Objective Adenosine deaminase (ADA) may be multifunctional, regulating adenosine levels and adenosine receptor (AR) agonism, and potentially modifying AR functionality. Herein we assess effects of ADA (and A1AR) deficiency on AR-mediated responses and ischaemic tolerance.
Methods Normoxic function and responses to 20 or 25min ischaemia and 45min reperfusion were studied in isolated hearts from wild-type mice and from mice deficient in ADA and/or A1ARs.
Results Neither ADA or A1AR deficiency significantly modified basal contractility, although ADA deficiency reduced resting heart rate (an effect abrogated by A1AR deficiency). Bradycardia and vasodilation in response to AR agonism (2-chloroadenosine) were unaltered by ADA deficiency, while A1AR deficiency eliminated the heart rate response. Adenosine efflux increased 10- to 20-fold with ADA deficiency (at the expense of inosine). Deletion of ADA improved outcome from 25min ischaemia, reducing ventricular diastolic pressure (by 45%; 21±4 vs. 38±3mm Hg) and lactate dehydrogenase (LDH) efflux (by 40%; 0.12±0.01 vs. 0.21±0.02U/g/min ischaemia), and enhancing pressure development (by 35%; 89±6 vs. 66±5mm Hg). Similar protection was evident after 20min ischaemia, and was mimicked by the ADA inhibitor EHNA (5μM). Deletion of ADA also enhanced tolerance in A1AR deficient hearts, though effects on diastolic pressure were eliminated.
Conclusions Deficiency of ADA does not alter sensitivities of cardiovascular A1 or A2ARs (despite markedly elevated [adenosine]), but significantly improves ischaemic tolerance. Conversely, A1AR deficiency impairs ischaemic tolerance. Effects of ADA deficiency on diastolic pressure appear solely A1AR-dependent while other ARs or processes additionally contribute to improved contractile recovery and reduced cell death.
This article is referred to in the Editorial by W.R. Law (pages 8–9) in this issue.
Despite knowledge of its enzymatic properties and cellular localisation, the specific physiological functions of adenosine deaminase (ADA; EC 188.8.131.52) in different tissues are unclear . This contributes, in part, to an incomplete picture regarding the genesis of the disease phenotype with human ADA deficiency . It has become apparent that 2 quite distinct forms of ADA exist–ADA1 and ADA2. The ADA1 enzyme is a Zn-binding protein of ∼35kDa encoded by the ADA gene, and is expressed in all tissues. In contrast, ADA2 (∼100kDa) belongs to a novel family of ADA-related growth factors, is the predominant ADA in plasma of most mammals (but not mice), and is encoded by the cat eye syndrome critical region candidate 1 gene . These enzymes can regulate intra- and extracellular levels of adenosine through hydrolytic deamination to inosine. This, in turn, regulates levels of AR agonism , and may also facilitate provision of substrate for the potentially damaging xanthine oxidase reaction . The enzyme also catalyses deamination of 2-deoxyadenosine which itself lacks known physiological roles, yet its accumulation expands the deoxy-ATP pool, inhibiting DNA synthesis and triggering apoptosis. Furthermore, adenosine and 2-deoxyadenosine accumulation inhibits S-adenosylhomocysteine (SAH) hydrolase, which may modify APO-1-mediated death  and inhibit transmethylation reactions through SAH accumulation. Absence or reductions in ADA may therefore generate beneficial (protection via AR agonism; purine pool preservation; reduced radical generation) and/or detrimental actions (deoxy-ATP and SAH accumulation; apoptosis). In addition, ADA (specifically ADA1) has been implicated in non-enzymatic regulation of A1 and A2AR functionality [7–10]. Studies of non-cardiac tissues indicate cell surface ADA1 interacts with ARs , facilitating high-affinity ligand binding [9–12], and AR internalisation/desensitisation [13,14]. The global aim of this study was to undertake an initial characterisation of the cardiovascular phenotype arising from ADA1 deficiency. Specifically, we assessed the effects of ADA and/or A1AR deficiency on: (i) cardiovascular function; (ii) adenosine and purine catabolite formation; (iii) A1 and A2AR-mediated responses; and (iv) myocardial tolerance to ischaemia–reperfusion.
2 Materials and methods
Investigations conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).
2.1 Experimental animals
The ADA [15,16] and A1AR  deficient mice were generated as described previously. The ADA deletion here eliminates ADA1 [15,16], with ADA2 not normally expressed in mice. Dual ADA/A1AR deficiency was generated by mating ADA with A1AR deficient mice. Genotypes for all lines were determined by PCR analysis of genomic DNA in a manner similar to that outlined recently . All mice were on a mixed C57BL6/129sv background, and phenotypic comparisons were performed against wild-type littermates. All ADA and ADA/A1AR deficient mice were maintained for 2months on weekly intraperitoneal injections of 5 U polyethylene glycol (PEG) conjugated ADA (as in replacement therapy for ADA deficient humans) . The PEG-ADA supplement was then withdrawn and animals sacrificed for heart studies after a further 2weeks. Preliminary zymogram analysis of blood samples revealed an absence of detectable ADA after 2weeks of PEG-ADA withdrawal (data not shown). This is consistent with data demonstrating that the half life of PEG-ADA in mice is ∼4days . For this study hearts were isolated from: ADA deficient (n=20); A1AR deficient (n=14); ADA/A1AR double deficient (n=21); wild-type littermate (n=24); and wild-type C57/BL/6J (n=17) mice.
2.2 Experimental protocol
Mice were anaesthetised with intraperitoneal tribromoethanol (Avertin), and hearts isolated and perfused on a Langendorff system at a pressure of 80mm Hg, as described previously [20,21]. Function was assessed via a fluid-filled intra-ventricular balloon connected to a pressure transducer [20,21]. After 15min stabilisation at intrinsic heart rate, hearts were paced at 7Hz and stabilised a further 15min. Baseline measurements were made, and hearts subjected to 20 or 25min of ischaemia and 45min reperfusion. Pacing was stopped on induction of ischaemia and resumed at 2min reperfusion [20,21]. We also examined effects of pharmacological inhibition of ADA with 5μM erythro-2-(2 hydroxy-3-nonyl)adenine (EHNA)–EHNA was infused into wild-type hearts for 10min prior to 20min ischaemia and for 10min of reperfusion. Responses to ischaemia–reperfusion were compared to those for untreated C57/Bl6 hearts.
Coronary venous effluent was collected prior to ischaemia (normoxic efflux), over the initial 2min of reperfusion (washout of extracellular purines accumulated during ischaemia) and for the remaining duration of reperfusion. Effluent samples were stored at − 80 °C until enzymatic analysis for lactate dehydrogenase (LDH) activity , or HPLC analysis of purine metabolites (adenosine, inosine, hypoxanthine, xanthine, uric acid) as outlined by us previously . Efflux of LDH efflux throughout reperfusion was expressed as U/g wet weight per min of ischaemia. Coronary venous purine efflux was expressed as nmoles/min/g wet weight.
To assess functional A1 and A2 receptor sensitivities in wild-type and ADA deficient hearts, we acquired concentration–response curves for bradycardia and vasodilatation in response to 2-chloroadenosine. This stable analogue was specifically chosen to avoid complications associated with differing catabolism of adenosine in ADA deficient vs. wild-type hearts. After stabilisation hearts were subjected to incremental concentrations of 2-chloroadenosine (∼2min at each concentration), and changes in heart rate or flow assessed . Concentration–response data for individual hearts were fit to a single receptor site logistic equation for absolute values for flow or heart rate, using the Statistica program (Statsoft, Tulsa, OK, USA):
where A is the pre-infusion rate or flow, B is the value at infinite dose, and EC50 is the concentration generating 50% maximal responses.
Data are presented as mean±SEM. Baseline data, EC50 values, final functional recoveries, and purine and LDH effluxes were analyzed via multi-factorial ANOVA. When significant differences were detected, a Newman-Keuls post-hoc test was employed for specific comparisons. A P<0.05 was considered indicative of significance.
3.1 Effects of ADA and A1AR deficiency on normoxic function and purine levels
There were no significant differences in baseline contractile function or coronary flow between groups (Table 1), though there was a tendency for higher flow (by ∼20%) in ADA deficient and ADA/A1AR double deficient hearts. On the other hand, ADA deficiency reduced intrinsic heart rate (prior to ventricular pacing). This bradycardia was abrogated by dual deletion of A1ARs (Table 1).
Values are means±SEM. Function was measured after 30 min stabilisation, except for heart rate which was measured after 15 min stabilisation (prior to pacing). WT, wild-type. *P<0.05 vs. WT Litter mates; †P<0.05 vs. ADA deficiency alone.
Infusion of 2-chloroadenosine generated a comparable negative chronotropic response (A1AR-mediated) in wild-type and ADA deficient hearts (Fig. 1A). Simultaneous deletion of A1ARs abolished this bradycardia, confirming A1AR dependence. Coronary vasodilatation (A2AAR-mediated in this model ) in response to 2-chloroadenosine was comparable in wild-type, ADA deficient and ADA/A1AR double deficient hearts (Fig. 1A). Table 2 provides mean pEC50 values for the different groups, confirming no effects of ADA deficiency on functional A1 or A2AR sensitivities.
Effects of ADA and ADA/A1AR double deficiency on (A) bradycardia (A1AR-mediated), and (B) coronary dilation (A2AR-mediated) with 2-chloroadenosine. Concentration–response curves were acquired in hearts from wild-type litter mate (n=7), ADA deficient (n=6) and ADA/A1AR deficient (n=6) mice. Values are means±SEM. *P<0.05 vs. baseline; †P<0.05 vs. wild-type.
Effects of ADA deficiency on sensitivities of A1AR-mediated bradycardia and A2AR-mediated coronary dilation
WT litter mates (n=7)
ADA deficient (n=6)
ADA/A1AR double deficient (n=6)
Values are means±SEM. Functional sensitivities are depicted as pEC50 values, calculated from concentration-response data (shown in Fig. 1). Values for A1AR-mediated bradycardia were not calculated for ADA/A1AR double deficient hearts due to absence of detectable responses. There were no significant differences between pEC50 values between groups.
Normoxic venous adenosine efflux was increased ∼10-fold by ADA deficiency, accompanied by a significant reduction in efflux of inosine and its catabolites (hypoxanthine, xanthine, uric acid) (Fig. 2). Nonetheless, considerable quantities of these latter metabolites were still generated prior to ischemia in ADA deficient hearts.
Effects of ADA deficiency on venous efflux of adenosine and inosine (Ino)+hypoxanthine (Hyp)+xanthine (Xan)+uric acid (UA), under (A) normoxic conditions, and (B) over the initial 2min of reperfusion following 25min ischaemia. Measures were made in hearts from wild-type (n=7) and ADA deficient (n=6) mice. Values are means±SEM. *P<0.05 vs. normoxic; †P<0.05 vs. wild-type.
3.2 Effects of ADA and A1AR deficiency on ischaemic responses
Deletion of ADA improved functional tolerance to 20 and 25min ischaemia (Fig. 3), without substantially modifying final recovery of coronary flow (Fig. 3C). Effects of ADA deficiency included an improvement in diastolic pressure (Fig. 3A), leading to enhanced pressure development (Fig. 3B). Washout of LDH (reflecting oncosis) was also reduced (Fig. 4). Treatment of wild-type hearts with the ADA inhibitor EHNA mimicked cardioprotection evident with ADA deficiency (Fig. 5). Beneficial effects of ADA deficiency were still evident in the absence of functional A1ARs (Figs. 3 and 4). However, simultaneous deletion of A1ARs abolished the effects of ADA deficiency on diastolic pressure (Fig. 3A). Deficiency of A1ARs alone significantly impaired recovery from ischaemia, specifically repressing pressure development without major effects on diastolic pressure (Fig. 3). Efflux of LDH was almost doubled by A1AR deficiency (Fig. 4).
Protective actions of the ADA inhibitor EHNA in hearts subjected to 20min ischaemia and 45min reperfusion. Shown are (from left to right) recoveries for left ventricular diastolic and developed pressures, and total post-ischaemic efflux of LDH, respectively. Data were acquired in wild-type C57/Bl6 mice untreated (n=9) or treated with 5μM EHNA (n=8). Values are means±SEM. *P<0.05 vs. wild-type.
Post-ischaemic LDH efflux following 20 or 25min ischaemia in hearts from wild-type litter mate (n=9 and 8 for 20 and 25min, respectively), ADA deficient (n=7 and 8 for 20 and 25min, respectively), A1AR deficient (n=7 for both 20 and 25min), and ADA/A1AR double deficient (n=8 and 7 for 20 and 25min, respectively) mice. Values are means±SEM. *P<0.05 vs. wild-type; †P<0.05 ADA or A1AR deficient vs. ADA/A1AR double deficient.
Effects of ADA and A1AR deficiency on post-ischaemic recoveries of (A) left ventricular diastolic pressure, (B) left ventricular developed pressure, and (C) coronary flow. Recoveries after 45 min reperfusion are shown for hearts from wild-type litter mate (n=9 and 8 for 20 and 25min, respectively), ADA deficient (n=7 and 8 for 20 and 25min, respectively), A1AR deficient (n=7 for both 20 and 25min), and ADA/A1AR double deficient (n=8 and 7 for 20 and 25min, respectively) mice. Values are means±SEM. *P<0.05 vs. wild-type; †P<0.05 ADA or A1AR deficient vs. ADA/A1AR double deficient.
Cardioprotection with ADA deficiency was associated with a 20-fold increase in adenosine accumulation during ischaemia/early reperfusion (reflected by washout over the initial 2min of reperfusion) (Fig. 2B). While this was paralleled by reduced accumulation of inosine and its catabolites, substantial quantities of the latter were still generated during ischaemia–reperfusion in ADA deficient hearts (Fig. 2).
Almost nothing is known regarding the cardiovascular impact of ADA deficiency, and the functions of ADA in the heart remain incompletely defined. The physiological role of ADA is likely to centre on its ability to regulate intra- and extracellular [adenosine], thereby modulating intrinsic AR agonism. However, reduced levels of ADA1 may also adversely impact on AR-mediated responses [7–11], and low ADA has been linked to myocardial ischaemic disorders [23,24]. In this study we show that absence of ADA1: enhances normoxic and ischaemic adenosine levels 10- to 20-fold; generates A1AR sensitive reductions in heart rate; fails to modify A1 and A2AR sensitivities despite elevated endogenous adenosine; and enhances myocardial tolerance to ischaemia–reperfusion via A1AR dependent and independent processes. Our data also confirm a non-redundant role for A1ARs in enhancing ischaemic tolerance, and reveal an important role for IMP vs. AMP catabolism in ischaemic purine loss.
4.1 Effects of ADA deficiency on cardiovascular function and A1/A2AR sensitivity
Phenotypic effects of ADA deficiency may provide clues regarding physiological functions of the protein, and of endogenously generated adenosine. Intrinsic heart rate fell with ADA deficiency (Table 1), an effect abolished by A1AR deletion (which alone does not alter rate). These data confirm A1AR-mediated bradycardia in response to elevated endogenous adenosine, yet indicate that basal adenosine levels are insufficient to activate bradycardia (since A1AR deficiency failed to alter rate). This is consistent with concentration–response data (Fig. 1, Table 2), confirming relatively low sensitivity of A1AR-mediated bradycardia. In terms of the vasculature, we observe an insignificant increase in coronary flow (∼20%) in ADA and ADA/A1AR double deficient hearts (Table 1), consistent with vascular sensitivity to endogenous adenosine, yet inconclusive regarding the potential role of endogenous adenosine in controlling coronary vascular tone (admittedly in a denervated preparation). The lack of a major flow change in the isolated organ is not due to a restricted capacity to dilate since normoxic flow (20–25ml/min/g) is ∼50% of the maximum achievable in this model .
Curiously, we detect no shifts in functional A1 or A2AR sensitivities with ADA deficiency (Fig. 1, Table 2). This was unexpected since elevated adenosine (which generates bradycardia) is predicted to down-regulate ARs. This, in turn, suggests absence of ADA could interfere with/limit the AR down-regulation normally associated with enhanced AR agonism [25,26]. Preserved AR sensitivity may be relevant to human ADA1 deficiency, since recent studies suggest enhanced AR activity is important in genesis of the disease phenotype . There may be a failure in AR down-regulation with ADA1 deficiency, negating desensitisation that normally limits detrimental effects of prolonged agonism. There is evidence that membrane-bound ADA1 does facilitate AR internalisation [13,14].
The lack of a shift in A1/A2AR sensitivity is also relevant to proposed non-enzymatic functions of ADA in modifying AR functionality [7,11]. Recent work indicates ADA1 interacts with A1ARs, facilitating ligand binding and effector activation in neuronal , smooth muscle , fibroblast , and kidney cells . Similar effects on A2AAR activity have been detected in lymphocytes and CHO cells . Whether this action is relevant in hearts is unknown: immunohistochemical data suggest membrane ADA complexes occur predominantly within exocrine glands and absorptive epithelia , and cell membrane ADA1 protein interactions may differ in rodents vs. other species [28,29]. Nonetheless, extracellular ADA activity is detected in mammalian myocardium . Absence of ADA1 protein in the current model (where the majority of the ADA1 gene is ablated) would negate any such effect, reducing AR sensitivity. Thus, absence of a shift in A1/A2AR sensitivities with ADA deficiency tends to argue against a function of ADA1 in directly regulating AR sensitivity in cardiovascular tissue.
4.2 Effects of ADA deficiency on ischaemic tolerance and purine catabolism
In terms of harnessing adenosinergic protection, it is appropriate to target ADA since phosphorylation of adenosine by adenosine kinase is normally inhibited during ischaemia/hypoxia . Prior studies employing pharmacological agents support enhanced ischaemic tolerance when ADA activity is acutely suppressed [4,32,33]. This may arise from enhanced adenosine and AR activation , purine salvage , reduced xanthine oxidase activity [5,34], or non-specific effects of inhibitors (including phosphodiesterase inhibition) [1,35]. Our data are the first to describe protection with selective absence of ADA (Figs. 3 and 4). We also show this protection is mimicked by ADA inhibition with EHNA (Fig. 5), which (whilst not conclusive) indicates the effects of EHNA are indeed due to ADA inhibition vs. other non-specific effects [1,35].
The ∼20-fold elevation in adenosine during ischaemia/reperfusion with ADA deficiency (Fig. 2) should significantly enhance AR agonism and protective signalling [36,37]. This is evidenced by elimination of the effects of ADA deficiency on diastolic pressure by A1AR deletion (Fig. 3). We previously found that enhanced A1AR activity selectively limits diastolic injury in this model [4,38]. While inhibition of protection in ADA deficient hearts by simultaneous A1AR deletion agrees with pharmacological data supporting AR-dependent protection with ADA inhibition , other processes may contribute to protection, including reduced xanthine oxidase derived ROS [5,34] and enhanced purine salvage . In terms of xanthine oxidase, our data show this reaction remains active and will still generate considerable ROS in ADA deficient hearts (Fig. 2). Moreover, these data demonstrate that IMP hydrolysis (the major alternate source of inosine) must be an important source of myocardial purines. This agrees with prior findings of Chen and Gueron , who employed pharmacological inhibition of ADA to delineate contributions of AMP vs. IMP to purine washout in rat hearts. Based on data for wild-type and ADA deficient hearts, we estimate that 40–50% of normoxic purine washout and ∼30% of post-ischaemic washout stems from IMP hydrolysis. Given differences in species, insult, and form of ADA inhibition, this estimate is in excellent agreement with a 23% contribution from IMP for anoxic rat hearts . These estimates, of course, assume only minor potential contributions from alternate reactions generating inosine. For example, it is feasible AMP deaminase could deaminate some adenosine .
4.3 Effects of A1AR deficiency on ischaemic tolerance
Several studies indicate AR antagonism does not modify intrinsic ischaemic tolerance [41–43], and there is even evidence A1AR blockade enhances tolerance . Alternatively, there is support for protection of ischaemic myocardium [21,45–48] by endogenous adenosine. Several explanations may account for varied effects of antagonists, the most likely involving mixed selectivity and potency of agents employed, and the fact that antagonists may exaggerate adenosine formation. The latter has been verified previously [49,50]. However, it is nonetheless important to note that widely observed cardioprotection with exogenous AR agonists [36–38,47] does imply sub-maximal AR agonism by endogenous adenosine. An alternate approach to identifying the roles of endogenous adenosine involves genetic deletion of proteins impacting on adenosine levels (e.g., ADA) or responses (e.g., A1ARs). We recently reported on the ability of A1AR deletion to limit myocardial ischaemic tolerance . This is confirmed here in hearts subjected to periods of ischaemia producing ∼40 and 50% reductions in functional outcome (Fig. 3).
Effects of A1AR deficiency are evident in terms of improved contractility and reduced oncosis, with little to no effect on diastolic pressure (Fig. 3). Thus, intrinsic A1AR agonism selectively improves contractility and limits cell death, whereas only exaggerated A1AR agonism (e.g., with ADA deficiency) further limits the elevation in diastolic pressure. This agrees with preferential effects of A1AR antagonism on systolic vs. diastolic pressure in ischaemic hearts . Abolition of the effects of ADA deficiency on post-ischaemic diastolic pressure by A1AR deletion confirms this form of protection is solely A1AR-dependent, consistent with lack of effect of A3 or A2AAR agonism on diastolic pressure .
Only one other group has assessed the impact of A1AR deficiency in ischaemic myocardium . Schulte and colleagues found no shift in ischaemic tolerance in the absence of A1ARs (in a different knockout model). Reasons for differences between their observations and the current data are unclear. However, one issue relates to ischaemic outcomes in their study–despite significantly prolonged 40min ischaemia they observed an almost immediate and stable 90% recovery of function (not differing from pre-ischaemia) . This unusual recovery indicates hearts sustained no injury despite a prolonged ischaemic insult, at odds with their prior data  and studies from other groups [see data reviewed in ]. These unusual outcomes raise concerns regarding the validity of the model employed.
4.4 Study limitations
Though gene knockout offers distinct advantages, a global drawback to all knockout studies is that prolonged absence of a gene product may lead to compensatory changes in other proteins/pathways. While unpredicted compensatory responses may provide important insights into cellular physiology and signalling [54,55], they also complicate interpretation of phenotypic outcomes. This drawback is, to some extent, limited in the current ADA deficient model since animals are maintained on exogenous ADA until 2weeks prior to experimentation . Moreover, we show no compensatory shifts in A1/A2AR sensitivity which would cloud interpretation. Nonetheless, it is important to be aware of possible compensatory responses or changes in any knockout model.
It is also worth noting that rodent tissues may differ from other species in terms of ADA interactions with other proteins. On the cell surface ADA1 associates with the CD26/dipeptidyl peptidase IV glycoprotein, forming an ecto-ADA which may regulate local adenosine levels and thus AR activation. However, though CD26/dipeptidyl peptidase IV protein is expressed in cardiac tissue , and ecto-ADA activity is evident in mammalian myocardium , evidence suggests CD26 does not bind ADA in rodent tissues .
Finally, in terms of concentration–response analysis in ADA deficient vs. wild-type hearts, there is a possibility of slightly overestimating functional sensitivity to applied agonists owing to higher baseline adenosine levels in these hearts. However, this will only impact on initial ‘threshold’ responses to the lowest agonist concentrations, and becomes increasingly unimportant at higher applied concentrations. As shown in Fig. 1, there are no substantial differences in chronotropic or vascular responses to the lowest agonist concentrations. It is therefore unlikely this effect will substantially shift the position of the concentration–response curves and thus the derived EC50 values.
This initial description of the cardiac phenotype in ADA deficient mice reveals A1AR-mediated bradycardia under baseline conditions (associated with ∼10-fold elevations in adenosine), yet no shift in functional A1 or A2AR sensitivities. While the basis for unchanged A1/A2AR sensitivities in the face of sustained elevations in endogenous agonist is not known, this may be relevant to the disease phenotype in human ADA1 deficiency . Data also reveal protection against ischaemia–reperfusion with ADA deficiency, involving modulation of post-ischaemic diastolic pressure, contractile recovery, and cell death. Protection against elevated diastolic pressure is A1AR-dependent, whereas other AR sub-types or processes must additionally contribute to protection against cell death/contractile impairment. Our data also support a major role for IMP hydrolysis in myocardial purine catabolism, and confirm impaired ischaemic tolerance in the absence of functional A1ARs.
This work was supported by a NIH grant (AI-43572) to M.R. Blackburn, and National Health and Medical Research Council of Australia grants (231419 and 326222) to J.P. Headrick. J.P. Headrick was the recipient of a fellowship from the National Heart Foundation of Australia.
↵1 The first two authors contributed equally to this study and manuscript.
(2000) The use of enzyme therapy to regulate the metabolic and phenotypic consequences of adenosine deaminase deficiency in mice. Differential impact on pulmonary and immunologic abnormalities. J Biol Chem 275:32114–32121.
(1983) Effect of adenosine deaminase inhibitors on the hearts functional and biochemical recovery from ischemia: a study using isolated hearts adapted to 31P-nuclear magnetic resonance. J Cardiovasc Pharmacol 5:1040–1047.
(1996) AMP degradation in the perfused rat heart during 2-deoxy-d-glucose perfusion and anoxia: Part II. The determination of the degradation pathways using an adenosine deaminase inhibitor. J Mol Cell Cardiol 28:2175–2182.
(1996) Inhibition of glycolysis and enhanced mechanical function of working rat hearts as a result of adenosine A1 receptor stimulation during reperfusion following ischaemia. Br J Pharmacol 118:355–363.
LauraWillems, Melissa E.Reichelt, Jose G.Molina, Chun-XiaoSun, Janci L.Chunn, Kevin J.Ashton, JurgenSchnermann, Michael R.Blackburn, John P.HeadrickCardiovasc Res(2006)71 (1):
79-87DOI: http://dx.doi.org/10.1016/j.cardiores.2006.03.006First published online: 1 July 2006 (9 pages)