Abstract Adenosine is produced locally in the kidney. Accumulating data suggest that adenosine plays a role in regulating renal functions. Using a microdialysis technique, we monitored adenosine levels in cortical and medullary renal interstitial fluid and urine after 5 days of diets containing low (0.15%), normal (0.28%), and high (4.0%) sodium. Samples were collected from anesthetized rats (n=5 for each diet). Microdialysis fluid was infused at a rate of 1 μL/min. Adenosine, measured by radioimmunoassay, was stable in the dialysate. During normal sodium intake, renal interstitial fluid adenosine estimated from the concentration in dialysate leaving the cortex was 63±6 nmol/L, which was significantly lower than in dialysate leaving the medulla (157±6 nmol/L, P<.01). The concentration of interstitial medullary adenosine was estimated to be 190 nmol/L. In rats consuming a low sodium diet, renal cortical and medullary dialysate adenosine concentrations were significantly decreased (P<.01) by 62.6% and 64.9%, respectively. Rats consuming a high sodium diet had renal cortical and medullary dialysate adenosine concentrations that were increased 18.2- and 18.9-fold, respectively (P<.01), compared with levels in rats on a low sodium diet. Similar to changes in dialysate adenosine, urinary adenosine concentration decreased during low sodium intake (P<.01) and increased during high sodium intake (P<.01). The higher adenosine levels in renal medullary than in cortical interstitial fluid may reflect its major renal site of generation. The changes in renal adenosine generation with sodium intake may reflect renal energy expenditure.
Adenosine produced by the kidney and other tissues as a by-product of ATP metabolism is involved in the regulation of renal functions, including renal blood flow and glomerular filtration rate.1 2 Adenosine also inhibits renin release, sympathetic neurotransmission, platelet aggregation, lipolysis, and stimulation of glucose oxidation.3 Adenosine is produced intracellularly by 5′-nucleotidase, and its production is enhanced during hypoxia and ischemic and metabolic stress. 5′-Adenosylhomocysteine hydrolase also catalyzes adenosine production from 5′-adenosylhomocysteine.3 In the kidney, adenosine activates both A1 and A2a receptors.3 The A1 receptor acts to inhibit the activity of adenylyl cyclase and mobilize intracellular Ca2+, whereas A2 receptors (A2a and A2b) act to stimulate adenylyl cyclase.3 Both A1 and A2 receptors are present in the kidney.3 Measurements of renal interstitial adenosine and adenosine concentrations in urine and arterial and renal venous plasma4 5 have provided some important information, but adenosine has not been measured in renal extracellular fluid in response to physiological maneuvers.
In the present study, we evaluated the effects of sodium intake on intrarenal levels of adenosine in the rat. We used a renal interstitial fluid microdialysis technique to sample rat renal cortical and medullary adenosine levels in response to low, normal, and high dietary sodium intake.
Renal Microdialysis Technique
For the determination of renal interstitial fluid adenosine, we constructed a microdialysis probe as previously described.6 7 8 A single piece of 1-cm-long hollow fiber dialysis tubing (0.1-mm ID; molecular mass cutoff, about 5000 D) was inserted into a manually dilated (with a warm 26-gauge needle) end of a 30-cm-long hollow polyethylene (inflow) tube (0.12-mm ID, 0.65-mm OD; Bioanalytical Systems). The other end of the dialysis fiber was inserted into an end of a second polyethylene (outflow) tube (similar in characteristics to the inflow tube). The distance between the ends of the polyethylene tubes was 3 mm (dialysis area), and the dialysis fiber was sealed in place within the polyethylene tubes with cyanoacrylic glue. The dead volume of the dialysis tubing and outflow tube was 3.6 μL. The microdialysis probe was sterilized by a gas sterilization method.
In Vitro Microdialysis
In vitro recovery of adenosine was evaluated by immersing dialysis membranes of 10 probes in a beaker containing [3H]adenosine (1 000 000 cpm/mL). The inflow tube of each probe was connected to a gas-tight syringe filled with lactated Ringer’s solution and perfused at 1, 2, 3, and 5 μL/min. The effluent was collected from the outflow tube for 30-minute sample periods. The recovery of [3H]adenosine was calculated as %[adenosine in perfusate]/[adenosine in superfusate]. To show that negligible amounts of adenosine stick to the dialysis probe polyethylene tubes, a known amount of [3H]adenosine (1 000 000 cpm/mL) was perfused through these tubes. The perfusate and [3H]adenosine in the dialysis tubing were counted. To estimate in vitro relative recovery of inulin, this experiment was repeated with the use of [3H]inulin (1 000 000 cpm/mL) instead of [3H]adenosine.
In Vivo Equilibrium Microdialysis
To accurately estimate the renal interstitial concentration of adenosine, we used a technique in which adenosine was added to the dialysate, as previously described.5 The medullary microdialysis probe (n=5) was perfused with different concentrations of adenosine (0 to 500 nmol/L). Dialysate fluid was collected during perfusion at each concentration and its adenosine concentration was determined. A linear regression analysis was performed to determine the relationship between the net loss or gain of adenosine in the collected dialysate and initial adenosine concentration in the perfused fluid. The concentration at which there is no net flux of adenosine across the dialysis membrane can be considered an estimate of renal cortical interstitial adenosine concentration.
Animal Preparation and Renal Interstitial Fluid Collection
We conducted experiments in three groups of 6-week-old female Sprague-Dawley rats (n=5 each) (Harlan Sprague Dawley Inc). After arrival at the age of 5 weeks, the rats consumed a chow diet (Bioserve) containing either 0.15% (low), 0.28% (normal), or 4.0% (high) NaCl for 5 consecutive days. Twenty-four–hour urinary sodium excretion and body weights were monitored before and at the end of the 5 days of controlled sodium intake. Then, with rats under general anesthesia (80 mg/kg ketamine IM, Aveco Co Inc; and 8 mg/kg xylazine IM, Mobay Corp, Animal Health Division), the right kidney was exposed via a midline abdominal incision. The right renal capsule was penetrated with a 31-gauge needle that was tunneled into the outer renal cortex approximately 1 mm from the outer renal surface for 0.5 cm before it exited by penetrating the capsule again. The tip of the needle was inserted into one end of the dialysis probe, and the needle was pulled together with the dialysis tube until the dialysis fiber was situated in the renal cortex. The same procedure was repeated to place the dialysis probe into the right renal medulla, except the dialysis fiber was placed approximately 5 mm from the outer renal surface. The inflow tubes of the renal interstitial cortical and medullary dialysis probes were connected to gas-tight syringes filled with lactated Ringer’s solution and perfused at 1 μL/min (pump 22, Harvard Apparatus). At this perfusion rate, the concentration of dialysate [3H]adenosine was 53% of the concentration of [3H]adenosine in Ringer’s solution surrounding the dialysis tubing (Fig 1⇓).
Urine was collected from the right kidney by placing a 5-cm-long hollow PE-10 tube (0.28-mm ID, 0.61-mm OD; Clay Adams) in the right ureter. A 90-minute recovery period elapsed before the experimental protocol was initiated. This was a sufficient time for dialysate adenosine to reach a steady-state level. The effluent from the microdialysis outflow tube and the urine from the right kidney were collected for 270 minutes. The urine and interstitial samples were stored at −20°C until radioimmunoassay.9
To ensure that the dialysate was not contaminated by fluid from renal tubules, we infused [3H]inulin (1 000 000 cpm/200 μL) intravenously at the end of each study into the rats on the normal sodium diet. Urine and renal interstitial dialysate fluid were collected during inulin infusion, as well as blood at the end of the infusion period, and their [3H] content were counted. Also, each kidney was examined histologically at the end of each study to confirm the location of the dialysis fibers.
Radioimmunoassay of Adenosine
Samples of urine or renal dialysate (10 μL) were diluted to 100 μL in phosphate-buffered saline. Diluted samples or standards were then treated with 50 μL each of 0.3 mol/L ZnSO4 and 0.3 mol/L Ba(OH)2, centrifuged to remove a white precipitate containing adenine nucleotides, and assayed for adenosine by radioimmunoassay as described.9 The nanomolar concentration of adenosine in the original undiluted dialysate was then calculated and reported as renal interstitial adenosine concentration (nanomoles per liter) without correction for the concentration gradient between the interstitial fluid and dialysate (<twofold, Fig 1⇑). Since the dialysate flow rate was 1 μL/min, concentration units can be converted to picomoles per milliliter by dividing nanomoles per liter by 1000.
Data were examined by ANOVA using the general linear model procedures of the SAS Institute10 and by linear regression analysis. Values are given as mean±SEM. Statistical significance was identified at a level of P<.05.
In Vitro and In Vivo Validation of Renal Interstitial Microdialysis
Studies of in vitro recovery of tritiated adenosine (Fig 1⇑) demonstrated that adenosine concentration in the dialysate is inversely proportional to the flow rate through the dialysis tubing. Adenosine relative recovery was about 53±2% with a perfusion flow rate of 1 μL/min. [3H]Adenosine recovery was 99.5% when perfused through the microdialysis probe polyethylene tubes, suggesting that adenosine sticking to the tubes was not a significant problem and did not alter the amount recovered from the interstitium. Since adenosine inside the dialysis tubing probably does not completely equilibrate with renal interstitial adenosine, we sought to measure cortical interstitial adenosine by an equilibrium dialysis technique. A linear relationship was observed between the different adenosine concentrations in perfusate and the change in adenosine levels in dialysate (Fig 2⇓). The estimated concentration of endogenous interstitial adenosine was 190 nmol/L from renal medulla during a normal salt intake. In vitro [3H]inulin recovery was 65±5% with a perfusion flow rate of 1 μL/min. Studies of in vivo recovery of intravenously infused [3H]inulin demonstrated that [3H]inulin (10 000 cpm/mL) in urine was much higher than in blood (8000 cpm/mL) or renal interstitial dialysate (200 cpm/mL). These data suggest that the microdialysate is not contaminated by renal tubular fluid or blood and that adenosine in the dialysate is proportional to but somewhat less than adenosine outside the dialysis tubing.
Renal Interstitial Adenosine
The mean 24-hour urinary sodium excretion during a low sodium diet was 0.02±0.004 mEq, during a normal sodium diet was 0.6±0.08 mEq, and during a high sodium diet was 9.5±0.5 mEq. The average body weights of rats before and at the end of a low sodium diet were 243±8 and 251±8 g, respectively, before and at the end of a normal sodium diet were 230±9 and 239±9 g, and before and at the end of a high sodium diet were 249±8 and 260±7 g.
Adenosine concentrations, measured by radioimmunoassay, in dialysates from rats consuming a normal diet containing 0.28% sodium were 63±6 nmol/L from the cortex and significantly higher in the medulla (157±6 nmol/L, P<.01) (Fig 3⇓). In rats consuming a low sodium diet (0.15%), renal interstitial adenosine levels from both cortex and medulla were significantly decreased (23±3 and 55±5 nmol/L, respectively; each P<.01). In rats consuming a high sodium diet (4%), both cortical and medullary adenosine concentrations were markedly increased to 418±43 and 1040±37 nmol/L (each P<.01).
As shown in Fig 4⇓, the adenosine concentration in the urine of rats consuming a normal sodium diet was 281±8 nmol/L. Similar to renal interstitial adenosine levels, these levels were significantly decreased in rats consuming a low sodium diet (197±7 nmol/L, P<.01). In rats consuming a high sodium diet, urinary adenosine levels were significantly increased, to 589±12 nmol/L (P<.01 versus a normal diet).
In this study, we monitored changes in renal interstitial fluid adenosine levels in response to changes in dietary sodium intake. In rats on a normal sodium diet, we were able to measure basal renal cortical and medullary interstitial fluid adenosine levels. Renal interstitial adenosine levels associated with a normal intake of sodium in the diet are in the same range as previously reported.5 Adenosine was significantly higher in the renal medulla than in the cortex, suggesting that the medulla is the major site of adenosine production. Although the reason for the cortical-to-medullary adenosine gradient cannot be determined from this study, it is tempting to speculate that this reflects a gradient in renal energy utilization. Since the percentage decrease in urine flow and sodium excretion in response to adenosine exceeds by far the percentage decrease in glomerular filtration rate,11 12 the observed higher adenosine levels in renal medulla are consistent with the possibility that adenosine effects on renal function are mediated by medullary as well as cortical sites. Adenosine receptors have been reported to be present in renal glomeruli, in inner and outer medullas, and in membranes of the medullary thick ascending limb tubules and inner medulla.3
Switching rats from a normal to low sodium diet caused 2.7- and 2.9-fold decreases in cortical and medullary interstitial adenosine, respectively. Conversely, switching rats from a normal to high sodium diet caused a 6.6-fold increase in both renal cortical and medullary interstitial adenosine. The results of this study extend previous observations that arterial, renal venous, and urinary adenosine levels are higher in chronically sodium-loaded than in sodium-deprived dogs.13 14 Interstitial concentrations may be most relevant to the regulation of renal function by cellular adenosine receptors accessible to the interstitial fluid.
The kidneys produce and release adenosine into extracellular fluids.15 Compared with interstitial adenosine levels, plasma adenosine is found in low concentrations (<50 nmol/L) and has a very short half-life16 of about 1 to 3 seconds, since it is rapidly taken up by red blood cells17 and metabolized. The short half-life of circulating adenosine suggests that it acts largely as a paracrine substance, exerting its effect at or near the site of its production. The large gradient of adenosine between renal interstitial fluid and plasma underscores the value of measuring the nucleotide in the interstitial compartment.
The observation that adenosine levels in the renal interstitium exceed 60 nmol/L during normal salt intake suggests that adenosine plays a role in regulating renal function under normal physiological conditions. During a high salt intake, the increase in adenosine concentration may contribute to a reduction of macula densa–mediated renin secretion18 and enhance sodium excretion. Previous experiments using hypertonic saline infusion19 suggest that the tubules generate increased amounts of adenosine in response to an enhanced NaCl load and reabsorption.
In summary, the ability to monitor changes in renal interstitial adenosine levels during different physiological maneuvers would be a great advantage in clarification of the local mechanisms of adenosine that control renal function. Our results show an unexpected and extreme sensitivity of renal interstitial fluid adenosine to changes in dietary sodium. The changes in renal adenosine generation with sodium intake may reflect renal energy expenditure. The data suggest the possibility that endogenous renal adenosine plays a role in the regulation of renal fluid and electrolyte homeostasis.
This work was supported by grant HL-4766229 from the National Heart, Lung, and Blood Institute, National Institutes of Health, to Dr Siragy. Dr Helmy M. Siragy is the recipient of a Research Career Development Award (1 K04 HL-03006) from the National Institutes of Health, Bethesda, Md. We thank Toni Barbara of the University of Virginia RIA core for adenosine determinations.
Reprint requests to Helmy M. Siragy, MD, Department of Medicine, Box 482, University of Virginia Health Sciences Center, Charlottesville, VA 22908.
- Received November 11, 1995.
- Revision received December 7, 1995.
- Accepted December 7, 1995.
Osswald H, Nabakowski G, Hermes H. Adenosine as a possible mediator of metabolic control of glomerular filtration rate. Int J Biochem. 1980;12:262-267.
Pawlowska D, Granger JP, Knox FG. Effects of adenosine infusion into renal interstitium on renal hemodynamics. Am J Physiol. 1987;252:F678-F682.
Spielman WS, Arend LJ. Adenosine receptors and signaling in the kidney. Hypertension. 1991;17:117-130.
Ramos-Salazar A, Baines AD. Role of 5′-nucleotidase in adenosine-mediated renal vasoconstriction during hypoxia. J Pharmacol Exp Ther. 1986;236:494-499.
Baranowski RL, Westenfelder C. Estimation of renal interstitial adenosine and purine metabolites by microdialysis. Am J Physiol. 1994;267:F174-F182.
Siragy HM, Jaffa AA, Margolius HS. Stimulation of renal interstitial bradykinin by sodium depletion. Am J Hypertens. 1993;6:863-866.
Goodnight JH, Harvey WR. Least Square Means in the Fixed Effects General Linear Model. Cary, NC: SAS Institute; 1978. SAS technical report R-103.
Spielman WS. Antagonistic effect of theophylline on the adenosine-induced decrease in renin release. Am J Physiol. 1984;247:F246-F251.
Churchill PC, Bidani AK. Renal effects of selective adenosine receptor agonists. Am J Physiol. 1987;252:F299-F303.
Miller WL, Thomas RA, Berne RM, Rubio R. Adenosine production in the ischemic kidney. Circ Res. 1978;43:390-397.
Möser GH, Schrader J, Deussen A. Turnover of adenosine in plasma of human and dog blood. Am J Physiol. 1989;256:C799-C806.
Lorenz JN, Weihprecht H, He XR, Skott O, Briggs JP, Schnermann J. Effects of adenosine and angiotensin on macula densa-stimulated renin secretion. Am J Physiol. 1993;265:F187-F194.
Osswald H, Mühlbauer B, Schenk F. Adenosine mediates tubuloglomerular feedback response: an element of metabolic control of kidney function. Kidney Int. 1991;39:S128-S131.