- Email: [email protected]
Augmenting intracellular adenosine improves myocardial recovery
J
THORAC CARDIOVASC SURG
1990;99:469-74
Augmenting intracellular adenosine improves myocardial recovery The objective of this study was to determine if augmentation of myocardial adenosine levels during global ischemia improves functional recovery after reperfusion. Isolated adult rabbit bearts were subjected to 120 minutes of mildly hypothermic ischemia (34° C) with modified St. Thomas' Hospital cardioplegic solution used to provide myocardial protection. Myocardial adenosine levels were augmented during ischemia by providing exogenous adenosine in the cardioplegic solution or by inhibiting adenosine degradation with 2-deoxycoformycin, a noncompetitive inhibitor of adenosine deaminase. Four groups of bearts were studied: (1) control (n = 23)-cardioplegia alone; (2) adenosine group (n = 10)-adenosine 200 #LmoljL added to the cardioplegic solution; (3) 2-deoxycoformycin group (n = 8)-2-deoxycoformycin 1 #Lmol/L added to the cardioplegic solution; and (4) a combined adenosine/deoxycoformycin group (n = 10). Recovery of developed pressure 45 minutes after reperfusion in the control group averaged only 38 % ± 4 % of baseline values. Significantly better recovery was evident in the adenosine (66 % ± 7 %), deoxycoformycin (59 % ± 2 %), and adenosine j deoxycoformycin (75 % ± 2 %) groups. The slope of the relatio~hip between end-diast06c pressure and volume was used as an index of diast06c stiffness. The slope averaged 85 ± 2 rom Hgjml in the control group 45 minutes after reperfusion, significantly higher than that in the adenosine (31 ± 6), deoxycoformycin (75 ± 5), and adenosinejdeoxycoformycin (58 ± 5) groups; this suggests better Wastolic function in the adenosine-augmented groups. During ischemia, adenosine levels were significantly elevated iiI' the adenosine-augmented groups, whereas adenosine tripbosphate decreased equally in aU four groups, whicb indicates that augmenting myocardial adenosine had no effect on depletion of adenosine tripbosphate during ischemia. After reperfusion, adenosine tripbosphate levels were depressed in the control group but increased in the other groups above baseline values, whicb suggests that improvement in functional recovery was due to accelerated repletion of adenine nucleotide stores in the adenosine-augmented groups.
Steven F. Bolling, MD (by invitation), Laurence E. Bies, BS (by invitation), Edward L. Bove, MD, and Kim P. Gallagher, PhD (by invitation), Ann Arbor, Mich.
Many investigators have demonstrated that myocardial high-energy phosphate and adenosine levelsdecrease during ischemia-? and correlate with the incomplete re-
From the Thoracic Surgery Research Laboratory, Section of Thoracic Surgery, University of Michigan Medical School, Ann Arbor, Mich. Supported in part by National InstitutesofHealthGrant ROI HL32043. K.P. Gallagher is recipient of National Institutes of Health Research Career Development Award HL01420. Read at the Sixty-ninth Annual Meeting of The American Association for Thoracic Surgery, Boston, Mass., May 8-10, 1989. Address for reprints: Steven F. Bolling, MD, The University of Michigan Hospitals 2120D Taubman Center, Box 0344 1500 E. Medical Center Dr., Ann Arbor, MI 48109.
12/6/17783
covery of ventricular function observed after ischemia.v 4 Adenine nucleotides can be restored by two pathways. One is the slow, de novo synthesis pathway that can replace only 0.4% of the total nucleotide pool per hour." The second, more rapid means is the nucleotide salvage pathway, which begins with rephosphorylation of adenosine to adenosine monophosphate by the enzyme adenosine kinase but, however, requires that adenosine be available as a precursor. Experimental studies have shown that inadequate repletion of adenosine triphosphate (ATP) is due to unavailability of adenosine rather than inability of the mitochondria to phosphorylate precursors of ATP. 5,6 Since adenosine in cardioplegic solution improves functional recovery.?'? the objective of the present study was to address two questions by measuring functional recov-
469
The Journal of Thoraci c and Cardiovascular
4 7 0 Bolling et al.
Surgery
• P<.05
vs. control
80
~
60
40
20-'--.......-
Control
ADO
DCF ADO/DCF
Control
15 Min
ADO
DCF ADO/ DCF
30 Min
Control
ADO
DCF ADO/DCF
45 Min
Fig. 1. Recoveryof postischemicdeveloped pressurewith time after reperfusion, comparing control hearts with adenosine-augmentedgroups. % DP. Percent of developed pressure as compared with preischemic baseline level.
ery and adenine nucleotide levelsin globally ischemic and reperfused rabbit hearts. First, does adenosine in cardioplegic solution minimize the depletion of ATP during ischemia? Second, does it accelerate repletion of ATP after reperfusion of ischemic myocardium? In addition, the effectiveness of inhibiting adenosine deaminase and sustaining endogenous adenosine levels was compared against delivery of exogenous adenosine.
Material and methods Studies wereperformed in 51 isolated,perfusedrabbit hearts, with perfusionperformedat 80 mm Hg with a nonrecirculating oxygenated modified Ringer's bicarbonate solution, as described previously.' Left ventricular function (left ventricular developed pressure and its first derivative, dP/dt) was determined isovolumically before and after ischemia with an intraventricular balloon, inflated to a preischemic control end-diastolic pressure of 10 mm Hg. We estimated diastolicstiffness by increasing left ventricular volumein incrementsand measuring the slope of end-diastolic pressure versus end-diastolicvolume curves (.:lEDP/.:lEDV) for each heart. Coronary flow was measured volumetrically. A thermistor needleand an intramyocardial pH probe (Khuri, Vascular Technology, Inc., Clemsford, Mass.) were inserted into the mid-anterior left ventricleto record myocardial pH and temperature.' After a 3D-minute stabilizationperiod,control measurements of end-diastolic pressure, developed pressure (peak systolic pr~ure minus end-diastolic pressure), peak positive dP/dt, coronary flow, pH, and diastolic stiffness were made. Hearts were then rendered globallyischemic. The intraventricular bal-
loon was deflated and 60 ml of cardioplegic solution was administered. All hearts were maintained at 34° C during 120 minutes of total ischemia. All hearts received IS ml of cardioplegic solutionevery30 minutesduring ischemia. Hearts received modified St. Thomas' Hospital cardioplegic solution (controls, n = 23), cardioplegic solution containing adenosine 200 ILmol/L (ADO hearts, n = 10), 2-deoxycoformycin 1 ILmol/L (DCF hearts, n = 8), or a combinationof both adenosine 200 umol/L and 2-deoxycoformycin I ILmol/L (ADO/ DCF, n = 10). Reperfusionwasperformedat 37°C, with pressureat 80 mm Hg. During the initial IS minutes of reperfusion, the intraventricular balloonwas kept deflated to simulate the beating, nonworking condition. After this, the left ventricular balloon was refilled to the preischemic control volumeand measurementsof end-diastolic pressure, developed pressure, dP/dt, diastolic stiffness, pH, and coronary flow wereobtained after 15,30, and 45 minutes of reperfusion. Creatinine kinase (Sigma Diagnostics, St. Louis, Mo.) from coronary sinuseffluentwas also measured at these intervals. After 45 minutes of reperfusion, all hearts were removed from the perfusion apparatus. The myocardial water content was determined by the formula ([ I - dry weight/wet weight] X 100 = % H20). Furthermore, a parallel series of identical experiments (n = 12)weredone to determine recovery of adenine nucleotide levels. Biopsy specimens were taken from the left ventricle during the control period, at end ischemia, and at I and IS minutes of reperfusion. The specimens were assayed by high-performance liquid chromatography for myocardial cellular adenosine triphosphate, diphosphate, and monophosphate, adenosine, and phosphocreatine levels. Statistical analysis was performed with Student's two-tailed t test and analysis of variance. A p
Volume 99
Adenosine
Number 3 March 1990
100
" P<.05
<1
control
•
80
> 0 w
Y5 .
471
60
•
•
<,
0.
0
W
<1
40
20
o................
_~=
Control
ADO
DCF ADO/DCF
Control
15 Min
ADO
DCF ADO/DCF
30 Min
Control
ADO
DCF ADO/DCF
45 Min
Fig. 2. ~hange in.slope of end-diastolic pressure versus end-diastolic volume curves with time. comparing control hearts with adenosine-augmented groups. MDP/MDV. Change in end-diastolic pressure/change in end-diastolic volume. value of less than 0.05 was considered significant. All results are expressed as mean ± standard error of the mean. All animals received humane care in compliance with the National Society for Medical Research.
Results There were no significant differences in the prearrest developed pressure or dP /dt values in any group. After 120 minutes of hypothermic ischemia and 45 minutes of reperfusion, control hearts recovered to 38% ± 4% of developed pressure and 36% ± 5% to dP/dt. There was significantly better recovery of left ventricular function in the ADO hearts (Fig. 1), with 66% ± 7% recovery of developed pressure and 64% ± 3% recovery of dP /dt. DCF hearts recovered to 59% ± 2% of developed pressure and 65% ± 6%todP/dt, and ADO/DCFhearts recovered to 75% ± 2% of developed pressure and 68% ± 2% of dP/ dt (p < 0.05 versus control for all treated hearts but p = NS* versus each other). Left ventricular end-diastolic pressure was significantly lower at all times during reperfusion in the adenosine-augmented hearts than in control hearts. After 45 minutes of reperfusion, control hearts were characterized by an elevated end-diastolic pressure with a 39 ± 6 mm Hg rise in left ventricular end-diastolic pressure, "NS = Not significant.
whereas the ADO hearts had a 17 ± 4 mm Hg elevation in left ventricular end-diastolic pressure. The DCF hearts displayed a 28 ± 5 mm Hg rise in left ventricular end-diastolic pressure and the ADO/DCF hearts showed a 29 ± 10 mm Hg rise in left ventricular end-diastolic pressure. Diastolic stiffness in the adenosine-augmented hearts was significantly lower at all times during reperfusion than in the untreated hearts (Fig. 2). After 45 minutes of reperfusion the control hearts were characterized by a diastolic stiffness slope value of 85 ± 5 versus a diastolic stiffness value of 31 ± 6 in ADO hearts (p < 0.05). The DCF and ADO/DCF groups had diastolic stiffness values of 75 ± 5 and 58 ± 5, respectively (p < 0.05 compared with control). Loss of creatine kinase from the coronary sinus was significantly less when measured after reperfusion in the treated groups as compared with the control group. Control creatine kinase loss was 424 ± 17 IV /L as opposed to 262 ± 40 IV /L in the DCF group (p < 0.05 versus control). Additionally, both ADO and the ADOjDCF hearts had significantly less creatine kinase loss after reperfusion (l05 ± 17 and 104 ± 20, respectively) compared with control or DCF hearts. There were no differences in coronary flowat any time or in myocardial water content after reperfusion between control and treated hearts.
472
The Journal of Thoracic and Cardiovascular
Bolling et al.
Surgery
8
. . . . Control ADO DCF • P < .05 vs control
•. e
_
r:::
6
.-..-
Q)'Qi
_r:::0
en ..
o
Q.
r::: en
~ E
eli
-
...
4 ~
::::L
2
~
~
.
•
i
. ... ... ···
•
..
.-e
-..
--. --.
... ... . ...•
• . .... -_ ....•..•••.....•...•.• ... .. -
~.
f.. .........
•
•
..
O....--r------r-----~------r-----Baseline End 1 Min 15 Min Ischemia Post Post Reperfuslon Reperfuslon Fig. 3. Change inadenosine level measured duringischemia and reperfusion, expressed as micromoles of adenosine per milligram of protein. ADO, Adenosine-treated hearts; DCF, deoxycoformycin-treated hearts. The change in adenosine levels and ATP levels during ischemia and reperfusion are expressed as micromoles per milligram of protein and are demonstrated in Figs. 3 and 4, respectively. Baseline readings of adenosine and ATP were equivalent in all groups measured (control, ADO, and DCF). During ischemia, ATP levels fell equally in all groups, with end-ischemic ATP decreasing to approximately 0.8. However, control hearts had adenosine levels of only 2.0 ± 0.12 at end ischemia compared with 5.6 ± 0.17 and2.9 ± 0.09 in the ADO and DCFgroups, respectively (both p < 0.05 versus control). During reperfusion, at 1 and 15 minutes after reflow, control hearts were characterized by depressed adenosine levels (1.9 ± 0.5 and 1.5 ± 0.07) and depressed ATP levels averaging only 1.8 ± 0.05 and 1.3 ± 0.19 at 1 and 15 minutes after reperfusion. However, ADO hearts demonstrated augmentation of adenosine levels (5.9 ± 0.09 and 4.0 ± 0.11 at I and 15 minutes after reflow. The ADO hearts also had significantly higher ATP levels at I and IS minutes after reflow-5.9 ± 0.05 and 6.9 ± 0.07, respectively (p < 0.05 versus control). Although the DCF group did not achieve the high nucleotide augmentation observed in the ADO group, it was characterizedbyadenosinelevelsof3.1 ± 0.06 and 2.9 ± 0.05 and ATP levels of 2.4 ± 0.05 and 2.9 ± 0.06 at the same times, respectively (p < 0.05 versus control). ADO, DCF, and ADO/DCF hearts had better maintenance of intramyocardial pH during ischemia than did control hearts. At end ischemia, intramyocardial pH of
the control group was 6.7 ± 0.06, whereas in the ADO, DCF, and ADO/DCF groups it was significantly higher, averaging 7.14 ± 0.03, 7.10 ± 0.06, and 7.10 ± 0.04, respectively (p < 0.05 versus control). Preischemic intramyocardial pH and the pH of the cardioplegic solution did not differ significantly among groups.
Discussion In a previous study from this laboratory, we? demonstrated that cardioplegic solution containing adenosine significantly improved functional recovery in isolated adult rabbit hearts exposed to 2 hours of moderately hypothermic (32° C) ischemia. Control hearts receiving cardioplegic solution alone were characterized by 47% ± 3% recovery of developed pressure 45 minutes after reperfusion. Hearts receiving cardioplegic solution supplemented with adenosine in concentrations of 100, 200, or 400 rnmol/L displayed significantly better recovery of developed pressure, averaging 63% ± 4%, 78% ± 3%, and 70% ± 4%, respectively. The main objective of the present study was to investigate the mechanism of adenosine's beneficial action by testing the hypothesis that adding adenosine or sustaining endogenous adenosine levels decreased depletion of ATP during ischemia, as suggested by some investigators, 10 and/or accelerated repletion of ATP after reperfusion. Because ATP decreased equally during ischemia in all groups, we conclude that there was no effect of augmenting adenosine on ischemia-induced depletion of ATP. Compared with the control group, however, ATP levels
Volume 99 Number 3 March 1990
Adenosine
8
.5
..- ... ...... .. .. ••• ,
. ..
6
Gl
0
c. ..
....
Q.
-
4
~
:1
2
•
...
I I
.
••
...'
,
473
..... Control • • • ADO . - . DCF
• P < .05
YS
control
•••••••••••••• ",,<#
~~
.. ~~~ . . - - - - - - -_ _z • t/fJ'"'''' --
O...L...~-----.,.-----.....,.------r------
Baseline
End Ischemia
1 Min Post Reperfuslon
15 Min Post Reperfuslon
Fig. 4. Change in ATP level measured during ischemia and reperfusion, expressed as micromoles of ATP per milligram of protein. ADO. Adenosine-treated hearts; DCF. deoxycoformycin-treated hearts.
were rapidly restored to baseline or higher after reperfusion in the intervention groups, which suggests that the beneficial actionsof adenosineare due, at least in part, to accelerated repletion of ATP levels. This present work confirms that of others? who have demonstrated, in an isolated, working rat heart preparation,that the addition of adenosine 20 J.LmoljL or EHNA 80 umol/L (erythro-9,2-hydroxy-3-nonyl adenine hydrochloride, a reversible competitive adenosine deaminase inhibitor)was able to prevent a reduction in myocyteadenosine levels during 30 minutes of ischemia and was associatedwith improvedfunctional recoveryduring reperfusion, findings consistent with other reports 1,8. I I demonstrating that adenosine and/or inhibition of ATP catabolismwaseffective in maintaining tissueATP levels and enhancing functional recovery in models of global stunning. The beneficial effects of augmenting intracellular adenosine are in accordance with enzyme measurements indicatingthat the salvagerate of intracellular adenosine by intact myocytes is eight times the rate exhibited by myocytes dependent on extracellular adenosine.l? Rat cardiac myocytes':' demonstrate that at adenosine concentrationsof greater than 100J.Lmol/L there is a nonsaturable or "free flow" component of adenosine transport intomyocytes. Therefore, a 200 J.Lmol/L concentration of adenosine was used to augment nucleotide precursors available for salvage when reperfusion occurred. Adenosine deaminase regulatesintracellular adenosine by the irreversible deamination of adenosine. 2-Deoxyco-
formycin (Pentostatin, Warner-Lambert Pharmaceuticals, Morris Plains, N.J.) noncompetitively inhibits adenosine deaminase, and prevents the breakdown of adenosine to inosine. 14 After administration of 2deoxycoformycin, plasma and tissue levels of adenine nucleotides have been reported to be increased up to fivefold." Therefore a 1 J.Lmol/L concentration of 2deoxycoformycin was used to inhibit myocardial adenosine deaminase, augmenting endogenous nucleotide precursors available for salvage when reperfusion occurred. When adenosineconcentrations higher than 0.1 to 0.3 J.Lmol/L are achieved, there can be pronounced vasodilation.P However, in our model, a crystalloid-perfused isolated heart, maximal vasodilationwas evident at all times, probably because of low total oxygen delivery. Therefore any favorable action of adenosine or deoxycoformycin cannot be attributed to an increase in coronary flow. Interestingly, intramyocardial pH during ischemia was better maintained in adenosine-augmented hearts, which suggests another beneficial mechanism of adenosine. In summary, we conclude that adenosine and 2-deoxycoformycin exert beneficial actions on functional recovery after global ischemia.ATP levels were depleted to the same degree during ischemia in each group, but were restored to normal or greater levels after reperfusiononly in the treatment groups, which suggests that augmenting adenosineexogenously or endogenously enablesenhanced repletion of ATP in reperfused myocardium and underscores the ability of ischemic/reperfused myocardium to
474
The Journal of Thoracic and Cardiovascular Surgery
Bolling et al.
regenerate high-energy phosphate compounds, if crucial precursors (such as adenosine) are available. Although results from crystalloid perfused models can beapplied to patients only with caution, adenosine or 2-deoxycoformycin may prove useful in cardiac surgery, especially in those patients at high risk for myocardial injury during ischemia. REFERENCES 1. Foker JE, Einzig S, Wang T, Anderson RW. Adenosine metabolism and myocardial preservation. J THORAC CARDIOVASC SURG 1980;80:506-16. 2. Taegtmeyer H, Roberts AF, Raine AE. Energy metabolism in reperfused heart muscle: metabolic correlates to return of function. J Am Coil Cardiol 1985;6:864-70. 3. Reibel OK, Rovetto MJ. Myocardial adenosine salvage rates and restoration of ATP content following ischemia. Am J Physiol 1979;237:H247-52. 4. Reibel OK, Rovetto MJ. Myocardial ATP synthesis and mechanical function following oxygen deficiency. Am J Physiol I978;234:H620-4. 5. DeWitt OF, Jochim KE, Behrendt OM. Nucleoside degradation and functional impairment during cardioplegia: amelioration by inosine. Circulation 1983;67:171-8. 6. Ward HB, St. Cry SA, Cogordan JA, et al. Recovery of adenosine nucleotide levels after global myocardial ischemia in dogs. Surgery 1984;96:248-55. 7. Bolling SF, Bies LE, Gallagher KP, Bove EL. Enhanced myocardial protection with adenosine. Ann Thorac Surg 1989;47:809-15. 8. Ely SW, Mentzer RM, Lasley RD, Lee BK, Berne RM. Functional and metabolic evidence of enhanced myocardial tolerance to ischemia and reperfusion with adenosine. J THORAC CARDIOVASC SURG 1985;90:549-56. 9. Humphrey SM, Seelye RN. Improved functional recovery of ischemic myocardium by suppression of adenosine catabolism. J THORAC CARDIOVASC SURG 1982;84:16-22. 10. Buhl MR. The postanoxic regeneration of 5' -adenine nuc1eotides in rabbit kidney tissue during in vitro perfusion. Scand J Clin Lab Invest 1976;36:175-81. II. Humphrey SM, Holliss DG, Cartner LA. Influence of inhibitors of ATP catabolism on myocardial recovery after ischemia. J Surg Res 1987;43:187. 12. Sparks HV Jr, Bardenheuer H. Regulation of adenosine formation by the heart. Circ Res 1986;58:193-201. 13. Bowditch J, Brown AK, Dow JW. Accumulation and sal-
vage of adenosine and inosine by isolated mature cardiac myocytes. Biochim Biophys Acta 1985;844:119-28. 14. O'Dwyer PJ, Wagner B, Leyland-Jones B, Wittes RE, Cheson BD, Hoth OF. 2' Deoxycoformycin (Pentostatin) for lymphoid malignancies. Ann Int Med 1988;108:733. 15. Sollevi A, Torssell L, Owall A, Edlund A, Lagerkranser M. Level and cardiovascular effects of adenosine in humans: topics and perspectives in adenosine research. In: Gerlach E, Becker BF, eds. Proceedings of the Third International Symposium on Adenosine. Berlin: Springer-Verlag, 1987: 599-612.
Discussion Dr. Norman A. Silverman (Chicago. Ill.). Can the stoichiometric amount of adenosine that you give account for subsequent incorporation into the nucleotide levels? Do you give enough to raise the ATP levels 4 mgj dl? Dr. Bolling. We probably do. As you know, the normal serum levelof adenosine is between 0.2 and 0.3 JLmoljL.Thus the dose of adenosine that we give, 200 JLmoljL,is a thousand times the normal level. Although, stoichiometrically, it does not add up entirely correctly. I think there are other factors in adenosine augmentation. As you can realize from the functional recovery in all the treated groups, the beneficial effect was approximately the same, even though the ATP and adenosine augmentation in the treated groups was different. Dr. Silverman. Did you do any experiments to prove Koch's postulates and deplete adenosine ATP precursors to show that that was deleterious? Dr. Bolling. We have just completed such a series, but it is not presented in this paper. If adenosine deaminase is added to this preparation, the cells can be depleted of adenosine and all high-energy phosphate compounds at the time of end ischemia and those hearts return no function. Dr. Silverman. One final question. Your end ischemic ATP levels were all the same. Your end ischemic pH levelswere different. The acidosis during ischemia is usually attributed to the breakdown of ATP. You have not changed the rate of generation of the hydrogen ion. Did you measure either the pH or, more important, the buffering capacity of your intracoronary perfusates? Perhaps even though the pH is comparable, the buffering capacity of these adenolated moieties could account for almost all of your results. Dr. Bolling. We did measure the pHs of all our cardioplegic solutions, and they were similar. Also, the starting pHs at ischemia were similar for all hearts, but the buffering capacities were not measured.
THORAC CARDIOVASC SURG
1990;99:469-74
Augmenting intracellular adenosine improves myocardial recovery The objective of this study was to determine if augmentation of myocardial adenosine levels during global ischemia improves functional recovery after reperfusion. Isolated adult rabbit bearts were subjected to 120 minutes of mildly hypothermic ischemia (34° C) with modified St. Thomas' Hospital cardioplegic solution used to provide myocardial protection. Myocardial adenosine levels were augmented during ischemia by providing exogenous adenosine in the cardioplegic solution or by inhibiting adenosine degradation with 2-deoxycoformycin, a noncompetitive inhibitor of adenosine deaminase. Four groups of bearts were studied: (1) control (n = 23)-cardioplegia alone; (2) adenosine group (n = 10)-adenosine 200 #LmoljL added to the cardioplegic solution; (3) 2-deoxycoformycin group (n = 8)-2-deoxycoformycin 1 #Lmol/L added to the cardioplegic solution; and (4) a combined adenosine/deoxycoformycin group (n = 10). Recovery of developed pressure 45 minutes after reperfusion in the control group averaged only 38 % ± 4 % of baseline values. Significantly better recovery was evident in the adenosine (66 % ± 7 %), deoxycoformycin (59 % ± 2 %), and adenosine j deoxycoformycin (75 % ± 2 %) groups. The slope of the relatio~hip between end-diast06c pressure and volume was used as an index of diast06c stiffness. The slope averaged 85 ± 2 rom Hgjml in the control group 45 minutes after reperfusion, significantly higher than that in the adenosine (31 ± 6), deoxycoformycin (75 ± 5), and adenosinejdeoxycoformycin (58 ± 5) groups; this suggests better Wastolic function in the adenosine-augmented groups. During ischemia, adenosine levels were significantly elevated iiI' the adenosine-augmented groups, whereas adenosine tripbosphate decreased equally in aU four groups, whicb indicates that augmenting myocardial adenosine had no effect on depletion of adenosine tripbosphate during ischemia. After reperfusion, adenosine tripbosphate levels were depressed in the control group but increased in the other groups above baseline values, whicb suggests that improvement in functional recovery was due to accelerated repletion of adenine nucleotide stores in the adenosine-augmented groups.
Steven F. Bolling, MD (by invitation), Laurence E. Bies, BS (by invitation), Edward L. Bove, MD, and Kim P. Gallagher, PhD (by invitation), Ann Arbor, Mich.
Many investigators have demonstrated that myocardial high-energy phosphate and adenosine levelsdecrease during ischemia-? and correlate with the incomplete re-
From the Thoracic Surgery Research Laboratory, Section of Thoracic Surgery, University of Michigan Medical School, Ann Arbor, Mich. Supported in part by National InstitutesofHealthGrant ROI HL32043. K.P. Gallagher is recipient of National Institutes of Health Research Career Development Award HL01420. Read at the Sixty-ninth Annual Meeting of The American Association for Thoracic Surgery, Boston, Mass., May 8-10, 1989. Address for reprints: Steven F. Bolling, MD, The University of Michigan Hospitals 2120D Taubman Center, Box 0344 1500 E. Medical Center Dr., Ann Arbor, MI 48109.
12/6/17783
covery of ventricular function observed after ischemia.v 4 Adenine nucleotides can be restored by two pathways. One is the slow, de novo synthesis pathway that can replace only 0.4% of the total nucleotide pool per hour." The second, more rapid means is the nucleotide salvage pathway, which begins with rephosphorylation of adenosine to adenosine monophosphate by the enzyme adenosine kinase but, however, requires that adenosine be available as a precursor. Experimental studies have shown that inadequate repletion of adenosine triphosphate (ATP) is due to unavailability of adenosine rather than inability of the mitochondria to phosphorylate precursors of ATP. 5,6 Since adenosine in cardioplegic solution improves functional recovery.?'? the objective of the present study was to address two questions by measuring functional recov-
469
The Journal of Thoraci c and Cardiovascular
4 7 0 Bolling et al.
Surgery
• P<.05
vs. control
80
~
60
40
20-'--.......-
Control
ADO
DCF ADO/DCF
Control
15 Min
ADO
DCF ADO/ DCF
30 Min
Control
ADO
DCF ADO/DCF
45 Min
Fig. 1. Recoveryof postischemicdeveloped pressurewith time after reperfusion, comparing control hearts with adenosine-augmentedgroups. % DP. Percent of developed pressure as compared with preischemic baseline level.
ery and adenine nucleotide levelsin globally ischemic and reperfused rabbit hearts. First, does adenosine in cardioplegic solution minimize the depletion of ATP during ischemia? Second, does it accelerate repletion of ATP after reperfusion of ischemic myocardium? In addition, the effectiveness of inhibiting adenosine deaminase and sustaining endogenous adenosine levels was compared against delivery of exogenous adenosine.
Material and methods Studies wereperformed in 51 isolated,perfusedrabbit hearts, with perfusionperformedat 80 mm Hg with a nonrecirculating oxygenated modified Ringer's bicarbonate solution, as described previously.' Left ventricular function (left ventricular developed pressure and its first derivative, dP/dt) was determined isovolumically before and after ischemia with an intraventricular balloon, inflated to a preischemic control end-diastolic pressure of 10 mm Hg. We estimated diastolicstiffness by increasing left ventricular volumein incrementsand measuring the slope of end-diastolic pressure versus end-diastolicvolume curves (.:lEDP/.:lEDV) for each heart. Coronary flow was measured volumetrically. A thermistor needleand an intramyocardial pH probe (Khuri, Vascular Technology, Inc., Clemsford, Mass.) were inserted into the mid-anterior left ventricleto record myocardial pH and temperature.' After a 3D-minute stabilizationperiod,control measurements of end-diastolic pressure, developed pressure (peak systolic pr~ure minus end-diastolic pressure), peak positive dP/dt, coronary flow, pH, and diastolic stiffness were made. Hearts were then rendered globallyischemic. The intraventricular bal-
loon was deflated and 60 ml of cardioplegic solution was administered. All hearts were maintained at 34° C during 120 minutes of total ischemia. All hearts received IS ml of cardioplegic solutionevery30 minutesduring ischemia. Hearts received modified St. Thomas' Hospital cardioplegic solution (controls, n = 23), cardioplegic solution containing adenosine 200 ILmol/L (ADO hearts, n = 10), 2-deoxycoformycin 1 ILmol/L (DCF hearts, n = 8), or a combinationof both adenosine 200 umol/L and 2-deoxycoformycin I ILmol/L (ADO/ DCF, n = 10). Reperfusionwasperformedat 37°C, with pressureat 80 mm Hg. During the initial IS minutes of reperfusion, the intraventricular balloonwas kept deflated to simulate the beating, nonworking condition. After this, the left ventricular balloon was refilled to the preischemic control volumeand measurementsof end-diastolic pressure, developed pressure, dP/dt, diastolic stiffness, pH, and coronary flow wereobtained after 15,30, and 45 minutes of reperfusion. Creatinine kinase (Sigma Diagnostics, St. Louis, Mo.) from coronary sinuseffluentwas also measured at these intervals. After 45 minutes of reperfusion, all hearts were removed from the perfusion apparatus. The myocardial water content was determined by the formula ([ I - dry weight/wet weight] X 100 = % H20). Furthermore, a parallel series of identical experiments (n = 12)weredone to determine recovery of adenine nucleotide levels. Biopsy specimens were taken from the left ventricle during the control period, at end ischemia, and at I and IS minutes of reperfusion. The specimens were assayed by high-performance liquid chromatography for myocardial cellular adenosine triphosphate, diphosphate, and monophosphate, adenosine, and phosphocreatine levels. Statistical analysis was performed with Student's two-tailed t test and analysis of variance. A p
Volume 99
Adenosine
Number 3 March 1990
100
" P<.05
<1
control
•
80
> 0 w
Y5 .
471
60
•
•
<,
0.
0
W
<1
40
20
o................
_~=
Control
ADO
DCF ADO/DCF
Control
15 Min
ADO
DCF ADO/DCF
30 Min
Control
ADO
DCF ADO/DCF
45 Min
Fig. 2. ~hange in.slope of end-diastolic pressure versus end-diastolic volume curves with time. comparing control hearts with adenosine-augmented groups. MDP/MDV. Change in end-diastolic pressure/change in end-diastolic volume. value of less than 0.05 was considered significant. All results are expressed as mean ± standard error of the mean. All animals received humane care in compliance with the National Society for Medical Research.
Results There were no significant differences in the prearrest developed pressure or dP /dt values in any group. After 120 minutes of hypothermic ischemia and 45 minutes of reperfusion, control hearts recovered to 38% ± 4% of developed pressure and 36% ± 5% to dP/dt. There was significantly better recovery of left ventricular function in the ADO hearts (Fig. 1), with 66% ± 7% recovery of developed pressure and 64% ± 3% recovery of dP /dt. DCF hearts recovered to 59% ± 2% of developed pressure and 65% ± 6%todP/dt, and ADO/DCFhearts recovered to 75% ± 2% of developed pressure and 68% ± 2% of dP/ dt (p < 0.05 versus control for all treated hearts but p = NS* versus each other). Left ventricular end-diastolic pressure was significantly lower at all times during reperfusion in the adenosine-augmented hearts than in control hearts. After 45 minutes of reperfusion, control hearts were characterized by an elevated end-diastolic pressure with a 39 ± 6 mm Hg rise in left ventricular end-diastolic pressure, "NS = Not significant.
whereas the ADO hearts had a 17 ± 4 mm Hg elevation in left ventricular end-diastolic pressure. The DCF hearts displayed a 28 ± 5 mm Hg rise in left ventricular end-diastolic pressure and the ADO/DCF hearts showed a 29 ± 10 mm Hg rise in left ventricular end-diastolic pressure. Diastolic stiffness in the adenosine-augmented hearts was significantly lower at all times during reperfusion than in the untreated hearts (Fig. 2). After 45 minutes of reperfusion the control hearts were characterized by a diastolic stiffness slope value of 85 ± 5 versus a diastolic stiffness value of 31 ± 6 in ADO hearts (p < 0.05). The DCF and ADO/DCF groups had diastolic stiffness values of 75 ± 5 and 58 ± 5, respectively (p < 0.05 compared with control). Loss of creatine kinase from the coronary sinus was significantly less when measured after reperfusion in the treated groups as compared with the control group. Control creatine kinase loss was 424 ± 17 IV /L as opposed to 262 ± 40 IV /L in the DCF group (p < 0.05 versus control). Additionally, both ADO and the ADOjDCF hearts had significantly less creatine kinase loss after reperfusion (l05 ± 17 and 104 ± 20, respectively) compared with control or DCF hearts. There were no differences in coronary flowat any time or in myocardial water content after reperfusion between control and treated hearts.
472
The Journal of Thoracic and Cardiovascular
Bolling et al.
Surgery
8
. . . . Control ADO DCF • P < .05 vs control
•. e
_
r:::
6
.-..-
Q)'Qi
_r:::0
en ..
o
Q.
r::: en
~ E
eli
-
...
4 ~
::::L
2
~
~
.
•
i
. ... ... ···
•
..
.-e
-..
--. --.
... ... . ...•
• . .... -_ ....•..•••.....•...•.• ... .. -
~.
f.. .........
•
•
..
O....--r------r-----~------r-----Baseline End 1 Min 15 Min Ischemia Post Post Reperfuslon Reperfuslon Fig. 3. Change inadenosine level measured duringischemia and reperfusion, expressed as micromoles of adenosine per milligram of protein. ADO, Adenosine-treated hearts; DCF, deoxycoformycin-treated hearts. The change in adenosine levels and ATP levels during ischemia and reperfusion are expressed as micromoles per milligram of protein and are demonstrated in Figs. 3 and 4, respectively. Baseline readings of adenosine and ATP were equivalent in all groups measured (control, ADO, and DCF). During ischemia, ATP levels fell equally in all groups, with end-ischemic ATP decreasing to approximately 0.8. However, control hearts had adenosine levels of only 2.0 ± 0.12 at end ischemia compared with 5.6 ± 0.17 and2.9 ± 0.09 in the ADO and DCFgroups, respectively (both p < 0.05 versus control). During reperfusion, at 1 and 15 minutes after reflow, control hearts were characterized by depressed adenosine levels (1.9 ± 0.5 and 1.5 ± 0.07) and depressed ATP levels averaging only 1.8 ± 0.05 and 1.3 ± 0.19 at 1 and 15 minutes after reperfusion. However, ADO hearts demonstrated augmentation of adenosine levels (5.9 ± 0.09 and 4.0 ± 0.11 at I and 15 minutes after reflow. The ADO hearts also had significantly higher ATP levels at I and IS minutes after reflow-5.9 ± 0.05 and 6.9 ± 0.07, respectively (p < 0.05 versus control). Although the DCF group did not achieve the high nucleotide augmentation observed in the ADO group, it was characterizedbyadenosinelevelsof3.1 ± 0.06 and 2.9 ± 0.05 and ATP levels of 2.4 ± 0.05 and 2.9 ± 0.06 at the same times, respectively (p < 0.05 versus control). ADO, DCF, and ADO/DCF hearts had better maintenance of intramyocardial pH during ischemia than did control hearts. At end ischemia, intramyocardial pH of
the control group was 6.7 ± 0.06, whereas in the ADO, DCF, and ADO/DCF groups it was significantly higher, averaging 7.14 ± 0.03, 7.10 ± 0.06, and 7.10 ± 0.04, respectively (p < 0.05 versus control). Preischemic intramyocardial pH and the pH of the cardioplegic solution did not differ significantly among groups.
Discussion In a previous study from this laboratory, we? demonstrated that cardioplegic solution containing adenosine significantly improved functional recovery in isolated adult rabbit hearts exposed to 2 hours of moderately hypothermic (32° C) ischemia. Control hearts receiving cardioplegic solution alone were characterized by 47% ± 3% recovery of developed pressure 45 minutes after reperfusion. Hearts receiving cardioplegic solution supplemented with adenosine in concentrations of 100, 200, or 400 rnmol/L displayed significantly better recovery of developed pressure, averaging 63% ± 4%, 78% ± 3%, and 70% ± 4%, respectively. The main objective of the present study was to investigate the mechanism of adenosine's beneficial action by testing the hypothesis that adding adenosine or sustaining endogenous adenosine levels decreased depletion of ATP during ischemia, as suggested by some investigators, 10 and/or accelerated repletion of ATP after reperfusion. Because ATP decreased equally during ischemia in all groups, we conclude that there was no effect of augmenting adenosine on ischemia-induced depletion of ATP. Compared with the control group, however, ATP levels
Volume 99 Number 3 March 1990
Adenosine
8
.5
..- ... ...... .. .. ••• ,
. ..
6
Gl
0
c. ..
....
Q.
-
4
~
:1
2
•
...
I I
.
••
...'
,
473
..... Control • • • ADO . - . DCF
• P < .05
YS
control
•••••••••••••• ",,<#
~~
.. ~~~ . . - - - - - - -_ _z • t/fJ'"'''' --
O...L...~-----.,.-----.....,.------r------
Baseline
End Ischemia
1 Min Post Reperfuslon
15 Min Post Reperfuslon
Fig. 4. Change in ATP level measured during ischemia and reperfusion, expressed as micromoles of ATP per milligram of protein. ADO. Adenosine-treated hearts; DCF. deoxycoformycin-treated hearts.
were rapidly restored to baseline or higher after reperfusion in the intervention groups, which suggests that the beneficial actionsof adenosineare due, at least in part, to accelerated repletion of ATP levels. This present work confirms that of others? who have demonstrated, in an isolated, working rat heart preparation,that the addition of adenosine 20 J.LmoljL or EHNA 80 umol/L (erythro-9,2-hydroxy-3-nonyl adenine hydrochloride, a reversible competitive adenosine deaminase inhibitor)was able to prevent a reduction in myocyteadenosine levels during 30 minutes of ischemia and was associatedwith improvedfunctional recoveryduring reperfusion, findings consistent with other reports 1,8. I I demonstrating that adenosine and/or inhibition of ATP catabolismwaseffective in maintaining tissueATP levels and enhancing functional recovery in models of global stunning. The beneficial effects of augmenting intracellular adenosine are in accordance with enzyme measurements indicatingthat the salvagerate of intracellular adenosine by intact myocytes is eight times the rate exhibited by myocytes dependent on extracellular adenosine.l? Rat cardiac myocytes':' demonstrate that at adenosine concentrationsof greater than 100J.Lmol/L there is a nonsaturable or "free flow" component of adenosine transport intomyocytes. Therefore, a 200 J.Lmol/L concentration of adenosine was used to augment nucleotide precursors available for salvage when reperfusion occurred. Adenosine deaminase regulatesintracellular adenosine by the irreversible deamination of adenosine. 2-Deoxyco-
formycin (Pentostatin, Warner-Lambert Pharmaceuticals, Morris Plains, N.J.) noncompetitively inhibits adenosine deaminase, and prevents the breakdown of adenosine to inosine. 14 After administration of 2deoxycoformycin, plasma and tissue levels of adenine nucleotides have been reported to be increased up to fivefold." Therefore a 1 J.Lmol/L concentration of 2deoxycoformycin was used to inhibit myocardial adenosine deaminase, augmenting endogenous nucleotide precursors available for salvage when reperfusion occurred. When adenosineconcentrations higher than 0.1 to 0.3 J.Lmol/L are achieved, there can be pronounced vasodilation.P However, in our model, a crystalloid-perfused isolated heart, maximal vasodilationwas evident at all times, probably because of low total oxygen delivery. Therefore any favorable action of adenosine or deoxycoformycin cannot be attributed to an increase in coronary flow. Interestingly, intramyocardial pH during ischemia was better maintained in adenosine-augmented hearts, which suggests another beneficial mechanism of adenosine. In summary, we conclude that adenosine and 2-deoxycoformycin exert beneficial actions on functional recovery after global ischemia.ATP levels were depleted to the same degree during ischemia in each group, but were restored to normal or greater levels after reperfusiononly in the treatment groups, which suggests that augmenting adenosineexogenously or endogenously enablesenhanced repletion of ATP in reperfused myocardium and underscores the ability of ischemic/reperfused myocardium to
474
The Journal of Thoracic and Cardiovascular Surgery
Bolling et al.
regenerate high-energy phosphate compounds, if crucial precursors (such as adenosine) are available. Although results from crystalloid perfused models can beapplied to patients only with caution, adenosine or 2-deoxycoformycin may prove useful in cardiac surgery, especially in those patients at high risk for myocardial injury during ischemia. REFERENCES 1. Foker JE, Einzig S, Wang T, Anderson RW. Adenosine metabolism and myocardial preservation. J THORAC CARDIOVASC SURG 1980;80:506-16. 2. Taegtmeyer H, Roberts AF, Raine AE. Energy metabolism in reperfused heart muscle: metabolic correlates to return of function. J Am Coil Cardiol 1985;6:864-70. 3. Reibel OK, Rovetto MJ. Myocardial adenosine salvage rates and restoration of ATP content following ischemia. Am J Physiol 1979;237:H247-52. 4. Reibel OK, Rovetto MJ. Myocardial ATP synthesis and mechanical function following oxygen deficiency. Am J Physiol I978;234:H620-4. 5. DeWitt OF, Jochim KE, Behrendt OM. Nucleoside degradation and functional impairment during cardioplegia: amelioration by inosine. Circulation 1983;67:171-8. 6. Ward HB, St. Cry SA, Cogordan JA, et al. Recovery of adenosine nucleotide levels after global myocardial ischemia in dogs. Surgery 1984;96:248-55. 7. Bolling SF, Bies LE, Gallagher KP, Bove EL. Enhanced myocardial protection with adenosine. Ann Thorac Surg 1989;47:809-15. 8. Ely SW, Mentzer RM, Lasley RD, Lee BK, Berne RM. Functional and metabolic evidence of enhanced myocardial tolerance to ischemia and reperfusion with adenosine. J THORAC CARDIOVASC SURG 1985;90:549-56. 9. Humphrey SM, Seelye RN. Improved functional recovery of ischemic myocardium by suppression of adenosine catabolism. J THORAC CARDIOVASC SURG 1982;84:16-22. 10. Buhl MR. The postanoxic regeneration of 5' -adenine nuc1eotides in rabbit kidney tissue during in vitro perfusion. Scand J Clin Lab Invest 1976;36:175-81. II. Humphrey SM, Holliss DG, Cartner LA. Influence of inhibitors of ATP catabolism on myocardial recovery after ischemia. J Surg Res 1987;43:187. 12. Sparks HV Jr, Bardenheuer H. Regulation of adenosine formation by the heart. Circ Res 1986;58:193-201. 13. Bowditch J, Brown AK, Dow JW. Accumulation and sal-
vage of adenosine and inosine by isolated mature cardiac myocytes. Biochim Biophys Acta 1985;844:119-28. 14. O'Dwyer PJ, Wagner B, Leyland-Jones B, Wittes RE, Cheson BD, Hoth OF. 2' Deoxycoformycin (Pentostatin) for lymphoid malignancies. Ann Int Med 1988;108:733. 15. Sollevi A, Torssell L, Owall A, Edlund A, Lagerkranser M. Level and cardiovascular effects of adenosine in humans: topics and perspectives in adenosine research. In: Gerlach E, Becker BF, eds. Proceedings of the Third International Symposium on Adenosine. Berlin: Springer-Verlag, 1987: 599-612.
Discussion Dr. Norman A. Silverman (Chicago. Ill.). Can the stoichiometric amount of adenosine that you give account for subsequent incorporation into the nucleotide levels? Do you give enough to raise the ATP levels 4 mgj dl? Dr. Bolling. We probably do. As you know, the normal serum levelof adenosine is between 0.2 and 0.3 JLmoljL.Thus the dose of adenosine that we give, 200 JLmoljL,is a thousand times the normal level. Although, stoichiometrically, it does not add up entirely correctly. I think there are other factors in adenosine augmentation. As you can realize from the functional recovery in all the treated groups, the beneficial effect was approximately the same, even though the ATP and adenosine augmentation in the treated groups was different. Dr. Silverman. Did you do any experiments to prove Koch's postulates and deplete adenosine ATP precursors to show that that was deleterious? Dr. Bolling. We have just completed such a series, but it is not presented in this paper. If adenosine deaminase is added to this preparation, the cells can be depleted of adenosine and all high-energy phosphate compounds at the time of end ischemia and those hearts return no function. Dr. Silverman. One final question. Your end ischemic ATP levels were all the same. Your end ischemic pH levelswere different. The acidosis during ischemia is usually attributed to the breakdown of ATP. You have not changed the rate of generation of the hydrogen ion. Did you measure either the pH or, more important, the buffering capacity of your intracoronary perfusates? Perhaps even though the pH is comparable, the buffering capacity of these adenolated moieties could account for almost all of your results. Dr. Bolling. We did measure the pHs of all our cardioplegic solutions, and they were similar. Also, the starting pHs at ischemia were similar for all hearts, but the buffering capacities were not measured.