Elevated cardiac

myocardial death
E. BILLMAN, GEORGE

calcium
BRIAN McILROY,

and
ANDJ.

its role
DAVID JOHNSON Biochemistry,

in sudden

Department of Physiology Ohio 43210, USA

and Department

of Medical

The Ohio

State University,

Columbus,

ABSTRACT The contribution of intracellular calcium to ventricular fibrillation (VF) was investigated using chronically instrumented dogs with healed myocardial infarctions. A 2-minute coronary occlusion was initiated during the last minute of exercise. Fourteen animals developed ventricular fibrillation (susceptible) whereas the remaining 12 did not (resistant) during this exercise plus ischemia test. The test was then repeated for the susceptible animals after pretreatment with the intracellular calcium chelator BAPTA-AM (1.0 mg/kg). BAPTA-AM significantly reduced left ventricular dp/dt max and prevented VF in 8 of 12 susceptible animals. Conversely, myocardial cytosolic calcium levels were increased in resistant animals using the calcium channel agonist Bay K 8644 (30 jig/kg) or phenylephrine (10 jig#{149} kg min 3-5 mm before occlusion). Bay K 8644 induced VF in all 5 resistant animals tested whereas phenylephrine induced VF in 8 of 12 resistant animals. BAPTA-AM pretreatment attenuated the hemodynamic effects of Bay K 8644 or phenylephrine and prevented VF in five of five Bay K 8644- and four of seven phenylephrine-treated animals. Finally, the endogenous level of calcium/calmodulin (Ca-CaM)-dependent phosphorylation of 170- and 55-kDa substrate proteins was measured (as an index of intracellular free calcium concentration). In the susceptible dog heart, the endogenous level of Ca-CaM-dependent phosphorylation was estimated to be two- to threefold higher than that observed in resistant dog heart. Treatment of resistant dog tissue with the calcium ionophore A23187 increased the level of Ca-CaM-dependent phosphorylation of these two proteins to the level observed in susceptible dog heart. These data suggest that elevated cytosolic calcium facilitates development of malignant arrhythmias and that elevated cytosolic calcium levels may be present in animals particularly susceptible to ventricular fibrillation. Billman, G. E.; Mcllroy, B.; Johnson, J. D. Elevated myocardial calcium and its role in sudden cardiac death. FASEBJ. 5: 2586-2592; 1991. Key Words: calcium a cardiac arrhythrnias ventricular and fi-adrenergic receptors fibrillation cytosolic

The mechanism by which these alterations in autonomic tone provoke disturbances in cardiac electrical stability, particularly at the cellular level, remains largely unknown. Ultimately, transmitter substances released from nerve terminals must bind with postsynaptic receptors on the cardiac cells to alter the concentration of second messengers (calcium, cyclic AMP, cyclic GMP, diacylglyceral, or ITP), protein phosphorylation, and myocardial contractility (7-10). It is probable that these alterations in second-messenger levels elicited by autonomic neural activity contribute significantly to the formation of malignant arrhythmias. In particular, large excesses in intracellular calcium can provoke oscillatory after-depolarization of cardiac membrane, and if sufficient to reach threshold, can provoke spontaneous action potential generation (10). This triggered activity is believed to be a potential mechanism for initiation of ventricular fibrillation (11). In fact, recent studies (32) using isolated rabbit hearts demonstrated that a slow inward calcium current was required for initiation and maintenance of ventricular fibrillation. In intact preparations, Billman (12) demonstrated that both organic (verapamil, nifedipine) and inorganic (magnesium) calcium channel antagonists could prevent VF elicited by the combination of exercise and ischemia whereas the calcium channel agonist Bay K 8644 provoked VF in animals resistant to malignant arrhythmias. These findins suggest that calcium entry and subsequent elevation in cytosolic calcium contribute significantly to the development of VF. Thus, this series of experiments investigates the effects of the intracellular calcium chelator, BAPTA-MA, on susceptibility to ventricular fibrillation. Preliminary results of some aspects of this work have been reported elsewhere (13).

METHODS Surgical preparation

A GROWING BODY OF EVIDENCE suggests that disturbances in the autonomic control of the heart play a critical role in the development of ventricular fibrillation (VF)2 In general, activation of the sympathetic nervous system tends to reduce cardiac electrical stability whereas parasympathetic activation may protect against malignant arrhythmias (1). Recent clinical (2, 3) and experimental studies (4-6) demonstrated that individuals with the greatest increases in sympathetic tone and/or reductions in parasympathetic tone after myocardial infarction also have the greatest propensity for sudden cardiac death.

Forty-two mongrel dogs weighing 11.6-20.7 kg were used in this study. The animals were anesthetized and instrumented to measure left circumflex coronary blood flow, left ventricular pressure, and ventricular electrogram as previously described (4-6, 12). The left anterior descending artery was ligated and an anterior wall myocardial infarction was then produced (4-6, 12).

To whom correspondence should be addressed, at: Department of Physiology, Ohio State University, 4196 Graves Hall, 333 W. 10th Ave., Columbus, Ohio 43210, USA. 2Abbreviations: VF, ventricular fibrillation; BAPTA-AM, the acetoxymethyl ester of 1,2 bis-(O-aminophenoxy ethane) N,N,N ,Ntetraacetic acid; LVP, left ventricular pressure; LVSP, left ventricular diastolic pressure; LVDP, left ventricular systolic pressure; CaM, calmodulin; DMSO, dimethyl sulfoxide.

2586

092-663811005-2586101.50.

FASEB

All leads to the cardiovascular instrumentation were tunneled under the skin to exit on the back of the animals neck. Pentazocine lactate (Talwin, Winthrop-Breon Lab., New York, N.Y., 30 mg., i.m.) was given to minimize postoperative pain. In addition, the long-acting local anesthetic bupivacaine HCI (Marcaine, Winthrop-Breon Lab.) was used to block the intercostal nerves (i.e., pain fibers) in the area of the incision to minimize discomfort to the animal. Each animal was placed on antibiotic therapy (penicillin G x 106 units, i.m., Burns Veterinary Supply, Oakland, Calif.) twice daily for 7 days. The animals were placed in an intensive care setting for the first 24 h and given antiarrhythmic drugs for the next 4 days, as previously described (4-7, 12). Twelve dogs died suddenly during surgery or within the first 72 h. Four additional animals could not be classified because of rupture of the coronary occluder (see below) and were also eliminated from the study. Of the original 42 animals, 26 entered the studies described below. The principles governing the care and treatment of animals as expressed by The American Physiological Society were followed throughout this study. Classification of animals: exercise plus ischemia test

reductions in heart rate. Atropine did not induce malignant arrhythmias in these animals. One week later the Bay K 8644 or phenylephrine studies were repeated after pretreatment with BAPTA-AM as described previously. A final control (DMSO) exercise plus ischemia test was repeated 1 wk after the BAPTA-AM studies. Biochemical assays

Three to four weeks after production of the myocardial infarction the studies began. The animals were walked on a motor-driven treadmill and adapted to the laboratory environment during this period. The susceptibility to ventricular fibrillation was tested as previously described (4-6, 12). Briefly, the animals ran on a motor-driven treadmill while work load increased every 3 mm. During the last minute of exercise, the left circumflex coronary artery was occluded, the treadmill was then stopped, and the occlusion was maintained for an additional minute. The occlusion lasted a total of 2 mm. Large metal plates (11-cm diameter) were placed across the animals chest so that electrical defibrillation could be achieved with a minimal delay but only after the animal was unconscious (10-20 s after VF began). This model has been shown to induce ventricular fibrillation reliably and reproducibly (4-6, 12) in susceptible animals. The control (untreated) exercise plus ischemia test was repeated as many as four times in each animal. The time to ventricular fibrillation was averaged for each animal. Experimental protocol

Five to seven days after the last exercise plus ischemia test animals were euthanized with an overdose of sodium pentobarbitol followed by saturated KC1. The heart was rapidly removed (within 5-10 mm) and tissue samples were frozen in a - 70#{176}C freezer for future biochemical analysis. Thirty-milligram samples of posterior papillary muscle obtained from susceptible (n = 5) and resistant (n = 5) animals were incubated for 10 mm in 1 ml of a physiological saline solution either in the presence or absence of 10 jiM A23187. After incubation, 100 mM NaPO4 and 10 mM NaF were added to inhibit protein phosphatases; the tissue was homogenized with ground glass homogenizer and the protein concentration was determined by the Bradford method (15). A portion of the homogenized tissue (250 jig total protein) was incubated in 0.2 ml of buffer containing 50 mM NaPO4, 5 mM NaF, 3 mM MgC12, pH 7.0, with either no calcium (+1 mM EGTA), calcium (2 mM CaC12), or calcium plus 5 jiM calmodulin. In each case the phosphorylation reaction was started by the addition of 1 jiM [32P]ATP (20,000 cpm/nmol) for 10 mm and stopped by addition of 40 jil of stop solution (1.2% DTE, 10% SDS, 40% glycerol). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was then performed on each reaction mixture using 7-15% gradient gels that were stained with Coomassie blue and exposed to Kodak X-AR film for determination of phosphate incorporation into substrate proteins. These autoradiographs were scanned and relative peak areas were determined using an LKB Ultrascan XL. The relative peak area for the 170- and 55-kDa protein in the resistant animals was divided by the corresponding areas obtained from susceptible animals and the increase in phosphorylation of these proteins was calculated. These data were averaged for the five susceptible and five resistant animals. Data analysis

The exercise plus ischemia test described previously was then performed after the following treatments. The susceptible (n = 12) dogs were treated with the intracellular calciumspecific chelator BAPTA-AM [the acetoxymethyl ester of 1,2 bis(O-aminophenoxy ethane)-N,N, -N ,N tetraacetic acid 1 mg/kg, Molecular Probes, Eugene, Oreg.]. The drug was dissolved in I ml of dimethyl sulfoxide (DMSO) and injected into a cephalic vein. Because the drug slowly incorporates into the cell, the exercise plus ischemia test was performed 45-60 mm after the BAPTA-AM injection. One week later the exercise plus ischemia test was repeated with vehicle alone (DMSO). The resistant animals (n = 12) were given either the calcium channel agonist Bay K 8644 (30 jig/kg dissolved in 1 ml 95% ethanol, n = 5)or phenylephrine HCI(10 jig.kg #{149}min, Sigma Chemical Co., St. Louis, Mo., n = 12) to increase intracellular calcium levels. The drugs were injected via a catheter placed in a cephalic vein 3-5 mm before the occlusion, i.e., while the animal was running. Atropine sulfate (50 jigS kg, iv., Lyphomed, Rosemont, Ill.) was given to the animal receiving the phenylephrine infusion to prevent reflex

All data were recorded on a Gould Model 2800S 8 channel chart recorder and a Teac Model MR-30 cassette tape recorder as previously described (12). The hemodynamic data were averaged over 5-s periods during the last minute of each exercise level, immediately before the coronary occlusion and at the 60-s time point (or immediately before VF) during the occlusion. The data were analyzed using analysis of variance for repeated measures. When the F ratio was found to exceed a critical value (P < 0.05), Scheffes test was used to compare the means (16). The effects of the various drug treatments on ventricular fibrillation were determined using a Chi-square test with Yates correction for continuity (17). All data are reported as the mean SEM. Cardiac arrhythmias were analyzed at a paper speed of 25 mm/s while P-R interval was determined at paper speed of 100 mm/s.

RESULTS The dogs were divided to the exercise plus malignant arrhythmias in two groups based on the response ischemia test: 14 animals exhibited (ventricular flutter that deteriorated

ELEVATED CELL CALCIUM

AND

VENTRICULAR

FIBRILLATION

2587

into ventricular fibrillation) and were classified as susceptible to sudden death whereas 12 animals did not exhibit arrhythmias and were designated resistant. The time to onset for malignant arrhythmias in the susceptible animals was 45.4 4.2 s (range 37.0-81.0 s); four animals developed VF shortly after the treadmill was stopped whereas the remaining 10 had VF while running. Two susceptible animals were not successfully resuscitated and were eliminated from the study. In agreement with previous studies, no animals developed VF in response to exercise alone (i.e., the coronary occlusion was necessary to induce the malignant arrhythmias). The arrhythmias were reproducibly induced during each presentation of control exercise plus ischemia tests in the susceptible animals. Conversely, malignant arrhythmias were never noted during control exercise plus ischemia tests for the resistant animals. Effects of the calcium ventricular fibrillation: chelator, BAPTA-AM, susceptible animals on

S.

4000
C

Cu
3000

>E
0
0-

-o

BAPTA-AM
Control

2000

r
0/0 4.8/0 6.4/0 6.4/4 6.4/8 6.4/12 6.4/16

EXERCISE Figure 1. Effect of BAPTA-AM pressure,

LEVEL response to exercise.

Speed (kph)/Grade (%)

on inotropic
S*P < 0.01 control

vs. BAPTA-AM d(LVP)/dt

for a given exercise level. LVP, served as an index of inotropic

left ventricular state.

The intracellular calcium chelator, BAPTA-AM, was given to 12 susceptible animals. This drug significantly (x2 = 9.2, P < 0.005) reduced the incidence of malignant arrhythmias preventing VF in 8 of 12 animals. The onset of VF was delayed in each of the remaining four animals (control 36.1 5.2 s, BAPTA-AM 70.7 17.5 s), with onset delayed to after the occlusion release in one animal. The vehicle, DMSO, failed to prevent VF in any susceptible animal. The hemodynamic effects of BAP1A-AM before and during the coronary occlusion are shown in Table 1. BAPTA-AM significantly reduced the preocclusion values of d(LVP)/dt as well as the change in d(LVP)/dt elicited by the coronary occlusion. In a similar manner, BAPTA-AM significantly reduced the d(LVP)/dt max response to exercise (Fig. 1) and significantly increased preexercise P-R interval 22.0 6.7% (control 84.5 6.9 ms BAPTA-AM 99.8 4.8 ms). Previous studies demonstrated that these variables reflect myocardial free calcium levels (18, 19). The coronary occlusion elicited significant increases in heart rate and left ventricular diastolic pressure (LVDP) where left ventricular systolic pressure (LVSP) and d(LVP)/dt maximum decreased significantly. The hemodynamic reTABLE 1. Hetnodynamic ejects of 1/ic intraerllular
calcium chelator,

sponse to coronary BAPTA-AM.

artery

occlusion

was

not

affected

by

Effect of the calcium channel agonist susceptibility to ventricular fibrillation

Bay

K 8644

on

The calcium channel agonist Bay K 8644 was used to increase calcium entry and thereby elevate cytosolic calcium in five resistant animals. Bay K 8644 elicited VF in all five resistant animals (x2 = 6.4 P < 0.025 time to onset 30.7 3.4 s) and provoked several significant hemodynamic changes as shown in Table 2. In contrast, pretreatment with BAPTA-AM prevented the VF induced by Bay K 8644 (x2 = 6.4, P < 0.025) and significantly attenuated the hemodynamic effects of this drug (Table 2). For example, this intracellular calcium chelator prevented Bay K 8644-induced increases in heart rate and d(LVP)/dt max and attenuated the increase in LVSP. As in susceptible animals, BAPTA-AM significantly reduced predrug (i.e., before Bay K 8644 injection) values of d(LVP)/dt max.
on susceptible animals

BAPTA-AM,

Preocclusion Heart Control BAPTA-AM 207.6 194.6

9.0

Occlusion rate (beats/mm) 252.0 249.8 systolic pressure, 10.3 13.3

Change

8.3 Left ventricular

44.6 54.1

14.8 15.7

mmHg 1l.0 7.8 mmHg 5.2 3.8 21.4 6.1 18.3 6.2 -46.7 -45.5 10.8 6.1

Control BAPTA-AM

153.2 146.3

4.7 4.9 Left ventricular diastolic

106.6 103.7 pressure, 27.9 24.3 d(LVP)/dt maximum, mmHg/s 2708 2404
control vs. BAPTA-AM;

Control BAPTA-AM

6.5 1.7 8.2 2.4

Control BAPTA-AM
VF P < 0.01 or last 5 s occlusion vs. preocclusion

4858 3635
values.

403.0

524.4

364.0k
5P < 0.01,

420.9a
LVP,

-2346 319.7 1407 l90.4 pressure, occlusion last seconds before

left

ventricular

before treadmill stopped.

2588

Vol. 5

August 1991

The FASEB Journal

BILLMAN

ET AL.

TABLE

2. Hemodynamic

response to Bay K 8644 before and after BAPTA-AM

Pre-Bay K Heart Control Bay K 8644 BAPTA-AM 223.6 201.6 191.4 16.0 8.4 6.4 systolic pressure, rate (beats/mm)

Ba y K

Occlusion

plus Bay K 8644

183.6 189.0 mmHg

19.3 0.3

216.4 249.2 196.2

23.2
169b

34.0

Left ventricular Control Bay K 8644 BAPTA-AM 133.0 135.3 137.0 8.6 3.3 10.9

plus Bay K 8644

184.6 153.6 diastolic pressure, mmHg

lO.5 6.5

116.1 130.1 130.5

8.9k 6.0

Left ventricular Control Bay K 8644 BAPTA-AM 8.2 1.6 8.5 1.0 12.8 3.6 d(LVP)/dt Control Bay K 8644 BAPTA-AM
Bay and

19.4
13.6

996
4.3

plus Bay K 8644

15.9 18.0 maximum, mmHg/min

5.8

19.2

6.9

plus Bay K 8644

3658 3761 2692

384.7 210.0 657.9

left

4924 3148 ventricular pressure.

+

225.5 754.8

bp

3043 2318.6 2436

362.4 3562k 561.5

drug

K 8644 was injected 3-5 mm for other groups) vs. occlusion.

before coronary occlusion. LVP, P < 0.01 Bay K 8644 vs. on

BAPTA-AM

Bay

K 8644.

< 0.01 preocclusion (i.e., predrug for control 1P < 0.01 pre-Bay K vs. Bay K 8644.

Effect of the a-adrenergic susceptibility to ventricular

agonist phenylephrine fibrillation

The a-adrenergic receptor agonist phenylephrine was used to increase intracellular calcium levels in resistant animals. Phenylephrine infusion induced ventricular arrhythmias in 10 of 12 resistant animals; 8 dogs had VF (x2 = 9.2, P < 0.005, time to VF onset 27.4 6.6 s). Atropine was used to prevent reflex changes in heart rate. Atropine alone

given during the exercise plus ischemia test did not induce VF in these animals. Phenylephrine elicited significant increases in LVSP, d(LVP)/dt max, and LVDP (Table 3). The intracellular calcium chelator BAPTA-AM was then given to seven resistant animals that had VF in response to phenylephrine, and the exercise plus ischemia test was repeated. BAPTA-AM prevented phenylephrine-induced VF in four animals and delayed the onset of VF in the remaining

TABLE

3. Hemodynamic

effects of the a-adrenergic

receptor agonist phenylephrine Pre-PE Heart rate (beats/mm)

-

PE

Occlusion

Control PE BAPTA-AM

plus PE

198 7.8 231.6 6.9 214.6 8.6 Left ventricular systolic pressure,

193.0 7.1 50. 254.1 216.0

12.4 776 9.O

235.3 190.9 mmHg

Control PE
BAPTA-AM

plus PE

143.8 134.4 143.8

4.2 6.4 4.3

115.7 12.3 756/ 188.5 126.8

466 14.36 8.9

230.0
158.7

Left ventricular Control PE BAPTA-AM 11.1 2.5 6.2 1.8 7.6 1.6 d(LVP)/dt Control PE
BAPTA-AM plus PE

diastolic

pressure,

mmHg
-

22.0

396 786

plus PE

21.1 15.1 maximum, mmHg/s

6.3 3.0

31.0 26.0

6.6k

4368 4846
4202 beginning preocclusion 3-5 (i.e.,

369.1 451.6
438.O continuing and drug

3143

3546b

6995
3760

553.4
384.9w occlusion. P LVP, < 0.01

5162
3236 left ventricular pre-PE vs. PE.

5037k
PE, 0.01

Phenylephrine
phenylephrine. 6p

was infused
< 0.01

mm

before the occlusion, predrug for control group

throughout the for other groups).

pressure;
dp <

phenylephrine

vs. phenylephrine

plus BAPTA-AM.

ELEVATED CELL CALCIUM

AND

VENTRICULAR

FIBRILLATION

2589

three animals (phenylephrine 32.9 15.1 s phenylephrine plus BAPTA-AM 65.3 0.3 s). BAFTA-AM prevented the phenylephrine-induced increases in d(LVP)/dt and LVSP. As noted previously, BAPTA-AM also reduced predrug (i.e., before phenylephrine infusion) values of d(LVP)/dt max. Calcium-calmodulin in susceptible and dependent protein resistant myocardium phosphorylation

The experiments described previously in which Bay K 8644 induced VF in resistant dogs whereas BAPTA-AM protected dogs from Bay K 8644 and phenylephrine induced VF, indicate that elevated cytosolic calcium might be responsible for susceptibility to VF. If so, one would predict that susceptible dog myocardium would have elevated free calcium levels relative to resistant dog myocardium. If susceptible dog myocardium has higher cytosolic calcium, then calcium-dependent kinases (such as the calmodulin-dependent kinases) should be more active in susceptible than in resistant dog heart. Therefore, proteins phosphorylated by calmodulin (CaM)dependent kinases should be more heavily phosphorylated (with cold-phosphate) in susceptible than in resistant dog heart. When susceptible dog myocardium is homogenized and exposed to calcium and CaM in the presence of [32P]TP, less hot ATP should be incorporated into these substrate proteins in susceptible than in resistant dog heart. Posterior papillary muscle tissues was obtained from five resistant and five susceptible animals. Figure 2 compares the ability of calcium-CaM to phosphorylate substrate proteins (i.e., incorporate 32P) in homogenized resistant (Fig. 2A) and susceptible (Fig. 2B) dog myocardium. In resistant dog myocardium, the addition of CaM results in a major increase in the phosphorylation of 170- and 55-kDa substrate proteins

(Fig. 2A) whereas CaM induced much less phosphorylation of these same two proteins dependent in the susceptible tissue (Fig. 2B). The average CaM-dependent 32P phosphorylation of the 170-kDa protein was 2.9 ( 0.4)-fold greater and the phosphorylation of the 55-kDa protein was 2.2 ( 0.5)fold greater in the resistant than in the susceptible dog myocardium. Pretreatment of the resistant dog myocardium with the calcium ionophore A23187 produced a dramatic reduction in the CaM-dependent phosphorylation of the homogenized resistant tissue (phosphorylation of the 170-kDa protein was reduced by 70% and phosphorylation of the 55-kDa protein was reduced by 57%, with A23187 pretreatment, see Fig. 2A). With susceptible dog myocardium, however, A23187 pretreatment produced only minor changes in the CaMdependent phosphorylation pattern (phosphorylation of the 170-kDa protein was not reduced whereas the 55-kDa protein phosphorylation was reduced by only 13% with A23I87 pretreatment (see Fig. 2B). Myocardial infarct size

Myocardial infarct size was not measured in this study. However, a previous study using this model (5) demonstrated a significantly larger infarction in the susceptible (17.1 2.7%) compared with resistant animals (8.9 2.1%). The basis for this observation is unclear. Discussion The membrane-permeable compound, BAPTA-AM, has been used extensively as an intracellular calcium-specific chelator (14). BAPTA-AM does not bind calcium until after it pene-

A
Resistant

B
Dog Myocardium
Susceptible Dog Myocordium

170
Co

55 Co

170

55

Co. CaM

Ca, CaM

Ca. CaM

Ca, CaM

A23187
I

A23187 I
97

I
66 44

200

116

31

I 200

116 97

66

44

31

Molecular

Weight

kD

Molecular Weight

kD

Figure 2. Calcium- and calmodulin-dependent phosphorylation of proteins in resistant (A) and susceptible (B) dog posterior papillary muscle. Note the greater calmodulin-dependent phosphorylation of 170 and 55 kDa in the resistant myocardium. Pretreatment with A23187 converts a resistant phosphorylation pattern to a susceptible one. Identical amounts of protein were present in each lane and phosphorylation (32P incorporation), gel electrophoresis, autoradiography, and densitometry were conducted as described in Methods.

2590

Vol. 5

August 1991

The FASEB Journal

BILLMAN

ET AL.

trates the cell and cytoplasmic esterases hydrolyze it into the active, acid form that has a high affinity for cytoplasmic calcium. This compound provides a unique way to buffer selectively intracellular calcium levels without affecting extracellular calcium concentrations. In the present study, BAPTA-AM significantly reduced the incidence of malignant arrhythmias in animals known to be susceptible to ventricular fibrillation. Conversely, both Bay K 8644, which promotes calcium entry into the myocardium, and the a-adrenergic receptor agonist phenylephrine, which has been shown to increase myocardial cytosolic calcium (20), induced lethal ventricular arrhythmias in animals previously shown to be resistant to ventricular fibrillation. Pretreatment with BAPTA-AM significantly attenuated the hemodynamic effects of Bay K 8644 or phenylephrine, preventing the positive inotropic effects of these drugs. In an analogous manner, BAPTA-AM pretreatment completely suppressed the Bay K 8644-induced arrhythmias and either prevented or delayed the onset of the phenylephrineinduced ventricular fibrillation. These data are consistent with a BAPTA-AM-induced reduction in myocardial free calcium which protects against malignant arrhythmias. As cytosolic calcium rises from the 0.1-1 cm range, calmodulin-dependent processes and kinases are activated (35). This suggests that calmodulin-dependent phosphorylation of its substrate proteins could serve as an index of intracellular free calcium over this concentration range. The endogenous levels of calcium/calmodulin-dependent kinase activity (phosphorylation of 170 and 55-kDa protein) were found to be higher in tissue obtained from susceptible dog hearts compared with tissue from resistant animals. Furthermore, treatment of cardiac tissue obtained from resistant animals with the calcium ionophore A23187 increased calcium-dependent phosphorylation to levels similar to that observed in susceptible animals. When considered together, these data suggest that increased intracellular calcium promotes development of ventricular fibrillation, and further, that susceptible animals may exhibit a preexisting elevation in cytosolic calcium levels. In agreement with the present findings, calcium loading can induce spontaneous after-depolarizations and aftercontractions of isolated myocardial cells (10, 11, 22). Clusin (11) proposed that ventricular fibrillation could represent an extreme form of abnormal automaticity resulting from intracellular calcium overload. In general, interventions that favor intracellular calcium loading enhance triggered activity and ventricular fibrillation (10, 11, 22). For example, aadrenergic agonists have been shown to elicit a large accumulation of cellular calcium, during both systole and diastole (20), to increase action potential duration promoting calcium entry via voltage-dependent calcium channels (31), to induce after-depolarizations (23) and provoke spontaneous arrhythmias in both isolated cells and intact hearts (1, 25). The calcium channel agonist Bay K 8644 promotes calcium entry (23), elicits after-depolarizations in isolated tissue (26, 28), and induces arrhythmias in intact animals (12). Conversely, agents that decrease calcium entry or cytosolic calcium prevent triggered activity and the accompanying arrhythmias (12, 26-30). Ryanodine inhibited calcium release from the sarcoplasmic reticulum, attenuated arrhythmias resulting from ischemia or digitalis toxicity (27-30), and BAPTA-AM prevented catecholamine or digitalisinduced after-depolarization and after-contractions in ferret heart (26). Increases in myocardial free calcium could therefore contribute to development of malignant arrhythmias. Although a BAPTA-AM-induced decrease in myocardial calcium probably explains its protective effect on cardiac rhythm, this calcium chelator may have other effects that
ELEVATED CELL CALCIUM AND VENTRICULAR FIBRILLATION

could decrease mortality. First, BAPTA-AM enters smooth muscle cells where chelation of calcium could result in vasodilation. A corresponding reduction in arterial pressure (afterload) and a possible increase in coronary perfusion through collateral vessels could occur. Because the coronary occlusion elicited similar hemodynamic changes (LVSP decrease, LVDP increase, and heart rate increase; see Table 1) before and after BAPTA-AM, and BAPTA-AM had minimal effects on preocclusion systolic pressure, it is unlikely that this drug significantly altered afterload or the extent of ischemia in the susceptible animals. In addition, an infusion of the vasodilators nitroprusside (10 tg#{149} kg1 . min) or nifedipine (10 zg/kg), which reduced systolic pressure by more than 50% and increased coronary blood flow, failed to prevent VF in susceptible animals (12). Nifedipine only prevented VF at higher doses, which reduced d(LVP)dt maximum by a direct effect on the myocardial calcium. Thus, it is unlikely that BAPTA-AM-mediated protection against VF in the susceptible animals results from this drugs effects on vascular smooth muscle. In contrast to susceptible animals, BAPTA-AM significantly reduced the increase in systolic pressure elicited by either Bay K 8644 or phenylephrine in the resistant animals. It is possible that this reduction in afterload and thereby cardiac work improved the oxygen balance such that less tissue became ischemic during the occlusion. The confounding influences of BAPTA-AM on vascular smooth muscle calcium levels, therefore, cannot be excluded as a factor in the protection from Bay K 8644 and phenylephrineinduced VF afforded the resistant animals. Finally, reductions in serum calcium have been shown to protect against ventricular fibrillation (34). At the concentration used in the present studies, BAPTA-AM could not protect against VF by lowering serum calcium levels. Further, BAPTA-AM does not chelate calcium until the acetoxymethyl ester group is removed by cellular esterases, which forms an impermeable free acid (14). This allows for trapping of a calcium chelator inside the cell. If serum esterases did, in fact, hydrolyze all the injected BAPTA-AM, only 0.015 mM chelator would be produced, a level that would not significantly alter serum calcium concentrations. Indeed, BAPTA-AM did not significantly alter either total serum calcium levels (control 10.5 0.4 mg/100 ml BAPTA-AM 10.6 0.4 mg/100 ml; SD n = 7) or ionized calcium levels (control 5.35 0.32 mg/100 ml BAFTA-AM 5.44 0.33mg/100 ml ii SD n = 4). The indirect effects of BAPTA-AM, therefore, probably did not contribute significantly to the cardioprotection noted for this drug. The mechanisms responsible for the apparent accumulation of cytosolic calcium in the susceptible animals remain to be determined. However, as noted previously, alterations in the autonomic neural control of the heart could contribute significantly to the calcium overload noted for the susceptible animals. Previous studies have established that physiological interventions such as exercise (4) or acute myocardial ischemia (5) elicit a greater sympathetic response coupled with a greater withdrawal of parasympathetic tone in susceptible vs. resistant animals. It is well established (1, 7, 8, 20) that stimulating cardiac sympathetic nerves acting via 13-adrenergic and a-adrenergic receptors activate a second-messenger cascade (cyclic AMP, ITP) which results in an increase in cytosolic calcium and thereby positive inotropic events. Conversely, parasympathetic nerve activation (1, 9) opposes the actions of sympathetic nerve stimulation, reduces cyclic AMP, and enhances cyclic GMP levels, which results in a reduction in cytosolic calcium levels. Thus, the combination of increased sympathetic activity and reduced parasympa-

2591

thetic activity noted for the susceptible animals would favor accumulation of cytosolic calcium and could thereby increase the propensity of malignant arrhythmias. In summary, data suggest that elevations in intracellu ir calcium contribute to the development of lethal arrhythmi and that these increases in myocardial calcium can buffered by the intracellular calcium chelator BAPTA-AM. These data further suggest that the animals particularly susceptible to VF may have preexisting defects in cellular calcium homeostasis that lead to elevated cytosolic calcium levels. Thus, the management of malignant arrhythmias in patients known to be at risk for sudden death could be achieved by agents that selectively modulate myocardial cytosolic calcium. Interventions designed to reduce myocardial calcium levels may be particularly effective in preventing sudden cardiac death. Indeed, a recent large Danish multicenter study of patients recovering from myocardial infarction found that verapamil therapy significantly reduced the incidence of major cardiac events, including a 30% reduction in sudden cardiac death (33). Supported by National Institutes DA05917, DK33727, and by a grant Association. of Health grants HL36336, from the Central Ohio Heart

D., eds) pp. 755-765, Alan R. Liss, New York 14. Tsein, R. Y. (1981) A non-disruptive technique for loading calcium buffers and indicators into cell. Nature (London) 290, 527-528 15. Bradfort, M. M. (1971) A rapid and sensitive method for the quantification of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72, 248-254 16. Winer, B: J. (1971) Statistical Principles in Experimental Design, pp. 514-603, McGraw-Hill, New York 17. Bruning, J. L., and Kuntz, B. L. (1977) Computational Handbook of Statishcs, pp. 230-233, Scott, Foresman and Co., Glenview, Illinois 18. Recter, M. (1988) Calcium mobilization and cardiac inotropic mechanisms. PharmacoL Rev. 40, 189-217 19. Akhtar, M., Tchou, P., and Jazayeri, M. (1989) Use of calcium channel entry blockers in the treatment of cardiac arrhythmias. Circulation 80 (Suppl. IV), IV-31-IV-39 20. Auffermann, W., Stefenelli, T., Wu, S. T., Parmley, W. W., Wikman-Coffett, J., and Mason, D. T (1988) Influence of positive inotropic agents on intracellular calcium transients. Part 1, normal rat heart. Am. Heart]. 118, 1219-1227 21. Schramm, M., Thomas, G., Towart, R., and Franckowiak, G. (1983) Activation of calcium channels by novel 1,4 dihydropyri-

ular fibrillation. N., and Wood,

In Frontiers in Smooth Muscle Research (Sperelakis,

J.

dines. Arzneim. Forsch. 33, 1268-1271 22. January, C. T, and Fozzard, H. A. (1988) Delayed afterdepolarizations in heart muscle: mechanisms and relevance. Pharmacol. Rev. 40, 219-227 23. Kimura, S., Carmeron, J. S., Kozlovskis, P. L., Bassett, A. L., and Myerberg, R. J. (1984) Delayed after-depolarizations and
triggered activity induced in feline Purkinje fibers by alpha-

REFERENCES 1. Corr, P. B., Yamada,

K. A., and Witkowski, P X. (1986) Mechanisms controlling cardiac autonomic function and their relationships to arrhythmogenesis. In The Heart and Cardiovascular System (Fozzard, H. A., Haber, E., Jennings, R. B., Katz, A. M., and Morgan, H. E., eds) pp. 1597-1612, Raven Press, New York 2. Bigger, J. T, Kleiger, R. E., Fleiss, J. L., Rolnitsky, L. M., Steinman, R. C., and Miller, J. P. (1988) Components of heart
measured during healing of acute myocardial infarction. Am. j Cardiol. 61, 208-215 3. Kleiger, R. E., Miller, J. P., Bigger, J. T., and Moss, A. J. (1987) Decreased heart rate variability and its association with inrate variability

adrenergic stimulation in the presence of elevated calcium levels. Circulation 70, 1074-1082 24. January, C. T., Riddle, J. M., and Salata, J. J. (1988) A model for early after-depolarizations: induction with the Ca2 channel agonist Bay K 8644. Circ. Res. 62, 563-571 25. Sheridan, D. J., Penkoske, P. A., Sobel, B. E., and Corr, P. B. (1980)
Alpha-adrenergic contributions to dysrhythmia during myocardial ischemia and reperfusion in cats. j Clin. Invest. 60, 161-171 Marban, E., Robinson, S. W., and Wier, W. G. (1986) Mechanisms of arrhythmogenic delayed and early after-depolarization in ferret ventricular muscle. J. C/in. Invest. 78, 1185-1192 Thandroyen, F. T, McCarthy, J., Burton, K. P., and Opie, L. H. (1988) Ryanodine and caffeine prevent ventricular arrhythmias during acute myocardial ischemia and reperfusion in rat heart. Circ. Res. 62, 306-314 Kahn, M. D., Wittingham, D. J., and Wiesner, K. (1964) Effects of ryanodine in normal dogs and in those with digitalis induced arrhythmias. Am. J. Cardiol. 14, 658-668 Golovina, V. A., Zarharov, S. I., Bogdarnov, K. Y., and

26.

27.

mortality after acute myocardial infarction. Am. j 59, 256-262 4. Billman, G. E., and Hoskins, R. S. (1989) Time-series analysis of heart rate variability during submaximal exercise: evidence for reduced cardiac vagal tone in animals susceptible to ventricular fibrillation. Circulation 80, 874-880
Cardiol. 5. Collins, M. N., and Billman, G. E. (1989) Autonomic response

creased

28.

29.

6.
7. 8. 9.

10.

to coronary occlusion in animals susceptible to ventricular fibrillation. Am. j Physiol. 257, H1886-H1894 Billman, G. E., Schwartz, P. J., and Stone, H. L. (1982) Baroreceptor reflex control of heart rate: a predictor of sudden cardiac death. Circulation 66, 874-880 Evans, D. B. (1986) Modulation of cAMP: mechanism for positive inotropic action.]. Cardiovasc. Pharmacol. 8 (Suppl. 9), S22-S29 Berridge, M. J. (1987) Inositol trisphosphate and diacylglycerol: two interacting second messengers. Annu. Rev. Bioc/zem. 56, 159-193 Watanabe, A. M. (1984) Cellular mechanisms of muscannic regulation of cardiac function. In Nervous Control of Cardiovascular Function (Randall, W. C., ed) pp. 130-164, Oxford University Press, New York Levy, M. N. (1989) Role of calcium in arrhythmogenesis. Circulation 80 (Suppl. IV), IV-23-IV-30

Rosenshtraukh, L. V. (1988) Analysis of anti-arrhythmic effects of ryanodine in guinea pigs. j Mol. Cell Cardiol. 20, 303-311 30. Boutjdir, M., El-Sherif, N., and Gough, W. B. (1990) Effects of caffeine and ryanodine on delayed after-depolarizations and sustained rhythmic activity in 1-day-old myocardial infarction in the dog. Circulation 81, 1393-1400
31. Tohse, N., Nakaya, H., Hattoui, Y., Endou, M., and Kanno, M.

(1990)

Inhibitory

effect

mediated

by a1-adrenoreceptors

on

transient outward current in isolated rat ventricular cells. Ifluegers Arch. 41, 575-581 32. Merillat,J. C., Lakatta, E. G., Hano, 0., and Guarnieri, T. (1990) Role of calcium and the calcium channel in the initiation and maintenance of ventricular fibrillation. Circulation 67, 1115-1123 33. Danish Study Group on Verapamil in myocardial infarction. (1990) Effect of verapamil on mortality and major events after

11. Clusin, W. T, Bristow, M. R., and Karagueuzian, E. S., Katzung, B. G., and Schroeder, J. S. (1982) Do calciumdependent ionic currents mediate ischemic ventricular fibrillation? Am. j Cardiol. 49, 606-612 12. Billman, G. E. (1989) Effect of calcium channel antagonists on susceptibility to sudden cardiac death: protection from ventricular fibrillation. j Pharmacol. Exp. Therap. 248, 1334-1342 13. Billman, G. E., Mcllroy, B., and Johnson, J. D. (1990) The role of elevated intracellular calcium in the susceptibility to ventric-

acute myocardial trial II - DAVIT

34. Clusin, W. T,

infarction. (The Danish varapamil II). Am. ]. Cardiol. 66, 779-785

Bristow, M. R., Bairn,

infarction

D. S., Schroeder,

J.

S.,

Jaillon, P., Brett, P., and Harrison, D. C. (1982) The effects of Diltiazem and reduced serum ionized calcium on ischemic ventricular fibrillation in the dog. Circ. Res. 50, 518-526 35. Johnson, J. D., and Mills, J. S. (1986) Calmodulin. Med. Res. Rev. 6, 341-363
Received for publication Acceptedfor publication April May 10, 1991. 17, 1991.

2592

Vol. 5

August

1991

The FASEB Journal

BILLMAN

ET AL.