INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Adenosine is an endogenous purine nucleoside intermediary in the pathway of purine nucleotide degradation and formed by the enzyme 5'-nucleotidase. Generally, if utilization of adenosine triphosphate (ATP) exceeds its production under hypoxic conditions, the accumulating adenosine monophosphate (AMP) is catabolized to a chain of purine nucleotide degradation products. Indeed, a number of studies have used catabolites of ATP such as hypoxanthine and uric acid as markers of tissue hypoxia (1, 2). Accordingly, the tissue or plasma level of adenosine, if correctly measured, may be used as a more sensitive marker of tissue hypoxia than the metabolites just mentioned, because adenosine is rapidly formed and catabolized prior to the formation of these metabolites. Adenosine is also a putative neuromodulator that plays various physiologic roles in ventilatory (3, 4) and cardiovascular (5, 6) control, many of which appear to be related to the balance between energy (or oxygen) supply and demand, particularly during hypoxia.

Although increased production of adenosine under ischemic conditions has been repeatedly demonstrated in a variety of tissues and organs, no conclusive data have demonstrated an increased adenosine concentration in tissue during moderate hypoxia in humans. In this study we aimed to examine whether the tissue concentration of adenosine is really increased in forearm and brain tissues during moderate hypoxia in humans. For this purpose, we measured plasma concentrations of adenosine during normoxia and isocapnic hypoxia (Sa_O2 = 80%, 20 min) in a vein and in an artery in the forearm tissue and the brain. Additionally, and more importantly, we repeated the experiment using pretreatment with dipyridamole, since it is thought that the measurement of adenosine in veins does not necessarily reflect the tissue level of adenosine, thus preventing estimation of the extracellular adenosine concentration simply from the arteriovenous difference (7). Dipyridamole is expected to block reuptake of adenosine, which is released into the peripheral circulation and then rapidly taken up by red blood cells and/or vascular endothelial cells (8, 9), and may therefore make the change in the tissue level of adenosine more visible in plasma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental Setup

Twelve healthy adult men (body weight 63 ± 5 kg, mean ± SD) served as subjects. All were medical students at the Hokkaido University School of Medicine who volunteered for the study and gave informed consent to it. The protocol of the study was approved by the Ethics Committee of the Hokkaido University School of Medicine. All experiments were done with the subjects in a supine position, distracted by music, and breathing spontaneously through a mouthpiece connected to a J valve. The subjects had refrained from consuming food or any beverages except water for at least 12 h before the experiment. A gas control system developed in our laboratory (10) was used to regulate arterial PO₂ and PCO₂ simultaneously and independently by changing the composition of inspired gas. Minute ventilation (E) was measured every 15 s by electrical integration of the flow signals obtained from a hot-wire respiratory flowmeter (Model RF-H; Minato Medical Products, Tokyo, Japan). Respiratory gases were continuously monitored by a mass spectrometer (Medical Gas Analyzer 1100A; Marquette, Milwaukee, WI). Sa_O2 was estimated with an oximeter applied to the finger (Biox 3740; Ohmeda, Louisville, CO). Before the study, drip infusion of physiologic saline was started via a 20-gauge catheter inserted in the median cubital vein, and was continued throughout the experiment for the intravenous injection and sampling of venous blood. A 22-gauge catheter was inserted in the radial artery of the other hand for the purpose of arterial blood sampling. In seven of the 12 subjects, venous blood from the brain was also sampled via a 20-gauge catheter placed in the internal jugular vein. Details of this technique were described in our previous publications (11, 12).

Protocol

All of the subjects were challenged with 6 min of isocapnic progressive hypoxia followed by 20 min of hypoxia. Sa_O2 was maintained at 80%. Physiologic saline (20 ml), used as a placebo, was infused intravenously over a period of 5 min just before the induction of progressive hypoxia. During the 20 min of hypoxia, saline (100 ml) was continuously drip-infused intravenously (2 ml/min). Blood samples for adenosine measurement were taken simultaneously from the radial artery and the median cubital vein during the control period; at 0, 10, and 20 min of hypoxia; and at 30 min in the recovery phase. In seven subjects, blood was also simultaneously taken from the internal jugular vein.

With an interval of at least 30 min for recovery, the same protocol was repeated with intravenous infusion of dipyridamole (0.6 mg/kg body weight). For this, 10 mg of dipyridamole was dissolved in 20 ml of saline and infused over a period of 5 min prior to the induction of progressive hypoxia. The remainder of the dipyridamole was dissolved in 100 ml of saline and was drip-infused intravenously (4 ml/ min) throughout the experiment except for the period of blood sampling. Blood samples were taken in a manner similar to that in the placebo run. Arterial and venous blood gases and pH were checked during normoxia and at 20 min of hypoxia in both the placebo and dipyridamole runs. Blood was analyzed soon after sampling (pH/ blood gas analyzer type 1303; Instrumentation Laboratory, Barcelona, Spain).

The three normal volunteers who did not participate in the study just described received dipyridamole alone in a manner similar to that in the dipyridamole run, but did not inhale any hypoxic gas. Sa_O2 was thus maintained at the baseline level of 97 to 98%. Blood was sampled from the median cubital vein before the infusion of dipyridamole and at 10, 20, 30, and 60 min after the infusion.

Measurement of Adenosine

We measured the plasma concentration of adenosine following the protocol of Zhang and colleagues (13) A blood sample (3.6 ml) was rapidly collected into a 5-ml syringe containing 0.4 ml of a stopping solution (13, 14) consisting of five drugs; 1 mM dilazep to inhibit adenosine uptake into and release from red blood cells, 10 µM erythro-9-(2-hydroxy-3-nonyl) adenine (EHNA) to block adenosine deaminase activity, 2 µg/ml indomethacin to inhibit nucleotide release from platelets, and 40 µM ethylenediamine tetraacetic acid (EDTA) and 40 µM O,O'-bis(2-aminoethyl)ethyleneglycol-N,N,N'-N'-tetraacetic acid (G-EDTA) to inhibit platelet aggregation and release of adenosine from platelets. EDTA is also thought to inhibit purine-5'- nucleotidase, which catabolizes AMP to adenosine (14). Samples were immediately put into microcentrifuge tubes and centrifuged at 14,000 × g for 1 min. Aliquots (1 ml) of plasma were then placed into separate tubes and deproteinated with 100 µl of 50% trichloroacetic acid. Samples were centrifuged again for 5 min, and 750 µl of clear supernatant was immediately neutralized with 100 µl of 3.3 N potassium hydroxide. The adenine nucleotides were extracted by adding 500 µl of 1 M zinc sulfate and 1,000 µl of saturated barium hydroxide, and were then subjected to vortex mixing for 10 s and centrifugation at 14,000 × g for 5 min. For high-pressure liquid chromatographic (HPLC) analysis, adenosine was finally converted to ethenoadenosine by mixing the sample with chloroacetaldehyde to a final concentration of 440 mM and incubating it at 80° C for 1 h. Derivatized samples were kept at 4° C until used for HPLC analysis. Ethenoadenosine levels were stable for at least 48 h (13).

HPLC-Fluorometric Analysis

HPLC-fluorometric analysis was conducted according to the method of Zhang and colleagues (13), but in our study the sample volume injected into the HPLC system was 100 µl. The HPLC-fluorometric system (Shimadzu, Osaka, Japan) consisted of an SIL-10A autosampler, an LC-9A pump with an automated gradient controller, an RF-535 fluorescence HPLC monitor, a C-R7Ae plotter-printer-integrator, and an SCL-10A system controller. Mobile phase A consisted of 88% 0.01 M KH₂PO₄ (pH 3.5) and 12% methanol (vol/vol), whereas mobile phase B contained 50% 0.01 M KH₂PO₄ (pH 3.5) and 50% methanol (vol/vol). Chromatography was conducted under conditions in which mobile phase A was pumped at the rate of 1.5 ml/min for 10 min, followed by a step gradient to 100% mobile phase B, which was maintained for an additional 16 min. After 1 min, the system reversed to the initial conditions and the column was allowed to reequilibrate for 10 min before a new sample was injected. Separation of adenosine was achieved on a Waters 3.9 mm × 30 cm microBondapak C18 10-µm column (Waters Associates, Milford, MA) with a retention time of 4.9 min, and the fluorescence activity was detected at an excitation wavelength of 280 nm and an emission wavelength of 380 nm. Identification and confirmation of the adenosine peak were done by the following three methods: retention time, enzymatic confirmation by adenosine deaminase, and coelution verification. The concentration of adenosine was calculated from the peak area, using the standard line. Pearson's correlation coefficient for the standard line of standard adenosine solution was more than 0.999 from 3.125 nM to 400 nM, and the recovery of plasma adenosine was 72.7 ± 1.5% (n = 16). Plasma adenosine concentrations were calculated by dividing the amount of adenosine in the samples by the volume of plasma assayed, which can be obtained by the following equation:

where Cp is the plasma concentration and Cs is the sample concentration of adenosine.

Materials

Adenosine and EHNA were purchased from Sigma Chemical (St. Louis, MO). Chloroacetaldehyde, indomethacin, adenosine deaminase, EDTA, G-EDTA, zinc sulfate, barium hydroxide, potassium hydroxide, potassium dihydrogenphosphate, and methanol were purchased from Wako Chemical (Tokyo, Japan). Dilazep was a generous gift of Kowa Co., Ltd. (Tokyo, Japan).

Statistical Analysis

Data are shown as means ± SEM unless otherwise specified. Comparison between the plasma concentration of adenosine in the placebo run and that in the dipyridamole run was done with the nonparametric Wilcoxon's signed rank test. Changes in plasma adenosine concentration over time were analyzed first with Friedman's test and then with Bonferroni's test for individual differences. A value of p < 0.05 was accepted as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The blood gas data and Sa_O2 values are shown in Table 1. There were no significant differences in the level of hypoxia between the placebo run and the dipyridamole run, although small but statistically significant differences between the two runs were found in Pa_CO2 and Pv_CO2 at 20 min of hypoxia.

Typical recordings of HPLC for blood samples from the median cubital vein in the dipyridamole run are shown in Figure 1. The adenosine peaks could be clearly identified in all blood samples. The mean group data for plasma adenosine are shown in Figure 2. The left, middle, and right panels show the data from the radial artery, the median cubital vein, and the internal jugular vein, respectively. Open circles represent the data in the placebo run and closed circles the data in the dipyridamole run. The plasma concentration of adenosine in normoxia was not significantly different in the radial artery (12.0 ± 0.9 nM), median cubital vein (17.7 ± 3.5 nM) and internal jugular vein (14.0 ± 2.8 nM). When subjects were not pretreated with dipyridamole, there were no significant changes in the plasma concentration of adenosine at any phase of hypoxia in either the artery or the vein. However, when subjects were pretreated with dipyridamole, the plasma concentration of adenosine increased markedly in the median cubital vein over that in the placebo run during 20 min of hypoxia. A time- dependent increase in plasma adenosine was evident during hypoxia (20.7 ± 2.5 nM, 25.0 ± 4.2 nM, 35.9 ± 7.7 nM, and 50.7 ± 10.7 nM at normoxia and at the beginning, 10, and 20 min of hypoxia, respectively) (p < 0.01 for overall changes and p < 0.05 for normoxia versus 20 min of hypoxia), again only in the median cubital vein, and an almost complete return to the normal range was noted in the recovery phase. By contrast, we found no significant changes in the plasma concentration of adenosine during hypoxia in either the radial artery or the internal jugular vein, even when subjects were pretreated with dipyridamole.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Typical HPLC recordings of adenosine in blood samples from the median cubital vein while the subject was pretreated with an intravenous infusion of dipyridamole. The peak of adenosine (arrows) was identified by the retention time, pretreatment with adenosine deaminase, and coelution verification. The peak area increased markedly at 20 min of hypoxia (middle panel ) and decreased at 30 min of the recovery phase.

View larger version (18K): [in this window] [in a new window]   Figure 2.   Plasma concentration of adenosine in the radial artery (left panel, n = 12), median cubital vein (middle panel, n = 12), and internal jugular vein (right panel, n = 7). The open circles are the data from the placebo run and the closed circles are the data from the dipyridamole run. Means ± SEM are shown. *p < 0.05 versus placebo.

In the three subjects to whom it was given, dipyridamole alone did not cause any time-dependent changes in the plasma concentration of adenosine in the median cubital vein during breathing of room air (18.9 ± 1.1 nM, 16.3 ± 3.1 nM, 16.8 ± 1.8 nM, 18.6 ± 0.7 nM, and 19.3 ± 1.4 nM at 0, 10, 20, 30, and 60 min, respectively, after the beginning of dipyridamole infusion).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we demonstrated that during moderate hypoxia (Sa_O2 = 80%), the plasma concentration of adenosine in the median cubital vein of human subjects was significantly and markedly increased over a period of 20 min only when subjects were pretreated with dipyridamole. This was not the case in the radial artery or in the jugular vein. Without dipyridamole pretreatment, there were no significant differences between the artery and vein or between normoxia and hypoxia in the plasma concentration of adenosine.

Our finding of a significant increase in the plasma concentration of adenosine in the vein only after pretreatment with dipyridamole can most likely be explained by inhibition of reuptake of adenosine by red blood cells or vascular endothelial cells. Adenosine that has accumulated in the extracellular space should enter the peripheral circulation and be taken up by red blood cells or vascular endothelial cells. The process of reuptake is so rapid (7, 8) that the extracellular adenosine concentration cannot be estimated simply from arteriovenous differences (7). Because dipyridamole is a potent inhibitor of this adenosine reuptake process, it is considered that pretreatment with dipyridamole makes the change in the tissue level of adenosine more visible in plasma (5, 7, 9). Accordingly, the findings in the present study indicate that the tissue production of adenosine is increased during moderate hypoxia in humans.

Our failure to find a significant increase in the plasma concentration of adenosine either in the artery or in the vein without dipyridamole pretreatment during moderate hypoxia is not in accord with prior reports that the arteriovenous difference in the concentration of adenosine is increased in human cardiac muscles (5, 6, 15) and forearm tissue (7) under ischemic conditions. This can most likely be explained by the difference in the tissue level of adenosine in ischemia and moderate hypoxia. During severe hypoxia, such as anoxia or ischemia, cell destruction may give rise to the release of intracellular ATP and its intermediates into the extracellular space. This causes an increase in the concentration of extracellular adenosine through the action of ecto-5' nucleotidase existing on the cell membrane. However, such cell damage does not occur in moderate hypoxia. The extracellular level of adenosine may also increase through two other mechanisms in hypoxia. First, hypoxia per se causes increased production of adenosine via breakdown of adenine nucleotides within cells according to the level of hypoxia. Although the ATP level is known to be maintained in moderate hypoxia, the adenine nucleotide pool is much larger than that of adenosine in cells (16), so that even a small change in the concentration of adenine nucleotides may cause a significant increase in adenosine. Second, enhanced release of adenosine from cells may occur in hypoxia without any changes in the intracellular production of adenosine. In ex vivo preparations of dog muscles, secretion of adenosine from cells was reported to be changed by pH, Pa_CO2, and lactate in arteries (17). However, this third possibility is unlikely to have been the case in the present study because there were no significant changes in either arterial pH or Pa_CO2 from normoxia to moderate hypoxia. In addition, we and others have previously shown that in humans (18, 19), the plasma level of lactate does not change with the level of hypoxia (Sa_O2 = 80%) used in the present study.

The absence of human studies in which adenosine was measured in moderate hypoxia before the present study can be partly explained by past technical ambiguity in the measurement of plasma adenosine. It used to be difficult to reliably measure the plasma concentration of adenosine because of its rapid metabolism (7, 8). Interlaboratory differences in plasma adenosine concentrations were as large as 35 to 779 nM for healthy humans until 1990 (7, 15, 20, 21). In most previous human studies, the plasma concentration of adenosine was measured with the HPLC-ultraviolet method or by radioimmunoassay, both of which are considered less sensitive than the HPLC-fluorometric method used in the present study (22). Moreover, the stopping solution used in earlier studies may not have completely inhibited adenosine metabolism in plasma. McCann and Katholi claimed that plasma adenosine could be accurately measured only with use of a proper stopping solution, and that the concentration would otherwise be easily overestimated (14). They concluded that the concentration of plasma adenosine in humans is less than 20 nM (10.2 ± 1.2 nM, n = 4). In the present study, we gave careful consideration to recent technical progress in the measurement of plasma adenosine.

The reasons we did not see any significant changes in the plasma concentration of adenosine in the radial artery or in the jugular vein even with dipyridamole pretreatment were not clear from this study. However, in the case of the artery, the blockade of adenosine uptake by dipyridamole might not have been complete, considering the short half-life of plasma adenosine. Additionally, if increased release of adenosine does not occur throughout the body, the local production of adenosine would not be reflected in the artery, owing to a dilution effect. In the case of the jugular vein, the level of hypoxia might not have been severe enough to have caused increased production of adenosine in brain tissue. The mean PO₂ in the jugular vein was, however, low in comparison with that of the median cubital vein. Another possibility is that the blood-brain barrier may mask the actual increase of adenosine in brain tissue.

The clinical implications of this study are twofold. First, since adenosine has a variety of physiologic roles, adenosine accumulated in tissue may work as a mediator for metaboreceptors (23) in the muscles and/or as a neuromodulator for cardiovascular (5, 6) and ventilatory control (3, 4) even in moderate hypoxia. Second, the finding of increased production of adenosine in tissue during moderate hypoxia may provide a basis for assessment of tissue hypoxia.

In conclusion, the present study provides evidence that tissue production of adenosine increases, at least in forearm muscle tissue, even during moderate hypoxia (Sa_O2 = 80%, 20 min) in humans. However, this increase is not reflected in the arterial or venous plasma concentration of adenosine unless the subject is pretreated with dipyridamole.