AJP - Heart Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 279: H1397-H1410, 2000;
0363-6135/00 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rubin, L. J.
Right arrow Articles by Jones, A. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rubin, L. J.
Right arrow Articles by Jones, A. W.
Vol. 279, Issue 3, H1397-H1410, September 2000

Selective transport of adenosine into porcine coronary smooth muscle

L. J. Rubin1,3, L. R. Johnson1, J. R. Dodam1, A. K. Dhalla1, L. Magliola2, M. H. Laughlin1,2,3, and A. W. Jones2,3

1 Department of Veterinary Biomedical Sciences and 2 Department of Medical Physiology, 3 Dalton Cardiovascular Research Center, University of Missouri-Columbia, Columbia, Missouri 65211


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Adenosine (ADO), an endogenous regulator of coronary vascular tone, enhances vasorelaxation in the presence of nucleoside transport inhibitors such as dipyridamole. We tested the hypothesis that coronary smooth muscle (CSM) contains a high-affinity transporter for ADO. ADO-mediated relaxation of isolated large and small porcine coronary artery rings was enhanced 12-fold and 3.4-fold, respectively, by the transport inhibitor, S-(4-nitrobenzyl)-6-thioinosine (NBTI). Enhanced relaxation was independent of endothelium and was selective for ADO over synthetic analogs. Uptake of [3H]ADO into freshly dissociated CSM cells or endothelium-denuded rings was linear and concentration dependent. Kinetic analysis yielded a maximum uptake (Vmax) of 67 ± 7.0 pmol · mg protein-1 · min-1 and a Michaelis constant (Km) of 10.5 ± 5.8 µM in isolated cells and a Vmax of 5.1 ± 0.5 pmol · min-1 · mg wet wt-1 and a Km of 17.6 ± 2.6 µM in intact rings. NBTI inhibited transport into small arteries (IC50 = 42 nM) and cells. Analyses of extracellular space and diffusion kinetics using [3H]sucrose indicate the Vmax and Km for ADO transport are sufficient to clear a significant amount of extracellular adenosine. These data indicate CSM possess a high-affinity nucleoside transporter and that the activity of this transporter is sufficient to modulate ADO sensitivity of large and small coronary arteries.

dipyridamole; S-(4-nitrobenzyl)-6-thioinosine; erythro-9-(2-hydroxy-3-nonyl)-adenine hydrochloride; 2-chloroadenosine


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

THE NUCLEOSIDE ADENOSINE has been proposed to provide a metabolic link between the myocardium and coronary vasculature during periods of increased metabolic demand (5, 9). Several selective nucleoside transporters have been identified kinetically, and seven transporters have been cloned, including one class (equilibrative) that transports nucleosides by facilitative diffusion (52). In the myocardium, this exchanger is thought to increase interstitial adenosine levels during increased metabolic activity when intracellular nucleotides undergo rapid hydrolysis (47). In this situation, adenosine serves as a metabolic signal, activating adenosine receptors on vascular smooth muscle and/or the endothelium to elicit vasodilation and increase blood flow in response to muscle metabolic demand (5, 9). Moreover, recent publications (12, 14) have proposed that production of extracellular adenosine from released cAMP provides important regulation of vascular smooth muscle nitric oxide synthesis and growth. Although the hypothetical scheme proposed by Dubey and co-workers (12, 14) included adenosine transport, no data were presented to substantiate the functional significance of adenosine in regulating vascular smooth muscle. Furthermore, reuptake or metabolism of interstitial adenosine must occur in order for metabolic regulation to operate in a reversible manner. Reuptake into cardiac muscle cells or vascular endothelium have been proposed to be the major routes for interstitial adenosine removal (15, 21, 48). For instance, indicator dilution studies with small rodent hearts indicate that endothelial cells accumulate more than 70% of the radiolabeled adenosine in the coronary vasculature during a single pass through the circulation (36).

Vascular smooth muscle uptake of adenosine had not been previously regarded as a significant contributor to the regulatory process. High-affinity uptake [i.e., a Michaelis constant (Km) <20 µM] was not observed either in intact arteries or in cultured porcine aortic cells (3, 33, 43). This latter study, however, identified a saturable high-affinity (Km = 3 µM) uptake into cultured endothelial cells, leading the authors to conclude that the endothelium plays a major role in removing extracellular adenosine, whereas smooth muscle does not. The relatively low uptake of adenosine into coronary smooth muscle compared with other cell types (3) also reinforced the conclusion that vascular smooth muscle played a minor role in adenosine regulation. One brief report (4), however, showed that cultured cerebral vascular smooth muscle accumulated adenosine with a high affinity, but data were not presented concerning sensitivity to selective blockade. It is not established whether such a transporter is expressed in vascular smooth muscle in native arteries or in acutely isolated cells not exposed to growth factors during cell culture.

Indirect evidence supports a potential role for adenosine uptake in vascular smooth muscle. Several groups (3, 24, 32, 37) have shown that the sensitivity of constricted coronary artery strips to adenosine relaxation was increased in the presence of selective blockers of adenosine transport. However, because the endothelium was not removed systematically, it was unclear whether the effect of the blockade was directly on smooth muscle or was secondary to altered uptake into the endothelium in these studies. Inhibition of adenosine transport, in turn, could promote the interaction of extracellular adenosine with endothelial receptors, which leads to the release of nitric oxide (1, 34). Furthermore, whereas indicator dilution studies clearly define the endothelium as the "adenosine sink" during bolus infusion of adenosine into the coronary circulation, these same studies show ~20% of the label could not be attributed to endothelial cell uptake (38) and may represent uptake by other cellular compartments such as the cardiac myocyte, abluminal endothelial cell surfaces (36), or, potentially, the vascular smooth muscle cell itself.

Considering that a major physiological source of adenosine, the myocardium, is abluminal and not luminal, vascular smooth muscle adenosine uptake and metabolism could play a more significant role in regulation of interstitial adenosine than previously considered, particularly in the vicinity of smooth muscle adenosine receptors. In the present study, we tested the hypothesis that selective adenosine uptake occurs in both freshly dissociated coronary smooth muscle cells and endothelium-denuded coronary arteries and that uptake is sufficient to limit adenosine availability at receptors. Portions of this work have appeared in abstract form (11, 49).


    METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Coronary arteries were obtained from hearts of either female farm pigs (4-6 mo old, 39 ± 5 kg) or female Yucatan miniature swine (6-8 mo old, 37 ± 3 kg, Charles River). Yucatan swine were housed in animal facilities at the University of Missouri for 6-12 wk. Female farm pigs were obtained locally the day before the experiment. Animals were treated in accordance with institutional guidelines for humane animal care and use. Swine were sedated with ketamine hydrochloride (2.25 mg/kg) and then anethestized with phenobarbital sodium (20 mg/kg). Heparin sodium was administered intravenously (1,000 U/kg). Swine were then euthanized by removing the heart, which was immediately placed in cold (4°C) modified Krebs bicarbonate buffer solution (Krebs buffer, see Solutions and Drugs).

Functional Measure

Conduit rings. The left anterior descending (LAD) artery, the left circumflex coronary artery (LCCA), and, for some experiments, the right coronary artery (RCA) were dissected from the heart and maintained in cold Krebs buffer. Coronary segments were trimmed of fat, heart tissue, and connective tissue with the aid of a microscope. For measurements of isometric tension, conduit rings (~3-4 mm in axial length without branches) were cut from the arteries. Axial length, inside diameter, and outside diameter (1.95 ± 0.05 mm) were measured for each ring using a Fylar calibrated micrometer eyepiece mounted on a stereomicroscope. Isometric tension was measured using two stainless steel wires passed through the lumen of each ring. One wire was connected to a force transducer (Grass model FT03); the other wire was connected to a micrometer microdrive (Stoelting), which allowed for the stretching of each ring by known increments. Rings were submerged in Krebs buffer in a 20-ml tissue bath equilibrated at 37°C with a 95% O2-5% CO2 gas mixture throughout the experiment. Data of isometric tension development were continuously recorded and collected by the DATAQ computerized data acquisition system for analysis by CODAS software.

Conduit rings were stretched to 3 g resting tension and equilibrated in Krebs buffer for 1 h. Rings were then set to the individual optima of the length-developed tension relationship of the ring as described previously (40). At optimal stretch [~170% outer diameter (OD)], vascular rings were exposed two to three times to a maximal dose of KCl (80 mM) to determine maximal relative vasoreactivity. All studies were conducted at optimal stretch.

The involvement of the endothelium was evaluated by removing the endothelium by gently rubbing the lumen of the ring with the sharp edge of a scissor. Rings with an intact endothelium and denuded rings from the same animal were exposed to identical experimental protocols. Endothelial removal was verified by measuring bradykinin-induced (1 µm) relaxation of the maximal contractile response to PGF2alpha (30 µM). Relaxations of less than 5% of the PGF2alpha -developed tension were considered indicative of effective endothelial denudation, and only data from these preparations were included in the analysis of the adenosine-mediated relaxations in the absence of endothelium.

After bradykinin assessment of endothelial function, we rinsed the rings repeatedly with Krebs buffer (2× every 20 min) until baseline levels of tension were achieved. Concentration-dependent relaxation responses of coronary rings to adenosine were then determined by cumulative additions of adenosine (10-10-10-4 M) to the tissue baths after a second maximal PGF2alpha (30 µM) precontraction. In some experiments, coronary rings were relaxed by cumulative addition of a nonselective adenosine receptor agonist, 5'-(N-ethylcarboxamido)-adenosine (NECA, 10-10-10-4 M) or 2-chloroadenosine (2-CAD, 10-10-10-4 M), or a selective A2a receptor agonist, CGS-21680 HCl (10-10-10-4 M). In all cases, after the last concentration of dilatory agonist, vessels were rinsed with Krebs buffer for at least 1 h or longer until resting tension was reattained.

To assess the effect of adenosine transport inhibition or adenosine deaminase inhibition on coronary ring responsiveness to adenosine, each ring was recontracted with PGF2alpha (30 µM) and, after attaining a steady level of developed tension, an adenosine transport inhibitor [dipyridamole (DIP, 0.1 µM) or S-(4-nitrobenzyl)-6-thioinosine (NBTI, 0.3 µM)], an adenosine deaminase inhibitor [erythro-9-(2-hydroxy-3-nonyl)-adenine hydrochloride (EHNA, 0.1 µM)], or vehicle was added to the bath. The concentration of the inhibitors used for these studies was previously determined to be that which elicited a minimal relaxation of coronary rings in the absence of adenosine. The concentration-dependent response of each ring to adenosine was reassessed 10 min after exposure to an inhibitor.

In select experiments, rings were exposed to DIP (0.1 µM) or NBTI (0.3 µM) without the addition of adenosine for the time required to complete a second adenosine dose-response curve (done on adjacent rings) to assess the effect of long-term transport inhibition. In addition, vehicle control experiments were conducted for both DIP (using an ethanol vehicle) and NBTI (using a DMSO vehicle), and the effect of the vehicle on the second dose-response curve was evaluated. The specificity of the transport inhibitors for adenosine was evaluated by determining the concentration-dependent relaxation responses of coronary rings to the nonselective adenosine receptor agonist 2-CAD after exposure to an adenosine transport inhibitor.

Coronary branches. First- and second-order branches (0.7-1.5 mm OD) of the three major conduit arteries were dissected and denuded of endothelium by rotating them over the sharp edge of a scissor. This procedure reduced the bradykinin-induced relaxation to 10 ± 2% (SD) of an endothelin-1 (ET-1) contraction, whereas intact rings relaxed 89 ± 17% (31). Rings were mounted on a force transducer (Grass) with an attached microdrive and underwent a series of four stretches (1.33 × resting length) to achieve a stable resting length. This was followed by incubation at 37°C in a physiological bicarbonate-buffered solution (PBBS), gassed with 97% O2-3% CO2. The rings were tested with high-K+ solution (K+ = 80 mM, substituted for Na) to establish a reference contraction. Only one concentration-response curve was done on each ring, and siliconized cups were used to hold solutions that contained ET-1.

After a 2-h equilibration period, ET-1 (3 nM, ~EC80) was added to the bath. When a steady contraction developed, either a vehicle (DMSO, 0.1% final concentration) or the selective adenosine transport inhibitor (NBTI, 1 µM) was added 10 min before cumulative additions of adenosine, followed by sodium nitroprusside (10 µM). NBTI at this concentration induced a small relaxation (14 ± 6%, n = 7, P = 0.05). The maximum relaxation to adenosine was 96 ± 1% (n = 7) of that produced by sodium nitroprusside. The individual responses to adenosine were normalized in terms of the maximal response to adenosine.

Adenosine Transport Measures

Adenosine transport was assessed by measuring [3H]adenosine accumulation of 1) isolated smooth muscle cells from conduit vessels and 2) intact endothelium-denuded coronary artery branches. Branches used for radioisotope measurements were cut open, and the endothelium was removed by rubbing the intimal side in several directions with moistened filter paper. The strips were then mounted on wire holders to facilitate transfers.

Coronary branches. Initial experiments determined the time course for washout of radioactivity from vascular strips incubated in PBBS plus [3H]adenosine (5 µCi/ml) and nonlabeled adenosine (10 µM) for 30 min. Tissues were rinsed for 1-2 s and then passed through a series of vigorously gassed tubes containing 5 ml of nonradioactive PBBS. This procedure resulted in very rapid (<1 s) dilution of the isotope as it leaves the tissue to ensure unidirectional movement. The residual activity in the tissues was extracted with 3% vol/vol perchloric acid. The tissue and washout tube counts (liquid scintillation counting, Packard Instruments) were added and normalized as a percentage of total counts (see Ref. 29 for details of analyses). The washout of [14C]sucrose (2 µCi/ml) and unlabeled sucrose (1 mM) were measured similarly. The adenosine transport inhibitor (NBTI, 10 µM) and the adenosine deaminase inhibitor (EHNA, 1 µM) were added to the appropriate isotope-loading solutions.

The uptake of [3H]adenosine was measured on strips that were equilibrated in PBBS (1 h) and, in selected cases, exposed to EHNA (1 µM) or EHNA plus NBTI (1 µM) for 10 min before transfer to the corresponding [3H]adenosine-containing PBBS. After the designated times, the tissues were rinsed 10 min in three large volumes of cold PBBS (1°C) to remove extracellular activity, blotted, weighed, and prepared for liquid scintillation counting. Standards were prepared from aliquots of [3H]adenosine PBBS to convert counts to picomoles. Uptake was expressed as picomoles per minute per milligram wet weight.

Sucrose is regarded to distribute mainly in the extracellular water of arteries, with little taken up by the cells or bound to other substances (29). Sucrose distribution was determined by incubating tissues in [14C]sucrose (1 µCi/ml) plus unlabeled sucrose (1 mM) for 15 min, followed by blotting and weighing. Radioactivity was extracted and counted with measured aliquots of the loading solutions to convert [14C]sucrose counts to milligrams solution. The sucrose distribution was represented as a percent wet weight. Total water content was measured on paired tissues that were blotted, weighed, dried for 24 h at 105°C, and reweighed. The loss in weight was taken as total water and presented as the percent wet weight. The differences between total water and sucrose distribution provide a reasonable estimate of cell water (29).

Isolated smooth muscle cells. Measures of smooth muscle adenosine uptake also were made using freshly dissociated cells from the LAD, LCCA, and RCA arteries and pooled. These myocytes were isolated according to the procedure of Stehno-Bittel et al. (51) with modifications. Briefly, vessels were removed from the heart as described above, placed in ice-cold low-Ca2+ medium (see Solutions and Drugs) containing 20 mM HEPES buffered to pH 7.4, and cleaned of myocardium, fat, and connective tissue. Segments of vessels (10-15 mm long) were opened longitudinally and pinned adventitial side down in a 1.5-oz vial containing a Sylgard rubber base. The endothelium was removed by gently abrading the luminal surface with moist filter paper. Each vascular segment was then incubated in low-Ca2+ medium containing BSA (0.2% wt/vol; Fraction V, Sigma), collagenase (294 U/ml, Worthington), elastase (1 U/ml, Worthington), and DNase (0.4 mg/ml, Sigma) for 30 min in a shaking water bath (37°C).

Residual endothelial cells dissociated from the vessel during this first 30-min enzyme incubation period, and the solution was discarded. The enzyme solution was then replaced with fresh enzyme, and the tissue was allowed to further dissociate for 45-60 min. The presence of isolated coronary vascular smooth muscle cells in the solution was verified by microscopic observation, the solution with cells was centrifuged at low speed (15 g), and the pellet was resuspended in fresh low-Ca2+ medium without enzyme. This process was repeated two to three times until the maximal yield of smooth muscle cells was obtained and before digestion of the adventitia, thereby minimizing fibroblasts in the cellular preparation. All smooth muscle fractions were then pooled and centrifuged, and the low-Ca2+ medium was replaced with HEPES-buffered Krebs solution (~2 ml) containing normal Ca2+ (2.0 mM).

Viability of cells in these preparations was assessed by trypan blue (Sigma) exclusion, and 95 ± 3% of all the cells excluded trypan blue. An aliquot (200 µl) of cells from each preparation was mixed with 5% BSA and adhered to clean glass slides by centrifugation in a Wescor Cytopro 7620 Cytofuge. Slides were then stained with Wright-Giemsa stain in a Hema-tec 100 autostainer and examined by conventional bright-field microscopy at ×100. Cells were identified by morphological characterization, and smooth muscle cells represented 96 ± 3% of all cells in a given field. Fibroblast contamination was minimized in these preparations by pinning vessel segments adventitial side down (~20-25 pins per 2-cm length of a vessel) and terminating the enzymatic processing before the digestion of the adventitia. Endothelial cell contamination was minimized by removing the endothelium mechanically and by discarding the first solution after enzyme addition. On the basis of morphological characteristics, endothelial cells are the primary cellular contaminant.

To assess adenosine uptake, smooth muscle cells were divided into 500-µl aliquots and placed in tubes containing either DIP (50 µM), EHNA, (10 µM), DIP plus EHNA, or vehicle. The concentrations of the inhibitors were chosen to provide maximal inhibition of uptake. In a subset of preparations, uptake was measured in the presence of EHNA and NBTI (0.1 µM). Adenosine uptake was initiated by addition of 10 µM adenosine containing 150,000 counts/min [3H]adenosine (NEN) as a tracer. Tubes were placed in a 37°C water bath, and adenosine uptake was terminated 1-160 min later by addition of 5 ml of a stop solution (room temperature) containing 10 µM adenosine, 50 µM DIP, and 10 µM EHNA.

Adenosine uptake kinetics (Vmax and Km) were determined by exposing smooth muscle cells to varying concentrations of adenosine (0.01-100 µM) spiked with 300,000 counts per minute [3H]adenosine in the presence of EHNA (10 µM) for 10 min; 10 min was chosen for these experiments because uptake is linear. After the stop solution was added, we centrifuged the samples at low speed to pellet the cells, removed the supernatant, and rewashed the cells with 5 ml of the stop solution. After centrifugation and supernatant removal, we extracted the cell pellet with 800 µl of ice-cold trichloroacetic acid. The extract was centrifuged (11,000 g at 4°C), an aliquot of the supernatant was counted by scintillation spectroscopy, and the protein pellet was quantified by the Pierce bicinchoninic acid (BCA) protein assay using BSA as the standard. Data are reported as picomoles adenosine per milligram cell protein.

Solutions and Drugs

Krebs buffer solution contained (in mM) 131.5 NaCl, 5.0 KCl, 1.2 NaH2PO4, 1.2 MgCl2, 2.5 CaCl2, 11.2 glucose, 13.5 NaHCO3, 0.003 propranolol, and 0.025 EDTA. Propranolol is added to block beta 2-adrenergic receptor-mediated vasorelaxation. Other experiments in the laboratory assess norepinephrine responsiveness and, for consistency across all experimental protocols, propranolol is added to Krebs buffer routinely. The average adenosine relaxation response in the current study in the presence of propranolol (EC50, 8.2 ± 1.6 µM) was not different from responses in the absence of propranolol (EC50, 6.7 ± 1.3 µM) (23) or from previously reported responses (40) with the same porcine model in the presence of propranolol. After equilibration with 95% O2-5% CO2, the pH of the solution was 7.4. The normal PBBS used for coronary branches had the following composition (in mM): 138 NaCl, 5 KCl, 1.5 CaCl2, 1.2 MgCl2, 1.2 NaH2PO4, 13.5 NaHCO3, and 11.2 glucose. The solution was gassed with 97% O2-3% CO2 to yield a pH of 7.4. Low-Ca2+ medium was prepared from Joklik's modification of MEM (Life Science Technology) with 0.5 mM CaCl2 added.

The adenosine solutions used to generate concentration-response curves were prepared as a stock solution made in distilled H2O. 2-CAD, NECA, CGS, and NBTI (all from Research Biochemical International) solutions were prepared in buffer diluted from stock solutions made in DMSO (100%). DIP (Research Biochemical International) was prepared as a stock solution in 95% ethanol. A concentrated stock solution of EHNA was made in dilute HCl (0.001 N) and diluted in distilled H2O before use. ET-1 was purchased from Peninsula Laboratory and diluted to 100 µM in 10% (vol/vol) acetic acid. All stock solutions except DIP were stored frozen at -20°C and freshly diluted the day of the experiment. DIP is light sensitive and readily breaks down in solution. Therefore, DIP was made fresh the day of the experiment. PGF2alpha was purchased from Eli Lilly (Lutylase) as a stock solution and stored refrigerated. Radioisotopes were purchased from DuPont-New England Nuclear and were stored at 4°C. All other chemicals and drugs were obtained from Sigma Chemical (St. Louis, MO).

Data Analysis

Before statistical analysis, data from identically treated rings from one animal were averaged and considered representative of that animal; thus each animal counted as one observation. Previous studies determined that rings from the LAD artery and LCCA responded similarly to the vasoconstrictor PGF2alpha , and no differences in vasodilatory responses of rings from these vessel sources were determined in the current study; therefore, no distinctions are made based on the source of the coronary ring. The responses to constrictors are expressed in grams of developed tension, determined as the tension developed above the resting tension due to stretch of the coronary ring. The responses to vasorelaxing agents are expressed as relative relaxation from precontracted levels. For calculation of EC50, the adenosine concentration-relaxation responses for each ring were fitted by regression analysis to a four-parameter sigmoidal equation using the SigmaPlot equations library and software (SPSS). IC50 is normally distributed according to a log rather than arithmetic scale; therefore, log values were used to make statistical comparisons. Statistical difference in the log of EC50 values was determined using a Student's t-test. For clarity, EC50 values are graphed arithmetically. The effect of an inhibitor on adenosine-mediated relaxation was assessed using a two-way repeated measures ANOVA, and significant differences between multiple pairs determined using a least squares means post hoc test. All functional data are presented as means ± SE. Significance was considered at P < 0.05.

Transport kinetics were determined using regression of a linearize form (Hanes-Woolf) of the Michaelis-Menten equation to derive the kinetic parameters for the Km and the Vmax. Log-linear (percent counts vs. time) regression between 10 and 40 min was used to determine the rate constant and initial percent counts for the radioisotope washout experiments of the intact vessels. The parameters were derived for each strip (using SigmaPlot software), and the arithmetic mean and means ± SE are presented. Data from isolated myocytes were obtained by pooling cells from all conduit vessels from two animals for each kinetic curve (each preparation, n = 1). Each time-course curve for adenosine uptake into isolated cells utilized coronaries obtained from five animals.


    RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Effect of Adenosine Transport Inhibitors on Adenosine-Mediated Relaxations

Cumulative increases in adenosine concentration in the tissue bath elicited a dose-dependent relaxation of PGF2alpha -precontracted large conduit coronary artery rings (Fig. 1). The responsiveness of these rings to adenosine exhibited an EC50 of 8.0 ± 0.9 µM, which was one to three orders of magnitude less than the responsiveness to the synthetic adenosine receptor agonists 2-CAD, NECA, or CGS-21680 (Fig. 1, inset). In fact, adenosine EC50 values for many conduit rings represent an apparent EC50, because it was not always possible to reach maximal relaxation. Maximal relaxation was always attained with the synthetic adenosine receptor agonists.


View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1.   Relaxation of porcine conduit coronary artery rings to adenosine and adenosine receptor agonists. Coronary rings from the left anterior descending (LAD) and left circumflex coronary arteries (LCCA) were precontracted with 30 µM PGF2alpha and then exposed to cumulative increases in adenosine (ADO), 2-chloroadenosine (2-CAD), 5'-(N-ethylcarboxamido)-adenosine (NECA), or CGS-21680 (CGS). Inset, EC50 values for ADO and agonists calculated using a linear interpolation software program. Data represent means ± SE for n = 9 animals.

To evaluate the role of adenosine transport in determining the relative sensitivity of coronary arteries to adenosine, we examined relaxation responses to adenosine in the same ring before and after exposure to the adenosine transport inhibitors DIP or NBTI. Relaxations induced by adenosine were significantly increased in the presence of either DIP (Fig. 2A, 0.1 µM) or NBTI (Fig. 2B, 0.3 µM). EC50 values were 6.9 ± 1.4 in controls and 1.3 ± 0.4 µM after exposure to DIP, representing a 5.4-fold increase in sensitivity. NBTI similarly increased sensitivity to adenosine from 11 ± 2.1 µM in controls to 0.9 ± 0.2 µM, a 12.6-fold increase in sensitivity (Fig. 2, A and B, inset). In the presence of the vehicle alone, a second adenosine dose-response curve was superimposed on the initial dose-response curve (data not shown). Similarly, repeated constrictions with PGF2alpha attained developed tension within 0.9 ± 3.7% of the initial PGF2alpha -developed tension for the first repeat and 9.2 ± 6.7% of the initial developed tension for the second repeat. A repeated measures ANOVA did not indicate significant differences.


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 2.   ADO-mediated relaxation of coronary conduit rings in the presence or absence of the ADO transport inhibitor dipyridamole (A, 0.1 µM) or S-(4-nitrobenzyl)-6-thioinosine (NBTI) (B, 0.3 µM). Insets, EC50 values. Data represent means ± SE for n = 9 animals (A) or 8 animals (B). Each animal represents averaged data from 4 to 6 rings per animal. *Statistical significance at P <=  0.05.

In the absence of added adenosine, DIP and NBTI alone at the concentrations used caused a small statistically insignificant basal relaxation, which was variable from ring to ring. DIP-mediated relaxations (with DIP alone) never exceeded 20% relaxation by the conclusion of the experimental protocol (~1.5 h), whereas NBTI never exceeded 10% relaxation in the absence of added adenosine.

Specificity of adenosine transport inhibition. To assess whether the enhanced relaxation observed in the presence of DIP or NBTI was specific to adenosine, we examined the effect of transport inhibitors on relaxations induced by the synthetic adenosine agonist 2-CAD, which is structurally similar to adenosine. For these experiments, coronary rings from the same animal were exposed to increasing concentrations of either adenosine or 2-CAD, followed by a second dose-response curve for the same agonist but in the presence of transport inhibitor. As shown in Fig. 3, NBTI did not alter 2-CAD-relaxation responses (EC50: 2-CAD 0.5 ± 0.2 µM, 2-CAD + NBTI 0.5 ± 0.3 µM). Similar results were obtained when responses to NECA were tested in the presence of DIP or NBTI (data not shown). Thus the ability of adenosine transport inhibitors to enhance relaxation is unique to adenosine.


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3.   Relaxation of coronary conduit rings induced by ADO (A) or the ADO agonist 2-CAD (B) in the presence or absence of the ADO transport inhibitor NBTI (0.3 µM). Insets, EC50 values. Data represent means ± SE for n = 4 animals. *Significant differences at P <=  0.05.

Role of endothelium in adenosine uptake. In the absence of the endothelium, DIP and NBTI shifted the adenosine dose-response relationship to the left. DIP significantly decreased the EC50 value from 9.4 ± 4.5 µM in control to 3.2 ± 2.3 µM, a 2.9-fold increase in sensitivity (P < 0.05). NBTI similarly decreased the EC50 from 11 ± 2 µM in control to 1.3 ± 0.3 µM, an 8.5-fold increase in sensitivity (Fig. 4A). Thus, in the presence (Fig. 2) or absence of the endothelium (Fig. 4A), neither basal adenosine sensitivity nor the enhancement by adenosine transport inhibitors differed significantly.


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 4.   ADO-mediated relaxation of endothelium-denuded conduit large (A) and small (B) coronary rings in the presence or absence of the ADO transport inhibitor NBTI (A, 0.3 µM; B, 1.0 µM). Large arteries were precontracted with PGF2alpha (30 µM); small coronary branches were precontracted with endothelin-1 (3 nM). Data represent means ± SE and n = 3 animals (A) and 7 animals (B). Insets, EC50 values. *Significant differences at P <=  0.05.

Large conduit coronary arteries are notoriously less sensitive to the vasodilatory effects of adenosine than are the smaller first- or second-order coronary branches. Therefore, we tested whether transport inhibitors would further increase the sensitivity of small coronary arteries to adenosine. Endothelium-denuded small coronary branches in the current study exhibited an EC50 for adenosine of 0.3 ± 0.1 µM (Fig. 4B). The transport inhibitor NBTI (1 µM) shifted the concentration-response relationship to the left by 0.53 log units to 0.09 ± 0.01 µM, representing a 3.4-fold increase in sensitivity. No significant changes occurred in the parameters associated with the maximum or minimum responses, nor were there significant changes in the slope of the response at the EC50.

Adenosine uptake by coronary smooth muscle. To determine whether coronary smooth muscle possess an adenosine transport mechanism, we measured [3H]adenosine uptake for endothelium-denuded small coronary branches and enzymatically dissociated conduit smooth muscle cells. In the absence of a transport inhibitor, freshly dissociated smooth muscle cells accumulated adenosine (10 µM outside) to a low level that reached steady state by 20 min (Fig. 5A). In contrast, in the presence of EHNA (10 µM), time-dependent uptake was significantly increased, and accumulation was near linear over the first 40 min and continued to increase for 160 min. In the presence of EHNA, adenosine uptake exhibited typical saturation kinetics from 0.1 to 100 µM adenosine (Fig. 5B). A Hanes-Woolf linear transformation of these data was used to determine the uptake kinetic parameters Vmax and Km (Fig. 5B, inset). These measures indicate that the uptake process exhibits a Vmax of 66.7 ± 7.0 pmol adenosine · mg protein-1 · min-1 and a Km of 10.5 ± 5.8 µM (means ± SE, n = 6 preparations with each preparation represented by cells from two animals).


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 5.   [3H]ADO accumulation by freshly dispersed coronary smooth muscle cells. A: time dependence of ADO transport. Freshly dispersed smooth muscle from the LAD artery, LCCA, and right coronary artery of 4 animals were pooled and incubated for 1-160 min in [3H]ADO in the presence () or absence () of the ADO transport inhibitor dipyridamole (50 µmol/l) and/or the adenosine deaminase inhibitor erythro-9-(2-hydroxy-3-nonyl)-adenine hydrochloride (EHNA) (triangle , 1 µmol/l). Data represent means ± SD for n = 2 experiments with 4 animals supplying myocytes for each experiment. Each experiment represents data from duplicate or triplicate replicas. B: saturation-kinetic curve showing concentration dependence for ADO transport into coronary smooth muscle. Data represent means ± SE of 5 experiments with myocytes from 2 pigs in each experimental preparation. Inset, Hanes-Woolf linear transformation of data (B) for calculation of maximum uptake (Vmax) and Michaelis constant (Km). r2, Correlation coefficients.

Adenosine uptake into smooth muscle cells was sensitive to inhibition by submicromolar concentrations of NBTI. In a subset of samples, NBTI (0.1 µM) decreased smooth muscle cell accumulation of adenosine from 27 ± 6 to 2.4 ± 0.6 pmol · mg-1 · min-1. Maximal concentration of DIP (50 µM) added to aliquots of cells from the same preparations reduced adenosine uptake to 1.2 ± 0.5 pmol · mg-1 · min-1 (n = 3 preparations). EHNA (10 µM) was present in all samples.

2-Chloroadenosine is structurally similar to adenosine and has been reported to both block transport and be transported in other cell systems (4, 44). Smooth muscle cells from the same preparation were exposed to either 10 µM adenosine (spiked with [3H]adenosine) or 10 µM 2-CAD (spiked with 2-chloro-[3H]adenosine), and uptake was measured in the presence and absence of NBTI after 10 min incubation. Under these conditions, smooth muscle accumulated 32.3 ± 8.9 pmol · mg-1 · min-1 adenosine but only 3.3 ± 2.6 pmol · mg-1 · min-1 of 2-CAD (n = 2 preparations), which was not blocked by NBTI. These data support functional data demonstrating the lack of effect of transport inhibitors on relaxation responses mediated by 2-CAD.

Adenosine sensitivity of small coronary arteries was 10-fold greater than that of large conduits, and the enhanced sensitivity caused by transport inhibition, although significant, was less in small arteries (Fig. 4). To test whether the adenosine transport process differed in small arteries, we measured [3H]adenosine uptake into intact endothelium-denuded small arteries. Because arteries are a multicellular tissue with a relatively large extracellular space, to measure adenosine uptake required that we determine the kinetics for adenosine movements through this extracellular barrier. Cellular components are traditionally identified as being relatively slowly exchanging, temperature sensitive, and subject to selective transport inhibitors (29). However, the extracellular component is identified by comparing the kinetics to that of a similar-sized molecule that distributes mainly (>95%) in the extracellular space, e.g., sucrose. As shown in Fig. 6 and Table 1, adenosine and sucrose exhibited both fast and slow components; however, the magnitudes differed greatly. Over 80% of the adenosine was characterized by a slow rate of washout (<0.3% per minute), whereas less than 2% of sucrose exhibited this behavior. The computed diffusional coefficients for the fast exchange (total exchange minus slow exchange) indicated that sucrose moved at about one-half the rate for free diffusion, with adenosine being somewhat slower (28). Lowering the temperature reduced the rate constant for the slow adenosine component significantly (0.0026 ± 0.0002 vs. 0.0007 ± 0.0002 min-1, P < 0.005), whereas temperature reduction had relatively little effect on the fast component. The fast components are consistent with a relatively large and open extracellular space as reported for other arteries (29). This was confirmed in pig coronary arteries by measurements of water content and sucrose space. The total water was 79.7 ± 0.4% wet wt (n = 13) and the sucrose space 59.0 ± 1.3% wet wt, yielding a computed cell water of 20.7 ± 1.3% wet wt (n = 13).


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 6.   Time course for washout of [3H]ADO and [14C]sucrose from intact porcine small coronary branches. One group of ADO-loaded vessels was treated during radioisotope loading with EHNA (1 µM), and another group was treated with EHNA + NBTI (10 µM). Means ± SE of % total counts are plotted for washout at 37°C for 4 pigs.


                              
View this table:
[in this window]
[in a new window]
  		Table 1.   Parameters derived for the washout of [3H]adenosine and [14C]sucrose from porcine coronary arteries

Arteries were exposed to the adenosine transport inhibitor NBTI to test whether the size and rate of the slow component was regulated by a membrane transport mechanism. As shown in Fig. 6 and in Table 2, pretreatment with NBTI significantly reduced the size and rate associated with the slow component, which followed a single exponential between 10 and 40 min washout. The adenosine deaminase inhibitor EHNA slightly increased the percentage of slowly exchanging adenosine without effecting the rate constant. The kinetic analysis indicates that fast and slow components for adenosine can be separated. Subsequent protocols employed a 10-min rinse in cold PBBS (1°C) to remove greater than 98% of the extracellular adenosine while retaining greater than 95% of the cellular content.

                              
View this table:
[in this window]
[in a new window]
  		Table 2.   Effects of EHNA and NBTI on parameters for [3H]adenosine washout from porcine coronary arteries

Adenosine (10 µM) accumulation by endothelium-denuded coronary branches was linear for both controls and EHNA-treated arteries over the first 10 min, whereas linear accumulation continued in NBTI-treated arteries for 30 min (Fig. 7A). Controls and EHNA-treated tissues exhibited similar accumulation, whereas the addition of NBTI reduced the accumulation by 90%. This provides evidence for an NBTI-sensitive component. In subsequent experiments, the uptake of adenosine was measured over a 10-min period. On the basis of the washout analysis, less than 3% of the isotope accumulated over a 10-min period left the tissue; thus backflux corrections were not necessary. The half-time for extracellular diffusion (20 s, Table 1) was also short compared with 10 min; thus diffusional delays could be neglected. Uptake was computed by dividing the adenosine accumulated at 10 min by 10 and the tissue wet weight, yielding values of picomoles per minute per milligram wet weight.


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 7.   [3H]ADO accumulation by endothelium-denuded small coronary branches. A: time course for [3H]ADO accumulation by coronary branches. Means ± SE for 3 pigs are plotted. ADO concentration was 10 µM. One group (triangle ) was treated 10 min before and during radioisotope exposure with EHNA (1 µM), and another group () with EHNA + NBTI (10 µM). B: ADO concentration dependence of [3H]ADO uptake into coronary branches. Means ± SE for 6 pigs are plotted. Data represent the calculated difference (NBTI sensitive) between controls and NBTI-treated arteries in the presence (triangle ) or absence () of EHNA (1 µM) 10 min before and during the 10-min radioisotope exposure. *Significantly different from control at P < 0.05. C: Hanes-Woolf linear transformation of data (B), and parameters for the Michaelis-Menten kinetic model applied to NBTI-sensitive [3H]ADO uptake in coronary branches. open circle , Presence of 1 µM EHNA.

The NBTI sensitivity of adenosine uptake was determined by exposing each artery to a selected concentration of NBTI 10 min before and during exposure to adenosine (10 µM). The solution also contained EHNA to inhibit deamination of adenosine. Under these conditions, the threshold for inhibition of influx by NBTI was 10 nM, and >80% of the influx was NBTI sensitive with an IC50 of 42 ± 3 nM. Blockade appeared to be maximal at 1 µM, which was used subsequently to study the dependence of uptake on extracellular adenosine concentration.

Uptake in the presence of NBTI was linearly dependent on adenosine concentration between 1 and 100 µM (data not shown). In contrast, uptake in controls exhibited a relatively steep dependence on adenosine between 1 and 10 µM, followed by a more gradual rise between 30 and 100 µM. Subtracting the NBTI-insensitive (linear) component from controls yielded an NBTI-sensitive component, which exhibited saturation kinetics (Fig. 7B). The NBTI-sensitive component determined in the presence of EHNA exhibited increased uptake at adenosine between 1 and 5 µM, whereas uptake was decreased between 30 and 100 µM compared with controls (Fig. 7B). As shown in Fig. 7C, these data were fitted using the Hanes-Woolf linear transformation to derive the parameters Vmax and Km. The Vmax was statistically greater in controls (5.1 ± 0.5 pmol · min-1 · mg wet wt-1) than in the presence of EHNA (3.0 ± 0.3 pmol · min-1 · mg wet wt-1, n = 6). Importantly, EHNA reduced the Km almost threefold from 17.6 ± 2.6 to 5.9 ± 1.7 µM (P < 0.01). Parallel experiments conducted on vessels with an intact endothelium indicated that the endothelium did not alter either the Vmax or Km for adenosine uptake. The correlation coefficients (r2) (0.99 and 0.95 for control or EHNA, respectively) indicate that the kinetic model provides an acceptable fit of the data.

Effect of deaminase inhibition on adenosine-mediated relaxations. The adenosine deaminase inhibitor EHNA significantly increased the adenosine uptake of the isolated smooth muscle cells and increased the Km while reducing the Vmax for adenosine uptake in coronary branches. Therefore, we evaluated the effect of EHNA on coronary ring function. Conduit coronary rings from the same hearts were subjected to an adenosine dose-response curve followed by a second dose-response curve in the presence of either EHNA (0.1 µM), NBTI (0.3 µM), or both EHNA and NBTI. EHNA at a concentration that did not directly cause vasorelaxation elicited no significant shift of the adenosine vasorelaxation responses (Fig. 8). EHNA caused a small, not statistically significant decrease in the EC50 from 6.9 ± 1.9 µM for controls to 4.3 ± 0.8 µM, a 1.6-fold increase in sensitivity. In the presence of NBTI, EHNA treatment had no further enhancement of relaxation than did NBTI alone (Fig. 8, A and C). Although the EC50 in the presence of both NBTI and EHNA increased 18.6-fold compared with a 10.9-fold increase in the presence of NBTI alone, these differences did not reach statistical significance. Increasing the concentrations of EHNA alone caused a dose-dependent relaxation of coronary rings. The EC50 for relaxation was ~100 µM (n = 7 animals) in the presence or absence of a threshold dose of adenosine (0.9 µM). Dose-dependent relaxations to EHNA were shifted slightly left and parallel by this concentration of adenosine, which alone caused a 10% relaxation. Similarly, EHNA had no significant effect on the adenosine sensitivity of the coronary branches. Nor did EHNA affect the NBTI-induced increase in sensitivity (2.2-fold increase, 0.35 ± 0.12 log units, n = 7, P < 0.05).


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 8.   Effect of adenosine deaminase inhibition on ADO-mediated relaxations. Conduit coronary rings from the same animals were exposed to either NBTI (A, 0.3 µM), EHNA (B, 0.1 µM), or both NBTI and EHNA (C) before generation of a second ADO dose-response curve (at least 2 rings from each animal were allocated to each group). Insets, EC50 values. *Significant differences at P <=  0.05; n = 7 animals.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Data presented in the current paper demonstrate, using both functional measures and biochemical transport kinetic analysis, that coronary vascular smooth muscle expresses a high-affinity adenosine transporter that is sensitive to inhibition by low concentrations of NBTI. The transporter characteristics are consistent with an es (equilibrative sensitive)-type transporter (52). Although considerable indirect evidence supported the existence of a transport process in smooth muscle, our data are the first to measure transport kinetics in both isolated cells and intact, endothelium-denuded vessels. Furthermore, neither adenosine transport kinetics nor adenosine-mediated relaxations appear to be influenced by the presence of the endothelium.

Adenosine Sensitivity

Conduit coronary arteries exhibit a relatively modest sensitivity for adenosine-mediated relaxation in vitro compared with responses to the more selective synthetic agonists such as NECA, 2-CAD, and CGS-21680. Because selective agonists are not readily metabolized, these data suggest adenosine metabolism and/or transport may contribute to the reduced responsiveness to adenosine of conduit arteries in vitro. Consistent with this interpretation, inhibition of adenosine transport by either DIP or NBTI shifted the dose-response curve to the left and decreased the EC50 for adenosine 10-fold in conduit arteries and 3.5-fold in first- and second-order coronary branches. Similar effects of the transport inhibitors were described for bovine coronary artery strips (32), hog carotid arteries (24), and rabbit aortas (37). Although this response was initially attributed to inhibition of smooth muscle adenosine uptake, subsequent data demonstrating a significant adenosine accumulation by endothelial cells confounded these early results (8, 33, 36, 38, 43, 48, 50). Indicator dilution studies in both coronary (36, 54) and skeletal circulations (18) argue strongly that the endothelium accumulates adenosine. Furthermore, identification of adenosine receptors on endothelial cells that are coupled to activation of nitric oxide synthase activation (34, 39) led to the proposal that significant adenosine vasorelaxation was endothelium dependent and associated with nitric oxide release (39).

In the current study, we did not observe a significant difference in either adenosine sensitivity or the enhancement by transport inhibitors between intact or endothelium-denuded vessels from either coronary conduit or branch arteries. Oltman and co-workers (40) also failed to observe changes in the adenosine responses after endothelium denudation of porcine conduit vessels. We also did not detect any difference in [3H]adenosine uptake into segments of coronary branches that had an intact endothelium compared with denuded vessels. These data indicate that, under these conditions, the endothelium plays little role in determining the adenosine responsiveness of coronary conduit rings and that endothelial cell uptake of adenosine is likely not responsible for the regulation of adenosine sensitivity of small or large arteries. Although adenosine-mediated relaxations of porcine coronary arteries were independent of the presence of the endothelium, these data should not be interpreted as evidence that adenosine is not accumulated by endothelial cells. Rather, the relatively smaller volume of endothelial cells compared with smooth muscle cells in conduit rings (30) likely shifts the emphasis for metabolic regulation more to the smooth muscle cell. Thus it is possible that the endothelial transport of adenosine plays a more significant role as artery size decreases. Alternatively, the endothelial cell phenotype may be different in the microcirculation (2, 35) or in conduit endothelial cells in culture. We conclude from these data that, in porcine coronary arteries from 0.7 to 2.0 µm OD, endothelial cell uptake of adenosine does not influence vasorelaxation in vitro.

Adenosine Transport

Biochemical evidence that coronary vascular smooth muscle accumulates significant quantities of adenosine was obtained from studies of [3H]adenosine uptake into endothelium-denuded vascular segments and freshly dissociated smooth muscle cells. Nucleoside transport is a fairly ubiquitous process and, thus far, seven different transporters have been cloned. These transporters generally fall into two categories: concentrative, sodium-dependent transporters and equilibrative, nonsodium-dependent transporters (52). The equilibrative transporters move nucleosides by facilitated diffusion down their concentration gradient and have been characterized for multiple cell types (52). We calculated the Km of this transporter to be 10.5 µM in freshly dissociated cells and 5.9 µM in coronary branches under similar conditions (EHNA). These values are comparable to the affinity for adenosine of es-type transporters reported for other cell types [MCF7 cells, 21.7 µM (see Ref. 17); cultured cerebral vascular smooth muscle cells, 10 µM (see Ref. 4); rat brain synaptoneurosomes, 9.4 µM, (see Ref. 20), and cardiac myocytes, 6.2 µM (see Ref. 48)].

The accumulation of adenosine by both isolated smooth muscle cells and small artery branches was inhibited by submicromolar concentrations of NBTI, consistent with an es-type NBTI-sensitive nucleoside transporter subtype. The equilibrative types of adenosine transporters are subclassified into those sensitive to NBTI [inhibitor constant (Ki) <100 nM] and those that are insensitive (Ki for NBTI >1 µM) (45)(52). The uptake of [3H]adenosine into coronary smooth muscle was maximally inhibited at submicromolar concentrations of NBTI, and coronary branches exhibited 80-90% inhibition at the Km. The NBTI-insensitive uptake was linearly dependent on adenosine concentration and did not exhibit properties of a facilitated transport process under the conditions studied. In contrast, the NBTI-sensitive uptake exhibited the characteristics of carrier-mediated transport, as noted above, and may result from the presence of the es-type transporter in porcine coronary branches as well as in conduit smooth muscle.

Maximal adenosine transport activity was 66.7 pmol · min-1 · mg protein-1 in coronary smooth muscle cells. The Vmax of coronary rings was 3.0 pmol · min-1 · mg wet wt-1 in the presence of EHNA and 5.1 pmol · min-1 · mg wet wt-1 in the absence of EHNA. On the basis of our calculations of cell water and known protein content (7% wet wt), these value calculate to a Vmax between 42 and 73 pmol · min-1 · mg cell protein-1, respectively. Thus maximal transport of isolated smooth muscle cells is similar to that of intact vessels and comparable to the transport values of other cell types (cardiac myocytes, 57.6 pmol · mg protein-1 · min-1) (48). All of these transport values are modest, however, relative to transport activity of rat endothelial cells (4.4 nmol · mg protein-1 · min-1) (48). These observations are relevant to the maintenance of coronary artery tone because the smooth muscle cells in these arteries are located between the myocardium and the endothelium. An active uptake mechanism could therefore reduce the adenosine concentration as it diffuses across the smooth muscle layers. Our radioisotope washout studies showed the half-time for diffusion to be ~20 s. This time is sufficient for the adenosine uptake mechanism to clear a significant amount of extracellular adenosine at physiologically relevant concentrations (e.g., 1 µM) (10). For instance, in vascular segments, 1 µM adenosine was associated with an uptake of 0.4 pmol · min-1 · mg wet wt-1 in the presence of EHNA (Fig. 7B). The equilibrium extracellular content of adenosine in the artery would be about 0.5 pmol/mg wet wt (1 µM = 1 pmol/mg solution; 1 pmol/mg solution × 0.5 mg sucrose space/mg wet wt). From this estimate, the smooth muscle uptake during 20 s would remove about 25% of the extracellular adenosine content. If the transporter was colocalized with the adenosine receptors, then the reduced concentration would be even greater in this microenvironment.

The interplay of extracellular diffusion and smooth muscle adenosine uptake could contribute to differences in the adenosine sensitivity of relaxation exhibited between the thicker coronary conduit arteries and the coronary branches. Under control conditions, the EC50 for conduits was ~11 µM compared with 0.3 µM for branches. The increase in sensitivity induced by maximal inhibition with NBTI was greater in the conduits (~12-fold) compared with the branches (~3-fold). Our kinetic data suggests a similar relative expression of transporter protein in these two sized vessels. Thus adenosine uptake by vascular smooth muscle could establish a concentration gradient within the muscle layer that effectively decreases adenosine availability at adenosine receptors in deeper muscle layers. Establishment of such a gradient could account for the reduced sensitivity of large coronary arteries to adenosine compared with coronary branches from the same animals.

Alternatively, our data do not exclude the possibility that pharmacological differences between large and small arteries exist relative to adenosine receptor subtypes, adenosine metabolism, or selective coupling between metabolism and receptor activation. Multiple reports exist documenting enhanced receptor-dependent relaxation in small versus large arteries (6), and small arteries are more sensitive than large arteries to synthetic adenosine agonists (22, 53), which are not metabolized or transported. In addition, there appear to be size-dependent differences in adenosine metabolism, with small arteries expressing less adenosine deaminase and more 5' nucleotidase than large arteries (19, 35). These differences would promote higher interstitial concentrations of adenosine in small arteries and enhance relaxation. Colocalization of the receptors and mechanisms for removal of neurotransmitters by either transport or degradation is well documented (25, 26). Similarly, adenosine deaminase colocalizes with adenosine A1-receptors and modulates G protein signaling through these receptors (7, 16). Thus, in addition to diffusion differences between small and large artery muscular walls, we cannot exclude the possibility that branches differ from conduits in either the number or subtype of adenosine receptors or the selective interaction of receptors with transporter or metabolic enzymes.

In isolated vascular smooth muscle cells, the maximal adenosine transport was 67 pmol · min-1 · mg cell protein-1 but only in the presence of EHNA. In the absence of EHNA, the maximal adenosine uptake was significantly reduced. Because adenosine transport is concentration dependent, enzymes that metabolize adenosine will alter both the rate and net direction of transport. For example, cytosolic adenosine deaminase reduces intracellular adenosine levels, thereby facilitating inward transport. Inhibition of adenosine deaminase thus reduces adenosine transport into many cells, including endothelial cells (33), and may be significant in the reduced Vmax of intact vessels (discussed below). In contrast, extracelluarly located adenosine deaminase (ecto-adenosine deaminase) metabolizes adenosine to inosine, which may or may not be transported into cells (33). Inhibition of ecto-adenosine deaminase effectively increases extracellular adenosine levels (3) and most likely accounts for our observation of increased [3H]adenosine accumulation by isolated porcine coronary smooth muscle cells in the presence of EHNA. Adenosine deaminase was not inhibited in experiments using aortic smooth muscle (33, 36, 43) and may explain the lack of transport into smooth muscle in those preparations. Alternatively, aortic smooth muscle may not express adenosine transporters with high selectivity.

Ecto-adenosine deaminase has been reported to exist as a tethered plasmalemmal enzyme (16) or a released, intracellular enzyme (3, 42). The inability of EHNA to alter the responsiveness of isolated conduit coronary rings to adenosine in the presence or absence of a transport inhibitor suggests extracellular adenosine deaminase activity is low in intact vessels and/or present only in enzymatically dissociated cells. We cannot eliminate the possibility that: 1) the enzymatic dissociation procedure increases exposure or activates an ecto-adenosine deaminase in these cells, and/or 2) the 5% of cells that accumulate trypan blue continuously leak intracellular adenosine deaminase. Concentrations of EHNA above 1 µM relax coronary vascular rings in a concentration-dependent manner even in the absence of added adenosine, possibly due to nonspecific inhibition of phosphodiesterases (46) or an endogenous release of adenosine from smooth muscle (13, 14, 27).

Adenosine deaminase could also play a modulatory role in intact small coronary branches. We observed an increased uptake of adenosine at concentrations below Km and a reduced Vmax when the deaminase inhibitor EHNA was present. Extracellular deaminase activity has been measured in other smooth muscle preparations (3) and could be sufficient to reduce the availability of adenosine for transport. At high adenosine concentrations, this extracellular effect may be saturated, but the ability of the cells to convert intracellular adenosine to inosine and thus retain transported adenosine and maintain Vmax may be limited by the effect of EHNA. As in the case of the conduit coronary arteries, we did not observe any effects of EHNA on the relaxant sensitivity to adenosine. Possibly, the enzyme is more closely associated with the adenosine transporter than the receptors.

In summary, the data presented here indicate that the smooth muscle cells of porcine coronary arteries express an adenosine transporter with characteristics of an equilibrative, NBTI-sensitive subtype. The activity of this transporter is sufficient to modify the sensitivity to the relaxant effect of adenosine in vitro and, in theory, can significantly reduce the extracellular adenosine concentration within the artery. The uptake of adenosine was low (~1%) compared with published values for rates of ATP synthesis (41), and the washout of [3H]adenosine products was slow, contributing only a few percent during uptake measures. Thus intracellular metabolism of adenosine should not be rate limiting in these experiments, and kinetic values reflect transporter activity. Although the endothelium has been shown to take up and convert significant amounts of adenosine, this was a minor factor under our experimental conditions. We conclude that adenosine uptake into smooth muscle cells contributes significantly to the control of coronary tone and potentially to the regulation of smooth muscle biochemical synthesis and growth.


    ACKNOWLEDGEMENTS

We thank Drs. Rovetto, Huxley, Parker, and Hale for scientific input during the development of this study and for the advice on preparation of this manuscript. We also thank Dana Schmitz and Paula Du Tan for expert technical assistance.


    FOOTNOTES

This work was supported by funds from the National Heart, Lung, and Blood Institute Grant HL-52490.

Address for reprint requests and other correspondence: L. J. Rubin, Dept. of Veterinary Biomedical Sciences E102, Univ. of Missouri- Columbia, Columbia, MO 65211 (E-mail: RubinL{at}missouri.edu).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 30 August 1999; accepted in final form 27 March 2000.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

1.   Abebe, W, Hussain T, Olanrewaju H, and Mustafa SJ. Role of nitric oxide in adenosine receptor-mediated relaxation of porcine coronary artery. Am J Physiol Heart Circ Physiol 269: H1672-H1678, 1995[Abstract/Free Full Text].

2.   Andries, LJ, Brutsaert DL, and Sys SU. Nonuniformity of endothelial constitutive nitric oxide synthase distribution in cardiac endothelium. Circ Res 82: 195-203, 1998[Abstract/Free Full Text].

3.   Baer, HP, and Vriend R. Measurement of adenosine metabolism and uptake in smooth muscle and effects of adenosine transport inhibitors. J Pharmacol Exp Ther 229: 564-570, 1984[Abstract/Free Full Text].

4.   Beck, DW, Vinters HV, Moore SA, Hart MN, and Cancilla PA. Uptake of adenosine by cultured cerebral vascular smooth muscle cells. J Neurochem 41: 939-941, 1993[ISI][Medline].

5.   Berne, RM. The role of adenosine in the regulation of coronary blood flow. Circ Res 47: 807-813, 1980[Free Full Text].

6.   Bund, SJ, Oldham AA, and Heagerty AM. Influence of arterial diameter on vasomotor responses in the porcine coronary vasculature. Cardiovasc Res 28: 695-699, 1994[Abstract/Free Full Text].

7.   Ciruela, F, Saura C, Canela EI, Mallol J, Lluis C, and Franco R. Adenosine deaminase affects ligand-induced signaling by interacting with cell surface adenosine receptors. FEBS Lett 380: 219-223, 1996[ISI][Medline].

8.   Deussen, A, Bading B, Kelm M, and Schrader J. Formation and salvage of adenosine by macrovascular endothelial cells. Am J Physiol Heart Circ Physiol 264: H692-H700, 1993[Abstract/Free Full Text].

9.   Deussen, A, and Schrader J. Cardiac adenosine production is linked to myocardial PO2. J Mol Cell Cardiol 23: 495-504, 1991[ISI][Medline].

10.   Deussen, A, Stappert M, Schäfer S, and Kelm M. Quantification of extracellular and intracellular adenosine production. Circulation 99: 2041-2047, 1999[Abstract/Free Full Text].

11.   Dhalla, AK, Dodam J, Jones AW, and Rubin LJ. Identification of a high-affinity adenosine transporter in porcine coronary smooth muscle cells (Abstract). Biophys J 76: A184, 1999.

12.   Dubey, RK, Gillespie DG, and Jackson EK. Cyclic AMP adenosine pathway induces nitric oxide synthesis in aortic smooth muscle cells. Hypertension 31: 296-302, 1998[Abstract/Free Full Text].

13.   Dubey, RK, Gillespie DG, Mi ZC, and Jackson EK. Adenosine inhibits growth of human aortic smooth muscle cells via A2B receptors. Hypertension 31: 516-521, 1998[Abstract/Free Full Text].

14.   Dubey, RK, Gillespie DG, Shue H, and Jackson EK. A2B receptors mediate antimitogenesis in vascular smooth muscle cells. Hypertension 35: 267-272, 2000[Abstract/Free Full Text].

15.   Ford, DA, and Rovetto MJ. Rat cardiac myocyte adenosine transport and metabolism. Am J Physiol Heart Circ Physiol 252: H54-H63, 1987[Abstract/Free Full Text].

16.   Franco, R, Casadó V, Ciruela F, Saura C, Mal