Catalase and hydrogen peroxide cytotoxicity in cultured cardiac myocytes

J Mol Ceil Cardiol 27, 909-915 (1995)

Catalase and Hydrogen Peroxide Cytotoxicity in Cultured Cardiac

MyocNes Lawrence D. Horwitz and Jonathan A. Left The Divisions of Cardiology and of Puhnonary Diseases and The Webb-Waring Institute for Biomedical Research, University of Colorado Health Sciences Center, Denver, Colorado, 80262, USA (Received 11 February 1994, acceptedin revisedform 23 August 1994) L. D. HORWITZANDJ. A. LEft. Catalase and Hydrogen Peroxide Cytotoxicity in Cultured Cardiac Myocytes. Journal of Molecular and Celhdar Cardiology (1995) 27, 909-915. We examined the role of in tracellular catalase activity in modulating hydrogen peroxide (H_,O_,)-inducedcytotoxicity in cultured chick embryo cardiac myocytes. Injury was quantitated by release of lactate dehydrogenase (LDH). Application of 1.5 mM H20., to myocytes caused LDH release beginning at 2 h. Inactivation or inhibition of catalase with aminotriazole or sodium azide increased LDH release but did not cause earlier release. Free catalase which entered or became associated with myocytes, but not catalase bound to agarose beads, which did not enter or become associated with myocytes, was protective. Separate experiments demonstrated that myocyte catalase activity decreased by 27% between 1 and 4 h of H_,O2 exposure. Treatment with aprotinin, a protease inhibitor, prevented the H2Oa-induced fall in catalase activity at 4 h but treatment with deferoxamine, an iron chelator, had no effect on catalase activity. Thus, with exposure of cardiac myocytes to H_,O_,,the magnitude of the cytotoxicity is modulated by endogenous or cell associated exogenous catalase. It is proposed that in addition to excessive accumulation of H202, a reduction in intracellular catalase activity may be required before substantial cell injury occurs during H202 exposure. Activation of proteases may cause the reduction in catalase activity in this setting. KEY WoRDs:.Free Radicals; Catalase: Aminotriazole.

Introduction

drogen peroxide (H202) (Boveris and Chance, 19 73; Deisseroth and Dounce, 19 70). The significance of cellular catalase levels in preventing H202-mediated cytotoxicity has not been clear, particularly because other endogenous defenses, such as glutathione peroxidase, may be protective against peroxides. We sought to evaluate the importance of catalase activity for protecting cultured chick embryo cardiac myocytes against H202. We assessed the protective effect of endogenous catalase and the modifying role of exogenous catalase administration. Finally, we investigated mechanisms by which endogenous catalase activity could be altered during exposure to H202.

Reactive oxygen species are normally produced in small quantities in tissue but under certain conditions levels capable ofcytotoxicity accumulate (del Maestro et al., 1980; Guarnieri et al., 1980). To prevent injury there are extensive cellular defense mechanisms against reactive oxygen metabolites, including scavenger enzymes such as catalase and superoxide dismutase (Left et al., 1992; McCord et al., 1969). The balance between a scavenger enzyme and the reactive oxygen species which is its substrate is likely to determine whether tissue injury occurs. Catalase is an enzyme which inactivates by-

Please address all correspondence to: Lawrence D. Horwitz, M.D., CardiologyB130, University of Colorado Health Sciences Centre 4200 East Ninth Avenue,Denver, Co 80262, USA.

0022-2828/95/030909+07 $08.00/0

909

© 1995 Academic Press Limif~Bd

910

L.D. Horwitz and 1. A. Left

Methods Cell culture Monolayer cultures of spontaneously beating chick embryo myocytes were prepared by minor modifications of the method of Barry et al. (Barry and Smith, 1982; Byler et al., 1994). Briefly, hearts from 10 day-old chick embryos (White Leghorn) were dissected free in a sterile manner, minced, and placed in calcium- and magnesium-free Hanks' balanced salt solution (HBSS). As we have described previously (Byler et al., 1994), the tissue was trypsinized and centrifuged after which the cell pellets were placed in culture medium containing 54% balanced salt solution (116mu NaCI, 1.0raM Nail,P04, 0.8 mM MgSO4, 1.18 mM KC1, 26.2 mM NaHCO3, 0.87 mM CaCl2 and 5.6 m u glucose), 40% Medium 199 (Gibco Laboratories, Grand Island, NY USA) with Hanks' salts, 6% heat inactivated fetal bovine serum (Hyclone Laboratories, Logan, Utah USA), and 10 U penicillin/ml and 10 pg streptomycin/ml. The cell suspension was diluted to 6 x 10 s cells/ml and plated in six-well plates (35ram well diameter) at 3 m 1/well for measurements of lactate dehydrogenase (LDH) release. Twenty-four-well plates (15 mm well diameter) at 1 ml/well were used for catalase assays. Cultures were incubated in 5% C02 in ambient air at 37°C.

Experimental protocols All studies were performed on spontaneously beating myocyte monolayers after 4 days of primary culture as described previously (Byler et al., 1994). Some experiments were performed with modified Tyrode's solution (118 mM NaCI, 11.9 mM NaHCO3, 0.0004 mM Nail,P04, 4.4 mM KC1, 1.0 mM MgC12, 0.9 mM CaC12.2H20, and 11 mM D-Glucose) instead of medium. Cytotoxicity was measured by taking aliquots of medium or Tyrode's solution at various time points over 22 h and measuring the release of lactate dehydrogenase (LDH) into the medium. All conditions were replicated in three wells and all experiments were repeated at least once to ensure reproducibility. Wells with medium alone, HaOa alone, cell lysis with 1% Triton X-IO0 at zero time and experiment specific controls were included in each experiment.

LDH measurement To assess cell damage, LDH activity was measured

by m~nor modification of the fluorometric technique

of Green et al. (1984). Release of LDH was standardized with a cell injury index defined at (A-B)/ (C-B) xl00 where A =LDH activity in the test sample, B = LDH activity in samples containing untreated cells in medium (0% control), and C = LDH activity in samples from wells in which cells were lysed with Triton X-IO0 (100% control).

Catalase measurement Catalase activity in washed, lysed myocytes was measured polarographically as the rate of production of 02 from H202 (Left et al., 1991). One unit of catalase activity was defined as the quantity of enzyme that consumed 1.0 #mole H202 per rain at 25°C, pH 7.0. Catalase measurements were norrealized by measurement of cellular protein by the method of Bradford (Bradford, 1976).

Measurement of H202scavenging capacity A modification of the method of Thurman et al. (1972) was used to determine the capacity of catalase to scavenge under cell-free conditions. This utilizes the conversion of ferrous to ferric ion by H20v Ferric ion then reacts with thiocyanate to form ferric thiocyanate which is measured at 4 8 0 n m with a spectrophotometer (Gilford Instruments, Model 250). Agents in saline were incubated with and without 1.5 mM H202 for 30 rain at 37°C. After addition of trichloroacetic acid (TCA) and centrifugation, the supernatants were diluted with saline and any remaining reactive H202 was measured by adding ferrous ammonium sulfate (10rain, ambient temperature) followed by potassium thiocyanate (5 rain, ambient temperature). Results with H202 alone were compared to H202 and catalase.

Materials Reagents added to wells during experiments included sodium azide, 3-amino-l,2,4-triazole, Triton X-100, aprotinin, and beef liver catalase bound to 4% agarose beads, which were purchased from Sigma Chemical Co. (St Louis, MO, USA); and beef liver catalase which was purchased from Boehringer-Marmheim Corp. (Indianapolis, IN, USA). Deferoxamine mesylate was purchased from CIBA-Geigy Corp. (Woodbridge, N-J, USA).

Hydrogen Peroxide Cytotoxicity in Myocytes

911

• AMT+ H202

50 - Aminotriazole25 mM n=6 40 i ~ -~ 30

OH202 "ANT

i~ 2o

[]Tyrode's

C9 0 ,

0

I

,

2

4 Time(h)

30

I

,

I

,

6

.

8 I • NaAzide+H202

9.o-

~

• N a Azide

i0

oTyrode's

~

0

0

2

I

4 Time(h)

6

8

Figure 1 Effectof inactivation or inhibition of catalase on H202 cytotoxicity. Top frame compares cell injury index calculated from LDH release (see Text) in myocytes in Tyrode's IN, in Tyrode's with aminotriazole II, in Tyrode's with 1.5 mM H_,O.,O, and in Tyrode's with H_,O.,and AMT Q. There was significantly more LDH release when aminotriazote was added to H.,O., v H_,O2alone. Bottom frame depicts a similar experiment with sodium azide (Na Azide) instead of aminotriazole (Tyrode's I-'1.Tyrode's with Na Azide II; Tyrode's with 1 mM H20_, O; Tyrode's with H_,O2and Na Azide Q).

Statistics All statistical comparisons were done by analysis of variance using the Scheff6 test of multiple comparisons. A P<0.05 was considered to be significant.

Results Role of endogenous defense mechanisms in preventing or delaying H202-mediatedcardiac myocyte injury Myocytes were pretrcated for 1 h with 25 mM aminotriazole (to inactivate catalase) or for 30 rain with I mM sodium azide (to inhibit catalase) resuiting in complete elimination of measured catalase activity with each agent (data now shown). With aminotriazole, LDH release was significantly greater than occurred when untreated myocytes were exposed to the same dose (1.5 raM) of H20_, at 2, 4 and 7 h (P<0.01 for each) (Fig. 1, top panel). With sodium azide administration there was significantly more damage at 4 and 7 h (P<0.01 for each) but no difference in cell injury index at 2 h compared to untreated cells exposed to H20~ (Fig. 1, bottom panel). Therefore, the magnitude of LDH

release was increased by inactivation or inhibition of catalase.

Capacity of free and agarose-boundcatalase to decrease H202concentrations under cell-free conditions We examined the capacity of free catalase and agarose-bound catalase to scavenge H202 under cell-free conditions. Each agent was added to 1.5mM H202 in doses used in the experiments described below. Free catalase and catalase bound to agarose beads were both highly effective scavengers of H202, inactivating at least 99% of the H202 with 30 rain incubation at 37°C (Table 1).

Effects of diffusible and nondiffusible catalase We compared the protective effects of free and agarose bead-bound catalase. Free catalase (800 U) was incubated with cells for 1 h, after which the cells were washed and 1.5 mM I-I202 added. The free catalase protected against H202-induced IDH release (Fig. 2, top panel). There were: significant reductions in cell injury index at 4, 7 and 2 ! h. When the same dose of agarose-botmd catalase was

912

L. D. Horwitz and J. A. Left Table 1

In vitro H202 scavenging capacity of various agents

Incubation with 1.5 mM H_,O., compound, concentration

Dilution assayed

Catalase, 800 U/ml Agarose-Catalase, 800 U/ml Heat Inactivated Catalase 800 U/ml

Absorption at 480 nm

~tM Reactive H.,O_, remaining in undilute sample

inactivated 100.0

% H202

1:2

0

0

1:2

0.033

4.4:

99.7

1:100

0.236

1411.8

5.9

1.5 mM H202 was incubated with tile listed compounds as described in Methods. The heat inactivation of catalase was for 30 min at 65°C. The dilution assayed determined the reactive H.,O2 remaining after a 30min. 37°C incubation. The absorption values for the given dilution are corrected for blank readings, Blanks, which were compounds incubated without H20.,. were assayed at the same dilutions and were in all cases comparable to saline alone.

80

• Agarose-Catalase t h e n H202 a H202

~ _ 60

• Catalase t h e n H202

,o

o Medium

I

5

I

I

10 15 Time (h)

I

20

25

80

q • H202 + Aga-Cat at 45 min

~1 ~ 60

] tx H202 + Aga-Cat at 60 min

.~

a H202 + Cat at 60 min ~ ~

• H202 + Cat a t 45 min o Medium

0 I

5

I

I

10 15 Time (h)

20

25

Figure 2 Effectof catalase on H202 cytotoxicity. In the top frame n = 6 for each group. In the top frame simultaneous administration of free catalase prevented cell injury from H_,02 but catalase bound to agarose beads had no effect. In the bottom frame administration of free catalase 45 or 60 min after H202 exposure began prevented or attenuated cell injury but catalase bound to agarose beads had no effect.

administered in this manner it was not protective (Fig. 2, top panel). Both free and agarose-bound catalase scavenged similar quantities of H202 under cell-free conditions (Table 1). However, whereas incubation of free catalase with cells for i h resulted in an increase in catalase in washed cells (l.74-1-0.21U/mg protein control v 4 . 0 1 + 0 . 2 1 free catalase, P<0.01), incubation with agarosebound catalase did not alter cellular catalase levels (1,744-0.21 control v 2.03 + 0 . 3 4 agarose-bound catalase, P=N,S.) ( n = 6 for each group).

In another experiment H202 w a s applied to cells and either free or agarose-bound catalase added 45 or 60 min later. The free catalase reduced LDH release at 4, 7 and 24 h (Fig. 2, bottom panel). However, the agarose-bound catalase did not reduce LDH release in myocytes exposed to H202. Thus when catalase was administered in a form in which it could enter or become associated with cells it was protective, but when it was administered in a form which could not enter or become associated with ceils it was ineffective.

Hydrogen Peroxide Cytotoxicity in Myocytes

4

913

3o

3 2

Tyrode's o

Tyrode's 30 rain H202 1 h H202 4 h H202

Figure 3 Time course of myocyte catalase activity during exposure to H_,O2. Catalase activity in myocytes in Tyrode's solution alone and after addition of 1.5 mM H_~O2 for 30 min. 1 h or4 h. Bars are mean values plus standard error shown by brackets. *Depicts significant difference (P<0.05) from Tyrode's.

H20~.

Def.

Def. +H202

Figure 4 Effector treatment with deferoxamine on catalase activity after 4h H20, exposure. Shown are mean_+s.s, for catalase activity in untreated wells not exposed to H20, (Tyrode's). untreated wells which were exposed for 4 h to 1.5 mM H_,O2(H_,O_,),wells treated with 0.25 mM deferoxamine but not exposed to H20_,(De0 and wells pretreated with deferoxamine for 2 h and then exposed to H20_, for 4 h (Def + H.,O2). There were signillcant reductions (P
Measurements of myocyte catalase during exposure to H202 To determine whether H_,O2 influences myocyte catalase activity, we measured activity of this enzyme at various times during continuous exposure to 1.5 mM H202. Myocyte catalase activity was not significantly changed at 30 or 60 min. Surprisingly, catalase activity fell substantially ( - 2 7%, P
~ 4

~Et2

Tyrode's

H 2 0 2 Aprotinin Aprotinin +H202

Figure 5 Effectof treatment with aprotinin on catalase activity after 4 h H202 exposure. Shown are mean_+s.E. for catalase activity in wells with cells untreated and not exposed to H202 (Tyrode's),untreated but exposed to H202 for 4 h (H202), treated with aprotinin 6 rnl/ml but not exposed to H_,O_,(Aprotinin), and treated with aprotinin and exposed to 1.5 mM H202 for 4 h (Aprotinin+H,O2) n = 6 each group. Aprotinin prevented the reduction in catalase activity which occurred with exposure to H,.O2 in untreated ceils. In contrast, aprotinin prevented the fall in catalase activity during I-I202 exposure (Fig. 5). There was no significant difference between wells treated with aprotimn alone or control wells and wells treated with aprotinin and exposed to H202. However, there was a significant difference between wells treated with aprotinin and exposed to I-I202 and wells with untreated myocytes exposed to H202 (P<0.05). Therefore, proteases may have contributed to the decrease in catalase activity.

Discussion In our cultured myocyte system H202 (1.5mM) resulted in substantial release ofLDH beginning.2:h

914

L.D. Horwitz and J. A. Left

after exposure began. H202 is extremely diffusible and should reach high intracellular concentrations within seconds. Quaife et al. (1991) demonstrated that damaged cultured fetal chick myocytes release substantial amounts of LDH within 15 min. Therefore, there appears to be a delay between the time of arrival of H202 intracellularly and the onset of cell injury. The enzyme catalase appears to be critical in defending cultured fetal chick cardiac myocytes against the cytotoxicity of H202. Total inactivation or inhibition of cellular catalase activity with aminotriazole or sodium azide markedly increased the magnitude of the cell injury during H202 exposure. Our results support the hypothesis that H _,0 .~exerts its critical cytotoxic effects intracellularly. We found no evidence that extracellular scavenging influenced the intracellular reactions. Pretreatment with free catalase prevented myocyte injury from H202. In contrast, pretreatment with agarose-bound catalase, which remained extracellular, was not protective. We demonstrated by direct assay that the exogenous free catalase either enters or becomes associated with myocytes, Similar efficacy of free catalase administered to cultured cardiac myocytes has been reported with hypoxia and reoxygenation (Quaife et d., 1991). In other experiments we found that administration of free catalase 45 to 60 rain after initiation of exposure to H202 attenuated injury. However, agarose-bound catalase given at this time was ineffective. It appears that extracellular catalase is ineffective. Therefore, it is likely that during conditions in which H202 exposure is due to adherent leukocytes (Entman et al., 1990) or to intracenular generation of this oxidant (Boveris et tfl., 19 73; Thurm a n et d., 19 72), only intracellular catalase is likely to protect against cytotoxicity. However, it is not clear what is the source of the H202 neutralized by the late administration of free catalase. It would be expected, in view of the high diffusibility of H202, that either bound or free catalase would reduce intracellular H202 by removing H202 extracellularly. However, some exogenously administered H20~ may remain sequestered intracellularly or associated with the cell membrane where free but not bound catalase may scavenge it. Another possibility is that there is de novo generation of H202 that cannot be neutralized by endogenous scavengers, which may have been overwhelmed by the prior exposure to exogenous H202, and cytotoxicity is prevented by the exogenous free catalase. We considered the possibility that the delay of more than 2 h before substantial I ~ H release

occurred was due, at least in part, to the time required to overcome catalase and other endogenous cellular defenses. In support of this concept, we made the surprising observation that catalase activity in cardiac myocytes decreased over time with exposure to exogenous H202. The process by which endogenous cellular defenses are overcome may be more complex than simply the accumulation of concentrations of oxidants which exceed the capacity of catalase and other enzymes to catalyse their decomposition. It is possible that, because of the considerable efficiency of the enzymecatalysed process, a reduction in catalase activity, as well as excessive accumulation of H202, may be necessary before substantial cell injury occurs. Since H202 is a substrate for catalase, it is unlikely that it inactivated the enzyme in these experiments (Deisseroth and Dounce, 19 70). In addition, there is ample evidence that the concentrations of H202 in the range we utilized do not denature catalase (Davies, 19 8 7; Davies and Lin, 19 8 8; Deisseroth and Dounce, 1970). One possibility is that other reactive oxygen metabolites which are derived intracellularly from H202 are capable of denaturing the enzyme or reducing its activity, another possibility is that activation of proteases may contribute to decreasing catalase activity. Degradation of catalase by .OH leading to susceptibility to proteolysis has been reported (Davies, 1987; Davies and Lin, 1988). To gain more information about the possible rote of oxidative or proteolytic causes of the decline in catalase activity, we studied the effects of treatment with deferoxamine, an iron chelator that prevents iron-mediated oxidant injury (Byler et al., 1994) and aprotinin, a protease inhibitor (Vincent and Lazdunski, 1972). Since aprotinin prevented the fall in catalase activity, protease activity, in the presence of H202, probably played a critical role. H202 may slowly activate proteases which synergisticaUy with H202 denature catalase. However, prevention or attenuation of iron-mediated reactions, presumably including .OH production, did not prevent the decline in catalase. The significance of these observations with regard to prevention of cytotoxicity from H202is not clear. One possibility is that the mechanisms of the decline in catalase and the occurrence of cell injury may differ. However, it is conceivable that a fall in catalase is necessary for injury to occur with H202 exposure. We cannot exclude the possibility that the moderate decline in catalase activity did not affect susceptibility to H202, although larger changes due to inhibition of catalase with aminotriazole and sodium azide clearly increased cytotoxicity from this reactive oxygen species.

Hydrogen Peroxide Cytotoxicity in Myocytes

Acknowledgements The a u t h o r s wish to t h a n k ] u l i a n n S. Wallner, N a n c y A. S h e r m a n , Mark E. Bodman, and Jennifer M. Kirkanan for their superb technical assistance with this study. This w o r k was supported by g r a n t H L 4 8 1 7 7 , Dr Left is the recipient of a C l i n i c i a n Scientist Award from the A m e r i c a n Heart Association, BAI~Y WH, and SMITHTW, 1982. Mechanisms of transmembrane calcium movement in cultured chick embryo ventricular cells. J Physiol 325: 243-260. BOVERIS A and CHANCE B, 1973. The mitochondrial generation of hydrogen peroxide: General properties and effect of hyperbaric oxygen. Biochem J 134: 707716. BRADFORDMJVI, 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 72: 248-254. BYLERRM, SHERMANNA, WALLNERJS, HORWITZLD, 1984. Hydrogen peroxide cytotoxicity in cultured cardiac myocytes is iron-dependent. Am J Physiology 266: H121-H127. DAVIES KJA, 1987. Protein damage and degradation by oxygen radicals. I. General aspects. ] Bid Chem 262: 9895-9901. DAVIESKJA and LIN SW, 1988. Degradation of oxidatively denatured proteins in Escherichia coil Free Radic Bid Med 5: 215-223. DEISSEROTH A and DOUNCEAL, 1970. Catalase; physical

915

and chemical properties, mechanism of catalysis, and physiological role. Physiol Rev 50 319-375. DEL MAESTRORF, THAW RE BIORK HH. PLANKERM and ARFORS K. 1980. Free radicals as mediators of tissue injury. Acta Physiol Scand [Suppl 492]: 43-57. ENTMANM].,, YOUKERK, SHAPFELLSB, SIEGELC, ROTHLEIN R, DREYER WJ, SCHMALSTIEGFC, SMITH CW, 1990. Neutrophil adherence to isolated adult canine myocytes. Evidence for a CD18-independent mechanism. ] Clin Invest 85: 1497-1506. GREENHI, FRASERIG, RANNEVDA, 1984, Male and female differences in enzyme acitivities of energy metabolism in vastus lateralis muscle. ] Neurol Sci 65: 323-331. GUARNIERIC, FLAMIGNIF, CALDARERACM, 1980. Role of oxygen in the cellular damage induced by reoxygenation of the hypoxic heart. ] Mol Cell Cardiol 12: 797-808. LEFFJA, OPPEGARDMA, TERAOAL~, MCCARTYEC, REPINE JE, 1991. Human serum catalase decreases endothelial cell injury from hydrogen peroxide. J Appl Physiol 71: 1903-1906. McCoRD/M and FRIDOVICH[, ] 969, Superoxide dismutase: an enzymic function for erythro-cuprein(hemocuprein). J Biol Chem 244: 6049-6055. QUAIFERA, KOHMOTOO, BARRYI~A/I-I,1991. Mechanisms of reoxygenation injury in cultured ventricular myocytes. Circzdation 83: 566-577. THtmMAN RG, LEY HG, SCHOLZ R, 1972. Microsomal ethanol oxidation: hydrogen peroxide formation and the role of catalase. Eur J Biochem 25: 420-430. VINCENT ]-E LAZDONSKIM, 1972. Trypsin-pancreatic trypsin inhibitor association. Dynamics of the interaction and role of disulfide bridges. Biochemistry 11: 2697-2977.