- Email: [email protected]
Reperfusion-induced calcium gain after ischemia
Reperfusion-Induced Calciu Gain After lschemia Jennifer S. Hz, 5Sc (l-ions), Sianna Panagiotopoules, and Winifred G. Nayler, DSC
Reperfurion-induced calcium gain provides a marker of irreversible injury, but whether the ceils gain calcium because of irreversible injury caused by the ischemic episode, or whether it is the reperfusioninduced calcium gain that triggers the irreversible injury has yet to be established. Using isolated rat hearts made ischemic for either 30 or 60 minutes, and reperfusing with KrebsHenseieit buffer or Krebs-Henseieit buffer containing either 2,3-butanedione monoxime (to inhibit contractile activity) or 2,4-dinitrophenoi or nitrogen-gassed substrate-free Krebs-Henseieit buffer (to inhibit oxidative phosphoryiation), the effect of reperfusion was monitored in terms of calcium gain and ultrastructural changes including loss of sarcolemma1 integrity. The results establish that the routes of calcium entry during postischemic reperfusion are complex. The calcium gain can occur in the absence of mitochondriai oxidative phosphoryiation and is moduiated by interventions introduced at the moment of reperfusion which affect the contractile state. There are at least 2 routes of calcium entry: contractiondependent and contraction independent. The former is probably associated with the development of sarcoiemmai discontinuities. The results also establish that when sarcoiemmai integrity has been destroyed, the cells can gain excess calcium under conditions that prevent mitochondriai calcium uptake. Accordingly, the mitochondria cannot be the only intracellular organeiies that accumulate caiciurn under these conditions. Additional studies are needed to identify the other sites of calcium bindin under conditions of adenosine triphosphate deprivation. (Am J Cardioi f989;63:
From the Department of Medicine, University of Melbourne, Austin Hospital, Heidelberg, Victoria 3084, Australia. This study was supported in part by a grant from the National Health and Medical Research Council of Australia. Address for reprints: Winifred 6. Nayler, DSc, Department of Medicine, University of Melbourne, Austin Hospital, Heidelberg, Victoria 3084. Australia.
Es.2(HOGS),
ells of hearts that are made permanently ischemic1>2 do not accumulate calcium. However, reperfusion1,2 after 20 or more minutes of normothermic ischemia causes an uncontrolled gain in calcium, which in turn can precipitate or hasten cell death and tissue necrosis. Although there is no doubt that ischemic tissue must be reperfused if it is to survive, what remains questionable is whether the fate of the cells is predetermined by the ischemic episode, with reperfusion and the associated gain in calcium merely expressing this ultimate fate, or whether reperfusion itself can be damaging.3
C
REPERFUSlON
INJURY:
OES I+ EXIST?
After an ischemic episode 2 populations of cells exist-those that are reversibly injured and those that are irreversibly injured. After only a short period of ischemia, cells are reversibly injured and reperfusion results in their recovery.4 After extended periods of ischemia, however, irreversible cellular injury may occur and reperfusion at this time can be of no benefit5 Of particular importance is the condition under which reperfusion fails to result in recovery, yet irreversible damage to the cells is not evident before reperfusion. There are at least 2 explanations for such a condition. The cells may have been lethally injured by the ischemic episode, with reperfusion merely expressing this lethal damage and possibly exacerbating it. Alternatively, the cells may only have been reversibly injured during the ischemic episode; however, reperfusion, with its attendant reintroduction of an unlimited supply of calcium, causes further damage that results in lethal injury. If the first possibility is correct and the fate of the cells is already determined during the ischemic episode, then any intervention that may be introduced during reperfusion cannot be beneficial. Arguing against this and in favor of the second possibility are reports in which recovery after an ischemic episode is favored by altering the composition of the reperfusion buffer.6-8 In 1 such report,6 reperfusion with substrate-enriched blood cardioplegic solution after 6 hours of regional ischemia resulted in recovery, whereas reperfusion with blood after this or even shorter periods of ischemia failed to restore any contractile activity, and structural damage was evident. Other studies7x8have shown that recovery is favored if the initial reperfusion is with a solution containing a low calcium concentration. PORTANC
The reports7J previously mentioned suggest that calcium entry during reperfusion may be deleterious, and THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7. 1989
A SYMPOSIUM:
MYOCARDIAL
ISCHEMIA
there is no doubt that cellular calcium overloading can occur during reperfusion. Cellular calcium will increase whenever calcium influx is greater than calcium efflux. This will occur if calcium influx increases while calcium efflux remains the same or decreases, or if calcium efflux decreases while calcium influx remains constant. If any of these situations continues for an extended period, calcium overloading will occur. Calcium efflux from the cell occurs primarily in exchange for sodium and via the sarcolemma1 adenosine triphosphatase pump. Both of these processes of calcium extrusion depend either directly or indirectly on the availability of adenosine triphosphate, and because only limited energy supplies remain after an ischemic episode, it is not hard to envisage that cytosolic calcium will increase. A raised cytosolic calcium will in turn hasten the breakdown of any remaining adenosine triphosphate, and activation of the various phospholipases and proteases. Even the generation of free radicals is, to some extent, calcium dependent. Although a raised cellular calcium is obviously detrimental to the cells, the question that remains unanswered is whether calcium entry is the primary event resulting in lethal injury or whether some other lethal event has already occurred to which calcium entry is a secondary consequence.
ROUTES OF CALCIUM ENTRY Calcium entry during reperfusion may occur through physiologic or nonphysiologic channels, or both. It is unlikely that the predominant route of entry is through the slow channels because slow channel blockers added at the time of reperfusion fail to diminish the calcium gain or aid in recovery.‘0 Entry of calcium by passive perfusion down its concentration gradient or in exchange for sodium are 2 alternative and viable possibilities. Entry through nonphysiologic routes, including through sarcolemma1 disruptions, may also play a role.” However, if this latter route is the predominant route of calcium entry, then this must be secondary to the loss of sarcolemma1 integrity. Some investigators have suggested that entry through a discontinuous sarcolemma is the predominant route of calcium entry during reoxygenation after hypoxia.12 Others believe that mitochondrial calcium uptake is of primary importance.13 The present experiments were designed to investigate these possibilities with reference to calcium gain during postischemic reperfusion. EXPERIMENTAL MODEL In the following experiments, isolated adult SpragueDawley rat hearts were perfused in a nonrecirculating Langendorf mode with Krebs-Henseleit buffer at 37’C
30
kZi KH q BDM l
20
p < 0.01
10
FIGURE 1. Eftect of 10 mM 2,3-butanedione monoxime on the reperlusion-induced cakium gain after 30 minutes of global ischemia. Each column represents the mean f standard error of 6 separate experiments. Statistics refer to the reduction of calcium content in the presence of 2,3-butanedione monoxime compared with that found in its absence. BDM = 2,3-butanedione monoxime; KH = KrebsHenseleit buffer.
0$ 15min
30 min
aerobic
repeifusion
TABLE Creatine
I Effect of 10 mM 2,3-Butanedione Monoxime Phosphate After 30 Minutes of lschemia
on the l?epetfusion-induced
Recovery
of Adenosine
Triphosphate
and
Repetfusion Experimental
Aerobic
lschemic
ATP Olmol/g dry wt) CP Olmol/g dry wt) * p
15.76 f 0.69
2.47 f 0.20
22.22 f 0.81
2.83 f 0.54
KH
BDM
5.73 f 0.57 12.89f
1.12
Each result is the mean f standard error of 6 separate experiments. Statistics refer to the enhancement of adenosine triphosphate minutes reperfusion in the presence of 2.3.butanedione monoxime compared with those found in its absence. ATP = adenosine triphosphate; EIOM = 2,3-butanedione monoxime: CP = creatine phosphate: KH = Krebs-Henseleit buffer.
8E
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
63
9.31 f 0.63” 27.16 f 2.35’
and creatine
phosphate
levels during 15
and at a constant flow rate of 10 ml/min. These hearts were allowed to beat spontaneously to avoid the release of norepinephrine. Global ischemia was induced by the cessation of all flow to the heart. This model was chosen rather than other in vitro models involving muscle strips, lo because the latter do not have an intact circulation and ischemia can only be mimicked by surrounding the tissue with nitrogen-saturated solutions or by sealing them in nitrogen-gassed containers. Reperfusion is then accomplished by diffusion from the bathing solution. Other models using hypoxia12J3 were also not used, because the continual flow involved in these models does not allow for the accumulation of end products, as occurs in ischemia. An in vivo model4 was not chosen, although this would obviously be more representative of the human situation, since not all the conditions can be controlled. In our in vitro model, all cells were made uniformly ischemic, whereas in an in vivo model collateral flow of varying degrees is available, resulting in varying degrees of ischemia. However, care must be taken when extrapolating results from 1 model to another. Although we believe that our results are applicable to the in vivo situation, it is possible that the time changes observed under our experimental conditions may differ from that obtained under other conditions. After each perfusion the hearts were perfusion-fixed for electron microscopy or they were analyzed for calcium content as previously described.7 Briefly, the coronary vasculature of hearts to be analyzed for calcium content was flushed free of calcium with an ice-cold sucrose/
histidine solution, and the ventricles then dried to constant weight at 100°C. The dried ventricles were digested in concentrated nitric acid and an aliquot of this extract diluted in a calcium blank solution containing potassium chloride and lanthanum chloride. The calcium content was analyzed by atomic absorption spectrophotometry7 and compared to standards prepared by serial dilution of a calcium stock solution. CONTRACTILE ACTIVITY AND REPERFUSION-INDUCED CALCIUM AFTER 30 MINUTES OF ISCHEMIA
GAIN
To investigate the role of contractile activity during reperfusion on the gain in calcium, isolated perfused rat hearts were made globally ischemic for 30 minutes before reperfusion in the presence or absence of 10 mM 2,3butanedione monoxime. To inhibit contractile activity, 2,3-butanedione monoxime was used. With the addition of 10 mM 2,3-butanedione monoxime to aerobically perfused hearts, contractile activity ceases almost immediately but is restored when it is removed. The presence of 2,3-butanedione monoxime has a negative inotropic effect on both intact and chemically skinned cardiac muscle.14 It reduces the sensitivity of the myofibrils to calcium and inhibits cross-bridge formation, while having no major effect on the electrical activity of the heart or calcium release from the sarcoplasmic reticulum.14-l6 In the absence of 2,3-butanedione monoxime, the hearts rapidly gained calcium on reperfusion after 30 minutes of ischemia, so that after 30 minutes’ reperfusion
FIGURE 2. Electron micrograph from a heart reperfused for 30 minutes with Krebs-Henseleit buffer after 30 minutes of ischemia. Note the contraction bands. Magnification X 7,000, reduced by 30%.)
FIGURE 3. Electron micrograph from a heart reperfused for 30 minutes with Krebs-Henseieit buffer containing 10 mM 2,3-butenedione monoxime after 30 minutes of ischemia. The myofibrils are intact and sarcolemmal integrity maintained. Note also the presence of glycogen. (Magnification X 7,OGG, reduced by 309/o.)
THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989
9E
A SYMPOSIUM:
MYOCARDIAL
ISCHEMA
with Krebs-Henseleit buffer they contained 27.85 f 1.94 pmol calcium/g dry weight (Fig. 1). The addition of 10 mM 2,3-butanedione monoxime during reperfusion reduced this calcium gain but did not completely abolish it (Fig. 1). This reduction of calcium gain can be explained neither in terms of an altered reflow area in the presence of 2,3-butanedione monoxime nor as an effect of the increased osmolarity of the reperfusion buffer due to its presence. In addition to reducing the gain in calcium, 2,3-butanedione monoxime enhanced the recovery of adenosine triphosphate and creatine phosphate (Table I) and reduced the ultrastructural damage to the cells during reperfusion, In the absence of 2,3-butanedione monoxime, the cells displayed extensively contracted myofibrils with contraction bands. The mitochondria were frequently swollen and sarcolemmal disruptions were apparent (Fig. 2). The formation of contraction bands was prevented by 2,3-butanedione monoxime and the sarcolemma remained intact. Most of the mitochondria showed a normal appearance and glycogen deposits were present in the cytoplasm (Fig. 3). This structural “protection” was only maintained while 2,3-butanedione monoxime was present. Removal of 2,3-butanedione monoxime and continued perfusion with Krebs-Henseleit buffer resulted in the formation of contraction bands, a loss of sarcolemmal integrity and swollen mitochondria-changes similar to those found with reperfusion in the absence of 2,3-butanedione monoxime. At the same time there was a rapid increase in calcium, to 24.47 f 1.51 r.Lmol/g dry weight. These results show that by inhibiting contractile activity during reperfusion after 30 minutes of ischemia, the reperfusion-induced gain in calcium is reduced but not abolished. This suggests that calcium entry during reper-
fusion may occur via routes that are contraction dependent and those that are contraction independent. In the presence of 2,3-butanedione monoxime, calcium entry via contraction-dependent routes is inhibited. The present results suggest that calcium entry via the contractiondependent routes may involve entry through a disrupted sarcolemma. Other investigators have argued that this is a major route of calcium entry during reoxygenation after hypoxia.12 However, after 30 minutes of ischemia, entry via sarcolemmal “holes” cannot be the sole route of calcium entry during reperfusion. Some calcium entry must be occurring across an intact sarcolemma through contraction-independent routes. This entry would be masked by the more rapid entry of calcium through a disrupted sarcolemma in the absence of 2,3-butanedione monoxime. The results also show that reducing the reperfusioninduced calcium gain does not result in the hearts resuming their preischemic state. This may imply that the calcium gain that still occurred is deleterious to the cells or that the fate of the cells is already determined by the ischemic episode and we have merely delayed the full expression of this damage. The latter hypothesis seems unlikely since we would not expect lethally injured cells to be capable of synthesizing adenosine triphosphate and creatine phosphate, nor would we expect to find glycogen deposits as seen in Figure 3. OXIDATIVE PHOSPHORYLATION AND REPERFUSION-INDUCED CALCIUM GAIN AFTER 30 MINUTES OF ISCHEMIA Oxidative phosphorylation was uncoupled during reperfusion after 30 minutes of ischemia, by adding 1 mM 2,4-dinitrophenol. This not only prevents the synthesis of
l
pco.01 FIGURE 4. Eflect of 1 mM ?,Cdinitrophenol on the reperk&n-induced calcium gain after 30 minutes of global ischemia. Each column represents the mean f standard error of 6 separate experiments. Statistics refer to the reduction of calcium content in the presence of 2,4-dinitrophenol compared with that found in its absence. DNP = 2,4-dinitrophend; KH = Krebs-lfensekit bulfer.
15 min
30 min
aerobic
repetfusion
1OE
THE AMERICAN
JOURNAL
OF CARDIOLbGY
VOLUME
63
adenosine triphosphate, but also inhibits the uptake of calcium by the mitochondria. The addition of 1 mM 2,4dinitrophenol during reperfusion attenuated the reperfusion-induced gain in calcium throughout 30 minutes of reperfusion (Fig. 4). It also prevented the ultrastructural damage normally seen with Krebs-Henseleit buffer reperfusion. The formation of contraction bands was inhibited and the sarcolemma remained intact (Fig. 5). Because active tension development cannot occur in the absence of adenosine triphosphate, we would expect that 2,4-dinitrophenol would prevent calcium entry via contraction-dependent routes. Hence, we may expect similar results with respect to reperfusion-induced calcium gain as we found when 2,3-butanedione monoxime was used to inhibit contractile activity. However, in contrast to the results obtained when 2,3-butanedione monoxime was
present, 2,4-dinitrophenol abolished the calcium gain. This suggests that 2,4-dinitrophenol inhibits not only the contraction-dependent routes of calcium entry, but also the contraction-independent routes. It would therefore appear likely that these latter routes involve calcium uptake by the mitochondria. OXIDATIVE PHOSPHORYLATION AND REPERFUSION-INDUCED CALCIUM GAIN AFTER 60 MINUTES OF ISCHEMIA
These experiments were designed to investigate whether inhibiting the production of adenosine triphosphate by oxidative phosphorylation attenuates the reperfusion-induced gain in calcium after a longer period of ischemia. 2,4-dinitrophenol was used to uncouple oxidative phosphorylation during reperfusion, or the hearts
FiGURE 5. Electron micrograph from the heart reperfused for 30 minutes with KrebsHenseleit buffer containing 1 mM 2,4-dinitrophenol after 30 minutes of ischemia. Note the intact sarcolemma and lack of contraction bands. (Magnification X 11,400, reduced by
30%.)
FIGURE 6. Effect of 1 mM 2,4-dinitrophenot or substrate-free hypoxic reperfusion on the reperfusion-induced calcium gain after 60 minutes of global ischemia. Each column represents the mean i: standard error of 6 separate experiments. Statistics refer to a comparison to the calcium content with Krebs-Henseleit buffer reperfusion. DNP = 2,4-dinitrophenol; glue-free/ nit-glucose-free Krebs-Henseleit buffer gassed with ~~trQ~~~~ KH = Krebs-Hen lieit buffer.
15 min
30 min
aerobic
reperksion
i-HE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989
A SYMPOSIUM:
MYOCARDIAL
ISCHEMIA
FIGURE 7. Electron micrograph from a heart reperked for 30 minutes with Krebs-Henseleit buffer containing 1 mM 2.4.dinitroohenol after SO minutes ofkchemia.‘Note theioss of sarcolemmal integrity. (Magnification X 13,200, reduced by 30%)
were reperfused with substrate-free Krebs-Henseleit buffer gassed with 95% nitrogen/5% carbon dioxide after 60 minutes of ischemia. Because the results of both of these interventions were similar they will be discussed together. Reperfusion with Krebs-Henseleit buffer after 60 minutes of ischemia resulted in cellular calcium overloading similar to that found during reperfusion after 30 minutes of ischemia (Fig. 6). The addition of 1 mM 2,4-dinitrophenol coincident with reperfusion, or reperfusion with substrate-free hypoxic buffer, significantly reduced the calcium gain during the first 15 minutes of reperfusion (Fig. 6). However, after 30 minutes of reperfusion there was no difference, suggesting that after 60 minutes of ischemia the reperfusion-induced calcium gain has been delayed but not abolished (Fig. 6). Why is it that 2,4-dinitrophenol can abolish the reperfusion-induced calcium gain after 30 minutes of ischemia but only delay it after 60 minutes of ischemia? One possible explanation emerged from studies of the ultrastructure. After 30 minutes of ischemia, reperfusion for 30 minutes in the presence of 1 mM 2,4-dinitrophenol resulted in cells that displayed an intact sarcolemma and lacked contraction bands (Fig. 5). A similar picture was found after reperfusion for 10 minutes in the presence of 1 mM 2,4-dinitrophenol after 60 minutes of ischemia. However, when reperfusion after 60 minutes of ischemia was continued beyond 10 minutes, sarcolemmal disruptions became apparent (Fig. 7), although contraction bands were still not formed. The number of cells showing a loss of sarcolemmal integrity increased with increasing length of reperfusion. The late gain in calcium can be explained, therefore, in terms of its entry through a disrupted sarcolemma. These results indicate that the cell can accumulate calcium even though calcium uptake by the mitochondria is inhibited. This provides further support for the conclusion that the sarcolemma was intact during reperfusion in the presence of 2,4-dinitrophenol after 30 minutes of ischemia; otherwise, we would have expected to see an increase in the calcium content. The reason for the development of sarcolemmal disruptions during reperfusion with 2,4-dinitrophenol after 60 but not 30 minutes of ischemia is not obvious. Possibly, 12E
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
63
it is due to the greater damage caused by the longer period of ischemia. More osmotically active breakdown products would have accumulated within the cells during the ischemic episode,i7 placing added stresson the membrane due to cell swelling. In the absence of 2,4-dinitrophenol this may not be important, since the resumption of contractile activity may sufficiently stress the membrane to cause its rupture after 30 or 60 minutes of ischemia. CONCLUSIONS
Reperfusion after 30 or 60 minutes of ischemia results in the cells becoming overloaded with calcium. The cells undergo an extensive contracture, with the formation of contraction bands and loss of sarcolemmal integrity (Fig. 1, 2 and 6). If contractile activity is inhibited during reperfusion with 2,3-butanedione monoiiime, after 30 minutes of ischemia the calcium gain is reduced but not abolished. This suggests that the reperfusion-induced calcium entry involves calcium entry via routes that are both contraction dependent and contraction independent. Reperfusion with 1 mM 2,4-dinitrophenol after 30 minutes of ischemia prevented calcium entry through both of these routes, suggesting a role for mitochondrial calcium uptake in the contraction-independent route of calcium entry. After 60 minutes of ischemia, reperfusion with 2,4dinitrophenol or a nitrogen-saturated, glucose-free buffer delayed but did not abolish reperfusion-induced calcium gain. This late gain in calcium can probably be explained in terms of calcium entry through sarcolemmal holes that deveiop independently of contractile activity and that are readily demonstrable by standard electron microscopic techniques. REFERENCES 1. Shen AC, Jennings RB. Myocardial calcium, and magnesium in acute ischemic injury. Am J Path01 1972;67:417-440. 2. Shen AC, Jennings RB. Kinetics of calcium accumulation in acute myocardial ischemic injury. Am J Pathol 1972;67:441-452. 3. Hearse DJ. Reperfusion of the ischemic myocardium. J Mel Cell Cardiol 1977:9:605-616. 4. Jennings RB, Schaper J, Hill ML, Steenbergen C Jr, Reimer KA. Effect of reperfusion late in the phase of reversible ischemic injury. Changes in cell volume, electrolytes, metabolites, and ultrastructure. circ Res 1985;565:262-278. 5. Reimer K.4, Jennings RB, Hill ML. Total ischemia in dog hearts, in vitro. High
energy phosphate depletion and associated defects in energy metabolism, cell volume regulation and sarcolemmal integrity. Circ Res 1981;49:901-911. 5. Alien BS, Okamoto F, Buckberg GD, Bugyi H, Young H, Leaf J, Beyersdorf F, Sjostrand F, Maloney JV. Studies of controlled reperfusion after &hernia. XV. Immediate functional recovery after six hours of regional ischemia by careful control of conditions of reperfusion and composition of reperfusate. J Thorac Cardiouasc Surg 1986,92:621-635. 7. Nayler WG. The role of calcium in the ischemic myocardium. Awl J Pathol 1981;102:262-270. 8. Shine Kl, Douglas AM. Low calcium reperfusion of ischemic myocardium. J Mel Cell Cardiol 1983:15:251-260. 3. Nayler WG, Elz JS. Reperfusion injury: laboratory artifact or clinical dilemma. Circulation 1986;74:215-221. 10. Bourdilon PD, Poole-Wilson PA. The effects of verapamil quiescence and cardioplegia on calcium exchange and mechanical function in ischemic rabbit myocardium. Circ Res 1982;50:360-368. 11. Jennings RB, Stecnbergen C, Kinney RB, Hill ML, Reimer KA. Comparison
of the effect of ischaemia and anoxia on the sarcolemma of the dog heart. Bur Heart J 1983;4:123-137. 12. Ganote CE, Kaltenbach JP. Oxygen-induced enzyme release: early events and a proposed mechanism. J Ma1 Cell Cardiol 1979;1/:389-406. ta. Nakanishi T, Nichioka K, Jarmakani JM. Mechanism of tissue CaZ+ gain during reoxygenation after hypoxia in rabbit myocardium, Am J Physiol 1982;242:H437mH449. 14. Li T, Sperclakis N, TenEick RE, Solaro JR.,Effects of diacetyl monoxime on cardiac excitation-contraction coupling. J Pharm Exp Ther 1985;232:688-695. 15. Blanchard EM, Mulieri LA, Alpert NR. The effect of 2,3-butanedione monoxime (BDM) on the relation between initial heat and mechanical output of rabbit papillary muscle (abstr). Siophys J 1984;45:48a. 16. Mulieri LA, Alpert NR. Differential effects of 2,3-butanedione monoxime (BDM) on activation and contraction (abstr). Biophys J 1984;45:47a. 17. Jennings RB, Reimer KA, Steenbergen C. Myocardial ischemia revisited. The osmolar load, membrane damage, and reperfusion. J Mel Cell Cardiol 1986;18:769-780.
THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989
Reperfurion-induced calcium gain provides a marker of irreversible injury, but whether the ceils gain calcium because of irreversible injury caused by the ischemic episode, or whether it is the reperfusioninduced calcium gain that triggers the irreversible injury has yet to be established. Using isolated rat hearts made ischemic for either 30 or 60 minutes, and reperfusing with KrebsHenseieit buffer or Krebs-Henseieit buffer containing either 2,3-butanedione monoxime (to inhibit contractile activity) or 2,4-dinitrophenoi or nitrogen-gassed substrate-free Krebs-Henseieit buffer (to inhibit oxidative phosphoryiation), the effect of reperfusion was monitored in terms of calcium gain and ultrastructural changes including loss of sarcolemma1 integrity. The results establish that the routes of calcium entry during postischemic reperfusion are complex. The calcium gain can occur in the absence of mitochondriai oxidative phosphoryiation and is moduiated by interventions introduced at the moment of reperfusion which affect the contractile state. There are at least 2 routes of calcium entry: contractiondependent and contraction independent. The former is probably associated with the development of sarcoiemmai discontinuities. The results also establish that when sarcoiemmai integrity has been destroyed, the cells can gain excess calcium under conditions that prevent mitochondriai calcium uptake. Accordingly, the mitochondria cannot be the only intracellular organeiies that accumulate caiciurn under these conditions. Additional studies are needed to identify the other sites of calcium bindin under conditions of adenosine triphosphate deprivation. (Am J Cardioi f989;63:
From the Department of Medicine, University of Melbourne, Austin Hospital, Heidelberg, Victoria 3084, Australia. This study was supported in part by a grant from the National Health and Medical Research Council of Australia. Address for reprints: Winifred 6. Nayler, DSc, Department of Medicine, University of Melbourne, Austin Hospital, Heidelberg, Victoria 3084. Australia.
Es.2(HOGS),
ells of hearts that are made permanently ischemic1>2 do not accumulate calcium. However, reperfusion1,2 after 20 or more minutes of normothermic ischemia causes an uncontrolled gain in calcium, which in turn can precipitate or hasten cell death and tissue necrosis. Although there is no doubt that ischemic tissue must be reperfused if it is to survive, what remains questionable is whether the fate of the cells is predetermined by the ischemic episode, with reperfusion and the associated gain in calcium merely expressing this ultimate fate, or whether reperfusion itself can be damaging.3
C
REPERFUSlON
INJURY:
OES I+ EXIST?
After an ischemic episode 2 populations of cells exist-those that are reversibly injured and those that are irreversibly injured. After only a short period of ischemia, cells are reversibly injured and reperfusion results in their recovery.4 After extended periods of ischemia, however, irreversible cellular injury may occur and reperfusion at this time can be of no benefit5 Of particular importance is the condition under which reperfusion fails to result in recovery, yet irreversible damage to the cells is not evident before reperfusion. There are at least 2 explanations for such a condition. The cells may have been lethally injured by the ischemic episode, with reperfusion merely expressing this lethal damage and possibly exacerbating it. Alternatively, the cells may only have been reversibly injured during the ischemic episode; however, reperfusion, with its attendant reintroduction of an unlimited supply of calcium, causes further damage that results in lethal injury. If the first possibility is correct and the fate of the cells is already determined during the ischemic episode, then any intervention that may be introduced during reperfusion cannot be beneficial. Arguing against this and in favor of the second possibility are reports in which recovery after an ischemic episode is favored by altering the composition of the reperfusion buffer.6-8 In 1 such report,6 reperfusion with substrate-enriched blood cardioplegic solution after 6 hours of regional ischemia resulted in recovery, whereas reperfusion with blood after this or even shorter periods of ischemia failed to restore any contractile activity, and structural damage was evident. Other studies7x8have shown that recovery is favored if the initial reperfusion is with a solution containing a low calcium concentration. PORTANC
The reports7J previously mentioned suggest that calcium entry during reperfusion may be deleterious, and THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7. 1989
A SYMPOSIUM:
MYOCARDIAL
ISCHEMIA
there is no doubt that cellular calcium overloading can occur during reperfusion. Cellular calcium will increase whenever calcium influx is greater than calcium efflux. This will occur if calcium influx increases while calcium efflux remains the same or decreases, or if calcium efflux decreases while calcium influx remains constant. If any of these situations continues for an extended period, calcium overloading will occur. Calcium efflux from the cell occurs primarily in exchange for sodium and via the sarcolemma1 adenosine triphosphatase pump. Both of these processes of calcium extrusion depend either directly or indirectly on the availability of adenosine triphosphate, and because only limited energy supplies remain after an ischemic episode, it is not hard to envisage that cytosolic calcium will increase. A raised cytosolic calcium will in turn hasten the breakdown of any remaining adenosine triphosphate, and activation of the various phospholipases and proteases. Even the generation of free radicals is, to some extent, calcium dependent. Although a raised cellular calcium is obviously detrimental to the cells, the question that remains unanswered is whether calcium entry is the primary event resulting in lethal injury or whether some other lethal event has already occurred to which calcium entry is a secondary consequence.
ROUTES OF CALCIUM ENTRY Calcium entry during reperfusion may occur through physiologic or nonphysiologic channels, or both. It is unlikely that the predominant route of entry is through the slow channels because slow channel blockers added at the time of reperfusion fail to diminish the calcium gain or aid in recovery.‘0 Entry of calcium by passive perfusion down its concentration gradient or in exchange for sodium are 2 alternative and viable possibilities. Entry through nonphysiologic routes, including through sarcolemma1 disruptions, may also play a role.” However, if this latter route is the predominant route of calcium entry, then this must be secondary to the loss of sarcolemma1 integrity. Some investigators have suggested that entry through a discontinuous sarcolemma is the predominant route of calcium entry during reoxygenation after hypoxia.12 Others believe that mitochondrial calcium uptake is of primary importance.13 The present experiments were designed to investigate these possibilities with reference to calcium gain during postischemic reperfusion. EXPERIMENTAL MODEL In the following experiments, isolated adult SpragueDawley rat hearts were perfused in a nonrecirculating Langendorf mode with Krebs-Henseleit buffer at 37’C
30
kZi KH q BDM l
20
p < 0.01
10
FIGURE 1. Eftect of 10 mM 2,3-butanedione monoxime on the reperlusion-induced cakium gain after 30 minutes of global ischemia. Each column represents the mean f standard error of 6 separate experiments. Statistics refer to the reduction of calcium content in the presence of 2,3-butanedione monoxime compared with that found in its absence. BDM = 2,3-butanedione monoxime; KH = KrebsHenseleit buffer.
0$ 15min
30 min
aerobic
repeifusion
TABLE Creatine
I Effect of 10 mM 2,3-Butanedione Monoxime Phosphate After 30 Minutes of lschemia
on the l?epetfusion-induced
Recovery
of Adenosine
Triphosphate
and
Repetfusion Experimental
Aerobic
lschemic
ATP Olmol/g dry wt) CP Olmol/g dry wt) * p
15.76 f 0.69
2.47 f 0.20
22.22 f 0.81
2.83 f 0.54
KH
BDM
5.73 f 0.57 12.89f
1.12
Each result is the mean f standard error of 6 separate experiments. Statistics refer to the enhancement of adenosine triphosphate minutes reperfusion in the presence of 2.3.butanedione monoxime compared with those found in its absence. ATP = adenosine triphosphate; EIOM = 2,3-butanedione monoxime: CP = creatine phosphate: KH = Krebs-Henseleit buffer.
8E
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
63
9.31 f 0.63” 27.16 f 2.35’
and creatine
phosphate
levels during 15
and at a constant flow rate of 10 ml/min. These hearts were allowed to beat spontaneously to avoid the release of norepinephrine. Global ischemia was induced by the cessation of all flow to the heart. This model was chosen rather than other in vitro models involving muscle strips, lo because the latter do not have an intact circulation and ischemia can only be mimicked by surrounding the tissue with nitrogen-saturated solutions or by sealing them in nitrogen-gassed containers. Reperfusion is then accomplished by diffusion from the bathing solution. Other models using hypoxia12J3 were also not used, because the continual flow involved in these models does not allow for the accumulation of end products, as occurs in ischemia. An in vivo model4 was not chosen, although this would obviously be more representative of the human situation, since not all the conditions can be controlled. In our in vitro model, all cells were made uniformly ischemic, whereas in an in vivo model collateral flow of varying degrees is available, resulting in varying degrees of ischemia. However, care must be taken when extrapolating results from 1 model to another. Although we believe that our results are applicable to the in vivo situation, it is possible that the time changes observed under our experimental conditions may differ from that obtained under other conditions. After each perfusion the hearts were perfusion-fixed for electron microscopy or they were analyzed for calcium content as previously described.7 Briefly, the coronary vasculature of hearts to be analyzed for calcium content was flushed free of calcium with an ice-cold sucrose/
histidine solution, and the ventricles then dried to constant weight at 100°C. The dried ventricles were digested in concentrated nitric acid and an aliquot of this extract diluted in a calcium blank solution containing potassium chloride and lanthanum chloride. The calcium content was analyzed by atomic absorption spectrophotometry7 and compared to standards prepared by serial dilution of a calcium stock solution. CONTRACTILE ACTIVITY AND REPERFUSION-INDUCED CALCIUM AFTER 30 MINUTES OF ISCHEMIA
GAIN
To investigate the role of contractile activity during reperfusion on the gain in calcium, isolated perfused rat hearts were made globally ischemic for 30 minutes before reperfusion in the presence or absence of 10 mM 2,3butanedione monoxime. To inhibit contractile activity, 2,3-butanedione monoxime was used. With the addition of 10 mM 2,3-butanedione monoxime to aerobically perfused hearts, contractile activity ceases almost immediately but is restored when it is removed. The presence of 2,3-butanedione monoxime has a negative inotropic effect on both intact and chemically skinned cardiac muscle.14 It reduces the sensitivity of the myofibrils to calcium and inhibits cross-bridge formation, while having no major effect on the electrical activity of the heart or calcium release from the sarcoplasmic reticulum.14-l6 In the absence of 2,3-butanedione monoxime, the hearts rapidly gained calcium on reperfusion after 30 minutes of ischemia, so that after 30 minutes’ reperfusion
FIGURE 2. Electron micrograph from a heart reperfused for 30 minutes with Krebs-Henseleit buffer after 30 minutes of ischemia. Note the contraction bands. Magnification X 7,000, reduced by 30%.)
FIGURE 3. Electron micrograph from a heart reperfused for 30 minutes with Krebs-Henseieit buffer containing 10 mM 2,3-butenedione monoxime after 30 minutes of ischemia. The myofibrils are intact and sarcolemmal integrity maintained. Note also the presence of glycogen. (Magnification X 7,OGG, reduced by 309/o.)
THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989
9E
A SYMPOSIUM:
MYOCARDIAL
ISCHEMA
with Krebs-Henseleit buffer they contained 27.85 f 1.94 pmol calcium/g dry weight (Fig. 1). The addition of 10 mM 2,3-butanedione monoxime during reperfusion reduced this calcium gain but did not completely abolish it (Fig. 1). This reduction of calcium gain can be explained neither in terms of an altered reflow area in the presence of 2,3-butanedione monoxime nor as an effect of the increased osmolarity of the reperfusion buffer due to its presence. In addition to reducing the gain in calcium, 2,3-butanedione monoxime enhanced the recovery of adenosine triphosphate and creatine phosphate (Table I) and reduced the ultrastructural damage to the cells during reperfusion, In the absence of 2,3-butanedione monoxime, the cells displayed extensively contracted myofibrils with contraction bands. The mitochondria were frequently swollen and sarcolemmal disruptions were apparent (Fig. 2). The formation of contraction bands was prevented by 2,3-butanedione monoxime and the sarcolemma remained intact. Most of the mitochondria showed a normal appearance and glycogen deposits were present in the cytoplasm (Fig. 3). This structural “protection” was only maintained while 2,3-butanedione monoxime was present. Removal of 2,3-butanedione monoxime and continued perfusion with Krebs-Henseleit buffer resulted in the formation of contraction bands, a loss of sarcolemmal integrity and swollen mitochondria-changes similar to those found with reperfusion in the absence of 2,3-butanedione monoxime. At the same time there was a rapid increase in calcium, to 24.47 f 1.51 r.Lmol/g dry weight. These results show that by inhibiting contractile activity during reperfusion after 30 minutes of ischemia, the reperfusion-induced gain in calcium is reduced but not abolished. This suggests that calcium entry during reper-
fusion may occur via routes that are contraction dependent and those that are contraction independent. In the presence of 2,3-butanedione monoxime, calcium entry via contraction-dependent routes is inhibited. The present results suggest that calcium entry via the contractiondependent routes may involve entry through a disrupted sarcolemma. Other investigators have argued that this is a major route of calcium entry during reoxygenation after hypoxia.12 However, after 30 minutes of ischemia, entry via sarcolemmal “holes” cannot be the sole route of calcium entry during reperfusion. Some calcium entry must be occurring across an intact sarcolemma through contraction-independent routes. This entry would be masked by the more rapid entry of calcium through a disrupted sarcolemma in the absence of 2,3-butanedione monoxime. The results also show that reducing the reperfusioninduced calcium gain does not result in the hearts resuming their preischemic state. This may imply that the calcium gain that still occurred is deleterious to the cells or that the fate of the cells is already determined by the ischemic episode and we have merely delayed the full expression of this damage. The latter hypothesis seems unlikely since we would not expect lethally injured cells to be capable of synthesizing adenosine triphosphate and creatine phosphate, nor would we expect to find glycogen deposits as seen in Figure 3. OXIDATIVE PHOSPHORYLATION AND REPERFUSION-INDUCED CALCIUM GAIN AFTER 30 MINUTES OF ISCHEMIA Oxidative phosphorylation was uncoupled during reperfusion after 30 minutes of ischemia, by adding 1 mM 2,4-dinitrophenol. This not only prevents the synthesis of
l
pco.01 FIGURE 4. Eflect of 1 mM ?,Cdinitrophenol on the reperk&n-induced calcium gain after 30 minutes of global ischemia. Each column represents the mean f standard error of 6 separate experiments. Statistics refer to the reduction of calcium content in the presence of 2,4-dinitrophenol compared with that found in its absence. DNP = 2,4-dinitrophend; KH = Krebs-lfensekit bulfer.
15 min
30 min
aerobic
repetfusion
1OE
THE AMERICAN
JOURNAL
OF CARDIOLbGY
VOLUME
63
adenosine triphosphate, but also inhibits the uptake of calcium by the mitochondria. The addition of 1 mM 2,4dinitrophenol during reperfusion attenuated the reperfusion-induced gain in calcium throughout 30 minutes of reperfusion (Fig. 4). It also prevented the ultrastructural damage normally seen with Krebs-Henseleit buffer reperfusion. The formation of contraction bands was inhibited and the sarcolemma remained intact (Fig. 5). Because active tension development cannot occur in the absence of adenosine triphosphate, we would expect that 2,4-dinitrophenol would prevent calcium entry via contraction-dependent routes. Hence, we may expect similar results with respect to reperfusion-induced calcium gain as we found when 2,3-butanedione monoxime was used to inhibit contractile activity. However, in contrast to the results obtained when 2,3-butanedione monoxime was
present, 2,4-dinitrophenol abolished the calcium gain. This suggests that 2,4-dinitrophenol inhibits not only the contraction-dependent routes of calcium entry, but also the contraction-independent routes. It would therefore appear likely that these latter routes involve calcium uptake by the mitochondria. OXIDATIVE PHOSPHORYLATION AND REPERFUSION-INDUCED CALCIUM GAIN AFTER 60 MINUTES OF ISCHEMIA
These experiments were designed to investigate whether inhibiting the production of adenosine triphosphate by oxidative phosphorylation attenuates the reperfusion-induced gain in calcium after a longer period of ischemia. 2,4-dinitrophenol was used to uncouple oxidative phosphorylation during reperfusion, or the hearts
FiGURE 5. Electron micrograph from the heart reperfused for 30 minutes with KrebsHenseleit buffer containing 1 mM 2,4-dinitrophenol after 30 minutes of ischemia. Note the intact sarcolemma and lack of contraction bands. (Magnification X 11,400, reduced by
30%.)
FIGURE 6. Effect of 1 mM 2,4-dinitrophenot or substrate-free hypoxic reperfusion on the reperfusion-induced calcium gain after 60 minutes of global ischemia. Each column represents the mean i: standard error of 6 separate experiments. Statistics refer to a comparison to the calcium content with Krebs-Henseleit buffer reperfusion. DNP = 2,4-dinitrophenol; glue-free/ nit-glucose-free Krebs-Henseleit buffer gassed with ~~trQ~~~~ KH = Krebs-Hen lieit buffer.
15 min
30 min
aerobic
reperksion
i-HE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989
A SYMPOSIUM:
MYOCARDIAL
ISCHEMIA
FIGURE 7. Electron micrograph from a heart reperked for 30 minutes with Krebs-Henseleit buffer containing 1 mM 2.4.dinitroohenol after SO minutes ofkchemia.‘Note theioss of sarcolemmal integrity. (Magnification X 13,200, reduced by 30%)
were reperfused with substrate-free Krebs-Henseleit buffer gassed with 95% nitrogen/5% carbon dioxide after 60 minutes of ischemia. Because the results of both of these interventions were similar they will be discussed together. Reperfusion with Krebs-Henseleit buffer after 60 minutes of ischemia resulted in cellular calcium overloading similar to that found during reperfusion after 30 minutes of ischemia (Fig. 6). The addition of 1 mM 2,4-dinitrophenol coincident with reperfusion, or reperfusion with substrate-free hypoxic buffer, significantly reduced the calcium gain during the first 15 minutes of reperfusion (Fig. 6). However, after 30 minutes of reperfusion there was no difference, suggesting that after 60 minutes of ischemia the reperfusion-induced calcium gain has been delayed but not abolished (Fig. 6). Why is it that 2,4-dinitrophenol can abolish the reperfusion-induced calcium gain after 30 minutes of ischemia but only delay it after 60 minutes of ischemia? One possible explanation emerged from studies of the ultrastructure. After 30 minutes of ischemia, reperfusion for 30 minutes in the presence of 1 mM 2,4-dinitrophenol resulted in cells that displayed an intact sarcolemma and lacked contraction bands (Fig. 5). A similar picture was found after reperfusion for 10 minutes in the presence of 1 mM 2,4-dinitrophenol after 60 minutes of ischemia. However, when reperfusion after 60 minutes of ischemia was continued beyond 10 minutes, sarcolemmal disruptions became apparent (Fig. 7), although contraction bands were still not formed. The number of cells showing a loss of sarcolemmal integrity increased with increasing length of reperfusion. The late gain in calcium can be explained, therefore, in terms of its entry through a disrupted sarcolemma. These results indicate that the cell can accumulate calcium even though calcium uptake by the mitochondria is inhibited. This provides further support for the conclusion that the sarcolemma was intact during reperfusion in the presence of 2,4-dinitrophenol after 30 minutes of ischemia; otherwise, we would have expected to see an increase in the calcium content. The reason for the development of sarcolemmal disruptions during reperfusion with 2,4-dinitrophenol after 60 but not 30 minutes of ischemia is not obvious. Possibly, 12E
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
63
it is due to the greater damage caused by the longer period of ischemia. More osmotically active breakdown products would have accumulated within the cells during the ischemic episode,i7 placing added stresson the membrane due to cell swelling. In the absence of 2,4-dinitrophenol this may not be important, since the resumption of contractile activity may sufficiently stress the membrane to cause its rupture after 30 or 60 minutes of ischemia. CONCLUSIONS
Reperfusion after 30 or 60 minutes of ischemia results in the cells becoming overloaded with calcium. The cells undergo an extensive contracture, with the formation of contraction bands and loss of sarcolemmal integrity (Fig. 1, 2 and 6). If contractile activity is inhibited during reperfusion with 2,3-butanedione monoiiime, after 30 minutes of ischemia the calcium gain is reduced but not abolished. This suggests that the reperfusion-induced calcium entry involves calcium entry via routes that are both contraction dependent and contraction independent. Reperfusion with 1 mM 2,4-dinitrophenol after 30 minutes of ischemia prevented calcium entry through both of these routes, suggesting a role for mitochondrial calcium uptake in the contraction-independent route of calcium entry. After 60 minutes of ischemia, reperfusion with 2,4dinitrophenol or a nitrogen-saturated, glucose-free buffer delayed but did not abolish reperfusion-induced calcium gain. This late gain in calcium can probably be explained in terms of calcium entry through sarcolemmal holes that deveiop independently of contractile activity and that are readily demonstrable by standard electron microscopic techniques. REFERENCES 1. Shen AC, Jennings RB. Myocardial calcium, and magnesium in acute ischemic injury. Am J Path01 1972;67:417-440. 2. Shen AC, Jennings RB. Kinetics of calcium accumulation in acute myocardial ischemic injury. Am J Pathol 1972;67:441-452. 3. Hearse DJ. Reperfusion of the ischemic myocardium. J Mel Cell Cardiol 1977:9:605-616. 4. Jennings RB, Schaper J, Hill ML, Steenbergen C Jr, Reimer KA. Effect of reperfusion late in the phase of reversible ischemic injury. Changes in cell volume, electrolytes, metabolites, and ultrastructure. circ Res 1985;565:262-278. 5. Reimer K.4, Jennings RB, Hill ML. Total ischemia in dog hearts, in vitro. High
energy phosphate depletion and associated defects in energy metabolism, cell volume regulation and sarcolemmal integrity. Circ Res 1981;49:901-911. 5. Alien BS, Okamoto F, Buckberg GD, Bugyi H, Young H, Leaf J, Beyersdorf F, Sjostrand F, Maloney JV. Studies of controlled reperfusion after &hernia. XV. Immediate functional recovery after six hours of regional ischemia by careful control of conditions of reperfusion and composition of reperfusate. J Thorac Cardiouasc Surg 1986,92:621-635. 7. Nayler WG. The role of calcium in the ischemic myocardium. Awl J Pathol 1981;102:262-270. 8. Shine Kl, Douglas AM. Low calcium reperfusion of ischemic myocardium. J Mel Cell Cardiol 1983:15:251-260. 3. Nayler WG, Elz JS. Reperfusion injury: laboratory artifact or clinical dilemma. Circulation 1986;74:215-221. 10. Bourdilon PD, Poole-Wilson PA. The effects of verapamil quiescence and cardioplegia on calcium exchange and mechanical function in ischemic rabbit myocardium. Circ Res 1982;50:360-368. 11. Jennings RB, Stecnbergen C, Kinney RB, Hill ML, Reimer KA. Comparison
of the effect of ischaemia and anoxia on the sarcolemma of the dog heart. Bur Heart J 1983;4:123-137. 12. Ganote CE, Kaltenbach JP. Oxygen-induced enzyme release: early events and a proposed mechanism. J Ma1 Cell Cardiol 1979;1/:389-406. ta. Nakanishi T, Nichioka K, Jarmakani JM. Mechanism of tissue CaZ+ gain during reoxygenation after hypoxia in rabbit myocardium, Am J Physiol 1982;242:H437mH449. 14. Li T, Sperclakis N, TenEick RE, Solaro JR.,Effects of diacetyl monoxime on cardiac excitation-contraction coupling. J Pharm Exp Ther 1985;232:688-695. 15. Blanchard EM, Mulieri LA, Alpert NR. The effect of 2,3-butanedione monoxime (BDM) on the relation between initial heat and mechanical output of rabbit papillary muscle (abstr). Siophys J 1984;45:48a. 16. Mulieri LA, Alpert NR. Differential effects of 2,3-butanedione monoxime (BDM) on activation and contraction (abstr). Biophys J 1984;45:47a. 17. Jennings RB, Reimer KA, Steenbergen C. Myocardial ischemia revisited. The osmolar load, membrane damage, and reperfusion. J Mel Cell Cardiol 1986;18:769-780.
THE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
7, 1989