythropoietin secretion (abstract). Kidney Int 1987;31: Ginsborg BL, Hirst GDS: The effect of adenosine on the release of transmitter from the phrenic nerve of the rat. J Physiol (Lond) 1972;224: Hedqvist P, Fredholm BB: Effects of adenosine on adrenergic neurotransmission: Prejunctional inhibition and postjunctional enhancement. Naunyn Schmiedcbergs Arch Pharmacol 1976;293: Hedqvist P, Fredholm BB, Olundh S: Antagonistic effects of theophylline and adenosine on adrenergic neuroeffector transmission in the rabbit kidney. Ore Res 1978;43: Snyder SH: Adenosine as a neuromodulator. Annu Rev Neuwsci 1985;8: Fredholm BB, Dunwiddie TV: How does adenosine inhibit transmitter release? Trends Pharmacol Sci 1988;9: Silinsky EM: Inhibition of transmitter release by adenosine: Are Ca 2+ currents depressed or are the intracellular effects of Ca 2+ impaired? Trends Pharmacol Sci 1986;7: Miyamoto M, Yagil Y, Larson T, Robertson C, Jamison RL: Effect of intrarenal adenosine on renal function and medullary blood flow in the rat. Am J Physiol 1988;255: F1230-F Yagil Y, Schabel T, Jamison RL: The effects of adenosine on renal excretory function (abstract). Kidney Int 199O;37: Dillingham MA, Anderson RJ: Purinergic regulation of basal and arginine vasopressin-stimulated hydraulic conductivity in rabbit cortical collecting tubule. J Membr Biol 1985;88: Anand-Srivastava MB, Vinay P, Genest J, Cantin M: Effect of atrial natriuretic factor on adenylate cyclase in various nephron segments. Am J Physiol 1986;251:F417-F Bell PD, Franco M, Higon M: The effects of adenosine A, agonist on transepithelial potential difference and cytosolic calcium concentration in the cortical thick ascending limb (abstract). Kidney Int 199O;35:3O9 93. Arend LJ, Sonnenberg WK, Smith WL, Spielman WS: A, and A 2 adenosine receptors in rabbit cortical collecting tubule cells: Modulation of hormone-stimulated camp. / Clin Invest 1987;79: Arend LJ, Burnatowska-Hledin MA, Spielman WS: Adenosine receptor-mediated calcium mobilization in cortical collecting tubule cells. Am J Physiol 1988;255:C581-C Burnatowska-Hledin MA, Spielman WS: Regulation of hormonal responses in medullary thick ascending loop (MTAL) by adenosine (abstract). Kidney Int 1989;35: Arend LJ, Gusovsky F, Handler JS, Rhim JS, Spielman WS: Adenosine-sensitive phosphoinositide turnover in a newly