Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Extracellular purine nucleosides and nucleotides mediate a wide range of pharmacological effects through two main families of purine receptors, P₁, which is sensitive to adenosine, and P₂, which is sensitive to ATP, expressed on many mammalian cell types, including enterocytes (Fredholm et al., 1994). Extracellular nucleosides present in body fluids also correspond to a pool of nucleotide precursors of the salvage pathway. Nucleoside transporter (NT) proteins on cell membrane are necessary for cellular uptake of physiological nucleosides (Cass et al., 1998). Two classes of NT processes have been recognized. The first one, called equilibrative, exhibits the typical features associated with facilitated diffusion driven by the concentration gradient of nucleosides. Two subtypes of equilibrative NT (es and ei) have been identified according to their sensitivity to nitrobenzylmercaptopurine (NBMPR), and their genes (ENT₁ and ENT₂) have been cloned in rat and humans (Griffith et al., 1997a,b; Yao et al., 1997; Crawford et al., 1998). The second class of NT process is concentrative and driven by transmembrane Na⁺ gradients. Concentrative NT thus are Na⁺-dependent but insensitive to NBMPR. Six concentrative NT proteins have been characterized (Cass et al., 1998), and two genes have recently been cloned (Che et al., 1995; Huang et al., 1994; Ritzel et al., 1997, 1998). Equilibrative processes can transport a wide variety of permeants from both sides of the membrane, whereas concentrative ones display a relative selectivity for the influx of purine or pyrimidine nucleosides (Wang et al., 1997). The nature of NT proteins expressed on plasma membrane varies among cell types, although both equilibrative and concentrative transporters can coexist on the same cell (Belt, 1983). Moreover, the transport rate of nucleoside can change during cell cycle or after cell activation (Smith et al., 1989; Kichenin et al., 2000a), in correlation with the level of equilibrative or concentrative NT protein expression.

The concentration of free nucleosides present in body fluids can change with physiological (Chagoya de Sanchez et al., 1983) or pathological (Burnstock, 1993; Driver et al., 1993) situations or with the purine content of the diet (Clifford and Story, 1976, LeLeiko et al., 1983; Porcelli et al., 1995). This influences the transport rate of nucleosides via equilibrative NT processes, which depends on the extracellular purine concentration and accounts for pharmacological effects mediated by P₁ and P₂ purinoceptors.

In this study, we explored whether the oral administration of natural purines over a long period of time could affect purine metabolism and transport in rats. ATP was chosen because it is susceptible to generate all of the purine metabolites of the salvage pathway and is the substrate of ectonucleotidases and ecto-protein kinases (Ziganshin et al., 1994). Results reveal an adaptive metabolic response in chronically supplemented rats characterized by an increased nucleoside influx in gut and erythrocytes associated with an increased ability to release ATP and purine nucleosides. Whether this response is dependent on and mediates pharmacological signals through P₁ or P₂ receptors is discussed.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male Sprague-Dawley rats (250 g) were obtained from the Center d'Elevage Depre (Doulcharel, France) and maintained in our animal facilities under standard conditions. Treatment was performed with the daily administration of ATP or adenosine (Sigma, Saint Quentin Fallavier, France) in 0.1 ml of sterile distilled water with a gastric cannula. All experiments were carried out in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC).

Chemicals. [¹⁴C]ATP (1.85-2.29 GBq/mmol) and [¹⁴C]adenosine (1.85-2.29 GBq/mmol) were purchased from Amersham (Saclay, France). [¹⁴C]Sucrose (2409 GBq/mmol), [³H]H₂O (3.7 GBq/ml), and Ecolume scintillation fluid were obtained from ICN (Orsay, France). Hanks' balanced salt solution (HBSS), NBMPR, dipyridamole, adenosine, ATP, and all other reagent-grade drugs were obtained from Sigma.

Purine Extraction and Characterization. Immediately after the collection of biological samples, purines were extracted with trichloroacetic acid (7% w/v) for 30 min at 4°C. Samples were then centrifuged, and supernatants were neutralized with a tri-N-octylamine/trichlorotrifluoroethane solution (0.5 M) before analysis, according to Rapaport (1988). ATP dosage was determined by luminescence using a luciferin-luciferase kit assay (Roche Diagnostics, Meylan, France) according to the manufacturer's recommendations. Luminescence was counted with a Top-Count luminometer (Packard Instruments, Rungis, France). Radiolabeled purine metabolites were analyzed after HPLC fractionation on a Licrosphere RP18 column (125 × 4 mm; Merck, Nogent sur Marne, France) using a Waters 600E pump and a UV detector (Waters model 486). The mobile phase consisted of a mixture of solution A (50 mM KH₂PO₄-KOH, pH 6) and methanol (B). For the first 10 min, solution A was passed at a flow rate of 0.5 ml/min. A linear gradient was then applied to achieve 80% A/20% B after 5 min. Fractions were collected every 30 s. Then 2 ml of scintillation fluid was added, and radioactivity was counted on a Rack-beta counter (LKB, Saclay, France). For analysis of cold purine metabolites, samples were first treated with chloroacetaldehyde to obtain the N⁶ ethene-derived metabolites (Nithipatikom et al., 1994). Derivatives were then separated by HPLC fractionation, and fluorescence was detected using a Fluostar fluorimeter (Bio-Tek Instruments, Winooski, VT).

Purine Absorption In Vivo. Fasted rats were anesthetized with 6 mg/kg sodium pentobarbital i.p. On dissection, a fraction of the jejunum was isolated by introducing a proximal catheter (i.d., 0.86 mm; Biotrol no. 6; Biotrol, Chennevières-les-Louvres, France), 2 cm under the stomach and a distal catheter (i.d., 1.014 mm; Biotrol no. 7) 10 cm below the stomach. The gut section was rinsed with 0.1 M sterile NaCl equilibrated at 37°C. A heparin-treated catheter (i.d., 0.3 mm; Biotrol no. 1) was introduced into one of the secondary mesenteric venules up to the portal vein, which was clamped before the liver. Then 4 ml of a solution providing 5 mg/kg cold purine mixed with 74 kBq of the corresponding ¹⁴C-labeled molecule was injected into the isolated section of the gut. Samples of portal blood and of intraluminal content were collected every 90 s over a period of 10 min, after which animals were sacrificed by an overdose of sodium pentobarbital. Samples were immediately extracted for analysis of purine metabolites.

Adenosine Metabolism in Erythrocytes. Blood from control or treated rats was collected on sodium citrate (3.8% w/v). After centrifugation for 10 min at 1200g, erythrocytes were washed and suspended in HBSS to reconstitute the initial hematocrit. A series of microtubes containing 200 µl of erythrocyte suspension were incubated for 10 min at 37°C with 20 µl of a mixture of cold and [¹⁴C]adenosine (666 kBq) at a final concentration of 1 µM. Individual tubes were centrifuged (30 s, 1000g) at various time over a period of 10 min, and free purines were extracted from the washed pellet with 200 µl of 7% (w/v) trichloroacetic acid. Total radioactivity in the supernatant and pellet was counted. Radioactivity corresponding to the different purine metabolites was determined after HPLC separation as described above.

ATP Exportation. One milliliter of an erythrocyte suspension in HBSS corresponding to the hematocrit was incubated with adenosine (1 µM) for 10 min at 37°C. Erythrocytes were rapidly washed and resuspended in HBSS supplemented with dipyridamole (10 µM) to inhibit nucleoside transport. Then 200-µl aliquots were collected at various times and centrifuged. Free purines were immediately extracted from supernatant and pelleted for analysis.

Adenosine Transport. Adenosine transport was determined according to the method described by Harley et al. (1982). One hundred microliters of erythrocyte suspension in transport medium (100 mM NaCl, 2 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, pH 7.5) was mixed at 25°C with 100 µl of [¹⁴C]adenosine (1 µM, 1.85 kBq). The reaction was stopped by the addition of 200 µl of transport medium supplemented with 20 µM NBMPR and dipyridamole. After extraction, total radioactivity in the supernatant and pellet was counted.

NBMPR Binding Measurement. Experiments were performed according to Cass et al. (1981). Briefly, 200 µl of erythrocyte suspension was incubated with increasing [³H-benzyl]NBMPR concentrations in a total volume of 1 ml for 30 min at 37°C. They were then centrifuged (30 s, 1000g) and washed, and the radioactivity in the pellet and medium was counted. In parallel tubes, 10 µM excess cold NBMPR was added to evaluate nonspecific binding.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Intraluminal ATP Metabolism in Gut of Normal and ATP-Treated Rats. The in vivo evolution of the purine content in the intraluminal space of the gut was compared in control and ATP-treated rats after the local injection of [¹⁴C]ATP into an isolated jejunum section. HPLC analysis of the intraluminal content 1.5 or 9 min after injection revealed that ATP was rapidly transformed into ADP, AMP, adenine, and uric acid (Fig. 1, A and B). No qualitative or quantitative difference was observed between normal animals and rats treated with 5 mg/kg/day ATP per os for 30 days.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Intestinal metabolism of ATP in vivo in control (bold line) and ATP-treated (10 mg/kg/day for 30 days) rats. [14C]ATP (74 kBq) mixed with cold purine to a final concentration of 5 mg/kg was introduced into an isolated jejunum section. Metabolites were analyzed after HPLC fractionation in the intraluminal (A and B) and the plasma of portal blood (C and D) at 1.5 min (A and C) or 9 min (B and D) after administration. 1 indicates uric acid; 2, ATP plus ADP; 3, hypoxanthine; 4, AMP; 5, inosine; 6, adenosine; and 7, adenine. HPLC profiles are representative of independent determinations from four rats in each group.

Adaptive Response of Gut in ATP-Treated Rats. To evaluate the effect of oral treatment with ATP on intestinal metabolism, we then analyzed the ATP concentration in the draining portal blood stream immediately after intraluminal deposition of a mixture of cold and ¹⁴C-labeled ATP into an isolated jejunum section. In normal rats, small amounts of adenosine, hypoxanthine, and uric acid were detected in portal blood, in the absence of liver passage, 9 min after intraluminal injection of an amount of ATP corresponding to 5 mg/kg (Fig. 1C). On the contrary, all types of ¹⁴C-labeled ATP metabolites (i.e., ADP, AMP, adenosine, inosine) were detected in the portal blood of rats treated with ATP for 30 days, indicating an increased absorption of purines resulting from treatment (Fig. 1D). We then tested whether intestinal cells could release ATP into blood stream by using the highly sensitive luciferin-luciferase ATP assay. Unexpectedly, the basal plasma ATP level in ATP-treated animals was much lower than that in control animals (see below). Yet in these animals, a rapid 1000-fold increase in portal plasma ATP concentration was detected within 2 min after the introduction of ATP into a jejunum section (Fig. 2A). A lower (10×) and slower (>5 min) increase was observed in control rats under the same experimental conditions. This demonstrated that gut can rapidly export ATP toward the bloodstream. It also suggested an adaptive metabolic response of enterocytes to chronic oral ATP treatment characterized by an increased release of ATP and purine nucleosides toward blood. When 4 nmol of adenosine, corresponding to the amount of purines present in a dose of 5 mg/kg ATP, was again introduced into an isolated jejunum, a more important and rapid liberation of ATP was observed in the blood of ATP-treated rats compared with control animals (Fig. 2B). This indicated that the improved ATP delivery was likely the consequence of an increased absorption of adenosine and purine nucleosides from the lumen by enterocytes and an increased ATP synthesis from these precursors.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Plasma ATP levels in portal blood of control () and ATP-treated (10 mg/kg/day for 30 days, ) rats. Blood was collected from the portal vein at different times after the administration of [14C]ATP (A) or [14C]Ado (B) (74 kBq mixed with cold purine to a final concentration of 5 mg/kg) in an isolated jejunum section in vivo. After extraction, ATP concentration was measured by luciferin-luciferase luminescence assay. Each point represents the mean of four animals, representative of two independent experiments. Vertical bars represent the S.E. *P < .05 compared with control, calculated from paired Student's t test.

Adenosine Transport in Red Cells from ATP-Treated Rats. Changes in purine uptake and exportation observed in the gut after chronic ATP treatment motivated an analysis of NT processes in erythrocytes, which play an important role in the regulation of extracellular purine level and are dependent on the purine salvage pathway for ATP synthesis. We first compared the ability of erythrocytes from normal and ATP-treated rats to capture exogenous adenosine by incubating 200 µl of washed red cells suspended in HBSS with 6.6 kBq of [¹⁴C]adenosine at 37°C. Samples of supernatant were collected over a period of 5 min, and radioactivity was counted. A faster decrease in extracellular radioactivity was observed with erythrocytes from ATP-treated rats compared with control animals as illustrated in Fig. 3A, suggesting an accelerated uptake of adenosine. Direct adenosine transport rate was thus evaluated. In red cells from ATP-treated rats, a faster and higher level of adenosine transport was observed than in red cells from control animals when 100 µl of erythrocyte suspension corresponding to the hematocrit was incubated with 1 µM adenosine in the transport medium (Fig. 3B). To explain this increase in adenosine transport, the kinetic constants of ENT, which is the only process present on red cells, were determined in erythrocytes from control and ATP-treated animals. The V_max value for adenosine transport increased from 2.53 × 10³ in control animals to 6.53 × 10³ pmol/µl/s in erythrocytes from ATP-treated rats, but the K_m value remained unchanged.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Adenosine metabolism in erythrocytes from control () and ATP-treated (5 mg/kg/day for 30 days) () rats. A, capture of extracellular adenosine. Microtubes containing 200 µl of erythrocytes (in reconstituted hematocrit were incubated in HBSS for 10 min at 37°C. Then, [14C]Ado (0.66 kBq), mixed with cold purine to a final concentration of 1 µM, was added. Reaction was stopped at various times, and extracellular radioactivity was counted after centrifugation (30 s, 1200g). B, adenosine transport. Erythrocyte suspension (100 µl) in transport medium corresponding to the hematocrit was incubated with [14C]Ado (1.85 kBq, 1 µM) at 25°C. The reaction was stopped by the addition of NBMPR and dipyridamole (10 µM) and then centrifuged (30 s, 1200g). Radioactivity in the pellet was counted. C, determination of transport kinetic constants. Vi was determined in transport medium with various concentrations of [14C]Ado. Kinetic constants were deduced from Lineweaver-Burke representations. Each point represents the mean of four animals, representative of two independent experiments. Vertical bars represent S.E. *P < .05, **P < .01 compared with control, calculated from paired Student's t test.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   A, inhibition of adenosine transport by NBMPR in erythrocytes from control () and ATP-treated (10 mg/kg/day for 30 days) () rats. Erythrocytes suspended in transport medium were preincubated with NBMPR for 5 min before the addition of [14C]Ado (1.85 kBq, 1 µM). Adenosine uptake was then determined in triplicate after 5 min by measuring the radioactivity in cell pellet. Adenosine uptake in erythrocytes preincubated without NBMPR was arbitrarily taken as 100%. B, determination of NBMPR binding sites in erythrocytes from control () and ATP-treated (10 mg/kg/day for 30 days) () rats. Erythrocytes were incubated with [3H-benzyl]NBMPR for 30 min at 37°C. Parallel experiments were performed in the presence of 10 µM cold NBMPR to determine the nonspecific binding. Binding sites were determined after Scatchard representation of radioactivity bound to cells. Each point represents the mean of four animals.

Purine Metabolism in Red Cells from ATP-Treated Rats. To further explore the consequence of increased nucleoside transport in red cells from ATP-treated rats, intracellular purine metabolites were analyzed after extraction from erythrocytes as soon as 30 s after contact with extracellular [¹⁴C]adenosine. Adenosine, adenine, AMP, uric acid, and a small amount of ATP were detected in red cells from control rats, but most extracellular adenosine captured by erythrocytes from ATP-treated animals was rapidly converted into ATP (Fig. 5). Thus, although red cells synthesize ATP through the glycolysis pathway, adaptation to ATP treatment improved their ability to produce ATP from extracellular purine precursors. Accordingly, a 2-fold increase in total ATP content was observed 30 s after incubation with adenosine in erythrocytes from treated animals (Table 1). Conversely, when resuspended into fresh HBSS, red cells from ATP-treated animals exported about 50 times more ATP than those from control animals. This was assessed by incubating cells with 1 µM adenosine for 10 min. After washing, the time course of ATP concentration in supernatants was followed by luminescence ATP assay.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Adenosine metabolism in erythrocytes from control (bold line) and ATP-treated (5 mg/kg/day for 30 days) rats. Erythrocyte suspension (200 µl) corresponding to the hematocrit was incubated in HBSS for 10 min at 37°C. Then, [14C]Ado (0.66 kBq) mixed with cold adenosine to a final concentration of 1 µM was added. After 30 s, cells were centrifuged (1000g), and the intracellular purine content was analyzed after HPLC fractionation. 1 indicates uric acid; 2, ATP plus ADP; 3, hypoxanthine; 4, AMP; 5, inosine; 6, adenosine; and 7, adenine. Values are representative of four independent experiments.

Paradoxical Effect of ATP Treatment on ATP Plasma Level. The influence of ATP treatment on plasma ATP level was investigated in rats treated with 5 mg/kg ATP via the oral route over a period of 30 days. Blood samples were collected every fourth day. Plasma and red cell ATP concentrations were measured by luminescence ATP assay. Adenosine concentration was measured by fluorometric detection after derivation and HPLC fractionation. Surprisingly, a dramatic reduction in plasma ATP level was gradually observed down, after 21 days of treatment, to about 1% of the initial value or of the concentration found in water-treated counterparts (Fig. 6A). No significant variation of intraerythrocyte ATP concentration was detected (Fig. 6B). Extracellular ATP breakdown is accomplished by various ectoenzymes generating ADP and AMP (Ziganshin et al., 1994), which are further catabolized by successive dephosphorylation into adenosine by ecto-5'-nucleotidases (Zimmermann, 1996). We thus looked for variation in the adenosine level in both plasma and red blood cells from control and ATP-treated rats. No modification of the adenosine plasma concentration was seen (Fig. 6C), but a small increase was observed in erythrocytes conversely related to the decrease of plasma ATP concentration. Hence, long-term oral treatment with ATP led to a paradoxical and dramatic decrease in the plasma ATP level correlated with a moderate accumulation of adenosine into erythrocytes.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   ATP and adenosine levels in plasma and erythrocytes from rats (n = 6) treated with oral ATP (5 mg/kg/day). A, plasma ATP level. B, erythrocyte ATP level. C, plasma and erythrocyte adenosine level. ATP concentration was measured using luciferin-luciferase luminescence assay. Adenosine concentration was measured by fluorescence after derivation with chloroacetaldehyde and HPLC fractionation. Each point represents the mean ± S.E. *P < .05 compared with day 0, calculated from paired Student's t test.

Influence of ATP Dose on Plasma ATP Level. The influence of the dose of ATP used for treatment on plasma ATP level was evaluated by treating animals with 1, 5, and 10 mg/kg ATP over a period of 20 days. A strong decrease in ATP plasma level was observed (Fig. 7A) for the three doses. This suggested that the effect of chronic ATP supplementation was not strictly dependent on the dose of free purines delivered during treatment. Adenosine is a pharmacologically active ATP metabolite that, contrary to ATP, can be uptake by nucleoside transporters in the gut and used for nucleoside salvage. We thus tested whether adenosine could reproduce the effect of ATP on plasma ATP level in chronically treated animals. As seen in Fig. 7B, an important decrease in plasma ATP level was observed in rats after 20 days of oral treatment with 1, 5, or 10 mg/kg adenosine. This effect could be compared with that obtained in ATP-treated rats and was not clearly dose-dependent.

View larger version (39K): [in this window] [in a new window]   Fig. 7.   Influence of the dose of ATP (A) or adenosine (B) administered on plasma ATP level after 30 days. Plasma ATP level was determined using luciferin-luciferase luminescence assay. Each point represents the mean ± S.E. of six animals. *P < .05 versus control calculated from paired Student's t test.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Results reported herein demonstrate that the daily oral administration of 5 mg/kg ATP during 1 month leads to a modulation of purine transport and metabolism in rats. An adaptive response is observed characterized by 1) an increased uptake of nucleosides in the intestine accompanied by an increased exportation of ATP and nucleosides toward portal blood, 2) an increased purine metabolism in red cells, and 3) a dramatic reduction in plasma ATP level.

The rapid degradation of ATP observed in the gut is consistent with the existence of many enzymatic activities on the brush barrier of the small intestine that are responsible for the breakdown of nucleotides of dietary origin (Mohamedali et al., 1993). It is also well known that most of the radioactivity found in animals fed radiolabeled nucleotides is found in the gut (Salati et al., 1984), which is in good agreement with the important salvage activity of this organ (LeLeiko et al., 1983). However, the increased exportation of nucleosides and ATP from gut toward blood observed in ATP-treated rats compared with controls after the introduction of a same amount of radioactive ATP in the lumen was not anticipated. It suggests that in the small intestine, chronic ATP treatment regulates nucleoside transporters and stimulates the nucleoside salvage pathway. The mechanisms that could account for this regulation are still unknown. However, they might involve purine receptors on enterocytes (Burnstock, 1993). ATP-sensitive P₂ and adenosine sensitive P₁ receptors have both been described on these cells (Fredholm et al., 1994), which can deliver signals responsible for the response observed.

A second consequence of repeated oral administration of ATP is a change in erythrocyte purine metabolism characterized by improved adenosine uptake, ATP synthesis, and ATP exportation by red cells. This might result from an increased availability of extracellular nucleotide precursors in plasma originating from the gut. Adenosine has been reported to activate glycolysis in rat erythrocytes (Gutierrez-Juarez et al., 1992) and to stimulate ATP synthesis (Marinez-Valdez et al., 1982) within 30 min after intraperitoneal injection in vivo. In our experiments, red cells were collected from treated rats 4 to 6 h after the last oral administration. Moreover, plasma adenosine level remained constant throughout the treatment in ATP supplemented rats. It thus seems unlikely that the improved uptake of adenosine and ATP synthesis are simply the consequence of a recent administration of ATP, which would be rapidly degraded into adenosine and other purine nucleosides by ectonucleotidases present on the plasma membrane of many cell types. They could possibly reflect a regulation of ATP metabolism in red cells resulting from the induction of enzymes by a pharmacological effect of extracellular purines on erythrocyte precursors.

Interestingly, like the adenosine plasma level, the concentration of adenosine in red cells was stable throughout the treatment. The adenosine gradient driving the equilibrative transport rate of nucleosides was thus theoretically equivalent in normal and ATP-treated rats. However, adenosine influx was higher in red cells from treated animals. Two equilibrative transporters (es and ei) are expressed on erythrocytes and allow nucleoside capture. The es transporter, which is the predominant and major adenosine transporter on rat erythrocytes, is sensitive to NBMPR inhibition, whereas the ei transporter is not (Jarvis and Young, 1986). The es transporter density on red cells and the K_m value were not significantly affected by treatment, but the V_max value for adenosine transport was strongly increased. This improved transport rate was correlated with the increased ATP synthesis and exportation observed in red cells from ATP-treated rats. The affinity for NBMPR was reduced. A conformational change of the es transporter might account for these effects. This change might be induced by ATP itself. Indeed, Delicado et al. (1994) recently showed on chromaffin cells that ATP can modulate nucleoside transporter function at both intracellular and extracellular levels. The consequence of this regulation is an increased nucleoside transport rate with an increased V_max value and no modification of the K_m value for the permeant. These effects resemble those reported here in ex vivo experiments with red cells from ATP-treated rats. Experiments to be published on human cells confirm that extracellular ATP or analogs can modulate nucleoside transport in vitro. Hence, oral ATP administration may progressively stimulate an autocrine regulation of adenosine uptake and ATP synthesis in red cells and possibly in enterocytes and other cell types.

Repeated oral administration of ATP also led to a progressive diminution of plasma ATP level. This seems quite paradoxical because the administration of doses corresponding to few micromoles reduces normal plasma ATP level, which is in the range of 1 to 3 µM, below 100 nM despite an increased ATP delivery from the gut. In preliminary experiments, we observed a similar phenomenon in cynomolgus monkeys treated with ATP. Moreover, this diminution was clearly not dependent on the dose administered and was rather similar in rats given 1 or 10 mg/kg/day ATP. It may be accounted for by the induction in ectonucleotidases on plasma cell membranes. Our attempts to demonstrate this induction on blood vessels and liver led to inconclusive results due to the high ecto-ATPase activity already present in these tissues, which degrades ATP within a few milliseconds (Ziganshin et al., 1994). Conversely, this activity is low on erythrocytes from mammals (Bencic et al., 1997) and was not stimulated by the treatment (data not shown). The diminution of plasma ATP level seems a progressive physiological reaction to control the potential pharmacological effects via P₂ receptors that would result from an increased delivery of ATP to the blood by enterocytes and red cells. A similar diminution was observed in adenosine-treated rats, demonstrating that the biological response observed results from metabolic changes and not simply from a pharmacological effect of ATP on the intestine. This progressive change in ATP concentration did not affect the uric acid level in biological fluids at the end of the treatment (data not shown).

It is questionable whether purine nucleosides and nucleotides present in the food can have a similar effect. It is noteworthy, however, that in food and meat, consistent with the very short half-life of ATP, most of the free purine nucleotides and nucleosides are present as IMP and inosine (LeLeiko et al., 1983), which are not effective ligands of purinoceptors but can be transported into cells via various NT systems (Huang et al., 1993). These metabolites may not have regulatory effects on the purine metabolism of enterocytes and represent, in addition, nucleoside precursors that can be used by intestinal microorganisms that have no ectonucleotidases on their surface. A direct coupling of these enzymes with nucleoside transporters on the surface of enterocytes might explain why the direct administration of ATP could specifically induce a pharmacological effect that would not normally occur in the gut with food intake.

Several experimental or clinical attempts have been made to use ATP or adenosine for the treatment of pain (Sawynok, 1998), cardiovascular diseases (Daval et al., 1991), cancer (Rapaport, 1993), and brain disorders (Williams, 1993). Most of these experiments have been done using high doses of purines by the i.v. route. Our present results demonstrate that it is possible to obtain a biological response to purines by chronic oral treatment. Experiments from our laboratory support this conclusion. Indeed, an important peripheral vascular response characterized by peripheral vasodilatation and increased paO₂, is observed in rabbits after 14 days of treatment with ATP, which, again, is not seen with a single oral administration (Kichenin et al., 2000b). This treatment also profoundly affects the metabolism of skeletal muscles (K. Kichenin and M. Seman, unpublished findings). The repeated oral administration of pharmacologically active purines, such as ATP or adenosine, seems thus able to modify purine metabolism in vivo and to regulate physiological parameters that are known to be under the control of purinoceptors.