93

Clinica Chimica Actu, 92 (1979) 93-100 @ Eisevier/North-Holland Biomedical Press

CCA 9998

ACTION OF PURINE NUCLEOSIDES ON THE RELEASE OF INTRACELLULAR ENZYMES FROM RAT LYMPHOCYTES

ARTHUR

W.G. COLE * and T. NORMAN

Department of Biochemistry, London; W6 8RF (U.K.) (Received

August

PALMER

Charing Cross Hospital

Medical School,

Fulham Palace Road,

31st, 1978)

Summary Rat lymphocytes incubated under hypoxic conditions in vitro show a timedependent release of intracellular enzymes. As reported previously, enzyme release (lactate dehydrogenase) is decreased by metabolites, notably ATP and glucose, that contribute towards lymphocyte energy metabolism. The action of purine nucleosides in relation to enzyme release was investigated. Inosine was shown to decrease significantly lactate dehydrogenase release, whereas adenosine exerted a supportive action only at concentrations less than 0.5 mmol/l. Inosine decreased enzyme efflux maximally at concentrations of at least 2-3 mmol/l. The mechanism of inosine action was deduced to be primarily ribose 5-phosphate formation and its subsequent metabolism by energy-yielding pathways. Inosine was presumed also to enter the purine ‘salvage pathway’ and thereby maintain intracellular adenine nucleotide pools.

Introduction Work notably in the laboratory of the late J.H. Wilkinson showed that the structural and functional integrity of cell membranes is dependent upon the energy content of the cell [l-6]. Irreversible loss of membrane integrity results in net release of intracellular enzymes. Enzyme release, thus, is an index of cellular integrity and energy metabolism and provides a readily measurable serum parameter to monitor damage to specific cell types and, as such, has wide clinical application. Enzyme release appears to be related to the intracellular adenine nucleotide concentration. Specifically, as the intracellular ATP concentration falls so enzyme release is increased [ 21. Incorporation of ATP or glucose (metabolism of which generates ATP intracellularly) in the medium is * To whom correspondence

should be addressed.

94

known to reduce markedly enzyme leakage. Conversely, metabolic inhibitors and respiratory uncouplers increase significantly the rate of enzyme release [4]. Therefore, study of the kinetics of release of intracellular enzymes, in this instance lactate dehydrogenase, provides a simple experimental model for the study of cellular damage. In hypoxia or ischaemia, the consequent decrease in the intracellular ATP concentration results in a concomitant increase in the intracellular concentrations of adenosine, inosine and hypoxanthine, the products of adenine nucleotide catabolism. These non-phosphorylated compounds diffuse readily across the cell membrane, resulting in net loss of purine base from the cell. In certain instances, it is possible to relate the extent of loss of purine base to the degree of hypoxic or ischaemic cell damage [ 71. Most cell types possess the complement of enzymes necessary to ‘salvage’ the purine nucleosides and reconvert them back to adenine nucleotides after an ischaemic or hypoxic episode, The purpose of the studies reported herein was to investigate whether non-phosphorylated precursors in adenine nucleotide biosynthesis via ‘salvage pathways’ could decrease enzyme release from hypoxic rat lymphocytes in vitro and to investigate the mechanism(s) by which this supportive function is achieved. Aside from providing a general insight into purine nucleoside metabolism in relation to hypoxic cell damage, the action of inosine and adenosine on lymphocytes is of interest in view of the fact that adenosine deaminase deficiency in man results in severe combined immunodeficiency [ 8,9]. We conclude that purine nucleosides, particularly inosine, exert a concentration-dependent protective action on lymphocyte metabolism. Enzyme release is decreased. However, it appears that action is mediated not by adenine nucleotide biosynthesis per se, but is attributed primarily to metabolism of the ribosome moiety of purine nucleosides via energy-yielding pathways [lo]. Materials and methods Chemicals and reagents ATP (adenosine 5-triphosphate) was obtained from the Boehringer Corporation (London) Ltd., Lewes, Sussex. Inosine and mannitol were purchased from the Sigma Chemical Co., Kingston-upon-Thames, Surrey and adenosine, adenine and hypoxanthine from the Aldrich Chemical Co. Ltd., Gillingham, Dorset. All other chemicals were of reagent grade. Krebs-Ringer bicarbonate pH 7.4 [ 1 l] was freshly prepared prior to each ‘experiment, Determination of lactate dehydrogenase activity Lactate dehydrogenase activities were determined at 25°C by the method of Wroblewski’and La Due [12] in a Gilford 240 Spectrophotometer. Activities are expressed as International Units (I.U.), one unit being equivalent to 1 pmol NADH oxidized/min at 25” C and pH 7.4. Rat lymphocyte preparatioits Lymphocytes were isolated from cervical, mesenteric, axillary and inguinal lymph nodes of adult Wistar mats. All steps in the isolation procedure were performed in Krebs-Ringer bicarbonate at 0-4°C. All vessels were of plastic con-

95

struction. Lymph nodes were washed and gently teased free of adhering fat before being gently forced through a strainer-filter to produce a crude lymphocyte suspension. The suspension was allowed to stand, on ice, in a conical tube for 5 min. The resulting cell suspension was decanted into a centrifuge tube thereby removing large aggregations of tissue which had settled out. The suspension was centrifuged at 800 rpm, 10 min. The pellet was washed once and then resuspended in Krebs-Ringer bicarbonate to provide the final lymphocyte preparation. Lymphocyte suspensions prepared in this way contained between 0.83 and 3.51 X 10’ cells/cmj. Cell viability, as assessed by eosin exclusion was routinely performed 1 h after cell preparation and found to be between 50 and 70%. Enzyme release from lymphocytes in hypoxia Lymphocyte suspension (1 ml) was added to Krebs-Ringer bicarbonate, pH 7.4 containing additions as specified in the text and equilibrated with 95%N2/ 5%COZ at 37°C (final volume 2.5 ml). The suspension was gassed gently with 95%N2/5%COZ, and the tubes sealed. Whenever samples were taken, tubes were gently gassed with 95%N2/5%C02 before resealing. At time intervals (up to eight hours), the suspensions were mixed gently and 0.3-ml samples removed and centrifuged (Beckman 152 microcentrifuge, 3 min). Supernatants were stored without loss of lactate dehydrogenase activity at 4” C for up to 4 h before enzyme assay. Total lymphocyte lactate dehydrogenase activities were determined in cell homogenates prepared by sonication (MSE Ultrasonic Disintegrator, 3 X 10 s at 0-4°C) of 0.5-ml samples of all incubations. Activities in supernatants were frequently expressed as percentages of total lactate dehydrogenase in cell homogenates. Results Effects of inosine, ATP and glucose on enzyme release In agreement with Wilkinson and Robinson [ 21, glucose and ATP decreased significantly the rate of lactate dehydrogenase release from lymphocytes in vitro. As shown in Fig. 1, inosine mimics the action of glucose or ATP.,to decrease enzyme release. As will be discussed shortly, the mode of action of ATP in modulating enzyme release from lymphocytes in vitro is problematical. In this regard, reference to Fig. 2 shows that enzyme release is decreased significantly by increasing the magnesium ion concentration. In other words, membrane integrity is influenced by factors other than lymphocyte energy metabolism. The specificity of purine nucleoside action in decreasing enzyme release from lymphocytes in vitro Given that inosine acts to maintain lymphocyte membrane integrity and thereby decreases enzyme release, the action of related purine analogues was investigated. Adenosine at concentrations up to 0.5 mmol/l decreased significantly lactate dehydrogenase release (Fig. 3). At concentrations of 1 mmol/l and above however, adenosine exerted a pronounced cytotoxic action, increas-

96

_ 6Ot

2

4

,

6 Incubation

6 time

(h)

I

4

2

5

I

fncubotion6 time (h?

Fig. 1. Effects of various additions on the release of lactate dehydrogenase from rat lymphocytes mcubated hypoxically at 37’C in Krebs-Ringer bicarbonate, pH 7.4: o, no addition: 4, in the presence of 10 mmolfl glucose; 0.8 mmol/l ATP; l, 2 mmol/l inoshre.

Fig. 2. Effects of magnesium on lactate dehydrogenase release from lymphocytes incubated hypoxically; 0, control (no addition); a. plus 10 mmolll glucose: 0, plus 10 mmol/l additional magnesium.

ing significantly enzyme release. Adenine and hypoxanthine only marginally decreased enzyme release even at high concentrations. The first presumed reaction in inosine metabolism, catalysedi by purine nucleoside phosphorylase, is phosphorolysis to yield ribose $-phosphate and hypoxanthine. Inorganic phosphate is a substrate in this reaction. To investigate whether inosine action on lymphocytes in vitro is dependent on the inorganic phosphate concentration, inosine (4 mmol/l) action was studied over a range of phosphate ‘cc~centrations. Addition of up to 1 mmol/l inorganic phosphate (final phosphate concentration 2.18 mmol/l) markedly decreased enzyme release. Higher phosphate concentrations complexed with available calcium in the Krebs-Ringer bicarbonate and were cytotoxic.

L

0

4

1 2 Adenosinc (mmol/l 1

Fi.k 3. Effects of varying concentrations of adenosine on lactate dehydrogenase release fr,cm lyrnphocytes incubated hypoxfcally at 37OC and pH 7.4 for 8 h (0). 7 h (A) and 5 h (a), respecttidy. Fig. 4. Effects on lactate.dehydrogenase release from lymphocytes fncubated hypoxic& at 87Oc and 7.4 of 5 mmol/l ghrcoae (A), 6 mapol/l inosine (0) and 5 mmol/l glucose slur 6 mmol/! inosfne (0). Controls (no addition) were examined shnultaneoualy.

PH

97

Experiments with various concentrations of mannitol (5-10 mmol/l) demonstrated no supportive effect on the lymphocytes. Given that inosine and glucose both decrease enzyme release in vitro, from a mechanistic standpoint it is pertinent to establish whether the modes of action of inosine and glucose are independent (additive) or otherwise. Fig. 4 shows clearly that their actions are in part additive even at an inosine concentration (5 mmol/l) selected on the basis that addition of further inosine per se would not increase significantly the supportive action of the purine nucleoside.

The concen tra tiondependency

of inosine action

As previously discussed, inosine exerts a protective action on lymphocytes in hypoxia. Fig. 5 s!nows a typical experiment designed to investigate the concentration dependency of inosine action.‘It is evident that inosine action is concentration dependent; the curves of enzyme release (7%)versus inosine tended consistently towards hyperbolicity. The hyperbolicity was taken to imply that the rate-limiting event in the mechanism of inosine action, a membrane carrier protein or an enzyme implicated in inosine metabolism, has a saturable binding site for inosine or a catabolite. Individual experiments designed to study the concentration dependency of inosine action, primarily due to significant variations in cell count and/or viability from one experiment to the next, were not directly comparable using the type of plot shown in Fig. 5. Accordingly, an alternative mode of data analysis and plotting was adopted. Values for percentage enzyme release were calculated. Enzyme release (%), r at incubation time, t: Xt -x0 rt =--H

where 3tt z enzyme release at incubation time, t; x0 = enzyme release at zero incubation time; H = total enzyme present as measured in cell homogenates. Plots of (xt -x,)/H versus incubation time, t are hyperbolic. Thus, log,,(xt x,)/(H) versus t plotsgive an approximate straight line relationship. When log,,@, --x0)/(H) = 2 then enzyme release is total (100%). Extrapolation of t 1oo, the time required for 100% enzyme release, is possible from plots of the type described above. The t 1oo of an incubation in the presence of inosine, glucose etc. divided by the tloo for control incubations (no addition) yields a ratio: t 1009addition/t,oo,

control

.

The ratio provides an index of the protective action of any specific addition. Ratios of one or less indicate that the addition increases enzyme leakage. Ratios above one indicate that the addition is protective; the higher the ratio, the greater the degree of protection. Compiled plots of the derived ratios versus addition concentration provide a finite means of assessing quantitatively concentration dependency. The concentration dependency of inosine action is shown in Fig. 6. Protective action was saturating at approximately 2 -3 mmol/l inosine at a ratio of 1.1-1.2. Analogous plots with ATP present showed no evidence of saturation at concentrations up to 10 mmol/l ATP.

.1

i; Qz

1 0

2

1

lnosine

I

1

4 (mmol/

I)

6

Fig. 5. The concentration dependency of inosine action on lactate dehydrogenase release from lymphocytes incubated hppoxically at 37’C and pH 7.4. Results from a typical experiment are shown in which lactate dehydrogenase release was determined after 6.5 h incubation. Fig. 6. The concentration dependency of inosine action in decreasing lactate dehydrogenase release from rat lymphocytesincubated hyDoxicaRy in vitro. The derived ratio tloo, inosine/t,oo. control (for explanation and derivation see text) plotted versus inosine concentration (mmol/l). The results are the compilation of four separate experiments.

Discussion The results obtained show clearly that purine nucleosides, particularly inosine, mimic the action of ATP or glucose in maintaining lymphocyte membrane integrity and, thereby decreasing lactate dehydrogenase release in hypoxia. The presumed mechanism of action of ATP has hitherto been thought to be via entry of the intact molecule into the cell to increase directly intracellular ATP concentrations [l-6]. This mechanism is unproven. ATP participates in reactions in vivo primarily as MgATP’-. The total magnesium concentration in Krebs-Ringer bicarbonate is 1.18 mmol/l. Addition of excess ATP“-, therefore, should complex with available magnesium to reduce free (Mg2’) to effectively zero. ATP4- and MgATPZ- will be the predominant nucleotide species present. Without documented evidence in direct support of entry of the intact ATP molecule into the lymphocyte, the action of ATP might be a function of specific or non-specific interaction of ATP with the cell membrane or by other mechanisms unrelated to ATP entry per se. Our own data show that decreased enzyme release is not an exclusive function of increased intracelluhu energy content, in that increased Mg2+ concentrations exerted a marked effect on membrane permeability. Despite these cautionary comments regarding the mode of action of extracellular ATP, there is no denying that intracellular energy metabolism is of central importance in maintaining membrane integrity. Glucose undoubtedly acts by providing an anaerobic glycolytic substrate. The specificity of purine nucleoside action supports the view that inosine (metabolized via purine nucleoside phosphorylase) and adenosine (metabolized

99

via adenosine kinase and adenosine deaminase) enter the lymphocyte via a membrane carrier protein(s) and are metabolized intracellularly. Adenine and hypoxanthin~ only mar~nally decreased enzyme release even at high concentrations, implying that membrane uptake is slow or metabolism via phosphoribosyl transferase reactions is limited by Sphosphoribosyl l-pyrophosphate (PRPP) availability. PRPP synthesis is ATP dependent and is presumably decreased in energydepleted lymphocytes in hypoxia. Adenosine decreased enzyme release at low concentrations. At high concentrations, adenosine was cytotoxic, a finding in line with its known action on hepatocytes in vitro [13]. In hepatocytes, it is thought the cytotoxicity results from the lack of stringent regulation of AMP synthesis via the adenosine kinase reaction, leading to overproduction of adenine nucleotides. Conceivably the same mechanism applies in reference to adenosine action on lymphocytes. Inosine was the most effective purine compound tested in terms of its supportive function in maintaining membrane integrity. Inosine via the purine nucleoside phosphorylase reaction, yields hypoxanthine and ribose l-phosphate. Ribose l-phosphate is converted to ribose 5-phosphate, which is either channelled into the pentose phosphate pathways and glycolysis or converted to PRPP via the ATP-dependent, ribosephosphate pyrophosphokinase reaction. PRPP and hypox~thine may be converted to adenine nucleotides via enzymes of the ‘salvage pathway’. It is’ concluded that inosine exerts its supportive action on lymphocyte metabolism primarily by providing carbohydrate (ribose 5-phosphate) to energy-yielding pathways, This conclusion is in accord with the findings of Nordeen and Young [lo] who studied purine nucleoside action on the functional integrity (RNA and protein biosynthesis) of rat thymic lymphocytes. Increased flux through the pentose phosphate and glycolytic pathways must presumably increase net PRPP synthesis permitting the channelling of hypoxanthine in part into adenine nucleotide biosynthesis. However, maintenance of intracellular adenine nucleotide concentrations is dependent upon and secondary to glycolytic metabolism of ribose 5-phosphate. The actions of glucose and inosine were shown to be additive implying either that inosine metabolism does not produce maximal glycolytic flux or that inosine action, at least in part, is by a pathway(s), possibly related to adenine nucleotide biosynthesis, not common to glucose metabolism. The supportive action of inosine was concentration dependent. Even allowing for the errors inherent in the determination of the saturating in vitro concentrations it is evident that inosine is effective in vitro at relatively low concentrations less than 3 mmol/l. The results obtained add weight to the view that inosine exerts a supportive action on tissues in ischaemia or hypoxia [ 14-161. Acknowledgements We wish to thank the Wellcome Trust for financial support of Simon Stewart is gratefully acknowledged, References 1 Wilkinson. 2 Wilkinson,

J.H. and Robinson, J.H. and Robinson,

J.M. (1974) J.M. (1974)

Nature 249.662 Clin. Chem. 20.1331

and the assistance

100 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Wilkinson, J.H., Robinson, J.M. and Johnson, K.P. (1975) Ann. Clin. Biochem. 12.58 Hallak. G.J. and Wilkinson, J.H. (1976) Clin. Chim. Acta 66, 251 Hallak. G.J. and Wilkinson, J.H. (1976) Clin. Chim. Acta 69,341 Hallak, G.J. and Wilkinson, J.H. (1977) Enzyme 22.361 Buhl. M.R. and JSrgenson. 5. (1975) Stand. J. Clin. Lab. Invest. 35.275 Knudsen, B.B. and Dissing. J. (1973) Clin. Genet. 4.344 Yount. J., Nichols, P.. Ocbs. H.D.. Hammar, S.P., Scott, C.R., Cben, S.H., Giblett, E.R. and Wedgwood, R.J. (1974) J. Pediatr. 84,173 Nordeen. 8.K. and Young. D.A. (1977) J. Biol. Chem. 252.5324 Krebs. H.A. and Hanseleit, K. (1932) 2. Pbysiol. Chem. 210.33 Wr5blewski. F. and La Due, J.S. (1955) Proc. Sot. Exp. Biol. Med. 90. 210 Ishii. K. and Green, H. (1973) J. Cell Sci. 13,429 Kingaby. R.O.. Lab. M.J. and Woollard, K.V. (1977) J. Physiol. 212.102 Bloom, D.S., Cole. A.W.G. and Palmer, T.N. (1977) J. Pbysiol. 270. 53P E.P.N., Perret. D.. D.M.G., Griffith% J.R., Hendry. W.F., Donogbue. Fernando, A.R., Armstrong, Ward. J.P. and Wickbam, J.E.A. (1976) Lancet i. 555