Computer simulation of ischemic rat heart purine metabolism. II. Model behavior MICHAEL
C. KOHN
AND
DAVID
GARFINKEL
The Moore School of Electrical Engineering Philadelphia, Pennsylvania 19104
and Department
of Medicine,
University
of Pennsylvania,
and FADH, (17). Using this observation, we may compute the rate of fatty acid activation, the first step in its oxidative mePhysiol.: Heart Circ. Physiol. l(4): H394-H399, 1977.-The tabolism, from the respiration rate, glycolytic flux, and behavior of a model for the partial depletion of adeninenucleo- rate of lactate production. Since the enzymes involved tides in the perfusedrat heart has been comparedfor ischemic in nucleotide degradation use only the cytosolic pools of
KOHN,MICHAELC.,ANDDAVIDGARFINKEL. Computersimulation of ischemic rat heart purine metabolism. II. Model behavior. Am. J. Physiol. 232(4):H39PH399, 1977or Am. J.
and high coronary flow anoxic conditions. The accumulation of noradrenaline in the interstitial fluid greatly activates adenylate cyclase ultimately resulting in the degradation of 11.02 pmollg dry wt of ATP to adenosine,inosine, and hypoxanthine in 30 min. The high coronary flow rate during anoxic perfusion promoteswashout of the noradrenaline from the interstitial fluid so that the hormone accumulatesto only one fifth of its highest level in ischemia.This results in only slight activation of adenylate cyclaseand in insignificant degradation of ATP in 2 min. The behavior of the model has been examined for two aerobic conditions-a transition from light to heavy work (2 min) and a transition from substrate-free to glucoseperfusion (12 min). In both casesadenylate cyclase was not activated above its basal activity, and insignificant depletion of adenine nucleotidesis predicted by the model. Adenosine hypothesis; adenine nucleotide degradation; cardiac noradrenaline
release; cyclic AMP;
vasodilation
acetyl-CoA producing 14.85 mol of NADH
the nucleotides
as substrates,
tissue
levels
of ADP
and
ATP were corrected for the amount of these nucleotides in the mitochondria using the algorithm of Klingenberg and Buchholz
(7). Most
of the cytosolic
ADP
is bound
to
F actin (26); the enzymes represented in this model react only with unbound ADP, which we have estimated by the position
pmol/g
of the equilibrium
binding
of ADP
to 2.87
dry wt of actin: ADP + actin II actin-ADP
complex
The equilibrium constant was considered as an adjustable parameter, since actin may have more than one molecular state. We used the position of the creatine kinase equilibrium: creatine-P
+ MgADP
+ H+ + creatine
+ MgATP
THE FIRST PAPER OF THIS SERIES (8)we described the construction of a model to stimulate the partial depletion of the adenine nucleotide pool in the ischemic rat heart. To be successful this model must be able to reproduce the depletion of adenine nucleotides during ischemia and also their preservation during short-term anoxia and aerobic metabolism. In order to examine the behavior of the model under each of these conditions, it is necessary to supply the simulation program with temporal profiles of cytosolic ADP, ATP, Mg2+, and H+ levels, because many processes not explicitly represented in this model (e.g., actomyosin ATPase activity) affect the levels of these chemicals. Since cytosolic MgPPi is produced by fatty acid activation, the temporal profile of this flux is also required. We will refer to the time derivative of these profiles (e.g., d[ATP]/& vs. time) as forcing functions. These profiles are read by the
to estimate the cytosolic pH. The enzyme phosphofructokinase uses MgATP as a substrate, while unchelated nucleotides (and other substances) are potent modifiers of the enzyme (13, 27). We computed the metabolically available Mg2+ level by finding that concentration of the metal ion which resulted in a distribution of chelated and unchelated nucleotide species which enabled us to reproduce the observed flux through phosphofructokinase using published nucleotide chelation and protonation equilibrium constants (16). Clearly, the ADP-binding to actin, creatine kinase, and nucleotide chelation equilibria are closely interrelated. We have, therefore, written a program to find the unbound ADP, Mg2+, and pH forcing function profiles by solving for all these quantities simultaneously by successive linear extrapolations starting from an initial guess. For the anoxic and aerobic cases no diffkulty was encountered, but for ischemia the computed cytosolic pH dropped from 6.94 initially to 5.92 at 30 min, which is
simulation program as tables-of the derivatives vs. time and are integrated along with the rest of the model thus reproducing the required concentrations of the above metabolites. Evaluation of forcing functions. On the average, each mole of endogenous fatty acid is oxidized to 8.48 mol of
below the range of values for intracellular pH reported for ischemic hearts (see Table 1). We have been able to prevent this great a fall in cytosolic pH by introducing a sigmoidal decline in the equilibrium constant for ADPbinding to actin during the interval from 12 to 16 min after the induction of ischemia. A sigmoidal function of
IN
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
IN
1. Intracellular
TABLE
Duration of Ischemia, min
0-t 6 16 26 36
O$ 1 6 15 60
pH in ischemic hearts
Extracellular Fluid pH
Intracellular
Intracellular
pH
7.25 7.05 6.87 6.84 6.79
7.04 6.96 6.87 6.83 6.80
6.84 6.72 6.62 6.61 6.58
7.05 6.96 6.66
7.05 6.99 6.87 6.63 6.47
6.72 6.70 6.50
6.46
O§
6.39
6.8 ~6.0
30 * Ref.
H395
ISCHEMIA
3.
t Ref.
15.
3: Ref.
19a.
8 Ref.
9.
pH
ml/min to an average rate of ca. 1.5 ml/min during the course of this experiment. We have, therefore, multiplied the rate constant for diffusion of small molecules between the perfusate and interstitial fluid by a factor of 0.1 to correct for the fact that the increased residence time of the perfusate in ischemic cardiac muscle results in local perfusate concentrations of metabolites which are greater than the concentrations in the bulk perfusate. This results in a greater rate of uptake of these metabolites by the tissue and hence slower washout from the interstitial fluid (8). The computed accumulation of noradrenaline in the interstitial fluid is shown in Fig. 1. The maximum concentration achieved in the extracellular fluid is 4.17 PM, a value slightly higher than the concentration for half-maximal activation of adenylate cyclase (25). This accumulation of noradrenaline activates the putative phosphoprotein phosphatase, while the fall in the MgATP level (due to decreasing ATP and Mg2+ levels) results in a reduced rate of phosphorylation of adenylate cyclase by the proposed protein kinase. Figure 1 shows the consequent increase in activity of the enzyme, which ultimately achieves 97% of full activity, and Fig. 2 shows the flux through the enzyme. After 20 min, the fall in cytosolic ATP and Mg2+ levels brings the MgATP concentration to 852 PM. This is below its effective K, for adenylate cyclase (1,780 PM including the effect of inhibition by unchelated ATP), and the computed flux through the enzyme chain resulting in loss of adenine nucleotides is greatly decreased. This is consistent with the experimental observation (14) that there is no significant decline in the nucleotide content of the heart between 20 and 30 min of ischemia. The flux through 3’ ,5’-cyclic AMP phosphodiesterase does not exactly match that for adenylate cyclase and
time was used since it produces a smooth, continuous change in the value of the equilibrium constant, K,,, and provides a time derivative, d&,/d& of zero at the beginning and end of the transition. We believe that this decline is not artifactual, because the final stages of contractile and respiratory failure occur during this period (14). We believe that our result is indicative of dramatic changes in conformation or the state of polymerization of the contractile proteins as suggested by Bing (1). We wish to emphasize that while qualitatively correct behavior is predicted without adjustment of this equilibrium constant, the numerical characteristics of the model itself require this adjustment in order to predict physiologically realistic results. One unexpected result of the above procedure was the calculation of a severe drop in cytosolic Mg2+ level during the course of ischemia to values well below those found for the aerobic hearts. While the parametric na- , 10 ture of this quantity suggests that its value not be taken 5 ‘Z too literally, we can make some qualitative interpretations. The fall in metabolically available Mg2+ may : reflect either export into the extracellular fluid, corn4 8 JZ partmentation in some subcellular organelle (e.g., the mitochondrion), binding to structural polymers, or pre$ c cipitation as an inorganic salt. Regardless of the cause, something fairly dramatic apparently is happening to w 6 ‘the ionic balances in the cardiac muscle, and this warrants further investigation. Purine metabolism during ischemia. Forcing functions for the ischemic rat heart were calculated based on the data of Neely et al. (14) and Rovetto et al. (21), and 30 min of ischemic metabolism were simulated by numerical integration of the model. In this preparation, a onerway ball valve in the aortic cannula restricted retrograde perfusion of the coronary arteries during diastole causing a 90% decrease in coronary flow, work E output, and respiration rates and leading to ventricular Z failure. The half-time for the initial drop in the respira0 0 tion rate is 1.5 min; this decrease probably reflects a drop in tissue PO, due to the rate of oxygen consumption FIG. exceeding its rate of delivery (5). We have, accordingly, adenylate set this value as the half-time for increasing the rate of ischemia.. noradrenaline release in the ischemic myocardium. The granules coronary flow rate dropped from the control value of 15 rate.
I
I
1
I
I
I
I
1 I 0 /-----/ //
2100
- 10 z 0
z
6
12
18
21
30
min
1. Noradrenaline in interstitial fluid cyclase in active (dephospho) form Rate of release of noradrenaline was made proportional to decrease in
(-) and percent (- - -) vs. duration from the presynaptic measured respiration
of of
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
M. C. KOHN
the resulting profiles for cyclic AMP and membranebound AMP are given in Fig. 3. Figure 2 shows the resulting flux through 5’.nucleotidase, and Fig. 4 shows the consequent depletion of the adenine nucleotide pool which is equal to the total amount of adenosine, inosine, and hypoxonthine produced. Given the large uncertainty in the sum of the adenine nucleotides (represented by the error bars for the data points in Fig. 4), we believe that our simulation adequately reproduces the observed degree of nucleotide degradation. We have included in Fig. 4 data points for the degradation of 2.0
’
I
I
I
I
I
I
I
//
I -0
-\\ 6
12
I
I
I
I
I
I
1
I 21
18
FIG. 2. Computed flux through adenylate nucleotidase (- - -> during ischemia.
10
,
I
I
adenine nucleotides observed for unperfused rat hearts (2) and for arrested rat hearts (4). It is apparent that nucleotide level changes in the ischemic myocardium are quite sensitive to the manner in which ischemia is induced. The computed fluxes through adenosine metabolizing enzymes in the muscle cytoplasm and vasculature are given in Figs. 5 and 6, respectively. According to our model, adenosine degradation occurs mainly in the vascular tissue, as suggested by the results of Rubio et al. (23), due to th e greater tissue capacity assumed for the permeases in the vasculature. This assumption is qualitatively supported by the data in Table 2; the relative levels of the various purines predicted by our model are similar to those observed in other experiments. Such a result would not have been obtained if the permease activity in the vasculature were not greatly increased and most of the adenosine deaminase activity were not assigned to this compartment. Again, it should be stressed that the biochemistry of ischemic cardiac tissue is apparently strongly dependent upon the experimental methods employed, and quantitative comparisons among these four experiments are not possible. Since the computed adenine nucleotide degradation rate becomes insignificant after 20 min of ischemia (Fig. 4), the slow washout of degradation products results in a decrease in the computed tissue purine level after this time (Table 2). This result was not observed in other experiments (Table 2), possibly because there is still a significant nucleotide degradation rate in these preparations atier 20 min. Olsson et al. (19), assuming Michaelis-Menten kinet-
30
cyclase (--)
I
AND D. GARFINKEL
and 5’-
1250
I
0
6
12
18
21
30
min 0.
/ 0
/
t 6
I
I
12
1 18
Y.-L.
, 24
30
mm FIG. 3. Computed cyclic AMP (-) G - -) vs. duration of ischemia.
and membrane-bound
AMP
FIG. 4. Quantity of adenine nucleotides degraded during ischemia computed on assumption that noradrenaline release is linked to decrease in respiration rate (- ) and that noradrenaline release commences only when new stable respiration rate is attained (- - -). Data from 4 experiments are from Ref. 14 (o), Ref. 21 (cI), Ref. 4 (o), and Ref. 2 (A).
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
IN
H397
ISCHEMIA
2. Unphosphorylated purine ischemic heart, pmollg dry wt
TABLE
Duration of Ischemia. min ’
5
Arrested Rat Heart?
Ado In0 Hx
10
Ado In0 Hx
20
Ado In0 Hx
0.170 0.340
0.100 0.100
0.185
0.050
0.550 2.000 0.450
levels in
~a~~~ryD$ Hea&
0.079 0.284 0.156
0.650 1.000 0.350
0.450 3.250 0.750
0.988 1.374 0.723
2.500 3.700 0.900
0.450 2.450 0.750
1.780 3.275 1.731
Ado 2.400 0.400 In0 4.300 3.400 Hx 1.200 0.800 Abbreviations: Ado, adenosine; Ino, inosine; Hx, thine. * Ref. 2. t Ref. 4. $ C. E. Jones, personal nication. 30
0
6
12
18
24
5. Computed flux through adenosine kinase (- - -) and adenin cardiac muscle cytoplasm during isosine deaminase (-) FIG.
0
I
6
I
I
12
FIG. 6. Computed flux through purine nucleoside phosphorylase ischemia.
1
min
I
18
I
0.058 2.802 1.528
hypoxancommu-
30
mln
I
This Work
I
24
I
30
adenosine deaminase (-) and (- - -) in vascular tissue during
its for both adenosine kinase and adenosine deaminase, compute that 68% of the adenosine imported by the heart muscle may be phosphorylated to AMP and the remainder deaminated to inosine. Our model for adenosine kinase includes competitive inhibition by unchelated ATP, a mechanism suggested by the data collected by these investigators. As a result, our model predicts that only 50% of the adenosine imported by the aerobic
heart is phosphorylated. There is experimental evidence that most of the (labeled) adenosine taken up by the heart is phosphorylated (6, 11). However, while the plasma membrane is impermeable to AMP, any inosine or hypoxanthine produced may be exported. The maximum rates of inosine and hypoxanthine production computed in this simulation are well within the maximal velocities for the respective permeases in the muscle and vascular tissue compartments. This suggests the possibility that the appearance of the predominance of phosphorylation over deamination of adenosine may be due to export of the deamination products and retention of the labeled AMP. One result of the decrease in the cytosolic pH during ischemia is that a greater fraction of the adenine nucleotides is unchelated than for the aerobic case. Our model predicts deamination to predominate over phosphorylation under these conditions, because adenosine kinase is more inhibited. Purine metabolism during anoxia. The forcing functions for this simulation were calculated from the data of Williamson (28). Since the residual oxygen is expected to be depleted after about 9 s of anoxic perfusion (lo), we have selected 0.1 min as the half-time for increasing the rate of noradrenaline release into. the interstitial fluid. (Even if a smaller value is chosen, the following results are not significantly affected.) The level of noradrenaline in the interstitial fluid predicted by the model is given in Fig. 7. The higher coronary flow rate in this experiment promotes washout of the noradrenaline from the interstitial fluid and the maximum calculated concentration of the hormone in this compartment is 0.88 PM, only 21% of its maximum concentration during ischemia. As a result, adenylate cyclase activity only rises from 11% to 14.7% of maximal activity (Fig. 7), and the computed cyclic AMP content rises from 1.84 to 1.94 nmol/g dry wt during the 2-min simulation. Membrane-bound AMP is essentially constant at about 1 nmol/g dry wt, and as a result, only 139 nmol/g dry wt of ATP is degraded to adenosine, inosine, and hypoxanthine, which is a negligibly small part (ca.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
H398
M. I
I
I
I
I
1
I
I
4 100
1
w v) a
80
"0
60
5
-w------
-ccc---------
---
b
01
0
I
I
0.4
I
I
I
0.8
I
1.2
I
I
1.6
I
IO
2.0
min FIG. 7. Noradrenaline in interstitial fluid (-> and percent of adenylate cyclase in active (dephospho) form (- - -) during COinduced anoxia. If this is extrapolated to 30 min, so as to be comparable with Fig. 1, noradrenaline peaks at about 3 min and then declines; activity of adenylate cyclase increases at longer times.
0.4%) of the total adenine nucleotide pool. In our model adenosine accumulates in the interstitial fluid to a concentration of 2.65 PM, which is sufEcient to promote vasodilation (19). During ischemia, in contrast, extracellular adenosine is computed to rise to 1.08 mM before utlimately falling to 30 LM. Such high concentrations would surely have a deleterious effect on the heart’s performance (19). Rovetto et al. (21) did find a large degree of depletion of adenine nucleotides during 30 min anoxic perfusion. Our model of noradrenaline release predicts that most of the noradrenaline will have been-washed out of the interstitial fluid before the decrease in total adenine nucleotides begins. We conclude that the failure of oxidative phosphorvlation results in a decrease in the ATP level, and that at this lower ATP level, less adenylate cyclase is phosphorylated to its inactive form. The resulting increase in adenylate cyclase activity would then, according to our model, lead to the production of substantial amounts of adenosine, inosine, and hypoxanthine and the concomitant depletion of the adenine nucleotide pool. Unfortunately, there were insufficient data from these experiments to generate the required forcing functions, so a quantitative comparison with 30 min of ischemia is not possible. The total adenine nucleotide level is constant during the first 2 min of anoxia andischemia. According to our model, the stability of this pool during anoxia is due to the fact that the- high coronary flow rate precludes buildup of noradrenaline to levels that would activate adenylate cyclase while the ATP level is still high. On the other hand, the respiration rate (and thus tissue Po2) falls much more slowly during ischemia than an-
C. KOHN
AND
D.
GARFINKEL
oxia. This delays noradrenaline release which, in turn, delays adenine nucleotide depletion. The proper function of the mechanism we have modeled may be inferred from findings such as that of L&se et al. (12) that there is local hypoxia in the aerobic dog heart in situ. The release of a small amount of adenosine would have a vasodilatory effect, which would tend to correct the deficiency. It seems likely, therefore, that the heart is well adapted to deal with short-term anoxia. Such is not the case in ischemia: release of adenosine will not relieve blockage of the coronary artery. Many of the deleterious effects of ischemia may be traced to a biochemical response which is ill-suited to deal with a structural abnormality. Purine metabolism in aerobic heart. Forcing func- . tions were computed for two aerobic conditions, the transitions from light to heavy work and from substrate-free to glucose perfusion, from the data of Opie et al. (20) and Safer and Williamson (24), respectively. The concentration of noradrenaline in the interstitial fluid in these aerobic hearts is constant at about 0.04 PM, which results in adenylate cyclase being activated to only ‘10% of its putative maximal activity during the course of these transitions. Consequently, only 77.5 and 83.7 nmol/g dry wt of ATP is predicted to be degraded after 2 min of perfusion during the work-jump and glucose-jump transitions, respectively (and 0.51 pmol/g dry wt after 12 min in the latter experiment, or about 2% of the total nucleotide pool). We believe that such an amount of degradation is small compared with the errors of measurement of the nucleotide levels themselves. The model predicts an interstitial adenosine concentration of 0.74 PM and 0.94 PM after 2 min perfusion in the work-jump and glucose-jump experiments respectively. The computed extracellular adenosine concentration after 12 min perfusion in the glucose-jump experiment is 1.27 PM. These values compare favorably with Rubio and Berne’s (22) observation that the concentration of adenosine in the pericardial fluid of the normal dog heart is 1.09 t 0.29 PM. The total tissue levels for adenosine, inosine, and hypoxanthine for these two simulations are summarized in Table 3. It should be noted that Gerl .ach et a1. (4) were unabl .e to detect any of these three purines in their aerobic rat hearts, even though they found other purines in amounts comparable to that found for adenosine by Degenring et al. (2). While our simulation predicts tissue levels of inosine and hypoxanthine similar to those found in rat hearts by Degenring et al. (2), the predicted tissue adenosine level is closer to that found for dog TABLE 3. Unphosphorylated purine aerobic heart, nmollg dry wt
levels in
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
H399
IN ISCHEMIA
hearts. This result suggests that either there is an addition .a1 adenosin .e-containi .ng compartment not included in our model, which does not seem very likely, or there is a defective rate law in our simulation. Since the maximal velocity found for “nucleoside permease” is less than that for adenosine kinase, the low K, for adenosine in the adenosine kinase rate law (0.909 PM) will control the tissue level of adenosine. If the maximal velocity of the permease were actually higher than that of adenosine kinase, the adenosine tissue level would -be
near the (hi .gher> K, for adenosine deaminase (‘43 PM), because the uptake of adenosine would be faster th .an its phosphorylation. This would result in adenosine levels close to those found for the rat heart (2). Since the kinetics of membrane transports are often controlled in a very complex way, we believe a reexamination of the kinetics of this process is indicated. Received for publication
15 December 1975.
REFERENCES 1. BING, R. J. Cardiac metabolism. Physiol. Rev. 45: 171-213, 1965. 2. DEGENRING, F. H., R. RUBIO, AND R. M. BERNE. Adenine nucleotide metabolism during cardiac hypertrophy and ischemia in rats. J. MOL. CeZZ.Cardiol. 7: 105-113, 1975. 3. EFFROS, R. M., B HAIDER, P. ETTINGER, H. OLDEWURTEL, S. AHMED, AND T. REGAN. In uiuo response of myocardial cell pH to changes of plasma pH and ischemia (Abstract). Federation Proc. 33: 396, 1974. 4. GERLACH, E., B. DEUTICKE, AND R. H. DREISBACH. Der Nucleotidabbau im Herzmuskel bei Sauerstoffmangel und seine moglithe Bedeutung fur die Coronardurchblutung. Nuturwissenschufien 6: 228-229, 1963. 5. HADDY, F. J., AND J. B. SCOTT. Metabolic factors in peripheral circulatory regulation. Federation Proc. 34: 2006-2011, 1975. 6. JACOB, M. I., AND R. M. BERNE. Metabolism of purine derivatives by the isolated cat heart. Am. J. Physiol. 198: 322-326, 1960. 7. KLINGENBERG, M., AND M. BUCHHOLZ. Localization of the glycerolphosphate dehydrogenase in the outer phase of the mitochondrial membrane. European J. Biochem. 13: 247-252, 1970. 8. KOHN, M. C., AND D. GARFINKEL. Computer simulation of ischemic rat heart purine metabolism. I. Model construction. Am. J. Physiol. 232: H386-H393, 1977. 9. KRUG, A. Alterations in myocardial hydrogen ion concentration after temporary coronary occlusions: a sign of irreversible cell damage. Am. J. Curdiol. 36: 214-217, 1975. 10. K~BLER, W., AND P. G. SPIEKERMANN. Regulation of glycolysis in the ischemic and the anoxic myocardium. J. MOL. CeZZ. Curdial. 1: 351-377, 1970. 1.1. LIU, M. S., AND H. FEINBERG. Incorporation of adenosine-8-14C and inosine-8-l*C into rabbit heart adenine nucleotides. Am. J. PhysioZ. 220: 1242-1248, 1971. 12. L&SE, B., S. SCHUCHHARDT, N. NIEDERLE, AND H. BENZING. The histogram of local oxygen pressure (Po2) in the dog myocardium and the PoZ behavior during transitory changes of oxygen administration. Aduun. Exptl. Med. BioZ. 37A: 535-540, 1973. 13. MANSOUR, T. E. Studies on heart phosphofructokinase: purification, inhibition, and activation. J. BioZ. Chem. 238: 2285-2292, 1963. 14. NEELY, J, R., M. J. ROVETTO, J. T. WHITMER, AND H. E. MORGAN. Effects of ischemia on function and metabolism of the isolated working rat heart. Am. J. Physiol. 225: 651-658,1973. 15. NEELY, J. R., J. T. WHITMER, AND M. J. ROVETTO. Effect of coronary blood flow on glycolytic flux and intracellular pH in isolated rat hearts. CircuZution Res. 37: 733-741, 1975. 16. NIHEI, T., L. NODA, AND M. F. MORALES. Kinetic properties and
equilibrium constant of the adenosine triphosphate-creatine transphosphorylase catalyzed reaction. J. BioZ. Chem. 236: 32033213, 1961. 17. OLSON, R. E., AND R. J. HOESCHEN. Utilization of endogenous lipid by the isolated perfused rat heart. Biochem. J. 103: 796-801, 1967. 18. OLSSON, R. A., AND J. A. SNOW. Tissue:blood purine nucleoside gradients in canine myocardium (Abstract). Federation Proc. 29: 586, 1970. 19. OLSSON, R. A., J. A. SNOW, M. K. GENTRY, AND G. P. FRICK. Adenosine uptake by canine heart. CircuZution Res. 31: 767-778, 1972. lga.OPrE, L. H. Effects of regional ischemia on the metabolism of glucose and fatty acids. Circulation Res. 38, Suppl. 1, 1-52-1-68, 1976. 20. OPIE, L. H., K. R. L. MANSFORD, AND P. OWEN. Effects of increased work on glycolysis and adenine nucleotides in the perfused heart of normal and diabetic rats. Biochem. J. 124: 475490, 1971. 21. ROVETTO, M. J., J. T. WHITMER, AND J. R. NEELY. Comparison of the effects of anoxia and whole heart ischemia on carbohydrate utilization in isolated working rat hearts. CircuZution Res. 32: 699-711, 1973. 22. RUBIO, R., AND R. M. BERNE. Release of adenosine by the normal myocardium in dogs and its relationship to the regulation of coronary resistance. CircuZution Res. 25: 407-415, 1969. 23. RUBIO, R., V. T. WIEDMEIER, AND R. M. BERNE. Nucleoside phosphorylase: localization and role in the myocardial distribution of purines. Am. J. Physiol. 222: 550-555, 1972. 24. SAFER, B., and J. R. WILLIAMSON. Mitochondrial-cytosolic interactions in perfused rat heart. Role of coupled transamination in repletion of citric acid cycle intermediates. J. BioZ. Chem. 248: 2570-2579, 1973. 25. SEVERSON, D. L., G. I. DRUMMOND, AND P. V. SULAKHE. Adenylate cyclase in skeletal muscle. J. BioZ. Chem. 247: 2949-2958, 1972. 26. SZENT-GY~RGYI, A. G., AND G. PRIOR. Exchange of adenosine diphosphate bound to actin in superprecipitated actomyosin and contracted myofibrils. J. MOL. BioZ. 15: 515-538, 1966. 27. UI, M. Multiple inhibitor sites for ATP on muscle phosphofructokinase as influenced by a change of pH: a computer analysis of “non-linear” kinetic data. Biochim. Biophys. Actu 159: 50-63, 1968. 28. WILLIAMSON, J. R. Glycolytic control mechanisms. II. Kinetics of intermediate changes during the aerobic-anoxic transition in perfused rat heart. J. BioZ. Chem. 241: 5026-5036, 1966.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
C. KOHN
AND
DAVID
GARFINKEL
The Moore School of Electrical Engineering Philadelphia, Pennsylvania 19104
and Department
of Medicine,
University
of Pennsylvania,
and FADH, (17). Using this observation, we may compute the rate of fatty acid activation, the first step in its oxidative mePhysiol.: Heart Circ. Physiol. l(4): H394-H399, 1977.-The tabolism, from the respiration rate, glycolytic flux, and behavior of a model for the partial depletion of adeninenucleo- rate of lactate production. Since the enzymes involved tides in the perfusedrat heart has been comparedfor ischemic in nucleotide degradation use only the cytosolic pools of
KOHN,MICHAELC.,ANDDAVIDGARFINKEL. Computersimulation of ischemic rat heart purine metabolism. II. Model behavior. Am. J. Physiol. 232(4):H39PH399, 1977or Am. J.
and high coronary flow anoxic conditions. The accumulation of noradrenaline in the interstitial fluid greatly activates adenylate cyclase ultimately resulting in the degradation of 11.02 pmollg dry wt of ATP to adenosine,inosine, and hypoxanthine in 30 min. The high coronary flow rate during anoxic perfusion promoteswashout of the noradrenaline from the interstitial fluid so that the hormone accumulatesto only one fifth of its highest level in ischemia.This results in only slight activation of adenylate cyclaseand in insignificant degradation of ATP in 2 min. The behavior of the model has been examined for two aerobic conditions-a transition from light to heavy work (2 min) and a transition from substrate-free to glucoseperfusion (12 min). In both casesadenylate cyclase was not activated above its basal activity, and insignificant depletion of adenine nucleotidesis predicted by the model. Adenosine hypothesis; adenine nucleotide degradation; cardiac noradrenaline
release; cyclic AMP;
vasodilation
acetyl-CoA producing 14.85 mol of NADH
the nucleotides
as substrates,
tissue
levels
of ADP
and
ATP were corrected for the amount of these nucleotides in the mitochondria using the algorithm of Klingenberg and Buchholz
(7). Most
of the cytosolic
ADP
is bound
to
F actin (26); the enzymes represented in this model react only with unbound ADP, which we have estimated by the position
pmol/g
of the equilibrium
binding
of ADP
to 2.87
dry wt of actin: ADP + actin II actin-ADP
complex
The equilibrium constant was considered as an adjustable parameter, since actin may have more than one molecular state. We used the position of the creatine kinase equilibrium: creatine-P
+ MgADP
+ H+ + creatine
+ MgATP
THE FIRST PAPER OF THIS SERIES (8)we described the construction of a model to stimulate the partial depletion of the adenine nucleotide pool in the ischemic rat heart. To be successful this model must be able to reproduce the depletion of adenine nucleotides during ischemia and also their preservation during short-term anoxia and aerobic metabolism. In order to examine the behavior of the model under each of these conditions, it is necessary to supply the simulation program with temporal profiles of cytosolic ADP, ATP, Mg2+, and H+ levels, because many processes not explicitly represented in this model (e.g., actomyosin ATPase activity) affect the levels of these chemicals. Since cytosolic MgPPi is produced by fatty acid activation, the temporal profile of this flux is also required. We will refer to the time derivative of these profiles (e.g., d[ATP]/& vs. time) as forcing functions. These profiles are read by the
to estimate the cytosolic pH. The enzyme phosphofructokinase uses MgATP as a substrate, while unchelated nucleotides (and other substances) are potent modifiers of the enzyme (13, 27). We computed the metabolically available Mg2+ level by finding that concentration of the metal ion which resulted in a distribution of chelated and unchelated nucleotide species which enabled us to reproduce the observed flux through phosphofructokinase using published nucleotide chelation and protonation equilibrium constants (16). Clearly, the ADP-binding to actin, creatine kinase, and nucleotide chelation equilibria are closely interrelated. We have, therefore, written a program to find the unbound ADP, Mg2+, and pH forcing function profiles by solving for all these quantities simultaneously by successive linear extrapolations starting from an initial guess. For the anoxic and aerobic cases no diffkulty was encountered, but for ischemia the computed cytosolic pH dropped from 6.94 initially to 5.92 at 30 min, which is
simulation program as tables-of the derivatives vs. time and are integrated along with the rest of the model thus reproducing the required concentrations of the above metabolites. Evaluation of forcing functions. On the average, each mole of endogenous fatty acid is oxidized to 8.48 mol of
below the range of values for intracellular pH reported for ischemic hearts (see Table 1). We have been able to prevent this great a fall in cytosolic pH by introducing a sigmoidal decline in the equilibrium constant for ADPbinding to actin during the interval from 12 to 16 min after the induction of ischemia. A sigmoidal function of
IN
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
IN
1. Intracellular
TABLE
Duration of Ischemia, min
0-t 6 16 26 36
O$ 1 6 15 60
pH in ischemic hearts
Extracellular Fluid pH
Intracellular
Intracellular
pH
7.25 7.05 6.87 6.84 6.79
7.04 6.96 6.87 6.83 6.80
6.84 6.72 6.62 6.61 6.58
7.05 6.96 6.66
7.05 6.99 6.87 6.63 6.47
6.72 6.70 6.50
6.46
O§
6.39
6.8 ~6.0
30 * Ref.
H395
ISCHEMIA
3.
t Ref.
15.
3: Ref.
19a.
8 Ref.
9.
pH
ml/min to an average rate of ca. 1.5 ml/min during the course of this experiment. We have, therefore, multiplied the rate constant for diffusion of small molecules between the perfusate and interstitial fluid by a factor of 0.1 to correct for the fact that the increased residence time of the perfusate in ischemic cardiac muscle results in local perfusate concentrations of metabolites which are greater than the concentrations in the bulk perfusate. This results in a greater rate of uptake of these metabolites by the tissue and hence slower washout from the interstitial fluid (8). The computed accumulation of noradrenaline in the interstitial fluid is shown in Fig. 1. The maximum concentration achieved in the extracellular fluid is 4.17 PM, a value slightly higher than the concentration for half-maximal activation of adenylate cyclase (25). This accumulation of noradrenaline activates the putative phosphoprotein phosphatase, while the fall in the MgATP level (due to decreasing ATP and Mg2+ levels) results in a reduced rate of phosphorylation of adenylate cyclase by the proposed protein kinase. Figure 1 shows the consequent increase in activity of the enzyme, which ultimately achieves 97% of full activity, and Fig. 2 shows the flux through the enzyme. After 20 min, the fall in cytosolic ATP and Mg2+ levels brings the MgATP concentration to 852 PM. This is below its effective K, for adenylate cyclase (1,780 PM including the effect of inhibition by unchelated ATP), and the computed flux through the enzyme chain resulting in loss of adenine nucleotides is greatly decreased. This is consistent with the experimental observation (14) that there is no significant decline in the nucleotide content of the heart between 20 and 30 min of ischemia. The flux through 3’ ,5’-cyclic AMP phosphodiesterase does not exactly match that for adenylate cyclase and
time was used since it produces a smooth, continuous change in the value of the equilibrium constant, K,,, and provides a time derivative, d&,/d& of zero at the beginning and end of the transition. We believe that this decline is not artifactual, because the final stages of contractile and respiratory failure occur during this period (14). We believe that our result is indicative of dramatic changes in conformation or the state of polymerization of the contractile proteins as suggested by Bing (1). We wish to emphasize that while qualitatively correct behavior is predicted without adjustment of this equilibrium constant, the numerical characteristics of the model itself require this adjustment in order to predict physiologically realistic results. One unexpected result of the above procedure was the calculation of a severe drop in cytosolic Mg2+ level during the course of ischemia to values well below those found for the aerobic hearts. While the parametric na- , 10 ture of this quantity suggests that its value not be taken 5 ‘Z too literally, we can make some qualitative interpretations. The fall in metabolically available Mg2+ may : reflect either export into the extracellular fluid, corn4 8 JZ partmentation in some subcellular organelle (e.g., the mitochondrion), binding to structural polymers, or pre$ c cipitation as an inorganic salt. Regardless of the cause, something fairly dramatic apparently is happening to w 6 ‘the ionic balances in the cardiac muscle, and this warrants further investigation. Purine metabolism during ischemia. Forcing functions for the ischemic rat heart were calculated based on the data of Neely et al. (14) and Rovetto et al. (21), and 30 min of ischemic metabolism were simulated by numerical integration of the model. In this preparation, a onerway ball valve in the aortic cannula restricted retrograde perfusion of the coronary arteries during diastole causing a 90% decrease in coronary flow, work E output, and respiration rates and leading to ventricular Z failure. The half-time for the initial drop in the respira0 0 tion rate is 1.5 min; this decrease probably reflects a drop in tissue PO, due to the rate of oxygen consumption FIG. exceeding its rate of delivery (5). We have, accordingly, adenylate set this value as the half-time for increasing the rate of ischemia.. noradrenaline release in the ischemic myocardium. The granules coronary flow rate dropped from the control value of 15 rate.
I
I
1
I
I
I
I
1 I 0 /-----/ //
2100
- 10 z 0
z
6
12
18
21
30
min
1. Noradrenaline in interstitial fluid cyclase in active (dephospho) form Rate of release of noradrenaline was made proportional to decrease in
(-) and percent (- - -) vs. duration from the presynaptic measured respiration
of of
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
M. C. KOHN
the resulting profiles for cyclic AMP and membranebound AMP are given in Fig. 3. Figure 2 shows the resulting flux through 5’.nucleotidase, and Fig. 4 shows the consequent depletion of the adenine nucleotide pool which is equal to the total amount of adenosine, inosine, and hypoxonthine produced. Given the large uncertainty in the sum of the adenine nucleotides (represented by the error bars for the data points in Fig. 4), we believe that our simulation adequately reproduces the observed degree of nucleotide degradation. We have included in Fig. 4 data points for the degradation of 2.0
’
I
I
I
I
I
I
I
//
I -0
-\\ 6
12
I
I
I
I
I
I
1
I 21
18
FIG. 2. Computed flux through adenylate nucleotidase (- - -> during ischemia.
10
,
I
I
adenine nucleotides observed for unperfused rat hearts (2) and for arrested rat hearts (4). It is apparent that nucleotide level changes in the ischemic myocardium are quite sensitive to the manner in which ischemia is induced. The computed fluxes through adenosine metabolizing enzymes in the muscle cytoplasm and vasculature are given in Figs. 5 and 6, respectively. According to our model, adenosine degradation occurs mainly in the vascular tissue, as suggested by the results of Rubio et al. (23), due to th e greater tissue capacity assumed for the permeases in the vasculature. This assumption is qualitatively supported by the data in Table 2; the relative levels of the various purines predicted by our model are similar to those observed in other experiments. Such a result would not have been obtained if the permease activity in the vasculature were not greatly increased and most of the adenosine deaminase activity were not assigned to this compartment. Again, it should be stressed that the biochemistry of ischemic cardiac tissue is apparently strongly dependent upon the experimental methods employed, and quantitative comparisons among these four experiments are not possible. Since the computed adenine nucleotide degradation rate becomes insignificant after 20 min of ischemia (Fig. 4), the slow washout of degradation products results in a decrease in the computed tissue purine level after this time (Table 2). This result was not observed in other experiments (Table 2), possibly because there is still a significant nucleotide degradation rate in these preparations atier 20 min. Olsson et al. (19), assuming Michaelis-Menten kinet-
30
cyclase (--)
I
AND D. GARFINKEL
and 5’-
1250
I
0
6
12
18
21
30
min 0.
/ 0
/
t 6
I
I
12
1 18
Y.-L.
, 24
30
mm FIG. 3. Computed cyclic AMP (-) G - -) vs. duration of ischemia.
and membrane-bound
AMP
FIG. 4. Quantity of adenine nucleotides degraded during ischemia computed on assumption that noradrenaline release is linked to decrease in respiration rate (- ) and that noradrenaline release commences only when new stable respiration rate is attained (- - -). Data from 4 experiments are from Ref. 14 (o), Ref. 21 (cI), Ref. 4 (o), and Ref. 2 (A).
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
IN
H397
ISCHEMIA
2. Unphosphorylated purine ischemic heart, pmollg dry wt
TABLE
Duration of Ischemia. min ’
5
Arrested Rat Heart?
Ado In0 Hx
10
Ado In0 Hx
20
Ado In0 Hx
0.170 0.340
0.100 0.100
0.185
0.050
0.550 2.000 0.450
levels in
~a~~~ryD$ Hea&
0.079 0.284 0.156
0.650 1.000 0.350
0.450 3.250 0.750
0.988 1.374 0.723
2.500 3.700 0.900
0.450 2.450 0.750
1.780 3.275 1.731
Ado 2.400 0.400 In0 4.300 3.400 Hx 1.200 0.800 Abbreviations: Ado, adenosine; Ino, inosine; Hx, thine. * Ref. 2. t Ref. 4. $ C. E. Jones, personal nication. 30
0
6
12
18
24
5. Computed flux through adenosine kinase (- - -) and adenin cardiac muscle cytoplasm during isosine deaminase (-) FIG.
0
I
6
I
I
12
FIG. 6. Computed flux through purine nucleoside phosphorylase ischemia.
1
min
I
18
I
0.058 2.802 1.528
hypoxancommu-
30
mln
I
This Work
I
24
I
30
adenosine deaminase (-) and (- - -) in vascular tissue during
its for both adenosine kinase and adenosine deaminase, compute that 68% of the adenosine imported by the heart muscle may be phosphorylated to AMP and the remainder deaminated to inosine. Our model for adenosine kinase includes competitive inhibition by unchelated ATP, a mechanism suggested by the data collected by these investigators. As a result, our model predicts that only 50% of the adenosine imported by the aerobic
heart is phosphorylated. There is experimental evidence that most of the (labeled) adenosine taken up by the heart is phosphorylated (6, 11). However, while the plasma membrane is impermeable to AMP, any inosine or hypoxanthine produced may be exported. The maximum rates of inosine and hypoxanthine production computed in this simulation are well within the maximal velocities for the respective permeases in the muscle and vascular tissue compartments. This suggests the possibility that the appearance of the predominance of phosphorylation over deamination of adenosine may be due to export of the deamination products and retention of the labeled AMP. One result of the decrease in the cytosolic pH during ischemia is that a greater fraction of the adenine nucleotides is unchelated than for the aerobic case. Our model predicts deamination to predominate over phosphorylation under these conditions, because adenosine kinase is more inhibited. Purine metabolism during anoxia. The forcing functions for this simulation were calculated from the data of Williamson (28). Since the residual oxygen is expected to be depleted after about 9 s of anoxic perfusion (lo), we have selected 0.1 min as the half-time for increasing the rate of noradrenaline release into. the interstitial fluid. (Even if a smaller value is chosen, the following results are not significantly affected.) The level of noradrenaline in the interstitial fluid predicted by the model is given in Fig. 7. The higher coronary flow rate in this experiment promotes washout of the noradrenaline from the interstitial fluid and the maximum calculated concentration of the hormone in this compartment is 0.88 PM, only 21% of its maximum concentration during ischemia. As a result, adenylate cyclase activity only rises from 11% to 14.7% of maximal activity (Fig. 7), and the computed cyclic AMP content rises from 1.84 to 1.94 nmol/g dry wt during the 2-min simulation. Membrane-bound AMP is essentially constant at about 1 nmol/g dry wt, and as a result, only 139 nmol/g dry wt of ATP is degraded to adenosine, inosine, and hypoxanthine, which is a negligibly small part (ca.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
H398
M. I
I
I
I
I
1
I
I
4 100
1
w v) a
80
"0
60
5
-w------
-ccc---------
---
b
01
0
I
I
0.4
I
I
I
0.8
I
1.2
I
I
1.6
I
IO
2.0
min FIG. 7. Noradrenaline in interstitial fluid (-> and percent of adenylate cyclase in active (dephospho) form (- - -) during COinduced anoxia. If this is extrapolated to 30 min, so as to be comparable with Fig. 1, noradrenaline peaks at about 3 min and then declines; activity of adenylate cyclase increases at longer times.
0.4%) of the total adenine nucleotide pool. In our model adenosine accumulates in the interstitial fluid to a concentration of 2.65 PM, which is sufEcient to promote vasodilation (19). During ischemia, in contrast, extracellular adenosine is computed to rise to 1.08 mM before utlimately falling to 30 LM. Such high concentrations would surely have a deleterious effect on the heart’s performance (19). Rovetto et al. (21) did find a large degree of depletion of adenine nucleotides during 30 min anoxic perfusion. Our model of noradrenaline release predicts that most of the noradrenaline will have been-washed out of the interstitial fluid before the decrease in total adenine nucleotides begins. We conclude that the failure of oxidative phosphorvlation results in a decrease in the ATP level, and that at this lower ATP level, less adenylate cyclase is phosphorylated to its inactive form. The resulting increase in adenylate cyclase activity would then, according to our model, lead to the production of substantial amounts of adenosine, inosine, and hypoxanthine and the concomitant depletion of the adenine nucleotide pool. Unfortunately, there were insufficient data from these experiments to generate the required forcing functions, so a quantitative comparison with 30 min of ischemia is not possible. The total adenine nucleotide level is constant during the first 2 min of anoxia andischemia. According to our model, the stability of this pool during anoxia is due to the fact that the- high coronary flow rate precludes buildup of noradrenaline to levels that would activate adenylate cyclase while the ATP level is still high. On the other hand, the respiration rate (and thus tissue Po2) falls much more slowly during ischemia than an-
C. KOHN
AND
D.
GARFINKEL
oxia. This delays noradrenaline release which, in turn, delays adenine nucleotide depletion. The proper function of the mechanism we have modeled may be inferred from findings such as that of L&se et al. (12) that there is local hypoxia in the aerobic dog heart in situ. The release of a small amount of adenosine would have a vasodilatory effect, which would tend to correct the deficiency. It seems likely, therefore, that the heart is well adapted to deal with short-term anoxia. Such is not the case in ischemia: release of adenosine will not relieve blockage of the coronary artery. Many of the deleterious effects of ischemia may be traced to a biochemical response which is ill-suited to deal with a structural abnormality. Purine metabolism in aerobic heart. Forcing func- . tions were computed for two aerobic conditions, the transitions from light to heavy work and from substrate-free to glucose perfusion, from the data of Opie et al. (20) and Safer and Williamson (24), respectively. The concentration of noradrenaline in the interstitial fluid in these aerobic hearts is constant at about 0.04 PM, which results in adenylate cyclase being activated to only ‘10% of its putative maximal activity during the course of these transitions. Consequently, only 77.5 and 83.7 nmol/g dry wt of ATP is predicted to be degraded after 2 min of perfusion during the work-jump and glucose-jump transitions, respectively (and 0.51 pmol/g dry wt after 12 min in the latter experiment, or about 2% of the total nucleotide pool). We believe that such an amount of degradation is small compared with the errors of measurement of the nucleotide levels themselves. The model predicts an interstitial adenosine concentration of 0.74 PM and 0.94 PM after 2 min perfusion in the work-jump and glucose-jump experiments respectively. The computed extracellular adenosine concentration after 12 min perfusion in the glucose-jump experiment is 1.27 PM. These values compare favorably with Rubio and Berne’s (22) observation that the concentration of adenosine in the pericardial fluid of the normal dog heart is 1.09 t 0.29 PM. The total tissue levels for adenosine, inosine, and hypoxanthine for these two simulations are summarized in Table 3. It should be noted that Gerl .ach et a1. (4) were unabl .e to detect any of these three purines in their aerobic rat hearts, even though they found other purines in amounts comparable to that found for adenosine by Degenring et al. (2). While our simulation predicts tissue levels of inosine and hypoxanthine similar to those found in rat hearts by Degenring et al. (2), the predicted tissue adenosine level is closer to that found for dog TABLE 3. Unphosphorylated purine aerobic heart, nmollg dry wt
levels in
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
PURINE
METABOLISM
H399
IN ISCHEMIA
hearts. This result suggests that either there is an addition .a1 adenosin .e-containi .ng compartment not included in our model, which does not seem very likely, or there is a defective rate law in our simulation. Since the maximal velocity found for “nucleoside permease” is less than that for adenosine kinase, the low K, for adenosine in the adenosine kinase rate law (0.909 PM) will control the tissue level of adenosine. If the maximal velocity of the permease were actually higher than that of adenosine kinase, the adenosine tissue level would -be
near the (hi .gher> K, for adenosine deaminase (‘43 PM), because the uptake of adenosine would be faster th .an its phosphorylation. This would result in adenosine levels close to those found for the rat heart (2). Since the kinetics of membrane transports are often controlled in a very complex way, we believe a reexamination of the kinetics of this process is indicated. Received for publication
15 December 1975.
REFERENCES 1. BING, R. J. Cardiac metabolism. Physiol. Rev. 45: 171-213, 1965. 2. DEGENRING, F. H., R. RUBIO, AND R. M. BERNE. Adenine nucleotide metabolism during cardiac hypertrophy and ischemia in rats. J. MOL. CeZZ.Cardiol. 7: 105-113, 1975. 3. EFFROS, R. M., B HAIDER, P. ETTINGER, H. OLDEWURTEL, S. AHMED, AND T. REGAN. In uiuo response of myocardial cell pH to changes of plasma pH and ischemia (Abstract). Federation Proc. 33: 396, 1974. 4. GERLACH, E., B. DEUTICKE, AND R. H. DREISBACH. Der Nucleotidabbau im Herzmuskel bei Sauerstoffmangel und seine moglithe Bedeutung fur die Coronardurchblutung. Nuturwissenschufien 6: 228-229, 1963. 5. HADDY, F. J., AND J. B. SCOTT. Metabolic factors in peripheral circulatory regulation. Federation Proc. 34: 2006-2011, 1975. 6. JACOB, M. I., AND R. M. BERNE. Metabolism of purine derivatives by the isolated cat heart. Am. J. Physiol. 198: 322-326, 1960. 7. KLINGENBERG, M., AND M. BUCHHOLZ. Localization of the glycerolphosphate dehydrogenase in the outer phase of the mitochondrial membrane. European J. Biochem. 13: 247-252, 1970. 8. KOHN, M. C., AND D. GARFINKEL. Computer simulation of ischemic rat heart purine metabolism. I. Model construction. Am. J. Physiol. 232: H386-H393, 1977. 9. KRUG, A. Alterations in myocardial hydrogen ion concentration after temporary coronary occlusions: a sign of irreversible cell damage. Am. J. Curdiol. 36: 214-217, 1975. 10. K~BLER, W., AND P. G. SPIEKERMANN. Regulation of glycolysis in the ischemic and the anoxic myocardium. J. MOL. CeZZ. Curdial. 1: 351-377, 1970. 1.1. LIU, M. S., AND H. FEINBERG. Incorporation of adenosine-8-14C and inosine-8-l*C into rabbit heart adenine nucleotides. Am. J. PhysioZ. 220: 1242-1248, 1971. 12. L&SE, B., S. SCHUCHHARDT, N. NIEDERLE, AND H. BENZING. The histogram of local oxygen pressure (Po2) in the dog myocardium and the PoZ behavior during transitory changes of oxygen administration. Aduun. Exptl. Med. BioZ. 37A: 535-540, 1973. 13. MANSOUR, T. E. Studies on heart phosphofructokinase: purification, inhibition, and activation. J. BioZ. Chem. 238: 2285-2292, 1963. 14. NEELY, J, R., M. J. ROVETTO, J. T. WHITMER, AND H. E. MORGAN. Effects of ischemia on function and metabolism of the isolated working rat heart. Am. J. Physiol. 225: 651-658,1973. 15. NEELY, J. R., J. T. WHITMER, AND M. J. ROVETTO. Effect of coronary blood flow on glycolytic flux and intracellular pH in isolated rat hearts. CircuZution Res. 37: 733-741, 1975. 16. NIHEI, T., L. NODA, AND M. F. MORALES. Kinetic properties and
equilibrium constant of the adenosine triphosphate-creatine transphosphorylase catalyzed reaction. J. BioZ. Chem. 236: 32033213, 1961. 17. OLSON, R. E., AND R. J. HOESCHEN. Utilization of endogenous lipid by the isolated perfused rat heart. Biochem. J. 103: 796-801, 1967. 18. OLSSON, R. A., AND J. A. SNOW. Tissue:blood purine nucleoside gradients in canine myocardium (Abstract). Federation Proc. 29: 586, 1970. 19. OLSSON, R. A., J. A. SNOW, M. K. GENTRY, AND G. P. FRICK. Adenosine uptake by canine heart. CircuZution Res. 31: 767-778, 1972. lga.OPrE, L. H. Effects of regional ischemia on the metabolism of glucose and fatty acids. Circulation Res. 38, Suppl. 1, 1-52-1-68, 1976. 20. OPIE, L. H., K. R. L. MANSFORD, AND P. OWEN. Effects of increased work on glycolysis and adenine nucleotides in the perfused heart of normal and diabetic rats. Biochem. J. 124: 475490, 1971. 21. ROVETTO, M. J., J. T. WHITMER, AND J. R. NEELY. Comparison of the effects of anoxia and whole heart ischemia on carbohydrate utilization in isolated working rat hearts. CircuZution Res. 32: 699-711, 1973. 22. RUBIO, R., AND R. M. BERNE. Release of adenosine by the normal myocardium in dogs and its relationship to the regulation of coronary resistance. CircuZution Res. 25: 407-415, 1969. 23. RUBIO, R., V. T. WIEDMEIER, AND R. M. BERNE. Nucleoside phosphorylase: localization and role in the myocardial distribution of purines. Am. J. Physiol. 222: 550-555, 1972. 24. SAFER, B., and J. R. WILLIAMSON. Mitochondrial-cytosolic interactions in perfused rat heart. Role of coupled transamination in repletion of citric acid cycle intermediates. J. BioZ. Chem. 248: 2570-2579, 1973. 25. SEVERSON, D. L., G. I. DRUMMOND, AND P. V. SULAKHE. Adenylate cyclase in skeletal muscle. J. BioZ. Chem. 247: 2949-2958, 1972. 26. SZENT-GY~RGYI, A. G., AND G. PRIOR. Exchange of adenosine diphosphate bound to actin in superprecipitated actomyosin and contracted myofibrils. J. MOL. BioZ. 15: 515-538, 1966. 27. UI, M. Multiple inhibitor sites for ATP on muscle phosphofructokinase as influenced by a change of pH: a computer analysis of “non-linear” kinetic data. Biochim. Biophys. Actu 159: 50-63, 1968. 28. WILLIAMSON, J. R. Glycolytic control mechanisms. II. Kinetics of intermediate changes during the aerobic-anoxic transition in perfused rat heart. J. BioZ. Chem. 241: 5026-5036, 1966.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.
