Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The preglomerular microcirculation represents an important site for the action of the endogenous nucleoside adenosine. In this regard, adenosine participates in the regulation of vascular resistance of the renal preglomerular microcirculation and in the modulation of renin release from juxtaglomerular cells that are located in the renal afferent arterioles. In the preglomerular microvessels (PGMVs), adenosine activates high-affinity A₁ receptors, resulting in direct vasoconstriction (Murray and Churchill, 1984, 1985) as well as augmentation of the vasoconstictor responses to angiotensin II (Munger and Jackson, 1994; Weihprecht et al., 1994; Traynor et al., 1998) and norepinephrine (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978). These actions of adenosine may participate in tubuloglomerular feedback (Franco et al., 1989; Schnermann et al., 1990), drug-induced neprhrotoxicity (Arakawa et al., 1996; Erley et al., 1997), and renal sympathetic neurotransmission (Mi and Jackson, 1999). In juxtaglomerular cells, adenosine also binds to high-affinity A₁ receptors, resulting in potent and efficacious inhibition of renin secretion (Jackson, 1997). This "adenosine brake" on renin release prevents hypersecretion of renin in response to several pathophysiological and pharmacological stimuli (Jackson, 1997). Given the biological significance of adenosine interactions with A₁ receptors in the preglomerular microcirculation, it is important to elucidate the mechanisms that determine adenosine biosynthesis in the PGMVs.

The "extracellular cAMP-adenosine pathway" may contribute to the local production of adenosine in a number of tissues, including the kidney (Jackson, 1991). The extracellular cAMP-adenosine pathway is defined as the egress of cAMP from cells during activation of adenylyl cyclase followed by the extracellular conversion of cAMP to adenosine. The extracellular cAMP-adenosine pathway has been identified in several tissues and cells, including the whole kidney (Mi et al., 1994; Mi and Jackson, 1995, 1998), cultured renal vascular smooth muscle cells (Jackson et al., 1997), cultured aortic vascular smooth muscle cells (Dubey et al., 1996, 1999, 2000b), cultured cardiac fibroblasts (Dubey et al., 2000a), and intact cerebral vessels (Hong et al., 1999).

It is conceivable that the extracellular cAMP-adenosine pathway contributes to the local production of adenosine in the preglomerular microcirculation. In support of this hypothesis, a recent report demonstrates that electrical stimulation of renal sympathetic nerves increases adenosine production in the perfused rat kidney and that this effect is blocked by the -adrenoceptor antagonist propranolol (Mi and Jackson, 1999). One interpretation of this finding is that activation of -adrenoceptors in the preglomerular circulation by norepinephrine triggers the extracellular cAMP-adenosine pathway, resulting in increased adenosine biosynthesis. Inasmuch as the kidney contains ₁-adrenoceptors (Lakhlani et al., 1994) that are responsive to norepinephrine, this hypothesis seems plausible.

The purpose of the present investigation was to evaluate more directly the hypothesis that intact PGMVs express a functioning extracellular cAMP-adenosine pathway. To assess this hypothesis we addressed four specific aims in freshly isolated rat PGMVs. These specific aims were the following: 1) to determine whether PGMVs metabolize exogenous cAMP to adenosine and to determine whether any such metabolism is blocked by inhibition of total phosphodiesterase or ecto-phosphodiesterase; 2) to determine the ability of -adrenoceptor activation to increase extracellular levels of cAMP; 3) to determine whether a relationship exists between intracellular and extracellular cAMP; and 4) to determine whether -adrenoceptor activation increases extracellular adenosine levels and to determine whether any such response is blocked by inhibition of total phosphodiesterase or ecto-phosphodiesterase. The results were highly consistent with the existence of an extracellular cAMP-adenosine pathway in the renal preglomerular microcirculation.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

PGMV Preparation. Adult male rats were housed at the University of Pittsburgh Animal Facility and fed Prolab RMH 3000 (PMI Feeds, Inc., St. Louis, MO) containing 0.26% sodium and 0.82% potassium. All studies received prior approval by the University of Pittsburgh Animal Care and Use Committee. Rats were anesthetized with Inactin (100 mg/kg i.p.), and the aorta below the renal artery was cannulated with polyethylene-190 tubing. The proximal aorta, mesentery artery, and small side branches of the aorta were ligated, and the kidneys were flushed (10 ml) with oxygenated Tyrode's solution (37°C). A 1% suspension of iron oxide (Aldrich Chemical Co., Milwaukee, WI) in oxygenated Tyrode's solution (10 ml; 37°C) was flushed into the kidneys. The kidneys were harvested; placed in oxygenated, ice-cold Tyrode's solution; and dissected by removing the renal medulla and interlobar arteries. The cortex was sliced into small pieces; suspended in oxygenated, ice-cold Tyrode's solution; and dispersed by pushing the cortical material through a series of increasingly small needle hubs (16, 18, 21, and then 23 gauge). The dispersed cortical material was suspended in ice-cold, oxygenated Tyrode's solution, and a magnet was applied to the tube to retrieve the iron oxide-laden PGMVs while the unwanted material was decanted. The glomeruli were removed from the microvessels by filtering the microvessel suspension through a 149-µm nylon mesh. The microvessels were retrieved from the nylon mesh and distributed into the wells of a 24-well culture plate. A sample of the PGMVs was examined by phase contrast microscopy to confirm that the preparation consisted of interlobular, accurate, and afferent arterioles without contaminating glomeruli or tubules. To each well was added 1 ml of Dulbecco's minimum essential medium containing 0.2% BSA, streptomycin (100 µg/ml), and penicillin (100 U/ml). PGMVs were incubated for 20 h in a water bath at 37°C under an atmosphere of 95% oxygen, 5% carbon dioxide. Pilot studies indicated that the cAMP responses to agonists were much greater and more reproducible if the PGMVs were allowed to recover from the isolation procedure for 20 h.

Protocol A: Conversion of Exogenous cAMP to Adenosine by PGMVs. After 20 h of incubation as described above, minimum essential medium was removed, and the PGMVs were washed twice with 1 ml of warmed (37°C), oxygenated (95% oxygen, 5% carbon dioxide) Tyrode's solution. In this regard, microvessels were retained with a magnet while the solution was decanted. Next, 0.6 ml of warmed, oxygenated Tyrode's solution was added, and the PGMVs were incubated for 30 min at 37°C under an atmosphere of 95% oxygen, 5% carbon dioxide. After 30 min, the medium was changed to 0.6 ml of warmed, oxygenated Tyrode's solution containing erythro-9-(2-hydroxyl-3-nonyl)adenine (10 µM; adenosine deaminase inhibitor) + iodotubercidin (1 µM; adenosine kinase inhibitor) with and without either cAMP (30 µM), 3-isobutyl-1-methylxanthine (IBMX, 1 mM; phosphodiesterase inhibitor), 1,3-dipropyl-8-p-sulfophenylxanthine (DPSPX, 1 mM; ecto-phosphodiesterase inhibitor), cAMP + IBMX, or cAMP + DPSPX. After 60 min the medium was collected and frozen at 40°C for subsequent analysis of adenosine levels as described below.

Protocol B: Release of cAMP into the Extracellular Compartment by PGMVs during Activation of Adenylyl Cyclase. Microvessels were prepared as described above and incubated for 60 min at 37°C under an atmosphere of 95% oxygen, 5% carbon dioxide with warmed, oxygenated Tyrode's solution (0.6 ml) containing IBMX (0.1 mM), ascorbic acid (0.1 mM), and thiourea (1 mM). The medium was collected, and the PGMVs were incubated for another 60 min with warmed, oxygenated Tyrode's solution (0.6 ml) containing IBMX, ascorbic acid, and thiourea or IBMX, ascorbic acid, and thiourea + isoproterenol (1 µM; -adrenoceptor agonist). Once again the medium was collected. Ascorbic acid and thiourea were added to the medium to protect isoproterenol from oxidation. Medium was analyzed for cAMP as described below.

Protocol C: Relationship between Intracellular and Extracellular cAMP in PGMVs. Microvessels were prepared as described above and incubated for 60 min at 37°C under an atmosphere of 95% oxygen, 5% carbon dioxide with warmed, oxygenated Tyrode's solution (0.6 ml) containing IBMX (0.1 mM) and isoproterenol (1 µM). Ascorbic acid and thiourea were not added to the medium in these experiments so as to avoid any possible interference with cAMP transport mechanisms. The medium and the PGMVs were collected separately and analyzed for cAMP as described below.

Protocol D: Conversion of Endogenous cAMP to Adenosine by PGMVs. The protocol was the same as described for protocol A, except that the treatment groups were isoproterenol (1 µM), IBMX (1 mM), DPSPX (1 mM), propranolol (1 µM; -adrenoceptor antagonist), isoproterenol + IBMX, isoproterenol + DPSPX, and isoproterenol + propranolol.

Analytical Methods. Adenosine and cAMP in the medium (extracellular compartment) and PGMVs (intracellular compartment) were measured using a previously described HPLC-fluorometric assay (Jackson et al., 1996). Protein content in PGMVs was determined by dissolving the PGMVs in 0.1% SDS and 0.1 N sodium hydroxide and measuring protein levels using the bicinchoninic acid method. Results are expressed as picomoles of adenosine or cAMP per milligram protein and all values are mean ± S.E. Statistical comparisons were performed with a two-tailed, unpaired Student's t test with a criterion of significance of P < .05.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Incubation of PGMVs for 1 h with 30 µM cAMP increased the amount of extracellular adenosine from 163 ± 18.6 (n = 18) to 9810 ± 604 (n = 12) pmol/mg of protein (P < 10⁶; Fig. 1, top). Both IBMX (1 mM) and DPSPX (1 mM) significantly (P < 10⁶ and P < 10⁵, respectively; Fig. 1, bottom) attenuated the increase in extracellular adenosine induced by 30 µM cAMP. In this regard, the cAMP-induced adenosine was 9640 ± 604 (n = 12), 868 ± 72.8 (n = 6), and 3220 ± 105 (n = 6) pmol/mg of protein in the control, IBMX-treated, and DPSPX-treated groups, respectively.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Top, extracellular (medium) adenosine levels generated by PGMVs after a 1-h incubation in the absence (basal) or presence (cAMP) of 30 µM exogenous cAMP added to the medium. Bottom, cAMP-induced extracellular adenosine generated by PGMVs in the absence (control) or presence of IBMX (1 mM) or DPSPX (1 mM). cAMP-induced adenosine was calculated by subtracting the mean extracellular level of adenosine in the absence of added cAMP (and in the absence or presence of IBMX or DPSPX) from the individual extracellular adenosine values in the cAMP-treated preparations (in the absence and presence of IBMX or DPSPX). Basal levels of extracellular adenosine were not changed by IBMX or DPSPX (data not shown). Values represent mean ± S.E. for the indicated number of preparations in parentheses. The P values were calculated with unpaired Student's t tests.

PGMVs were incubated for two consecutive 1-h periods in the presence of IBMX (0.1 mM), and the amount of extracellular cAMP was measured at the end of each 1-h incubation period (Fig. 2). In PGMVs not treated with isoproterenol (Fig. 2, left), the amount of extracellular cAMP did not change significantly from period 1 to period 2 (0.921 ± 0.051 and 0.816 ± 0.065 pmol/mg of protein, respectively; n = 41). In contrast, when PGMVs were incubated with isoproterenol (1 µM) during period 2 (Fig. 2, right) the amount of extracellular cAMP increased from 0.800 ± 0.047 to 22.3 ± 2.20 pmol/mg of protein (P < 10⁶; n = 41). In PGMVs incubated with isoproterenol (1 µM) + IBMX (0.1 mM) for 1 h, there was a significant (P < .0001) linear relationship between intracellular and extracellular amounts of cAMP (Fig. 3). In this regard, 60% of the variability in extracellular cAMP levels was explained by the intracellular cAMP levels (r² = 0.6)

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Extracellular cAMP levels generated by PGMVs during two consecutive 1-h incubation periods in the presence of IBMX (0.1 mM). In some preparations, no isoproterenol was added during either incubation period (left), whereas in other preparations isoproterenol (1 µM) was added during the second incubation period (right). Values represent mean ± S.E. for the indicated number of preparations in parentheses. The P values were calculated with unpaired Student's t tests.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Linear regression of intracellular cAMP versus extracellular cAMP in 89 PGMV preparations that were incubated with isoproterenol (1 µM) and IBMX (0.1 mM) for 1 h.

Incubation of PGMVs for 1 h with isoproterenol (1 µM) increased the amount of extracellular adenosine from 163 ± 18.6 (n = 18) to 297 ± 38.3 (n = 12) pmol/mg of protein (P = .002; Fig. 4, top). Propranolol (1 µM), IBMX (1 mM), and DPSPX (1 mM) significantly (P = .037, P = .015, and P = .026, respectively; Fig. 4, bottom) attenuated the increase in extracellular adenosine induced by 1 µM isoproterenol. In this regard, the isoproterenol-induced adenosine was 134 ± 38.3 (n = 12), 13.6 ± 20.7 (n = 7), 11.5 ± 39.7 (n = 14), and 7.31 ± 44.9 (n = 12) pmol/mg of protein in the control, propranolol-treated, IBMX-treated, and DPSPX-treated groups, respectively.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Top, extracellular (medium) adenosine levels generated by PGMVs after a 1-h incubation in the absence (basal) or presence (ISO) of 1 µM isoproterenol added to the medium. Bottom, isoproterenol-induced extracellular adenosine generated by PGMVs in the absence (control) or presence of propranolol (1 µM), IBMX (1 mM), or DPSPX (1 mM). Isoproterenol-induced adenosine was calculated by subtracting the mean extracellular level of adenosine in the absence of isoproterenol (and in the absence or presence of IBMX or DPSPX) from the individual extracellular adenosine values in the isoproterenol-treated preparations (in the absence and presence of IBMX or DPSPX). Basal levels of extracellular adenosine were not changed by propranolol, IBMX, or DPSPX (data not shown). Values represent mean ± S.E. for the indicated number of preparations in parentheses. The P values were calculated with unpaired Student's t tests.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study demonstrates that addition of exogenous cAMP to PGMVs markedly increases extracellular levels of adenosine generated by PGMVs. Because the membrane permeability of cAMP is low, this finding suggests that cAMP in the extracellular compartment is metabolized extracellularly to AMP and hence to adenosine. Indeed, most cells are richly endowed with ecto-5'-nucleotidase, a glycosylphosphatidylinositol-linked enzyme that converts AMP to adenosine (Zimmermann, 1992), and our previous studies demonstrate the existence of ecto-phosphodiesterase, an enzyme that converts cAMP to AMP, in a number of tissues and cell types (Mi and Jackson, 1995; Dubey et al., 1996, 2000a,b). The fact that generation of extracellular adenosine by extracellularly added cAMP is blocked by the well known phosphodiesterase inhibitor IBMX (Beavo and Reifsnyder, 1990) supports the conclusion that phosphodiesterase is involved in the conversion of cAMP to adenosine. However, the latter observation does not distinguish between ecto-phosphodiesterase versus intracellular phosphodiesterase because IBMX freely permeates cell membranes. However, DPSPX is not membrane permeable (Tofovic et al., 1991) and inhibits phosphodiesterase at appropriate concentrations (Mi and Jackson, 1995; Dubey et al., 1996). The present finding that DPSPX inhibits the conversion of cAMP to adenosine in PGMVs confirms the hypothesis that extracellular cAMP is converted in the extracellular compartment to adenosine in PGMVs.

The extracellular cAMP-adenosine pathway is defined as the egress of cAMP from cells during activation of adenylyl cyclase followed by the extracellular conversion of cAMP to adenosine. Studies with exogenous cAMP indicate that extracellular conversion of cAMP to adenosine occurs; however, this does not per se fulfill all the criteria necessary to establish the existence of an extracellular cAMP-adenosine pathway because by definition this pathway includes transport of cAMP out of cells during activation of adenylyl cyclase. In this regard, our studies demonstrating that isoproterenol, a -adrenoceptor agonist that stimulates adenylyl cyclase, causes a 28-fold increase in extracellular cAMP levels confirms robust egress of cAMP from PGMVs during adenylyl cyclase activation. Moreover, the linear relationship between intracellular and extracellular cAMP strongly supports transport of cAMP from the intracellular to the extracellular compartment in PGMVs.

The above-mentioned experiments support the existence of the extracellular cAMP-adenosine pathway in PGMVs. However, it is conceivable that even though exogenous cAMP is converted to adenosine and even though endogenous cAMP is exported from PGMVs, endogenous cAMP might not be converted to endogenous adenosine due to misalignment of cAMP transporters with ecto-phosphodiesterase. However, this is apparently not the case because in PGMVs isoproterenol increases endogenous extracellular levels of adenosine. The -adrenoceptor antagonist propranolol blocks the isoproterenol-induced increase in extracellular adenosine, suggesting that the effect of isoproterenol is mediated by -adrenoceptors coupled to adenylyl cyclase. Importantly, both IBMX and DPSPX block the isoproterenol-induced increase in extracellular adenosine levels. This latter finding indicates that activation of adenylyl cyclase leads to increases in extracellular adenosine via a pathway that involves phosphodiesterase and, more specifically, ecto-phosphodiesterase.

What are the physiological and pharmacological implications of the present findings? As noted above, adenosine importantly contributes to the regulation of preglomerular vascular tone. In this regard, adenosine causes direct vasoconstriction (Murray and Churchill, 1984, 1985) and potentiates the vasoconstrictor effects of angiotensin II (Munger and Jackson, 1994; Weihprecht et al., 1994; Traynor et al., 1998) and norepinephrine (Hedqvist and Fredholm, 1976; Hedqvist et al., 1978). Thus, the cAMP-adenosine pathway in the preglomerular microcirculation may importantly contribute to renovascular responses to angiotensin II, circulating catecholamines, and renal sympathetic nerve stimulation. Indeed, our previous work demonstrates that renal nerve stimulation releases adenosine and adenosine metabolites from the perfused rat kidney via a pathway that involves -adrenoceptors (Mi and Jackson, 1999). It is conceivable that norepinephrine released from renal sympathetic nerve terminals increases extracellular levels of adenosine by stimulating ₁-adrenoceptors, thereby activating the extracellular cAMP-adenosine pathway in the renal preglomerular microcirculation. Adenosine would then augment the renovascular response to renal sympathetic nerve activation. Although angiotensin II is usually associated with inhibition, not stimulation of adenylyl cyclase, our previously published results demonstrate that angiotensin II may synergize with -adrenoceptor agonists in the preglomerular microcirculation to increase cAMP production (Mokkapatti et al., 1998). Thus, the renal vascular response to angiotensin II may also be influenced by activation of the extracellular cAMP-adenosine pathway. However, whether the extracellular cAMP-adenosine pathway does indeed function to regulate renovascular resistance can only be established by comprehensive studies in intact kidneys.

In addition to regulation of preglomerular vascular tone, adenosine also importantly contributes to the regulation of renin release from juxtaglomerular cells (Jackson, 1997). Because juxtaglomerular cells reside in the preglomerular microcirculation and are themselves modified preglomerular vascular smooth muscles cells, the present findings suggest that the extracellular cAMP-adenosine pathway may exert a controlling influence on renin release. Indeed, many stimuli release renin by activating adenylyl cyclase (Jackson, 1991) and thus would concomitantly engage the extracellular cAMP-adenosine pathway. Because adenosine inhibits renin release, this mechanism would serve to limit the renin release response to a number of physiological and pharmacological stimuli. This concept is supported by studies demonstrating that blockade of adenosine receptors augments renin release in response to salt depletion (Kuan et al., 1989), systemic vasodilation (Tofovic et al., 1991), -adrenoceptor activation (Pfeifer et al., 1995), furosemide (Paul et al., 1989), and renal artery hypotension (Deray et al., 1989; Kuan et al., 1990).

The question as to whether the extracellular cAMP-adenosine pathway in the preglomerular microcirculation contributes substantially to total adenosine production by the kidney and to the "average" adenosine levels in the renal interstitium cannot be deduced from the present in vitro findings. It seems unlikely that this is the case because renal tubular epithelial cells, not PGMVs, are thought to be the most important source of renal adenosine (Miller et al., 1978; Spielman and Thompson, 1982; Navar et al., 1996). Another question not addressed by the present study is the role of the extracellular cAMP-adenosine pathway in the biosynthesis of adenosine by other renal elements such as tubular epithelial cells. It seems likely that the extracellular cAMP-adenosine pathway would contribute significantly to adenosine production by renal tubular epithelial cells because proximal tubular cells are well known to be richly endowed with both ecto-5'-nucleotidase and adenylyl cyclase. Additional studies are required to address both of these important questions.

It is critical to point out that the aforementioned questions are separate and distinct from the question as to whether the extracellular cAMP-adenosine pathway in the preglomerular microcirculation importantly contributes to the regulation of preglomerular vascular resistance and renin release. In this regard, we note that cAMP egress results in the direct placement of cAMP on the cell surface where it is converted to adenosine by two ecto-enzymes, ecto-5'-nucleotidase and ecto-phosphodiesterase. Thus, cAMP transport and extracellular metabolism to adenosine are highly spatially-linked processes that would provide large concentrations of adenosine in the biophase of the unstirred water layer of the cell membrane even though the absolute amount of adenosine produced may be very small. Most likely, the extracellular cAMP-adenosine pathway is a low-capacity/high-efficiency biochemical mechanism that functions to produce "autocoid" adenosine rather than "interstitial" adenosine.

Quantitative comparisons between the results with exogenous cAMP versus endogenous cAMP are instructive. In the experiments with exogenous cAMP, we added 30 µM cAMP, which was approximately 1125 nmol of cAMP/mg of protein (30 µM = 18 nmol/0.6 ml = 18 nmol/16 µg of protein = 1125 nmol/mg of protein), and during a 1-h incubation the adenosine levels in the presence of inhibitors of adenosine metabolism increased by approximately 10 nmol/mg of protein. Thus, the ratio of adenosine generated during the 1-h incubation to extracellular cAMP levels during the 1-h incubation was only 0.009 (approximately 1%). This poor efficiency is not surprising given the fact that most of the bulk medium was not in contact with the PGMVs. Nonetheless, the huge concentration of exogenous extracellular cAMP afforded a large enough adenosine signal to permit unambiguous detection of this low-capacity process in a small amount of tissue.

In contrast to exogenous cAMP, even in the presence of IBMX to block phosphodiesterases, the extracellular levels of cAMP during a 1-h incubation with isoproterenol increased by only 0.02 nmol/mg of protein. Yet, despite this modest increase in extracellular cAMP, isoproterenol increased extracellular adenosine levels by approximately 0.134 nmol/mg of protein, giving a ratio of adenosine generated during the 1-h incubation to extracellular cAMP levels during the 1-h incubation period of 6.7. This number is actually a lower limit of the ratio because the extracellular cAMP levels during incubations with isoproterenol were measured in the presence of IBMX to block phosphodiesterase (in the absence of IBMX the extracellular cAMP levels were below the assay detection limit) and antioxidants to inhibit degradation of isoproterenol, whereas the increases in extracellular adenosine levels were measured in the absence of IBMX (in the presence of IBMX isoproterenol did not generate adenosine) and in the absence of antioxidants. Thus, the conversion of endogenous extracellular cAMP to adenosine was much more than 700 times efficient compared with the conversion of exogenous extracellular cAMP (6.7 divided by 0.009). These calculations reveal two important aspects of the extracellular cAMP-adenosine pathway in PGMVs: 1) the endogenous pathway is highly efficient, most likely because of tight coupling of cAMP transport to metabolizing enzymes; and 2) cAMP egress is likely the rate-limiting step in the pathway.

In conclusion, the present findings demonstrate in isolated PGMVs that exogenous cAMP is converted to adenosine, that stimulation of adenylyl cyclase leads to the export of cAMP into the extracellular compartment, and that activation of adenylyl cyclase increases extracellular levels of adenosine by a mechanism that involves phosphodiesterase and, more specifically, ecto-phosphodiesterase. These results are highly consistent with the existence of a functional extracellular cAMP-adenosine pathway in the preglomerular microcirculation that may importantly contribute to the regulation of renovascular tone and renin release.