Introduction

Abstract Introduction Methods Results Discussion References

The protective effects of Ado in myocardial and cerebrovascular ischemia are well established (Ely and Berne, 1992; Foster et al., 1994; Lasley et al., 1994; Rudolphi et al., 1992). More recently, other important physiological roles of Ado, including anti-inflammatory and analgesic actions, have been found (Cronstein, 1991; Sawynok and Sweeney, 1989). During ischemic episodes in the myocardium, hippocampus and other tissues, endogenous Ado is produced, where it acts as a naturally protective metabolite; however, newly formed Ado is removed very quickly from tissues by the Ado-metabolizing enzymes ADA and AK. This rapid metabolism serves to reduce the cytoprotective benefits of Ado. Therefore, a number of concepts for the pharmacological elevation of Ado concentrations at ischemic foci have been attempted, including AICA-riboside (Acadesine, Protara) (Barankiewicz et al., 1990), AK inhibitors (Firestein et al., 1995), ADA inhibitors (Green, 1980), AMP deaminase inhibitors or nucleoside transport inhibitors (Van Belle, 1993). Some of these attempts were unsuccessful due to little, if any, ability of the compounds to elevate Ado concentrations, such as AICA-R (Barankiewicz et al., 1990), AK inhibitors (Firestein, et al., 1995) or AMP deaminase inhibitors.¹ Some compounds that proved able to elevate Ado concentrations but produced adverse effects were DCF (Agarwal, 1980; Paine et al., 1981) and DIP (Pennell et al., 1990; Van Belle, 1993). Also, exogenous Ado or synthetic Ado agonists cannot be used because they result in hypotension, bradycardia, atrioventricular block and a variety of other systemic side effects (Belloni et al., 1992; Habazette et al., 1992; Lee et al., 1992).

The most potent way to elevate Ado concentrations in biological systems is to block ADA activity with specific inhibitors because the suppression of the deamination reaction inhibits a major step in the pathway of ATP degradation in most cells. Indeed, in human heart ~70% of ATP catabolism proceeds via the ADA reaction (Smolenski et al., 1992). It has been shown in many different cells and tissues that inhibition of ADA is a potent way to increase the tissue half-life of Ado and elevate extracellular Ado concentrations when ATP degradation is accelerated by ischemia or during simulated ischemia (Achterberg et al., 1985; Hudspeth et al., 1994, Phillis et al., 1988; Van den Berghe et al., 1980; Zoref-Shani et al., 1988). In contrast, under normoxic conditions when there is no elevated cascade of ATP degradation, inhibition of ADA should not result in considerable elevation of Ado levels.

The specific ADA inhibitor DCF has been demonstrated to be cardioprotective in ischemic situations (Bolling et al., 1990; Hudspeth et al., 1995; McClanahan et al., 1993). In isolated, buffer-perfused rabbit hearts exposed to moderately hypothermic (34°C) global ischemia for 2 hr, perfusion with a DCF-enriched cardioplegia solution resulted in significantly improved functional recovery compared to non-DCF-treated hearts (Bolling et al., 1990). A particularly rigorous test of DCF was performed in dogs with global cardiovascular ischemia. Pretreatment with DCF (0.2 mg/kg) prevented postischemic dysfunction in canine hearts (Hudspeth et al., 1995), and a substantial reduction in mechanical stunning was observed as a result of DCF treatment (McClanahan et al., 1993). In all cases, hearts that were pretreated with DCF showed significantly better functional recovery after reperfusion and were associated with the accumulation of Ado in interstitial fluids during the ischemic episodes (Bolling et al., 1990; Hudspeth et al., 1994; McClanahan et al., 1993). In cerebral ischemia, DCF (2.5 mg/kg i.p.) markedly reduced ischemic injury in the rat (Gidday et al, 1995). Although DCF stimulated considerable elevation of Ado concentrations, it could not be considered for commercial development because it is a tight-binding ADA inhibitor that eventually results in immunosuppression and toxicity. Studies have shown that single doses of DCF can inhibit ADA to the extent that enzyme resynthesis is required to regenerate activity (Padua et al., 1992). The mechanism of DCF toxicity includes accumulation of 2-dAdo, which inhibits S-adenosylhomocysteine hydrolase and, consequently, the transmethylation reaction. The major route of toxicity is through phosphorylation of 2-dAdo to dATP, which inhibits ribonucleotide reductase and replicative DNA synthesis, especially in T lymphocytes (Gelfand and Cohen, 1983).

In contrast to DCF, which is practically an irreversible Ado deaminase inhibitor, EHNA produces a reversible inhibition of ADA. Therefore, EHNA, which also has been shown to be cardioprotective (Dhasmana et al.,1983; Dorheim et al., 1991; Sandhu et al., 1993; Zhu et al., 1994; Zughaib et al., 1993), should be less toxic than DCF because the cellular capacity for deamination of purines in the presence of this inhibitor is not destroyed.

We present the results of studies of novel ADA inhibitors that are derivatives of EHNA. Data are presented suggesting that these agents may represent a new class of compounds with applications in cardiovascular and cerebrovascular ischemia.

	Methods

Abstract Introduction Methods Results Discussion References

Chemicals. [methyl-³H]Thymidine (7 Ci/mmol) and [2,8-³H]Ado (27.4 Ci/mmol) were from Moravek Biochem (Brea, CA). ADA, tri-n-octyl-amine (alamine), DCF, 2-deoxy-D-glucose, EHNA, 2-dAdo and other nonradioactive nucleotides and nucleosides were from Sigma Chemical (St. Louis, MO). 5-Iodotubercidin was from ICN (Aurora, OH). 1,1,2-Trichlorotrifluoroethane (freon) was from Aldrich Chemical (Milwaukee, WI). Sephadex G-25 Quick Spin columns were from Boehringer-Mannheim Biochemical (Indianapolis, IN). Cellulose TLC sheets were from Eastman Kodak (Rochester, NY). Polyethyleneimine cellulose (PEI-cellulose) was from Macherey-Nagel (Doren, Germany), and RPMI 1640 medium and fetal bovine serum were from GIBCO (Grand Island, NY).

View larger version (14K): [in this window] [in a new window]

Characteristics of EHNA analogs. CPC-402 (9-hydroxy-EHNA), CPC-405 (9-chloro-EHNA), CPC-406 (9-phthalimido-EHNA), CPC-407 (8,9-didehydro-EHNA) and CPC-410 (9-butyldiphenylsilyloxy-EHNA) are ADA inhibitors synthesized by the Department of Chemistry at Cypros Pharmaceutical Corporation (United States Patent 5,491,146). Original synthesis and structural identification of these compounds took place in the laboratory of Dr. Elie Abushanab at the Department of Medicinal Chemistry at the University of Rhode Island, Kingston, Rhode Island.

Cells. Human astrocytoma cells (UC-11 MG) were obtained from the Department of Biochemistry and Molecular Biology, University of Cincinnati Medical School, Ohio. Bovine heart microvascular endothelial cells (B-88) were obtained from the Department of Pathology and Laboratory Medicine, University of Cincinnati, Cincinnati, Ohio. Jurkat T-lymphoblasts were obtained from the Department of Pharmacology, University of North Carolina, Chapel Hill, North Carolina. Cells were grown in RPMI-1640 medium containing 10% fetal bovine serum and were passaged every 3 days at 1:5 (UC-11 MG), 1:6 (Jurkat) and 1:10 (B-88) dilutions. Human RBCs were obtained from healthy human volunteers. Heparinized blood was centrifuged at 1700 × g for 4 min at room temperature, and plasma and the buffy coat were removed. RBCs were washed once and suspended in 5 volumes of 0.9% NaCl. In experiments, a 2.5% (v/v) suspension of RBCs was used.

Inhibition of purified ADA. Commercially available Type IX calf spleen ADA was used for evaluation of novel ADA inhibitors. ADA (50 µl) was purified in a Sephadex G-25 minicolumn in 50 mM phosphate buffer, pH 7.6. The enzyme was then diluted 40-fold with 50 mM phosphate buffer and kept on ice throughout the assay.

Inhibition of ADA in intact cells. Confluent monolayers of HACs or fresh RBCs (2 × 10⁶) were preincubated in RPMI medium containing 0 to 50 µM EHNA or EHNA analog for 1 hr. The cells were then incubated for an additional 30 min at 37°C with 10 µM 5-iodotubercidin to block the activity of Ado kinase and then for 30 min with 100 µM ³H-Ado. After a 30-min incubation, the amounts of radiolabeled inosine, hypoxanthine and Ado were measured in the cell medium after separation by cellulose TLC in a 1-butanol/methanol/water/ammonia (60:20:20:10) solvent (Barankiewicz et al., 1990). Tritium counts were measured in a scintillation counter.

Reversibility of ADA inhibition in intact cells. HACs were incubated in RPMI medium for 30 min at 37°C with 10 µM 5-iodotubercidin and then for 60 min with 1 µM DCF or 40 µM EHNA or analog. Cells were washed up to four times with 10 ml fresh medium and then incubated for 30 min with 100 µM ³H-Ado. Within this incubation time, the reaction catalyzed by ADA was linear. ADA activity was determined after separation by cellulose-TLC (Barankiewicz et al., 1990) of the radiolabeled inosine and hypoxanthine according to the above methods. ADA activity was measured as a sum of inosine and hypoxanthine. In this incubation, no more than 2% radioactivity was found in IMP or other purine metabolites.

Reversibility of cytotoxic effects on T-cells. Human T-lymphoblasts (Jurkat) were incubated in RPMI medium for 2 hr at 37°C with 1.0 or 20 µM DCF or 40 µM EHNA, CPC-405 or CPC-406. Lymphoblasts were then incubated 24 hr with 50 µM dAdo and 18 hr with 10 µCi of radiolabeled thymidine. Some of the cells were washed five times with 2 ml of incubation medium before the 24-hr incubation with dAdo. After incubation, the cells were centrifuged (1500 rpm, 2 min), the supernatant was discarded and the cells were extracted with 100 µl of 0.4 M PCA. The PCA-insoluble pellet was washed four times with 1 ml of 0.4 M PCA, and its radioactivity was measured in a scintillation counter.

Measurement of Ado kinase activity. HACs or bovine heart endothelial cells were incubated in RPMI with 20 µM DCF for 60 min at 37°C to block endogenous ADA activity. The incubation was continued for 60 min with 50 µM EHNA, CPC-405 or CPC-406 or 20 µM ITU as a positive control. Then, 10 µM ³H-Ado was added, and incubation continued for 60 min. The incubation medium was removed, and cells were extracted with 100 µM PCA for 5 min on ice. Extracts were collected, neutralized with alanine/freon (1:4 v/v) mixture and analyzed for radiolabeled adenine nucleotides using PEI-cellulose TLC (Barankiewicz et al, 1990). Activity of AK was determined as the sum of AMP + IMP + ADP + ATP.

Measurement of nucleoside transport. HACs were incubated for 60 min with 50 µM EHNA, CPC-405 or CPC-406 and then for 10 sec with 1 µM ³H-Ado. After incubation, cells were immediately washed with 10 ml of fresh medium, and intracellular radioactivity was measured after extraction with 0.4 M PCA. DIP (20 µM) was used as a positive control.

Ado release. To test EHNA and its analogs for their ability to enhance Ado release under simulated ischemia conditions, first the intracellular ATP pool was labeled during 1-hr incubation with 1 µM ³H-Ado. After washing off of unincorporated Ado, 1 µM EHNA or analog was added, and the cells were incubated for 5 hr in an anaerobic chamber in a glucose-free medium that had previously been bubbled with 95% nitrogen/5% CO₂ dioxide.

Langendorff ischemia-reperfusion injury in rat hearts. Male Sprague-Dawley rats (250-350 g) were anesthetized with sodium heparin and killed with CO₂. The heart was rapidly excised via thoracotomy and placed in PSS until contraction ceased. The heart was then mounted via the aortic root to a cannula and retrogradely perfused with PSS containing (in mM): NaCl (118), KCl (4.7), CaCl₂ (2.2), KH₂PO₄ (1.18), MgSO₄ (1.17), NaHCO₃ (25) and dextrose (11) at 80 mm Hg at 37°C. The perfusion solution was aerated with 95% O₂/5% CO₂ to maintain pH 7.4. Hearts were allowed to equilibrate for 15 min, during which a balloon-tipped catheter was introduced into the lumen of the left ventricle via a small incision in the left atrium. The catheter was connected to a pressure transducer and was used to measure left ventricular hemodynamic performance [i.e., left ventricular systolic pressure, left ventricular end-diastolic pressure, left ventricular developed pressure, +dP/dt_max (the rate at which pressure developed in the left ventricle during each contraction), dP/dt_max (the rate at which left ventricular pressure declined after each contraction) and heart rate. After placement of the balloon-tipped catheter, the pulmonary artery was cannulated to collect coronary effluent for measurements of coronary flow.

Hypoxic injury in rat hippocampal slices. The neuroprotective abilities of EHNA and its analogs against CA1 neuronal injury caused by hypoxia were examined in the rat hippocampal slice. Male Sprague-Dawley rats were briefly anesthetized with halothane and decapitated. Brains were quickly removed and placed in cold aCSF for 1 min. aCSF was composed of (in mM): NaCl, 126; KCl, 4; KH₂PO₄, 1.4; MgSO₄, 1.3; CaCl₂, 2.4; NaHCO₃, 26 and glucose, 4; pH 7.4, and saturated with 95% O₂/5% CO₂. Hippocampi were dissected free from brains, cut into 475-µm transverse sections and were placed in paired recording wells perfused with aCSF maintained at 37°C.

	Results

Abstract Introduction Methods Results Discussion References

Inhibition of ADA by CPC-compounds in purified enzyme and in intact cells. The novel EHNA analogs 9-hydroxy-EHNA (CPC-402), 9-chloro-EHNA (CPC-405), 9-phthalimido-EHNA (CPC-406), 8,9-didehydro-EHNA (CPC-407) and 9-butyldiphenylsilyloxy-EHNA (CPC-410) were evaluated for their ability to inhibit purified ADA and ADA in HACs and human RBCs (table 1). All these compounds except CPC-410 were potent inhibitors of ADA activity in cell-free extracts and in intact cells. The studies done with intact cells were performed in order to determine whether the CPC compounds could gain access to the cytoplasm. The most potent inhibitors of ADA activity in cells and of the purified enzyme were CPC-405 and CPC-406.

Effect of CPC compounds on Ado release from intact astrocytoma and heart endothelial cells. The compounds were evaluated for their ability to stimulate Ado release from intact HACs (fig. 1A) and bovine microvascular endothelial cells (fig.1B) under simulated ischemic and normoxic conditions. Control samples in these experiments were cells incubated under anaerobic conditions without drug treatment. Control cells released 400 of 500 cpm of Ado/sample. It was found that under normoxic conditions, no Ado release was detected when cells were incubated with EHNA or its analogs because the cpm were equal to background (30-40 cpm/sample). Under anaerobic conditions, however, inhibition of ADA by EHNA or analogs resulted in the release of Ado (fig. 1, A and B). These data correlate well with the data for inhibition of ADA in intact astrocytoma cells (table 1) and show that CPC-405 and CPC-406 appear more potent than EHNA in releasing Ado. When cells containing radiolabeled ATP were incubated in the presence of 5.5 mM deoxyglucose or 5.5 mM sodium azide to induce ATP degradation, massive increases in Ado release were observed in the presence of 10 µM CPC-405 (852% and 1443%, respectively) or CPC-406 (1743% and 1293%, respectively).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Effect of EHNA and analogs on Ado release from (A) HACs and (B) bovine heart microvascular endothelial cells. Cells (2 × 106) were incubated for 1 hr with 3H-Ado. After washing out unincorporated precursor, cells were incubated in glucose-free medium with 1 µM EHNA or analog for 5 hr under anaerobic conditions. Content of radiolabeled nucleosides and bases released from cells was analyzed in cell culture medium using TLC.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Dose-dependent effect of EHNA analog CPC-406 and AICA-riboside on Ado release from HACs. Cells (HACs) were incubated for 1 hr with 3H-Ado. After washing out unincorporated radioactivity, cells were incubated in glucose-free medium with various concentrations of CPC-406 or AICA-riboside for 2 hr under anaerobic conditions. Content of radiolabeled nucleosides and bases released from cells into the media was analyzed using TLC.

Reversibility of ADA inhibition and immunotoxicity. In contrast to DCF, the inhibition of ADA by EHNA and its analogs CPC-405 and CPC-406 was easily reversed by washing several times. This resulted in the recovery of initial ADA activity (fig. 3). The inhibition of ADA by EHNA is usually completely removed within five washes of the cells with 2 ml of medium, whereas inhibition of ADA by analogs CPC-405 and CPC-406 requires more washes, usually 10 washes for complete recovery of ADA activity (data not shown).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Reversibility of ADA inhibition by EHNA and its analogs. Cells (HACs) were incubated for 30 min with 10 µM 5-iodotubercidin and then for 60 min with 1 µM DCF or 40 µM EHNA or 40 µM EHNA analog (CPC-405 or CPC-406). After incubation, cells were washed four times and incubated for 30 min with 100 µM 3H-Ado. Radiolabeled Ado, inosine and hypoxanthine were measured after separation by TLC.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Reversibility of cytotoxic effects of EHNA analogs. Cells (human T-lymphoblasts) were incubated for 2 hr with (A) 1.0 µM DCF, (B) 20 µM DCF, (C) 40 µM EHNA, (D) 40 µM CPC-405 and (E) 40 µM CPC-406. Some of the cells were washed (w) 5 times before they were incubated for 24 hr with 50 µM dAdo and 18 hr with radiolabeled thymidine. After incubation, cells were extracted with PCA, and the radioactivity of the insoluble pellet was measured.

Specificity of EHNA analogs. EHNA, CPC-405 and CPC-406 were also evaluated for their effects on Ado kinase activity. ITU (20 µM), used as a positive control, inhibited 98.9% of AK activity in HAC cells and 92.5% of AK activity in endothelial cells. Neither CPC-405 nor CPC-406 in concentrations up to 50 µM affected AK activity in either cell type (table 2).

Effect of CPC compounds on the isolated ischemic rat heart. To examine possible functional cardioprotective properties of the novel EHNA derivatives, CPC-405 and CPC-406 were evaluated in the Langendorff ischemia-reperfusion injury rat heart model. Both compounds demonstrated significant cardioprotection as measured by the recovery of left ventricular developed pressure (fig. 5A), left ventricular end-diastolic pressure (fig. 5B), maximal rate of systolic increase in pressure (+dP/dt_max) (fig. 5C) and maximal rate of diastolic fall in pressure (dP/dt_max) (fig. 5D) in ischemic hearts.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Cardiac responses to CPC-405 and CPC-406 in Langendorff ischemia-reperfusion injury in rat hearts. A, Left ventricular developed pressure. B, Left ventricular end-diastolic pressure. C, Maximal rate of systolic increase in pressure. D, Maximal rate of diastolic fall in pressure. *, Significantly different from vehicle, P < .05 (n = 8).

Effect of EHNA and CPC-406 on hypoxic injury in rat hippocampal slices. EHNA and CPC-406 were evaluated for their ability to prevent hypoxic injury in rat hippocampal slices (fig. 6). The data show that EHNA and CPC-406 both protect hippocampal slices during a hypoxic event, as measured by the recovery of CA1 population spike. The EC₅₀ for the recovery of hippocampal hypoxic injury potentials was determined to be 10 µM for EHNA and 7.5 µM for CPC-406.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Neuroprotective effect of EHNA and CPC-406 during hypoxic injury in rat hippocampal slices. Rat hippocampal slices were treated with various concentrations of EHNA or CPC-406 for 30 min before hypoxia and through the first 15 min of return to oxygenated conditions. *, Significantly different from vehicle, P < .05.

	Discussion

Abstract Introduction Methods Results Discussion References

Ado, when it accumulates in ischemic tissue, plays two major roles. It can act on Ado receptors to regulate a number of specific physiological functions, including vasodilation in the heart or brain, and inhibition of glutamate release in nerve cells (Burke and Nadler, 1988). Accumulated Ado may also serve as a substrate for the regeneration of adenine nucleotides. Direct cardioprotective effects of Ado have been shown (Ely et al., 1985; Humphrey et al., 1982), although with variable success, probably due to its very rapid metabolism. Exogenous Ado is not optimal for therapeutic use because it produces hypotension, bradycardia and atrioventricular block. The physiological functions of Ado, including vasoregulation (Berne, 1963), are mediated by Ado receptors, which require relatively low Ado concentrations for receptor activation. In contrast, relatively high concentrations of Ado are needed for reconstitution of ATP pools in cells. Although the effect of Ado on Ado receptors has been extensively studied (Kollias-Baker et al., 1994; Phillis, 1990), the importance of Ado accumulation with regard to reconstitution of the ATP pool has been given little consideration. Both these functions may be important in regards to the cytoprotective effects of Ado.

The availability of Ado for both receptor activation and ATP synthesis is determined by the level of Ado accumulation (resulting from intracellular and extracellular formation), its release, reuptake and degradation. The majority of Ado is produced by cytoplasmic or extracellular 5-nucleotidases, especially under stress conditions. There is, however, very little or no accumulation of newly formed Ado due to the activity of the ubiquitous enzyme ADA. Because ADA is the major enzyme responsible for the degradation of Ado, the inhibition of its activity should represent one of the best ways to increase accumulation of Ado in tissues under stress conditions. This, then, could also be an important means for Ado accumulation in the human heart, where under ischemic conditions 70% of ATP is degraded via ADA activity (Smolenski et al., 1992). There have been many other attempts made to elevate Ado concentrations including inhibition of AK activity (Firestein et al., 1995), inhibition of Ado transport (Van Belle, 1993) and AICA-riboside treatment (Barankiewicz et al., 1990). However, with the exception of the nucleoside transport inhibition approach, they have not consistently produced a significant elevation of Ado levels.

Because ischemic tissues, and specifically myocardium, are rich sources of endogenous Ado, the rationale has developed that inhibition of ADA will augment or sustain local Ado levels during ischemic conditions. Indeed, the inhibition of ADA activity in a number of cell types and organs has been shown to result in a massive accumulation of Ado (Belloni et al., 1984; Green, 1980; Nimit et al., 1981; Phillis, 1989; Zetterstrom et al., 1982). Elevated Ado concentrations resulting from the inhibition of ADA would be restricted to the ischemic areas, minimizing potentially adverse side effects such as the systemic effects of Ado.

Interesting microdialysis experiments measuring Ado concentrations in the presence of ADA inhibitors have shown that in myocardium (Dorheim et al., 1991, Silva et al., 1995) or in brain (Scotti and Van Wylen, 1993) Ado levels in the presence of DCF can be increased 4-fold but still remain low (<0.8 µM) in nonischemic conditions. In contrast, in ischemic conditions in the heart, in the presence of ADA inhibitors (DCF or EHNA), a huge elevation of Ado concentrations (150 µM) is seen (Dorheim et al., 1991, Silva et al., 1995). Because a implantation of microdialysis probe is an invasive technique, it is difficult to evaluate to what degree Ado concentrations measured before ischemia can be extrapolated to unstressed conditions.

It has been shown in several models that ADA inhibition by DCF results in striking elevations of Ado within ischemic areas, fulfilling the criteria of site and event specificity and improving functional recovery, but only in reversible injury (Bolling et al., 1990; Hudspeth et al, 1994; Sandhu et al., 1993). In studies focused on myocardial infarction involving irreversible injury, however, ADA inhibition, while elevating Ado concentrations dramatically, has not been shown to be cardioprotective (Li and Kloner, 1993; Silva et al., 1995).

The concept of using ADA inhibitors as Ado releasing agents for cardioprotection or cerebroprotection has not been developed due to the immunotoxicity observed for the most potent and specific ADA inhibitor known, DCF (Gelfand and Cohen, 1983). In contrast to DCF, which is a tight-binding ADA inhibitor, EHNA and its analogs represent reversible type ADA inhibitors. This has led us to generate a series of novel EHNA analogs (United States Patent 5,491,146) that might possess enhanced pharmacokinetic properties suitable for commercial development. Indeed, in our studies, the inhibition of ADA by EHNA and its derivatives CPC-405 or CPC-406 can be easily reversed by simply removing them from the cell culture medium by washing (fig. 3), in much in the same manner as blood perfusion removes drugs from tissues. Even more important is the finding that removing EHNA or its analogs resulted in reversal of the ADA inhibitor-induced toxicity of T-cells (fig. 4). In contrast, DCF inhibition of ADA and the toxicity on T-cells was irreversible. Therefore, it appears that reversible ADA inhibitors display reduced immunotoxicity, a property that may better suit them to clinical use in both cardiovascular and cerebrovascular therapy.

Although there is little known about EHNA removal from tissues in vivo, there is some evidence that EHNA is extensively metabolized and cleared from the mammalian bloodstream (Lambe and Nelson, 1982; McConnel et al., 1980). Therefore, we expect that on the basis of the reversibility of ADA inhibition by EHNA analogs, these compounds can be therapeutically effective in vivo without irreversibly inactivating ADA.

CPC-405 and CPC-406 seem to be highly specific ADA inhibitors because they do not affect nucleoside transport or AK activity (table 2). They inhibit activity of ADA efficiently in different cells types: HACs and human RBCs (table 1). The potent inhibition of ADA in these cells resulted in extensive release of Ado (fig. 1). In contrast to other compounds used to elevate Ado concentrations, such as AICA-riboside, the compounds reported here are potent and effective Ado-releasing agents (fig. 2). Considering their ability to elevate Ado concentration and apparent lack of immunotoxicity, the protective effects of these compounds on ischemic tissues were studied in two models: Langendorff isolated heart and hippocampal slices (fig. 5 and fig. 6). In both models, EHNA derivatives showed protective effects. One plausible explanation for the observed protective effects is their action via Ado elevation. Use of ADA inhibitors results in not only the elevation of Ado concentrations but also the reduced loss of adenine bases from ischemic myocardium. It has been found that EHNA can improve the function of isolated rat hearts subjected to global ischemia, although this was not accompanied by an increase in myocardial ATP content or diminished loss of adenine bases during reperfusion (Dhasmana et al., 1983).

One important question that remains to be answered regarding protection in cerebral ischemia in vivo by EHNA analogs is their ability to cross the blood-brain barrier. Although we have good evidence for the protection of CNS tissue in vitro by the novel EHNA analogs (fig. 6), we do not yet have data for their activity in vivo. Evidence that ADA inhibitors can cross the blood-brain barrier was documented by Mendelson et al. (1983), who has showed that EHNA, injected intraperitoneally, inhibited brain ADA activity. Similar data have also been obtained by Geiger et al. (1987) for DCF.

In addition, inhibition of ADA can have protective effects on the ischemic tissues by preventing free radical-mediated injury (Xia et al., 1996). It has been reported that inhibition of ADA with EHNA reduced formation of hypoxanthine and xanthine (substrates for xanthine oxidase) and blocked free radical generation. This resulted in a >2-fold recovery of contractile function of postischemic heart.

Additional mechanistic questions remain regarding the cytoprotective and anti-inflammatory action of Ado. However, it is clear that specific and selective alteration of its degradation in metabolically stressed tissue should be of therapeutic benefit. The reversible ADA inhibitors described in this report may represent such a pharmacologic tool. The data thus far are preliminary in nature, but they do suggest that this class of reversible ADA inhibitor has less toxicity potential then DCF. This, coupled with the cardioprotective and neuroprotective activities observed with CPC-405 and CPC-406, suggests that this series of compounds may have clinical utility in a range of ischemic disorders.