Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Abundant evidence suggests that renal sympathetic tone is an important contributor to hypertension in several experimental models of high blood pressure (Katholi, 1985; DiBona, 1989, 1991). For instance, renal denervation reduces blood pressure in spontaneously hypertensive rats (Winternitz et al., 1980; Winternitz and Oparil, 1982) and in rats with deoxycorticosterone acetate (DOCA)-salt hypertension (Katholi et al., 1983; Takahashi et al., 1984). Moreover, renal sympathetic nerve activity is increased in spontaneously hypertensive rats (DiBona et al., 1996) and in rats with DOCA-salt hypertension (Fujita and Sato, 1984). Because renal sympathetic tone participates in the pathophysiology of hypertension, it is likely that factors that enhance the adverse effects of renal sympathetic activity on kidney function also contribute to the etiology of hypertension.

In this regard, adenosine may be an important modulator of the effects of renal sympathetic activity on kidney function. Although adenosine, acting at a prejunctional site, inhibits the release of norepinephrine from renal sympathetic nerves (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978; Ekas et al., 1981), a postjunctional action of adenosine enhances renal vascular responses to activation of renal sympathetic nerves (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978). Adenosine also augments the renal vascular response to angiotensin II (Jackson, 1997), and because renin release is increased by renal sympathetic nerve stimulation, adenosine-mediated enhancement of the renal effects of angiotensin II may participate in the overall renal response to renal sympathetic activation. Also, adenosine is well known to increase sodium chloride reabsorption (Jackson, 1997), particularly in the proximal tubule, and this effect may augment antinatriuresis induced by activation of renal sympathetic nerves. Finally, adenosine stimulates the sympathetic nervous system by activation of renal afferent nerves (Katholi et al., 1984), a process that may contribute to the pathophysiology of hypertension (Katholi, 1985; Janssen et al., 1989).

The likelihood that adenosine participates in the kidney response to renal sympathetic nerve activity would be increased if renal sympathetic nerve activation (RSNA) were shown to stimulate the renal production of adenosine. However, very limited data are currently available that address this important issue. A benchmark study by Fredholm and Hedqvist (1978) demonstrated nerve stimulation-induced release of radiolabeled purines from rabbit kidneys preloaded with [³H]adenine, but to our knowledge a thorough evaluation of the release of endogenous purines by the kidney in response to sympathetic nerve activation has not been reported. Accordingly, the purpose of this study was to characterize the effects of RSNA on renal purine secretion and to determine the relative contribution of - and -adrenoceptors in this response. Our results indicate that RSNA markedly increases renal purine production and that this response is inhibited by both - and -adrenoceptor blockade.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Sprague-Dawley rats (adult male) were obtained from Charles River (Wilmington, MA) and were maintained in the University of Pittsburgh Animal Facility. Animals were fed Prolab RMH 3000 (0.26% sodium and 0.82% potassium; PMI Feeds, St. Louis, MO).

Perfused Rat Kidney. Rats were anesthetized with sodium pentobarbital (45 mg/kg, i.p. injection). A midline incision was made, and the left kidney, renal artery and ureter, and aorta were cleared from surrounding tissue. The left ureter was cannulated with polyethylene-10 tubing and the abdominal aorta below the left kidney was cannulated (polyethylene-50 tubing). The suprarenal aorta was ligated, and the left kidney was flushed with oxygenated Tyrode's solution (2.5 ml/min) containing heparin (100 U/ml). While maintaining perfusion, the left kidney was isolated and mounted in a water-jacketed organ chamber that was maintained at 37°C with a thermostatically controlled water circulator (Thermocirculator; Harvard Apparatus, South Natick, MA). Kidneys were perfused with a Harvard model 1210 peristaltic pump at a rate of 5 ml/min (nonrecirculating). The perfusate was Tyrode's solution (NaCl, 137 mM; KCl, 2.7 mM; CaCl₂, 1.8 mM; MgCl ₂, 1.1 mM; NaHCO₃, 12 mM; NaH₂PO₄, 0.42 mM; D(+)-glucose, 5.6 mM) that was gassed with 95% O₂ and 5% CO₂, heated to 37°C with a warming coil, and passed through a bubble trap. Perfusion pressure was monitored continuously via a Statham pressure transducer (model P23ID; Statham Division, Gould Inc., Oxnard, CA) connected to a port located in the perfusion line just proximal to the kidney and recorded on a Grass model 79D polygraph (Grass Instruments, Quincy, MA). In pilot studies, we found that the renal purine secretion rate was high just after starting perfusion, yet decreased steeply with time. Consequently, a 3-h stabilization period was allowed before conducting the experiments so that a very low and stable baseline renal purine secretion rate was established.

Protocol A. Protocol A consisted of one 3-min period of RSNA. RSNA was accomplished by placing bipolar electrodes around the renal artery and electrically stimulating the periarterial sympathetic nerves with a Grass stimulator (model SD9E) using bipolar square wave pulses (frequency, 7 Hz; pulse duration, 1 ms; electrical potential, 35 V). One minute before and during the last minute of RSNA, perfusate exiting the renal vein was collected on ice and frozen at 40°C for later analysis of purines.

Protocol B. After 170 min of stabilization, either prazosin (₁-adrenoceptor antagonist; 0.03 µM), phentolamine (_1/2-adrenoceptor antagonist; 3 µM), or DL-propranolol (_1/2-adrenoceptor antagonist; 0.1 µM) was added directly to the perfusate. These receptor antagonists were obtained from the Sigma Chemical Co. (St. Louis, MO). Some kidneys (controls) received no antagonists. The concentrations of prazosin and phentolamine were selected to provide perfusate concentrations that were approximately 30- to 300-fold greater than their K_Bs (Williams and Lefkowitz, 1978; Watson and Arkinstall, 1994) for the -adrenoceptor to assure complete blockade of -adrenoceptors. The concentration of DL-propranolol was selected to provide a perfusate concentration that was approximately 10-fold greater than its K_B (Williams and Lefkowitz, 1978) for the -adrenoceptor to provide a high degree of -adrenoceptor blockade while avoiding higher concentrations that would cause membrane stabilizing/local anesthetic effects. Ten minutes later, RSNA was elicited at 1 Hz using the method described in protocol A. One minute before and during the last minute of RSNA, perfusate exiting the renal vein was collected on ice and frozen at 40°C for later analysis of purines. This procedure was repeated at 10-min intervals as the frequency was increased to 3, 5, 7, and then 9 Hz, with a new baseline sample taken before each 3-min stimulation period.

Sample Analysis. The concentrations of adenosine, inosine, hypoxanthine, xanthine, and uric acid in the perfusate were measured with an Isco (Lincoln, NE) high-pressure liquid chromatographic system (pump model 2350, gradient programmer model 2360, 4.6 × 250-mm C₁₈ column with 5-µm particle size; ChemResearch Data Management System) using UV detection as previously described (Mi and Jackson, 1995). The renal secretion rate of adenosine, inosine, hypoxanthine, xanthine, and uric acid was calculated by multiplying the concentration of each substance in the venous perfusate by the perfusion rate.

Data Analysis. In protocol A, the mean basal level of each parameter was compared with the mean level of that parameter during the 7-Hz RSNA using a two-tailed, paired Student's t test. In protocol B, the changes in each parameter in response to each level of RSNA for each of the four groups was calculated. Theses changes were then compared globally using a three-factor general linear model analysis of variance in which factor A was treatment level (fixed factor with four levels: vehicle, prazosin, phentolamine, or propranolol), factor B was kidney level (a factor nested under factor A with 20 levels representing each kidney in the protocol), and factor C was RSNA level (fixed factor with five levels: 1, 3, 5, 7, and 9 Hz). The overall analysis of variance was followed by one-term Fisher's least significant difference (LSD) tests to compare the overall effects of each treatment group versus the control group and by two-term Fisher's LSD tests to compare the effects of each treatment group versus the control group at each frequency of RSNA. All statistical analyses were performed using the Number Cruncher Statistical System (Kaysville, UT), and all values in the text and figures refer to means ± S.E.M.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

The objective of the first protocol was to carefully and completely characterize the effects of RSNA on purine release from the perfused rat kidney. To achieve a stable and low background release of purines, kidneys were perfused for 3 h before RSNA. Also, to avoid carryover effects, only a single frequency of RSNA was employed. Finally, to achieve a high level of statistical power, 13 kidneys were studied. As shown in Fig. 1, RSNA at 7 Hz markedly (2- to 5-fold) and significantly increased the renal venous secretion rate of adenosine (p = .0035), inosine (p = .0128), hypoxanthine (p = .0420), and xanthine (p = .0004). Although the renal venous secretion rate of uric acid was also significantly (p = .0202) increased by RSNA, this response was modest (uric acid secretion increased only approximately 25%). RSNA also significantly (p < .0001) increased renal perfusion pressure from 93 ± 8 to 199 ± 15 mm Hg.

View larger version (45K): [in this window] [in a new window]   Fig. 1.   Effects of RSNA (7 Hz) on renal venous secretion rate of adenosine, inosine, hypoxanthine, xanthine, and uric acid and on perfusion pressure. Values represent means ± S.E.M. for 13 experiments. The p values are for two-tailed, paired Student's t tests.

The objective of the second protocol was to determine the effects of - and -adrenoceptor blockade on purine secretion induced by RSNA. A range of frequencies of RSNA were explored to determine whether the effects of - and -adrenoceptor blockade on purine release were frequency dependent. Figures 2 to 6 illustrate the changes in purine secretion in response to RSNA in the absence and presence of propranolol, prazosin, and phentolamine. Statistical examination of the data using a global three-factor analysis of variance indicated that RSNA-induced secretions of adenosine (p = .0328; Fig. 2) inosine (p = .0006; Fig. 3), hypoxanthine (p = .0547; Fig. 4), xanthine (p < .0001; Fig. 5), and uric acid (p = .0079; Fig. 6) were diminished by one or more of the treatments. To determine which treatments were affecting secretion of purines, the overall effects of propranolol, prazosin, and phentolamine on renal purine secretion were analyzed by performing Fisher's LSD tests on the treatment factor, i.e., the frequency levels were collapsed together for each treatment group (one-term Fisher's LSD tests). This analysis indicated that all three treatments significantly (p < .05) inhibited the secretion of adenosine, inosine, hypoxanthine, and uric acid, with the exception that the effect of phentolamine on adenosine secretion did not reach statistical significance (p = .055). For inosine, hypoxanthine, xanthine, and uric acid, the global three-factor analysis of variance indicated a significant interaction between the effects of the various treatments and the frequency of RSNA. Therefore, these data were further analyzed using Fisher's LSD tests to explore the effects of each treatment at each level of RSNA (two-term Fisher's LSD tests). For inosine (Fig. 3), hypoxanthine (Fig. 4), xanthine (Fig. 5), and uric acid (Fig. 6) the results of this analysis were uniform, i.e., all three treatments significantly reduced purine secretion at 5, 7, and 9 Hz.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in renal venous adenosine secretion at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in renal venous inosine secretion at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

View larger version (39K): [in this window] [in a new window]   Fig. 4.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in renal venous hypoxanthine secretion at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in renal venous xanthine secretion at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in renal venous uric acid secretion at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

Figure 7 summarizes the effects of the treatments on changes in perfusion pressure induced by RSNA. A global three-factor analysis of variance indicated a significant (p < .0001) effect of one or more of the treatments on renovascular responses to RSNA. To determine which treatments were affecting renovascular responses, the overall effects of propranolol, prazosin, and phentolamine on vascular responses to RSNA were analyzed by performing Fisher's LSD tests on the treatment factor, i.e., the frequency levels were collapsed together for each treatment group (one-term Fisher's LSD tests). This analysis indicated an overall significant (p < .05) effect of prazosin and phentolamine. In contrast, the overall effects of propranolol were not significant. Because the interaction term was significant in the global three-factor analysis of variance, these data were further analyzed using Fisher's LSD tests to explore the effects of each treatment at each level of RSNA (two-term Fisher's LSD tests). This analysis indicated a significant effect of prazosin and phentolamine at 3, 5, 7, and 9 Hz and a significant, albeit small, effect of propranolol at 7 and 9 Hz. Quantitatively, prazosin and phentolamine nearly abolished renovascular responses to RSNA at 3, 5, 7, and 9 Hz, whereas propranolol had no effect on the 3-Hz response and slightly (by approximately 20%) diminished the responses to 7 and 9 Hz.

View larger version (43K): [in this window] [in a new window]   Fig. 7.   Effects of propranolol (0.1 µM), prazosin (0.03 µM), and phentolamine (3 µM) on RSNA-induced changes in perfusion pressure at stimulation frequencies of 1, 3, 5, 7, and 9 Hz. Values represent means ± S.E.M. for five experiments in each of the four groups. The p values for the global three-factor analysis of variance and for the one-term Fisher's LSD tests are shown in box. Asterisks over bars indicate significantly different from control group (p < .05) at indicated frequency (two-term Fisher's LSD tests).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study indicates that activation of renal sympathetic nerves in the perfused rat kidney markedly increases renal production of adenosine and adenosine metabolites. Moreover, these responses are abolished by either - or -adrenoceptor blockade. There are three major questions that the results of this study raise. First, what is the potential significance of RSNA-induced adenosine production with regard to renal physiology? Second, what is the potential significance of RSNA-induced adenosine production with regard to the antihypertensive mechanisms of - and -adrenoceptor blockers? Third, what is the mechanism of RSNA-induced adenosine production?

The secretion of adenosine by the kidneys in response to RSNA may importantly modulate the overall renal response to sympathetic stimulation. Activation of renal sympathetic nerves increases renal vascular tone (Janssen et al., 1997), renin release (Keeton and Campbell, 1980), and sodium chloride reabsorption in the proximal tubule (Zambraski, 1989). Although adenosine inhibits norepinephrine release from renal sympathetic nerve terminals (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978; Ekas et al., 1981), adenosine potentiates the postjunctional response to norepinephrine (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978) such that the overall effect of adenosine on noradrenergic neurotransmission in the kidney may be facilitatory (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978) or inhibitory (Ekas et al., 1981) or neutral (Yoneda et al., 1990). Therefore, adenosine released in response to sympathetic nerve activation may modulate the overall renovascular response to sympathetic nerve activation. With regard to the renin-angiotensin system, adenosine inhibits renin release (Jackson, 1991), but potentiates angiotensin II-induced vasoconstriction of intrarenal, preglomerular blood vessels (Jackson, 1997). Thus, adenosine released in response to sympathetic nerve stimulation may function to reduce the renin release response to renal sympathetic activation and at the same time potentiate the effects of circulating angiotensin II on preglomerular blood vessels. Finally, because adenosine enhances the reabsorption of sodium chloride in the proximal tubule, it is conceivable that adenosine released in response to RSNA participates in the increased rate of sodium chloride reabsorption that occurs in response to renal sympathetic stimulation.

To determine whether sympathetically induced adenosine release modulates the renal response to sympathetic nerve stimulation, it will be necessary to examine the effects of highly selective adenosine receptor antagonists on renal responses to sympathetic nerve activation. At present, published data in this regard are limited in scope. Ekas et al. (1981) and Yoneda et al. (1990) reported that theophylline, a nonspecific adenosine receptor antagonist, did not alter the vasoconstrictive response to sympathetic nerve stimulation in the perfused rat kidney and the dog kidney in vivo, respectively. However, Hedqvist et al. (1978) found that theophylline inhibited vasoconstrictive responses to sympathetic nerve activation in the perfused rabbit kidney. In both the dog kidney (Yoneda et al., 1990) and rabbit kidney (Hedqvist et al., 1978), theophylline increased renal nerve stimulation-induced release of norepinephrine. Clearly, additional studies using newer and better adenosine receptor antagonists and examining a wider range of renal function parameters are required to clarify the role of sympathetically induced adenosine release as a modulator of the renal response to sympathetic activation.

Additional studies are also required to clarify the time course and frequency response of adenosine release in the kidney. In the present study, we examined the release of purines in response to 3 min of sympathetic nerve stimulation; however, whether the same profile of purine metabolites would be released from the kidney during more prolonged periods of renal sympathetic nerve stimulation needs to be examined. Moreover, in the present study, although the relationship between frequency of nerve stimulation and the release of purines was strong for inosine, hypoxanthine, xanthine, and uric acid, this relationship was weak for adenosine per se. This probably reflects the extensive metabolism of released adenosine by the kidney, a hypothesis that could be tested by determining the frequency-response relationship with respect to adenosine release in the presence of inhibitors of adenosine metabolism.

The present study raises the possibility that RSNA-induced adenosine production may participate in the antihypertensive mechanisms of - and/or -adrenoceptor blockers. As noted, adenosine has several effects on the kidney that are prohypertensive such as facilitation of postjunctional renovascular responses to norepinephrine (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978) and angiotensin II (Jackson, 1997), enhancement of tubular sodium reabsorption (Jackson, 1997), and stimulation of renal afferent nerves (Katholi et al., 1984). Because blockade of either - or -adrenoceptors abolishes RSNA-induced adenosine production, this may contribute importantly to the antihypertensive mechanism of - and/or -adrenoceptor antagonists. This may be particularly important for -adrenoceptor blockers because interference with renal adenosine production would not be expected to enhance renin release, because renin release would be suppressed by the direct effects of -adrenoceptor antagonists on juxtaglomerular cells (Keeton and Campbell, 1980). Thus, any antihypertensive actions of adenosine due to inhibition of renin release would be minimal in the presence of -adrenoceptor blockade. Despite decades of intensive research, the antihypertensive mechanisms of action of -adrenoceptors have remained elusive (Prichard and Owens, 1990), and the results of the present study provide a new research lead in this regard.

The current dogma regarding the regulation of adenosine production is that the rate of adenosine biosynthesis is primarily determined by energy balance, i.e., whenever energy demand exceeds energy supply, ATP is degraded to adenosine (Schrader et al., 1995). Because RSNA causes vascular smooth muscle contraction, which would increase energy demand while decreasing energy supply in smooth muscle cells due to hypoperfusion, we anticipated that blockade of RSNA-induced vasoconstriction with the -adrenoceptor antagonists prazosin and phentolamine would block the release of purines in response to RSNA. Indeed, this prediction was confirmed. However, propranolol, a -adrenoceptor antagonist, did not alter renovascular responses to low frequencies of nerve stimulation and only slightly reduced renovascular responses to high frequencies of nerve stimulation, and yet propranolol completely prevented renal nerve stimulation-induced purine release. This result was not anticipated and strongly challenges the notion that renal nerve stimulation-induced purine release is due simply to energy depletion leading to ATP hydrolysis. Rather, our results are consistent with the notion of a pool of releasable purines that is converted to adenosine and subserves an autocrine/paracrine function to modulate the effects of RSNA on the kidneys.

A possible explanation for the ability of propranolol to abolish purine secretion in response to RSNA is that propranolol has membrane-stabilizing activity (Hoffman and Lefkowitz, 1996) and this could have accounted for the observed effect. However, this is unlikely on two accounts. First, the concentrations of propranolol used were low (0.1 µM), and second, membrane-stabilizing effects would have caused a marked reduction in renovascular responses to sympathetic nerve activation, which, as mentioned, was not observed.

A more probable mechanism for the effects of propranolol is that blockade of -adrenoceptors inactivates the "cyclic AMP (cAMP)-adenosine pathway". Previous studies from this laboratory have shown that in intact kidneys (Mi et al., 1994; Mi and Jackson, 1995) and in cultured preglomerular vascular smooth muscle cells (Jackson et al., 1997), mesangial cells (Dubey et al., 1997), aortic vascular smooth muscle cells (Dubey et al., 1996b, 1998), and cardiac fibroblasts (Dubey et al., 1996a), cAMP is converted to adenosine. This cAMP-adenosine pathway is mediated by the conversion of cAMP to AMP by phosphodiesterase, followed by the conversion of AMP to adenosine by 5'-nucleotidase (Mi and Jackson, 1995). Because in the rat kidney approximately 60% of the -adrenoceptors are ₁-adrenoceptors (Lakhlani et al., 1994) which, unlike ₂-adrenoceptors, have a high affinity for norepinephrine, it is likely that RSNA increases purine release in part by stimulating the cAMP-adenosine pathway via activation of ₁-adrenoceptors coupled to adenylyl cyclase. However, this hypothesis does not explain why -adrenoceptor blockade also inhibits purine secretion in response to RSNA. One possibility is that -adrenoceptor activation increases the activity of one or more of the enzymes involved in the cAMP-adenosine pathway, whereas ₁-adrenoceptor activation provides the prerequisite cAMP. In this regard, Kitakaze and coworkers (Kitakaze et al., 1995) demonstrated marked activation of ecto-5'-nucleotidase in cardiac myocytes by activation of ₁-adrenoceptors with methoxamine . The involvement of ₁-adrenoceptors in the present study is underscored by the fact that prazosin, an ₁-adrenoceptor antagonist, was as effective as phentolamine, an _1/2-adrenoceptor antagonist, in reducing the secretion of purines in response to RSNA.

Another possible explanation for the ability of propranolol to diminish renal nerve stimulation-induced purine release is that ₂-adrenoceptors facilitate the release of neurotransmitters from sympathetic nerves. It is well accepted that ATP, an adenosine precursor, is a cotransmitter with norepinephrine in sympathetic nerves (Sneddon and Westfall, 1984; Burnstock, 1986; von Kügelgen and Starke, 1991). Indeed, studies by Bohmann et al. (1997) demonstrate ATP release from isolated, perfused rat kidneys in response to sympathetic nerve stimulation. Moreover, recent evidence suggests that soluble nucleotidases are coreleased with ATP and norepinephrine from sympathetic nerves so that coreleased ATP is rapidly transformed to adenosine (Kennedy et al., 1997; Todorov et al., 1997). Because propranolol blocks prejunctional ₂-adrenoceptors and thereby reduces neurotransmitter release, this could explain the ability of propranolol to inhibit sympathetic nerve stimulation-induced purine release. Furthermore, ATP is also released from endothelial cells and vascular smooth muscle cells (Pearson and Gordon, 1979), and these cellular elements, in response to neurotransmitters, could also participate in the release of adenosine precursors in response to sympathetic nerve stimulation. In this regard, blockade of postjunctional ₁-adrenoceptors could diminish purine release by decreasing the stimulus for ATP release by endothelial cells and/or vascular smooth muscle cells.

In summary, this study is the first to profile the effects of RSNA on endogenous purine secretion in the rat kidney. Moreover, this study is the first to demonstrate the involvement of both - and -adrenoceptors in the purine release induced by RSNA. Although speculative, our data support that hypothesis that adenosine contributes to the renal response to sympathetic nerve activation and suggest a possible antihypertensive mechanism for -adrenoceptors, i.e., interference with sympathetically mediated renal adenosine production.