Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Renal insufficiency and failure are associated with hypoxemia (Finn et al., 1975; Myers and Moran, 1986). The mechanisms of hypoxia-induced renal dysfunction have been studied extensively but remain elusive. In animal models as well as in humans, it has been demonstrated that acute normocapnic hypoxia results in an immediate decrease in renal blood flow (RBF) and glomerular filtration rate (Busija, 1984; Wiesel et al., 1990; Pedrotti et al., 1992; Huet et al., 1997). Various mediators such as the sympathetic nervous system (Malpas et al., 1996), catecholamines (Hirakawa et al., 1997), and the renin-angiotensin system (Robillard et al., 1981; Ritthaler et al., 1997) have been implicated in renal hemodynamic changes during hypoxia; however, the mechanism by which this occurs is poorly understood.

Adenosine is a byproduct of normal ATP hydrolysis or cellular energy-dependent processes (Sparks and Bardenheuer, 1986; Meghji et al., 1988) and the adenosine production is increased by a reduction in O₂ availability for oxidative phosphorylation in a variety of renal cell types (Beach et al., 1991; Reyes et al., 1995). Indeed, adenosine rapidly increases during ischemia (Osswald et al., 1977; Miller et al., 1978; He et al., 1995) or hemorrhage in the kidney (Nagashima and Karasawa, 1996). Churchill and Bidani (1982) have proposed that a possible candidate mediator of hypoxia-induced renal dysfunction is adenosine. In isolated perfused rat kidneys, hypoxia-induced renal vasoconstriction is preventable by the administration of the adenosine A₁ receptor antagonist 1,3-diprophyl-8-(2-amino-4-chlorophenyl)xanthine (Ramos-Salazar and Baines, 1986). Gouyon and Guignard (1988) have shown that treatment with the nonselective adenosine receptor antagonist theophylline can prevent reductions of RBF and glomerular filtration rate induced by acute systemic hypoxia. These studies, which have used adenosine receptor blockade, suggest that adenosine production and/or release in the kidney may be involved in renal vasoconstriction during hypoxia. However, to date, no direct measurements of intrarenal adenosine during hypoxia have been carried out in an in vivo setting, due to the technical difficulties.

Several other researchers and also our department have recently measured the renal interstitial concentration of adenosine with a microdialysis method (Baranowski and Westenfelder, 1994; Siragy and Linden, 1996; Nishiyama et al., 1997; Zou et al., 1999). This method is well suited for studies of intrarenal adenosine under various conditions because the adenosine receptors are located on the surface of the cell membrane (Spielman and Thompson, 1982). Recently, we established an in vivo renal microdialysis method in anesthetized, mechanically ventilated rabbits (Nishiyama et al., 1999) and, therefore, were able to measure the dynamics of renal interstitial concentration of adenosine during systemic hypoxia.

The aim of the present study is to define the role of intrarenal adenosine and the putative protective effects of a selective adenosine A₁ receptor antagonist in hypoxia-induced renal hemodynamic changes. This study was, therefore, undertaken to determine if systemic hypoxia increases the renal interstitial concentration of adenosine, and to determine the ability of the highly selective adenosine A₁ receptor antagonist 8-(noradamantan-3-yl)-1,3-dipropylxanthine (KW-3902) (Shimada et al., 1992) to prevent renal vasoconstriction in anesthetized, mechanically ventilated rabbits under normocapnic systemic hypoxia.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

General Procedures

All surgical and experimental procedures were performed under the guidelines for the care and use of animals as established by the Kagawa Medical University.

Experiments were carried out with adult male New Zealand White rabbits weighing from 3.0 to 3.5 kg. Animals were housed in separate cages and maintained in a temperature-regulated room on a 12-h light/dark cycle for 1 week. The animals were anesthetized with sodium pentobarbital (25 mg/kg bolus and 5 mg/kg/h infusion) and were ventilated with room air by a mechanical ventilator after tracheotomy. Pancuronium bromide (0.25 mg/kg bolus and 0.125 mg/kg/h infusion), which had only minor effects on hemodynamics, was given i.v. for muscle relaxation. A catheter was inserted into the right femoral vein for the infusion of lactated Ringer's solution or drug solution, which were infused at a rate of 4 ml/kg/h throughout the experiment. Another catheter also was placed in the abdominal aorta via the right femoral artery, and mean arterial pressure (MAP) was continuously measured and recorded with a pressure transducer and a polygraph (Model 361; NEC-San-ei Co., Tokyo, Japan). The left kidney was exposed through a retroperitoneal flank incision. The kidney was carefully denervated by dissecting all visible nerve fibers as well as the tissue connecting the renal hilum cephalic to the renal artery. An electromagnetic flowmeter (MFV-1200; Nihon Kohden Co., Tokyo, Japan) was positioned around the renal artery and RBF was continuously monitored. Two microdialysis probes were gently implanted into the renal cortex and medulla. The probes were connected to a CMA/100 microinfusion pump (Carnergie Medicine, Stockholm, Sweden) and were perfused with isotonic saline solution with heparin (30 U/ml) containing iodotubercidine (10 µM), an inhibitor of adenosine kinase, and erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA) (100 µM), an inhibitor of adenosine deaminase, at a rate of 5 µl/min. The dialysates were collected in a chilled tube over a 5- or 10-min sample period and were analyzed for the concentration of adenosine. Samples were stored at 70°C before analysis. After surgery, the preparation was equilibrated for 90 min to allow for the stabilization of MAP, RBF, and the renal interstitial adenosine concentration, before initiation of test procedures.

At the end of the experiment, the animals were euthanized by giving an excess dose of sodium pentobarbital. The kidney was excised to confirm the position of the dialysis membrane. If the membranes were positioned incorrectly in either the cortex or the medulla, we omitted the data of these dialysates.

Experimental Protocols

Effects of Hypoxia on Renal Hemodynamics and Renal Interstitial Concentration of Adenosine. Following two 5-min control periods with room air, rabbits were ventilated with a gas mixture containing 8% O₂/92% N₂ (hypoxic gas) for 15 min (n = 11). MAP, heart rate, and RBF were measured continuously and the dialysates were collected at 5-min intervals. Three additional 5-min collection periods were performed at 15, 30, and 60 min after the cessation of the hypoxic exposure. In five rabbits, we performed a time control of this protocol in which samples were collected for 120 min.

Effects of Hypoxia on Renal Hemodynamics and Renal Interstitial Concentration of Adenosine during Treatment with KW-3902. We used KW-3902 (Kyowa Hakko Kogyo Co. Ltd., Tokyo, Japan), a highly selective adenosine A₁ receptor antagonist (Shimada et al., 1992), to investigate whether selective blocking of the adenosine A₁ receptors would modify renal hemodynamics and/or the renal interstitial concentration of adenosine during hypoxia. KW-3902 was dissolved in the lactated Ringer's solution containing 1% dimethylsulfoxide and 0.01 N NaOH. In preliminary experiments (n = 3), we had found that i.v. administration of KW-3902 (priming dose, 0.1 mg/kg; sustaining dose, 0.01 mg/kg/min) inhibited the renal vasoconstrictor response to the intrarenal arterial injection of exogenous adenosine (1 and 10 µg). It also was confirmed that the vehicle (lactated Ringer's solution containing 1% dimethyl sulfoxide and 0.01 N NaOH) did not affect any of the parameters studied (n = 2).

Microdialysis Probe

In this study, a newly developed microdialysis probe constructed in our laboratory was used (He et al., 1995). The dialysis membrane (Toyobo, Otsu, Japan) is made from cuprophan fiber, measuring 15 mm in length with a 0.22-mm o.d. and with a 5000-Da transmembrane diffusion cutoff. Steel needles were inserted into both sides of the cuprophan fiber. The efficiency of the microdialysis probe was determined as follows. The probe was placed in a beaker containing an isotonic saline solution into which different quantities of adenosine were added. We perfused the probes with isotonic saline solution with heparin (30 U/ml) containing iodotubercidine (10 µM) and EHNA (100 µM) at a rate of 5 µl/min. The dialysate was collected and the recovery rate of adenosine was calculated by dividing the concentration in the dialysate by the concentration in the medium. At a perfusion rate of 5 µl/min, the recovery rate of adenosine was 30 ± 4%. These recovery rates were higher than those obtained with a commercially available microdialysis probe measuring 2 mm in length and 0.65 mm in diameter, with a 10-kDa transmembrane diffusion cutoff. Based on these results, we considered that a perfusion rate at 5 µl/min was suitable for this experiment.

Analytical Procedures

Adenosine in the dialysate was measured according to the method developed by Zhang et al. (1991). The procedure is briefly described as follows. Twenty-five microliters of dialysate is transferred into a microcentrifuge tube and 72.5 µl of 1 mM acetate buffer (pH 4.0) and 2.5 µl of 40% chloroacetaldehyde are added. The preparation is incubated at 80°C for 1 h to allow for the conversion of adenosine to ethenoadenosine. For HPLC, a reversed-phase HPLC column (Develosil ODS HG-5, 150 × 4.6 mm i.d.) is maintained at 40°C with a column oven (655A-52; Hitachi Ltd., Tokyo, Japan). An isocratic elution with 7.5% acetonitrile in 50 mM potassium phosphate buffer (pH 3.0) at a flow rate of 1.0 ml/min is performed with an HPLC pump (L-600; Hitachi). Fifty microliters of sample is injected and the elution is monitored with a fluorescence spectrometer (F-1000; Hitachi) at an excitation wavelength of 310 nm and an emission wavelength of 400 nm. The chromatogram is recorded by a Hitachi recorder (Model 056).

Statistical Analysis

Values are presented as means ± S.E. Data were analyzed by one-way or two-way ANOVA as appropriate, followed by Scheffe's multiple comparison post hoc test. A p value <.05 was taken to indicate significant differences between means.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Hypoxia on Renal Hemodynamics and Renal Interstitial Concentration of Adenosine. Three minutes after changing to the gas mixture containing 8% O₂/92% N₂ (hypoxic gas), control arterial PaO₂ (98 ± 11 mm Hg) decreased significantly to 32 ± 6 mm Hg, but the control arterial pH (7.41 ± 0.05) and PaCO₂ (30.9 ± 3.4 mm Hg) was unchanged (7.43 ± 0.09 and 28.1 ± 3.3 mm Hg, respectively). These parameters remained at the same level during hypoxic exposure.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Changes in the renal interstitial concentration of adenosine during hypoxia. C, control; *p < .05; n = 11.

Effects of Hypoxia on Renal Hemodynamics and Renal Interstitial Concentration of Adenosine during Treatment with KW-3902. KW-3902 did not affect MAP or the concentration of adenosine, but increased RBF transiently (Table 2). At 25 min following KW-3902 administration, hypoxic exposure was initiated. The arterial PaO₂, pH, and PaCO₂ achieved in hypoxia after KW-3902 were not significantly different from those attained before KW-3902. During treatment with KW-3902, hypoxia caused a significant decrease in RBF, however, this reduction was significantly attenuated compared with that in the absence of KW-3902 (Fig. 2). At 5 min after beginning hypoxia, RBF returned to the control level and this recovery time also was shortened. The KW-3902 treatment did not affect the renal interstitial concentration of adenosine, but hypoxia caused a similar increase in adenosine levels compared with that in the absence of KW-3902 (Figs. 3 and 4).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Percentage of change in the reduction in RBF during hypoxia. Data are expressed as percentage of change of control values. Values are means ± S.E. *p < .05; n = 11.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Changes in the renal interstitial concentration of adenosine during hypoxia after the treatment with KW-3902. C, control; *p < .05; n = 11.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Percentage of change in the renal interstitial concentration of adenosine in the cortex and the medulla during hypoxia. Data are expressed as percentage of change of the control values. Values are means ± S.E; n = 11.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study demonstrates that normocapnic systemic hypoxia increased the renal interstitial concentration of adenosine and caused a renal vasoconstriction in mechanically ventilated anesthetized rabbits. KW-3902, a highly selective A₁ adenosine receptor antagonist (Shimada et al., 1992), did not affect the elevation of the adenosine level; however, it significantly attenuated the reduction in RBF during hypoxia. These results suggest that the intrarenal production and/or release of adenosine are involved in hypoxia-induced renal vasoconstriction.

Because endogenously produced adenosine is formed primarily by renal tubular epithelial cells and reaches the renal microvasculature via the interstitial space (Navar et al., 1996), measurements of interstitial adenosine levels could be critical. A microdialysis method has been suggested to be well suited to measure the dynamics of renal interstitial adenosine in the kidney (Baranowski and Westenfelder, 1994; Siragy and Linden, 1996; Nishiyama et al., 1997; Zou et al., 1999). Recently, we were able to minimize tissue injury by making a fiber type probe with a thinner diameter (0.22 mm). In addition, the length of the dialysis membrane is 1.5 cm, which is three to four times longer than that of a regular probe. As a result, the dialysis efficiency of the new probe was better than that of a regular probe (He et al., 1995; Nishiyama et al., 1999). We can perfuse our probe at a high perfusion rate (5 µl/min) and shorten the sampling time (5 min). Thus, this newly developed microdialysis probe appears to be a useful tool for monitoring the dynamics of adenosine in the renal interstitial space during hypoxia.

The basal interstitial concentration of adenosine in the cortex was 293 ± 70 nM, which was significantly higher than that in the medulla (170 ± 23 nM) in anesthetized rabbits. These findings are inconsistent with those reported by Siragy and Linden (1996) and Zou et al. (1999). They have reported that the renal interstitial concentration of adenosine is higher in the renal medulla than in the cortex in anesthetized rats. The reason for this discrepancy is not clear, however, it might be due to differences in species and experimental conditions. During ischemia, it has been reported that the AMP concentration in the cortex is the highest in the rabbit kidney, whereas the AMP concentration in the outer medulla is the highest in the rat kidney (Zager et al., 1990). Interspecies variation may provide a likely explanation for this discrepancy in adenosine levels. Moreover, we used perfusate containing iodotubercidine (10 µM) and EHNA (100 µM). Because adenosine metabolism is so fast, it is difficult to detect the exact adenosine level in the renal interstitial space without the use of inhibitors of adenosine deaminase and adenosine kinase (He et al., 1995). Therefore, the composition of the perfusate also might have contributed to this discrepancy.

Normocapnic systemic hypoxia resulted in an immediate decrease in RBF whereby levels initially fell to a minimum at 3 to 5 min after beginning hypoxic exposure (Table 1). These results are consistent with those of previous studies (Busija, 1984; Wiesel et al., 1990; Pedrotti et al., 1992; Huet et al., 1997). RBF tended to return to the control value during hypoxic exposure, whereas renal interstitial adenosine concentrations immediately increased and remained at the same level (Fig. 1). It is possible that the responsiveness of the renal vasculature to adenosine may have contributed to such differential changes because adenosine can act as either a vasoconstrictor or a vasodilator of the renal vasculature (Murray and Churchill, 1984; Navar et al., 1996). Furthermore, the fact that adenosine redistributes RBF with predominant actions in the inner cortex and medulla (Dinour and Brezis, 1991; Agmon et al., 1993; Navar et al., 1996) has led to confusion. Infusion of adenosine directly into the renal interstitium increased medullary PO₂ and decreased cortical PO₂, suggesting that elevating interstitial adenosine concentration redistributes cortical blood flow to the renal medulla and/or that adenosine decreased medullary oxygen consumption (Dinour and Brezis, 1991). Agmon et al. (1993) reported that the A₁ agonist N⁶-cyclopentyladenosine decreases both cortical and medullary blood flow, whereas the A₂ agonist CGS-21680C increases medullary blood flow without changing cortical blood flow significantly. Because four subtypes of adenosine receptors (A₁, A_2a, A_2b, A₃) are present in both the renal cortex and the medulla (Zou et al., 1999), changes in renal interstitial adenosine concentrations in the cortex as well as in the medulla may cause hypoxia-induced regional changes in intrarenal blood flow. Further studies are needed to determine the role of adenosine in regulating cortical and medullary blood flow.

It is known that various vasoactive factors such as the sympathetic nervous system (Malpas et al., 1996), catecholamines (Claustre et al., 1985; Schuijers et al., 1986; Hirakawa et al., 1997), and the renin-angiotensin system (Robillard et al., 1981; Hirakawa et al., 1997; Ritthaler et al., 1997) are activated during hypoxia. Recently, we investigated the role of adenosine in modulating the renal vasoconstrictor action of angiotensin II (ANG II) and norepinephrine (NE) (Aki et al., 1997). The ANG II or NE-induced reduction in RBF was attenuated by the administration of KW-3902. Furthermore, this renal vasoconstriction was enhanced by an intrarenal administration of adenosine, which, in turn, could be diminished by the administration of KW-3902 (Aki et al., 1997). These findings indicate that there was a relationship between adenosine A₁ receptor-mediated renal vasoconstriction and ANG II or NE. The evidence supporting a synergistic dependence on ANG II for adenosine to exert its renal hemodynamic effects also has reported by Deray et al. (1990) and Dietrich et al. (1991). Namely, it is possible that the synergistic effects of adenosine and ANG II or NE in renal blood vessels caused a significant decrease in RBF during hypoxemia. Although other factors were not assessed in the present experiments, this is a possible explanation for the role of adenosine in hypoxia-induced renal hemodynamic changes.

In summary, the present experiments demonstrate that hypoxia decreased RBF and significantly increased the renal interstitial concentration of adenosine. Pretreatment with KW-3902, a selective adenosine A₁ receptor antagonist, significantly attenuated hypoxia-induced reduction in RBF. These results suggest that intrarenal adenosine was involved in hypoxia-induced renal vasoconstriction and that the adenosine A₁ receptor antagonist KW-3902 had a partial protective effect against renal vasoconstriction during hypoxemia.