103,214-221(1990)
TOXlCOLOGYANDAPPLlEDPHARMACOLOGY
Cyanide-Induced Neurotoxicity: Calcium Mediation of Morphological Changes in Neuronal Cells E. U. MADUH, J. J. TUREK,* J. L. BOROWITZ, A. REBAR,* AND G. E. ISOM Department of Pharmacology and Toxicology, Veterinary Anatomy, School of Veterinary
Received
May
School of Pharmacy Medicine, Purdue
22, 1989; accepted
and Pharmacal Sciences, University, West Lafayette,
December
and *Department Indiana 47907
of
12, 1989
Cyanide-Induced Neurotoxicity: Calcium Mediation of Morphological Changes in Neuronal Cells. MADUH, E. U., TUREK, J. J., BOROWITZ, J. L., REBAR, A., AND ISOM, G. E. (1990). Toxicol. Appl. Pharmacol. 103, 2 14-22 1. Calcium channel blockade decreases the elevation of brain calcium as well as the tremors produced by cyanide in mice. To determine if cyanideinduced morphological changes could also be inhibited by calcium channel blockade, the effect of diltiazem was studied in cultured rat pheochromocytoma (PC12) cells, a neuronal model. Incubation with KCN (I to 10 mM for 1 to 2 hr) caused depletion of secretory granules, alignment of remaining granules along the plasma membrane. and mitochondrial swelling. All these effects were inhibited by pretreatment with 0.0 1 mM diltiazem. Scanning electron microscopy revealed that cyanide (I to IO mM for I to 2 hr) produced loss of microvilli and bleb formation in PC12 cells. These changes were partially inhibited by preincubation with 0.01 mM diltiazem. Incubation of cells with 10 mM cyanide increased release of lactic dehydrogenase (LDH) into the culture media at 60 and 120 min. A decrease in cell viability, as determined by trypan blue dye exclusion, paralleled the release of LDH. At 120 min of cyanide incubation, 24% of the cells excluded dye. Both the release of LDH and decreased cell viability were attenuated by pretreatment with diltiazem. The results indicate that the influx of extracellular calcium is an important factor mediating cyanide-induced morphologic changes in neuronal cells. o 1990 Academic
Press. Inc.
Permanent neurological damage by cyanide stems from inhibition of cellular metabolism and is preceded by changes in cell morphology (Brierly et al., 1976; Ashton et al., 1981; Funata et a/., 1984). Changes in cell morphology are therefore an important indication of neuronal damage by cyanide. Exact mechanisms by which cyanide causes neuronal structural alterations have not been clarified. Recently, Johnson et al. ( 1986) found cyanide treatment elevates total brain calcium in mice and increases cytosolic calcium in cultured rat pheochromocytoma (PC 12) cells (Johnson, 1987a). Peroxidation of brain lipid membranes also occurs in cultured neuronal 0041-008X/90
$3.00
Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
cells and in mouse brain after cyanide treatment (Johnson et al., 1987b). Both elevation of cytosolic calcium and lipid peroxidation induced by cyanide are inhibited by pretreatment with the calcium channel blocker, diltiazem (Johnson et al., 1987a,b). In hepatoma 1c 1c7 cells, KCN produced a rapid rise in cytosolic free Ca*+ which preceded blebbing of the plasma membrane and loss of cell viability (Nicotera et al., 1989). Cyanide-induced morphological changes in neuronal cells may be related to elevated cytosolic calcium and associated membrane damage. Therefore, the present study employs a calcium channel blocker to determine whether cyanide-in-
214
CYANIDE-INDUCED
MORPHOLOGICAL
duced changes in neuronal cell morphology and cell viability are related to extracellular calcium influx. METHODS C’eN culture. Rat pbeochromocytoma (PC 12) cells obtained from American Type Culture Collection were grown (37°C. 5% CO* in humidified air) attached to plastic culture flasks until confluency in a medium of 85% RPM1 1640, 10% v/v heat-inactivated horse serum, and 5% v/v fetal calf serum containing 1% penicillin-streptomycin. At the time ofthe experiment, cells were detached using a stream of media from a pipet and suspended in Krebs-Ringer (CaKR) solution consisting of(mM) NaCI, 125; Hepes-NaOH, 25; glucose. 6; NaHC03, 5; MgSO.,, 1.2; KH,PO,, 1.2; and CaC12, 1.O.Potassium cyanide solutions were prepared just prior to use. The millimolar concentrations of cyanide employed are the same as those used previously and correspond to amounts associated with acute intoxication in vivo (Johnson et al.. I987a). Diltiazem was added to the medium 15 min before KCN. Approximately IO6 cells were used for each sample. PC I2 cells for scanning electron microscopy were cultured at a density of 200.000 cells per chamber in glass chamber slides coated with poly-d-lysine. Media were refreshed every other day for lo- 14 days. At the time ofthe experiments, growth media were replaced with KrebsRinger solution and test agents added. LDH release was assessedaccording to the method of Higgins and Bailey (I 983) and Goldberg et al. (1987). Cells (2-4 X 106) were suspended in I ml CaKR media with or without treatments. After 30.60, or 120 min cells were rapidly pelleted ( 10,OOOgfor 10 set) and 100 ~1 of bathing media was assayed for LDH activity. Cyanide does not influence the measurement of lactate dehydrogenase activity (Higgings and Bailey, 1983). Cell viability was determined by the dye exclusion method in which cells were suspended in CaKR and incubated at 37’C (Schanne et ai.. 1979). At 30, 60. and 120 mitt after addition of KCN. 0.2-ml aliquots were withdrawn from the suspension and diluted five-fold with 0.2% trypan blue, and cells excluding dye were counted in a homocytometer. Each sample before addition of KCN served as its own control. RPM1 1640 and falcon plastic tissue culture flasks were purchased from GIBCO (Grand Island, NY); glass chamber slides were from Miles Scientific (Naperville, IL). Sera were obtained from HyClone (Logan, UT), penicillin G potassium from E. R. Squibb and Sons (Princeton, NJ), streptomycin from Eli Lilly &Co. (Indianapolis. IN), and poly-d-lysine from Sigma Chemical Co. (St. Louis, MO). Dr. Ronald Browne of Marion Lab-
CHANGES
315
oratory (Kansas City. MO) generously supplied diltiazem hydrochloride. Potassium cyanideand otherchemicals were of the highest chemical grade commercially available. Transmission electron microscopy. At the termination of the experiments, cells were pelleted at 1OOOgfor 4 min. Media were removed and 1.5% glutaraldehyde was added to fix the cells in the pellet. Fixative was removed and cells were imbedded in 2% agar. They were then postfixed with 1% osmium tetroxide and 1.5% potassium ferrocyanide for 1.5 hr at 4°C. Samples were dehydrated using graded ethanol and embedded in Poly/Bed 812 resin (Polysciences, Warrington. PA). After sectioning, cells were stained with uranyl acetate and lead citrate, and examined by transmission electron microscopy. Scanning electron microscopy. Media were drained off at the end ofthe experiments and replaced with 1.5% glutaraldehyde to fix the cells on the coated glass surface. They were then posttixed with 1% osmium tetroxide for I hr at 22°C. rinsed, dehydrated using graded ethanol solutions, and critical-point-dried with CO* as the transitional fluid. Samples were then mounted on aluminum stubs, sputter-coated with gold. and examined by scanning electron microscopy.
RESULTS Transmission electron microscopy. PC 12 cells in suspension were incubated (37°C 5% COz in humidified air) with KCN (0.01-10 mrvt) for 30,60, or 120 min and examined by transmission EM. Although lower concentrations were ineffective, I to 10 mM KCN for 1 to 2 hr produced three prominent effects: (a) secretory granule depletion, (b) alignment of remaining granules along the plasma membrane, and (c) mitochondrial swelling (Fig. 1). Depletion of secretory granules from PC 12 cells by KCN was not uniform in all cells. Generally KCN-treated cells had fewer secretory granules, but some cells retained their dark staining granules despite other evidence of cyanide’s effect (swollen mitochondria). Some cyanide-treated cells had almost no granules remaining deep within the cytoplasm, but numerous granules were located close to the surface membrane (Fig. 1). By contrast, alignment of granules along the plasma membrane was not observed in control cells. The most consistent finding in cells
216
MADUH
ET AL.
CYANIDE-INDUCED
MORPHOLOGICAL
CHANGES
217
d-lysine surface despite the presence of cyanide. These observations further support the conclusion that cyanide induces retraction of microvilli of PC 12 cells. Cell injury was assessed by measurement of the release of LDH into the incubation media. Incubation of cells with 10 mM KCN increased LDH at 60 min whereas pretreatment with diltiazem attenuated the release of the enzyme (Fig, 3). Diltiazem alone did not increase LDH release above controls (not shown). Cell viability paralleled the release of LDH in that 54% of the cells excluded dye at 60 min KCN incubation and pretreatment with diltiazem blocked cyanide-induced cell death (Fig. 4). However, at 120 min, significant cell death was detected in the presence ofthe diltiazem, although still less than in unprotected cells.
treated with cyanide was mitochondrial swelling. Addition of diltiazem (0.01 or 0.001 m&Q to the incubation medium 15 min prior to KCN prevented secretory granule depletion and alignment of granules along plasma membrane also was not observed under these conditions (Fig. I). Mitochondrial swelling, which appears to be the most sensitive indicator of cyanide’s action, was not evident in cells pretreated with diltiazem. Scanning electron microscopy. Since isolated neuronal cells treated with cyanide have not been previously studied using scanning electron microscopy, the photomicrographs taken in the present experiments were examined in detail. Control cells were characterized by abundant microvilli on their surface (Fig. 2). Addition of cyanide ( 1 to 10 mM for 1 to 2 hr) resulted in loss of microvilli and some blebbing of the plasma membrane (Fig. 2). Some cyanide-treated cells appeared rounded and exhibited fewer and shortened microvilli. Cells displaying blebs had few microvilli. Pretreatment with 0.01 mM diltiazem reduced but did not completely eliminate loss of microvilli and the bleb formation induced by cyanide (Fig. 2). However, microvilli remaining on cells protected with diltiazem were not distorted in shape as compared to those in the unprotected cells. Few cells remained attached to the poly-dlysine surface after cyanide exposure. Therefore the cyanide-treated cells in Fig. 2 were among the few survivors available for scanning electronmicroscopy. However, when either diltiazem or 1 mM EGTA was included in the medium prior to addition of KCN, cells generally remained attached to the poly-
DISCUSSION Since a calcium channel blocker inhibited cyanide-induced morphological changes in PC 12 cells, these changes probably are mediated by influx of extracellular calcium. Thus, the blebbing, loss of microvilli, mitochondrial swelling, and ruffling of the cell surface by cyanide all appear to result from excessive intracellular free calcium. Additionally, cell injury and eventual death are delayed by Ca2+ channel blockade. These findings are in line with the following reports: ( 1) a calcium channel blocker, flunarizine, is reported to be an effective antidote to cyanide intoxication (Dubinsky et al., 1984); (2) calcium channel blockers inhibit lipid peroxidation (and presumably any associated changes in membrane structure) in brain tissue after cyanide -
FIG. I. Effect of KCN and diltiazem on morphology of rat PC I2 cells. Transmission electron microscopy, magnification 7300X. Upper left, control incubated 2 hr; upper right, 1 mM KCN for I hr; lower left, 10 mM KCN for 2 hr (note that most ofthe remaining secretory granules are aligned along the plasma membrane in the treated cells); lower right, 15 min pretreatment 0.0 1 mM with diltiazem followed by 10 mM KCN for 2 hr; note preservation of mitochondrial structure.
218
MADUH
ET AL.
CYANIDE-INDUCED
MORPHOLOGICAL
219
CHANGES
:,oo.--0 2
80-
:: iu
60-
*
40-
*
L.¶
0
I
30
60 Time
90
I
120
IminI
FIG. 3. Etfect of KCN and diltiazem on LDH release by PC12 cells. Ceils were incubated in Ca-KR solution and at specific times. medium aliquots were assayed for LDH activity. Control-no treatment (A): KCN-10 mM added at zero time (0): diltiazem-10-..5 M added 15 min prior to KCN and 10 ItIM at zero time (A). Each value is the mean of four to six assaysand vertical bars are i SEM. Asterisks indicate significant difference from control or from diltiazemlcyanide treatment at specific time periods (p < 0.0 I ).
injection into mice (Johnson et al., 1987b); and (3) cyanide-induced lipid peroxidation in isolated PC12 cells is reduced by a calcium channel blocker or by calcium-free media. From these observations, it is concluded that calcium plays a critical role in disruption of neuronal cell morphology associated with cyanide intoxication. In support, Sher ( 1988) noted cell swelling and vacuolization l-2 hr after addition of I mM NaCN to cultured cells from cerebral cortex of 16- to 17-day-old fetal mice. These changes were prevented by previous addition of 10 mM MgClz to the medium. Since magnesium is capable of blocking calcium channels, these results parallel and are supportive of those of the present study. A proposed mechanism of structural damage resulting from disruption of cell metabo-
; 0 . E 0 0 c :
20
30
60 Time
90
Ii0
IminI
FIG. 4. Effect of KCN and diltiazem on cell viability. Cells were suspended in KR solution and at specific times cell viability, expressed as percentage of cells excluding trypan blue dye, was determined. At zero time 10 mM KCN was added alone (0) or to cells pretreated 15 min with 10m5M diltiazem (0). Each value represents the mean of three to four assaysand vertical bars are + SEM. Asterisks indicate significant difference from control (zero time) at p ( 0.01. Although there was some loss of viability in the diltiazem protected cells at 120 min, cell death was still significantly less (p < 0.05) than that caused by cyanide treatment alone.
lism suggests that ATP depletion inhibits active membrane transport of ions, producing altered ionic homeostasis and accumulation of intracellular sodium and water (Schwertschlag et al., 1986). Thus both cells and organelles swell and eventually ionic equilibrium across the plasma membrane is achieved. The present results suggest that a critical aspect of this process is entry of extracellular calcium, since calcium channel blockade prevents cyanide-induced morphological changes and manipulations limiting neuronal free calcium diminished membrane lipid peroxidation by cyanide (Johnson et al., 1987b). Mitochondrial swelling was the most sensitive morphological index of cyanide’s effect in PC1 2 cells. Two theories of anoxia-in-
FIG. 2. Effect of KCN and diltiazem on morphology of rat PC 12 cells. Scanning electron microscopy, magnification 1400X. Upper, control 2 hr incubation; middle, 2 hr incubation with 10 mM KCN (note loss of microvilli and blebbing): lower, 0.01 mM diltiazem added I5 min prior to IO mrvrKCN followed by 2 hr of incubation.
220
MADUH
duced mitochondrial swelling (Beatrice et al., 1984; Tagawa et al., 1985) suggest that release of bound intramitochondrial calcium leads to structural alterations in these organelles. However, the present study shows that diltiazem, which blocks calcium channels at the cell surface, prevents cyanide-induced mitochondrial swelling. Thus events at the plasma membrane may also be critical in the production of mitochondrial changes under conditions of impaired metabolism. The present study provides evidence that cyanide initiates functional changes in the PC12 cell, a neurosecretory cell line which secretes norepinephrine and dopamine (Greene and Tischler, 1976). Depletion of neurotransmitter granules and alignment of the remaining granules along plasma membranes indicate cyanide-activated exocytosis. Evoked release of neurotransmitters from storage granules is activated by elevation of cytosolic Ca *+. Cyanide-stimulated increases in cytosolic Ca2+ appear to activate neurotransmitter secretion as evidenced by the granule depletion and reversal of this phenomenon by diltiazem. These observations have important implications in cyanide toxicity, since the functional correlate would be release of central neurotransmitters producing excessive CNS firing. Cyanide-induced morphologic changes are probably initiated by decreased availability of ATP which impairs calcium and sodium extrusion processes. Cell swelling may occur due to increased cytosolic NaCl (van Reempts, 1984). Depolarization may also occur, resulting in activation of voltage-dependent calcium channels and subsequent release of neurotransmitters (Persson et al., 1985; Borowitz et al., 1988). Entering calcium is not removed and accumulates to toxic levels. Proteases are activated, oxygen radicals are generated, and destruction of cellular elements occurs (Halliwell, 1987; Braugler et al., 1985). Structures important for maintaining cell architecture (e.g., microfilaments and microtubules) are damaged
ET AL.
giving rise to retraction of microvilli and bleb formation. Unless metabolism is restored and repair processes are initiated, cell function ceases and death ensues. Such alterations may explain neuronal damage observed after cyanide intoxication (Uitti et al., 1985; Carella et al., 1988; Rosenberg et al., 1989). ACKNOWLEDGMENT This study was supported in part by PHS Grants 5T32ES07034, ES04140, and 507RR5586.
REFERENCES ASHTON,
D.,
VANREEMPTS,
J., AND
WAUQUIER,
A.
( 198 1). Behavioral, electroencephalographic and histological study of the protective effect of etomidate against histotoxic dysoxia produced by cyanide. Arch. Int. Pharmacodyn. 254,196-2 13. BEATRICE,
M. D., STIERS, D. L.. PFEIFFER.
D. R. (1984).
The role of glutathione in retention of Ca2+ by liver mitochondria. J. Biol. Chem. 259,1279-1287. BRAUGHLER, J. M., DUNCAN, L. A., ANDGOODMAN, T. ( 1985). Calcium enhances in vitro free radical induced damage to brain synaptosomes, mitochondria, and cultured spinal cord neurons. J. Neurochem. 45,12861293. BRIERLY,
J. B., BROWN,
A. W., AND CAVERLY,
J. (1976).
Cyanide intoxication in the rat; physiological and neuropathological aspects. J. Neurol. Neurosurg. Psychiat. 39, 129-140. BOROWITZ, J. L., BORN, G. S., AND ISOM, G. E. (1988).
Potentiation of evoked adrenal catecholamine release by cyanide: Possible role of calcium. Toxicology 50, 37-45. CARELLA, F., GRASSI, M. P., SAVOIARD~, M., CONTRI, P.. RAPUZZI, B., AND MANGONI, A. (1988). Dystonia-
parkinsonian syndrome after cyanide poisoning: Clinical and MRI finding. J. Neural. Neurosurg. Psychiat. 51, 1345-1348. DUBINSKY, RITCHIE,
B., SIERCHIO,
J. N., TEMPLE,
D. E., AND
D. M. (1984). Flunarizine and verapamil: Effects on central nervous system and peripheral consequences of cytotoxic hypoxia in rats. Life Sci. 34, 1299-1306.
FUNATA, SHINO,
N., SONG,
S.-Y.,
FUNATA,
M..
AND HAYA-
F. (1984). A study of experimental cyanide encephalopathy in the acute phase-physiological and pathological correlation. Acta Neuropathol. 64, 99107.
CYANIDE-INDUCED
MORPHOLOGICAL
GOLDBERG, M. P.. WEISS. J. H., PHAM, P.-C., AND CHOI, D. W. (1987). N-methyl-tr-aspartate receptors mediate hypoxic neuronal injury in cortical culture. J. Pharmacol. E.rp. Ther. 243, 784-79 1. GREENE. L. A., AND TISCHLER, A. S. (1976). Establishment of a noradrenergic clonal line of adrenal pheochromocytoma ceils which respond to nerve growth factor. Proc. Natl. .4cad. Sci. USA 73,2424-2428. HALI.IWELL, B. (1987). Oxidants and human disease: Some new concepts. F,4SEB J. 1,358-364. HIGGINS, T. J. C., AND BAILEY, P. J. (1983). The effects of cyanide and iodoacetate intoxication and ischemia on enzyme release from the perfused rat heart. Biochim. Biophy.s. Acta 762, 67-7 5, JOHNSON, J. D.. CONROY. W. G., AND ISOM, G. E. ( l987a). Alteration of cytosolic calcium levels in PC 12 cells by potassium cyanide. Touicol. .4ppl. Pharmacol. 88,2 17-224. JOHNSON. J. D.. CONROY, W. G.. AND ISOM, G. E. ( 1987b). Peroxidation ofbrain lipids following cyanide intoxication in mice. Toxicology46,21-28. JOHNSON, J. D., MEISENHEIMER. T. L.. AND ISOM, G. E. ( 1986). Cyanide-induced neurotoxicity: Role of neuronal calcium. To\-it,&. .4ppl. Pharmacol. 84, 464469. NK‘OTERA. P.. THOR. H.. AND ORRENIUS. S. (1989). Cytosolic-free Cal+ and cell killing in hepatoma lclc7 cells exposed to chemical hypoxia. F4SEB.J. 3,59-64.
CHANGES
221
PERSSON, A. R., CASSEL, G., AND SELLSTROM, A. ( 1985). Acute cyanide intoxication and central transmitter systems. Fundam. Appi. To.Gcol. S 150-S 155. ROSENBERG, N. L., MYERS, J. A., AND MARTIN, W. R. W. (1989). Cyanide-induced ParkinsonismClinical, MRI and h-fluorodopa PET studies. Neuralog?t39, 142-143. SCHANNE, A. X., KANE, A. B., YOUNG. E. E., ANL) FARBER, J. L. (1979). Calcium dependence of toxic cell death: A final common pathway. Science 206, 700-702. SCHWERTSCHLAG, U., SCHRIER. R. W., AND WILSON. P. (1986). Beneficial effects of calcium channel blockers and calmodulin binding drugs on in vitro renal cell anoxia. J. Pharmacol. Exp. Ther. 238, 1 19-124. SHER, P. K. (1988). Neurochemical correlates of cyanide-induced hypoxic neuronal damage in vitro. .Vl~jnrochem. Rex 13, 159-163. TAGAWA, K., NISHIDA. T., WATANABE, F., AND KOSEKI. M. (1985). Mechanism of anoxic damage of mitochondria: Depletion of intramitochondrial ATP and concomitant release of free Ca’+. Mol. Physiol. 8.5 15-524. UITTI, R. J.. RAJPIJT. A. H.. ASHENHIJRST. E. M.. AUL) ROZDILSKY, B. ( 1985). Cyanide-induced parkinsonism: A clinicopathological report. Nuzrvo/o~r35.92 l-925. VAN REEMPTS. J. (1984). The hypoxic brain: Histological and ultrastructural aspects. Behav. Brain Rev 14, 99-108.
TOXlCOLOGYANDAPPLlEDPHARMACOLOGY
Cyanide-Induced Neurotoxicity: Calcium Mediation of Morphological Changes in Neuronal Cells E. U. MADUH, J. J. TUREK,* J. L. BOROWITZ, A. REBAR,* AND G. E. ISOM Department of Pharmacology and Toxicology, Veterinary Anatomy, School of Veterinary
Received
May
School of Pharmacy Medicine, Purdue
22, 1989; accepted
and Pharmacal Sciences, University, West Lafayette,
December
and *Department Indiana 47907
of
12, 1989
Cyanide-Induced Neurotoxicity: Calcium Mediation of Morphological Changes in Neuronal Cells. MADUH, E. U., TUREK, J. J., BOROWITZ, J. L., REBAR, A., AND ISOM, G. E. (1990). Toxicol. Appl. Pharmacol. 103, 2 14-22 1. Calcium channel blockade decreases the elevation of brain calcium as well as the tremors produced by cyanide in mice. To determine if cyanideinduced morphological changes could also be inhibited by calcium channel blockade, the effect of diltiazem was studied in cultured rat pheochromocytoma (PC12) cells, a neuronal model. Incubation with KCN (I to 10 mM for 1 to 2 hr) caused depletion of secretory granules, alignment of remaining granules along the plasma membrane. and mitochondrial swelling. All these effects were inhibited by pretreatment with 0.0 1 mM diltiazem. Scanning electron microscopy revealed that cyanide (I to IO mM for I to 2 hr) produced loss of microvilli and bleb formation in PC12 cells. These changes were partially inhibited by preincubation with 0.01 mM diltiazem. Incubation of cells with 10 mM cyanide increased release of lactic dehydrogenase (LDH) into the culture media at 60 and 120 min. A decrease in cell viability, as determined by trypan blue dye exclusion, paralleled the release of LDH. At 120 min of cyanide incubation, 24% of the cells excluded dye. Both the release of LDH and decreased cell viability were attenuated by pretreatment with diltiazem. The results indicate that the influx of extracellular calcium is an important factor mediating cyanide-induced morphologic changes in neuronal cells. o 1990 Academic
Press. Inc.
Permanent neurological damage by cyanide stems from inhibition of cellular metabolism and is preceded by changes in cell morphology (Brierly et al., 1976; Ashton et al., 1981; Funata et a/., 1984). Changes in cell morphology are therefore an important indication of neuronal damage by cyanide. Exact mechanisms by which cyanide causes neuronal structural alterations have not been clarified. Recently, Johnson et al. ( 1986) found cyanide treatment elevates total brain calcium in mice and increases cytosolic calcium in cultured rat pheochromocytoma (PC 12) cells (Johnson, 1987a). Peroxidation of brain lipid membranes also occurs in cultured neuronal 0041-008X/90
$3.00
Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
cells and in mouse brain after cyanide treatment (Johnson et al., 1987b). Both elevation of cytosolic calcium and lipid peroxidation induced by cyanide are inhibited by pretreatment with the calcium channel blocker, diltiazem (Johnson et al., 1987a,b). In hepatoma 1c 1c7 cells, KCN produced a rapid rise in cytosolic free Ca*+ which preceded blebbing of the plasma membrane and loss of cell viability (Nicotera et al., 1989). Cyanide-induced morphological changes in neuronal cells may be related to elevated cytosolic calcium and associated membrane damage. Therefore, the present study employs a calcium channel blocker to determine whether cyanide-in-
214
CYANIDE-INDUCED
MORPHOLOGICAL
duced changes in neuronal cell morphology and cell viability are related to extracellular calcium influx. METHODS C’eN culture. Rat pbeochromocytoma (PC 12) cells obtained from American Type Culture Collection were grown (37°C. 5% CO* in humidified air) attached to plastic culture flasks until confluency in a medium of 85% RPM1 1640, 10% v/v heat-inactivated horse serum, and 5% v/v fetal calf serum containing 1% penicillin-streptomycin. At the time ofthe experiment, cells were detached using a stream of media from a pipet and suspended in Krebs-Ringer (CaKR) solution consisting of(mM) NaCI, 125; Hepes-NaOH, 25; glucose. 6; NaHC03, 5; MgSO.,, 1.2; KH,PO,, 1.2; and CaC12, 1.O.Potassium cyanide solutions were prepared just prior to use. The millimolar concentrations of cyanide employed are the same as those used previously and correspond to amounts associated with acute intoxication in vivo (Johnson et al.. I987a). Diltiazem was added to the medium 15 min before KCN. Approximately IO6 cells were used for each sample. PC I2 cells for scanning electron microscopy were cultured at a density of 200.000 cells per chamber in glass chamber slides coated with poly-d-lysine. Media were refreshed every other day for lo- 14 days. At the time ofthe experiments, growth media were replaced with KrebsRinger solution and test agents added. LDH release was assessedaccording to the method of Higgins and Bailey (I 983) and Goldberg et al. (1987). Cells (2-4 X 106) were suspended in I ml CaKR media with or without treatments. After 30.60, or 120 min cells were rapidly pelleted ( 10,OOOgfor 10 set) and 100 ~1 of bathing media was assayed for LDH activity. Cyanide does not influence the measurement of lactate dehydrogenase activity (Higgings and Bailey, 1983). Cell viability was determined by the dye exclusion method in which cells were suspended in CaKR and incubated at 37’C (Schanne et ai.. 1979). At 30, 60. and 120 mitt after addition of KCN. 0.2-ml aliquots were withdrawn from the suspension and diluted five-fold with 0.2% trypan blue, and cells excluding dye were counted in a homocytometer. Each sample before addition of KCN served as its own control. RPM1 1640 and falcon plastic tissue culture flasks were purchased from GIBCO (Grand Island, NY); glass chamber slides were from Miles Scientific (Naperville, IL). Sera were obtained from HyClone (Logan, UT), penicillin G potassium from E. R. Squibb and Sons (Princeton, NJ), streptomycin from Eli Lilly &Co. (Indianapolis. IN), and poly-d-lysine from Sigma Chemical Co. (St. Louis, MO). Dr. Ronald Browne of Marion Lab-
CHANGES
315
oratory (Kansas City. MO) generously supplied diltiazem hydrochloride. Potassium cyanideand otherchemicals were of the highest chemical grade commercially available. Transmission electron microscopy. At the termination of the experiments, cells were pelleted at 1OOOgfor 4 min. Media were removed and 1.5% glutaraldehyde was added to fix the cells in the pellet. Fixative was removed and cells were imbedded in 2% agar. They were then postfixed with 1% osmium tetroxide and 1.5% potassium ferrocyanide for 1.5 hr at 4°C. Samples were dehydrated using graded ethanol and embedded in Poly/Bed 812 resin (Polysciences, Warrington. PA). After sectioning, cells were stained with uranyl acetate and lead citrate, and examined by transmission electron microscopy. Scanning electron microscopy. Media were drained off at the end ofthe experiments and replaced with 1.5% glutaraldehyde to fix the cells on the coated glass surface. They were then posttixed with 1% osmium tetroxide for I hr at 22°C. rinsed, dehydrated using graded ethanol solutions, and critical-point-dried with CO* as the transitional fluid. Samples were then mounted on aluminum stubs, sputter-coated with gold. and examined by scanning electron microscopy.
RESULTS Transmission electron microscopy. PC 12 cells in suspension were incubated (37°C 5% COz in humidified air) with KCN (0.01-10 mrvt) for 30,60, or 120 min and examined by transmission EM. Although lower concentrations were ineffective, I to 10 mM KCN for 1 to 2 hr produced three prominent effects: (a) secretory granule depletion, (b) alignment of remaining granules along the plasma membrane, and (c) mitochondrial swelling (Fig. 1). Depletion of secretory granules from PC 12 cells by KCN was not uniform in all cells. Generally KCN-treated cells had fewer secretory granules, but some cells retained their dark staining granules despite other evidence of cyanide’s effect (swollen mitochondria). Some cyanide-treated cells had almost no granules remaining deep within the cytoplasm, but numerous granules were located close to the surface membrane (Fig. 1). By contrast, alignment of granules along the plasma membrane was not observed in control cells. The most consistent finding in cells
216
MADUH
ET AL.
CYANIDE-INDUCED
MORPHOLOGICAL
CHANGES
217
d-lysine surface despite the presence of cyanide. These observations further support the conclusion that cyanide induces retraction of microvilli of PC 12 cells. Cell injury was assessed by measurement of the release of LDH into the incubation media. Incubation of cells with 10 mM KCN increased LDH at 60 min whereas pretreatment with diltiazem attenuated the release of the enzyme (Fig, 3). Diltiazem alone did not increase LDH release above controls (not shown). Cell viability paralleled the release of LDH in that 54% of the cells excluded dye at 60 min KCN incubation and pretreatment with diltiazem blocked cyanide-induced cell death (Fig. 4). However, at 120 min, significant cell death was detected in the presence ofthe diltiazem, although still less than in unprotected cells.
treated with cyanide was mitochondrial swelling. Addition of diltiazem (0.01 or 0.001 m&Q to the incubation medium 15 min prior to KCN prevented secretory granule depletion and alignment of granules along plasma membrane also was not observed under these conditions (Fig. I). Mitochondrial swelling, which appears to be the most sensitive indicator of cyanide’s action, was not evident in cells pretreated with diltiazem. Scanning electron microscopy. Since isolated neuronal cells treated with cyanide have not been previously studied using scanning electron microscopy, the photomicrographs taken in the present experiments were examined in detail. Control cells were characterized by abundant microvilli on their surface (Fig. 2). Addition of cyanide ( 1 to 10 mM for 1 to 2 hr) resulted in loss of microvilli and some blebbing of the plasma membrane (Fig. 2). Some cyanide-treated cells appeared rounded and exhibited fewer and shortened microvilli. Cells displaying blebs had few microvilli. Pretreatment with 0.01 mM diltiazem reduced but did not completely eliminate loss of microvilli and the bleb formation induced by cyanide (Fig. 2). However, microvilli remaining on cells protected with diltiazem were not distorted in shape as compared to those in the unprotected cells. Few cells remained attached to the poly-dlysine surface after cyanide exposure. Therefore the cyanide-treated cells in Fig. 2 were among the few survivors available for scanning electronmicroscopy. However, when either diltiazem or 1 mM EGTA was included in the medium prior to addition of KCN, cells generally remained attached to the poly-
DISCUSSION Since a calcium channel blocker inhibited cyanide-induced morphological changes in PC 12 cells, these changes probably are mediated by influx of extracellular calcium. Thus, the blebbing, loss of microvilli, mitochondrial swelling, and ruffling of the cell surface by cyanide all appear to result from excessive intracellular free calcium. Additionally, cell injury and eventual death are delayed by Ca2+ channel blockade. These findings are in line with the following reports: ( 1) a calcium channel blocker, flunarizine, is reported to be an effective antidote to cyanide intoxication (Dubinsky et al., 1984); (2) calcium channel blockers inhibit lipid peroxidation (and presumably any associated changes in membrane structure) in brain tissue after cyanide -
FIG. I. Effect of KCN and diltiazem on morphology of rat PC I2 cells. Transmission electron microscopy, magnification 7300X. Upper left, control incubated 2 hr; upper right, 1 mM KCN for I hr; lower left, 10 mM KCN for 2 hr (note that most ofthe remaining secretory granules are aligned along the plasma membrane in the treated cells); lower right, 15 min pretreatment 0.0 1 mM with diltiazem followed by 10 mM KCN for 2 hr; note preservation of mitochondrial structure.
218
MADUH
ET AL.
CYANIDE-INDUCED
MORPHOLOGICAL
219
CHANGES
:,oo.--0 2
80-
:: iu
60-
*
40-
*
L.¶
0
I
30
60 Time
90
I
120
IminI
FIG. 3. Etfect of KCN and diltiazem on LDH release by PC12 cells. Ceils were incubated in Ca-KR solution and at specific times. medium aliquots were assayed for LDH activity. Control-no treatment (A): KCN-10 mM added at zero time (0): diltiazem-10-..5 M added 15 min prior to KCN and 10 ItIM at zero time (A). Each value is the mean of four to six assaysand vertical bars are i SEM. Asterisks indicate significant difference from control or from diltiazemlcyanide treatment at specific time periods (p < 0.0 I ).
injection into mice (Johnson et al., 1987b); and (3) cyanide-induced lipid peroxidation in isolated PC12 cells is reduced by a calcium channel blocker or by calcium-free media. From these observations, it is concluded that calcium plays a critical role in disruption of neuronal cell morphology associated with cyanide intoxication. In support, Sher ( 1988) noted cell swelling and vacuolization l-2 hr after addition of I mM NaCN to cultured cells from cerebral cortex of 16- to 17-day-old fetal mice. These changes were prevented by previous addition of 10 mM MgClz to the medium. Since magnesium is capable of blocking calcium channels, these results parallel and are supportive of those of the present study. A proposed mechanism of structural damage resulting from disruption of cell metabo-
; 0 . E 0 0 c :
20
30
60 Time
90
Ii0
IminI
FIG. 4. Effect of KCN and diltiazem on cell viability. Cells were suspended in KR solution and at specific times cell viability, expressed as percentage of cells excluding trypan blue dye, was determined. At zero time 10 mM KCN was added alone (0) or to cells pretreated 15 min with 10m5M diltiazem (0). Each value represents the mean of three to four assaysand vertical bars are + SEM. Asterisks indicate significant difference from control (zero time) at p ( 0.01. Although there was some loss of viability in the diltiazem protected cells at 120 min, cell death was still significantly less (p < 0.05) than that caused by cyanide treatment alone.
lism suggests that ATP depletion inhibits active membrane transport of ions, producing altered ionic homeostasis and accumulation of intracellular sodium and water (Schwertschlag et al., 1986). Thus both cells and organelles swell and eventually ionic equilibrium across the plasma membrane is achieved. The present results suggest that a critical aspect of this process is entry of extracellular calcium, since calcium channel blockade prevents cyanide-induced morphological changes and manipulations limiting neuronal free calcium diminished membrane lipid peroxidation by cyanide (Johnson et al., 1987b). Mitochondrial swelling was the most sensitive morphological index of cyanide’s effect in PC1 2 cells. Two theories of anoxia-in-
FIG. 2. Effect of KCN and diltiazem on morphology of rat PC 12 cells. Scanning electron microscopy, magnification 1400X. Upper, control 2 hr incubation; middle, 2 hr incubation with 10 mM KCN (note loss of microvilli and blebbing): lower, 0.01 mM diltiazem added I5 min prior to IO mrvrKCN followed by 2 hr of incubation.
220
MADUH
duced mitochondrial swelling (Beatrice et al., 1984; Tagawa et al., 1985) suggest that release of bound intramitochondrial calcium leads to structural alterations in these organelles. However, the present study shows that diltiazem, which blocks calcium channels at the cell surface, prevents cyanide-induced mitochondrial swelling. Thus events at the plasma membrane may also be critical in the production of mitochondrial changes under conditions of impaired metabolism. The present study provides evidence that cyanide initiates functional changes in the PC12 cell, a neurosecretory cell line which secretes norepinephrine and dopamine (Greene and Tischler, 1976). Depletion of neurotransmitter granules and alignment of the remaining granules along plasma membranes indicate cyanide-activated exocytosis. Evoked release of neurotransmitters from storage granules is activated by elevation of cytosolic Ca *+. Cyanide-stimulated increases in cytosolic Ca2+ appear to activate neurotransmitter secretion as evidenced by the granule depletion and reversal of this phenomenon by diltiazem. These observations have important implications in cyanide toxicity, since the functional correlate would be release of central neurotransmitters producing excessive CNS firing. Cyanide-induced morphologic changes are probably initiated by decreased availability of ATP which impairs calcium and sodium extrusion processes. Cell swelling may occur due to increased cytosolic NaCl (van Reempts, 1984). Depolarization may also occur, resulting in activation of voltage-dependent calcium channels and subsequent release of neurotransmitters (Persson et al., 1985; Borowitz et al., 1988). Entering calcium is not removed and accumulates to toxic levels. Proteases are activated, oxygen radicals are generated, and destruction of cellular elements occurs (Halliwell, 1987; Braugler et al., 1985). Structures important for maintaining cell architecture (e.g., microfilaments and microtubules) are damaged
ET AL.
giving rise to retraction of microvilli and bleb formation. Unless metabolism is restored and repair processes are initiated, cell function ceases and death ensues. Such alterations may explain neuronal damage observed after cyanide intoxication (Uitti et al., 1985; Carella et al., 1988; Rosenberg et al., 1989). ACKNOWLEDGMENT This study was supported in part by PHS Grants 5T32ES07034, ES04140, and 507RR5586.
REFERENCES ASHTON,
D.,
VANREEMPTS,
J., AND
WAUQUIER,
A.
( 198 1). Behavioral, electroencephalographic and histological study of the protective effect of etomidate against histotoxic dysoxia produced by cyanide. Arch. Int. Pharmacodyn. 254,196-2 13. BEATRICE,
M. D., STIERS, D. L.. PFEIFFER.
D. R. (1984).
The role of glutathione in retention of Ca2+ by liver mitochondria. J. Biol. Chem. 259,1279-1287. BRAUGHLER, J. M., DUNCAN, L. A., ANDGOODMAN, T. ( 1985). Calcium enhances in vitro free radical induced damage to brain synaptosomes, mitochondria, and cultured spinal cord neurons. J. Neurochem. 45,12861293. BRIERLY,
J. B., BROWN,
A. W., AND CAVERLY,
J. (1976).
Cyanide intoxication in the rat; physiological and neuropathological aspects. J. Neurol. Neurosurg. Psychiat. 39, 129-140. BOROWITZ, J. L., BORN, G. S., AND ISOM, G. E. (1988).
Potentiation of evoked adrenal catecholamine release by cyanide: Possible role of calcium. Toxicology 50, 37-45. CARELLA, F., GRASSI, M. P., SAVOIARD~, M., CONTRI, P.. RAPUZZI, B., AND MANGONI, A. (1988). Dystonia-
parkinsonian syndrome after cyanide poisoning: Clinical and MRI finding. J. Neural. Neurosurg. Psychiat. 51, 1345-1348. DUBINSKY, RITCHIE,
B., SIERCHIO,
J. N., TEMPLE,
D. E., AND
D. M. (1984). Flunarizine and verapamil: Effects on central nervous system and peripheral consequences of cytotoxic hypoxia in rats. Life Sci. 34, 1299-1306.
FUNATA, SHINO,
N., SONG,
S.-Y.,
FUNATA,
M..
AND HAYA-
F. (1984). A study of experimental cyanide encephalopathy in the acute phase-physiological and pathological correlation. Acta Neuropathol. 64, 99107.
CYANIDE-INDUCED
MORPHOLOGICAL
GOLDBERG, M. P.. WEISS. J. H., PHAM, P.-C., AND CHOI, D. W. (1987). N-methyl-tr-aspartate receptors mediate hypoxic neuronal injury in cortical culture. J. Pharmacol. E.rp. Ther. 243, 784-79 1. GREENE. L. A., AND TISCHLER, A. S. (1976). Establishment of a noradrenergic clonal line of adrenal pheochromocytoma ceils which respond to nerve growth factor. Proc. Natl. .4cad. Sci. USA 73,2424-2428. HALI.IWELL, B. (1987). Oxidants and human disease: Some new concepts. F,4SEB J. 1,358-364. HIGGINS, T. J. C., AND BAILEY, P. J. (1983). The effects of cyanide and iodoacetate intoxication and ischemia on enzyme release from the perfused rat heart. Biochim. Biophy.s. Acta 762, 67-7 5, JOHNSON, J. D.. CONROY. W. G., AND ISOM, G. E. ( l987a). Alteration of cytosolic calcium levels in PC 12 cells by potassium cyanide. Touicol. .4ppl. Pharmacol. 88,2 17-224. JOHNSON. J. D.. CONROY, W. G.. AND ISOM, G. E. ( 1987b). Peroxidation ofbrain lipids following cyanide intoxication in mice. Toxicology46,21-28. JOHNSON, J. D., MEISENHEIMER. T. L.. AND ISOM, G. E. ( 1986). Cyanide-induced neurotoxicity: Role of neuronal calcium. To\-it,&. .4ppl. Pharmacol. 84, 464469. NK‘OTERA. P.. THOR. H.. AND ORRENIUS. S. (1989). Cytosolic-free Cal+ and cell killing in hepatoma lclc7 cells exposed to chemical hypoxia. F4SEB.J. 3,59-64.
CHANGES
221
PERSSON, A. R., CASSEL, G., AND SELLSTROM, A. ( 1985). Acute cyanide intoxication and central transmitter systems. Fundam. Appi. To.Gcol. S 150-S 155. ROSENBERG, N. L., MYERS, J. A., AND MARTIN, W. R. W. (1989). Cyanide-induced ParkinsonismClinical, MRI and h-fluorodopa PET studies. Neuralog?t39, 142-143. SCHANNE, A. X., KANE, A. B., YOUNG. E. E., ANL) FARBER, J. L. (1979). Calcium dependence of toxic cell death: A final common pathway. Science 206, 700-702. SCHWERTSCHLAG, U., SCHRIER. R. W., AND WILSON. P. (1986). Beneficial effects of calcium channel blockers and calmodulin binding drugs on in vitro renal cell anoxia. J. Pharmacol. Exp. Ther. 238, 1 19-124. SHER, P. K. (1988). Neurochemical correlates of cyanide-induced hypoxic neuronal damage in vitro. .Vl~jnrochem. Rex 13, 159-163. TAGAWA, K., NISHIDA. T., WATANABE, F., AND KOSEKI. M. (1985). Mechanism of anoxic damage of mitochondria: Depletion of intramitochondrial ATP and concomitant release of free Ca’+. Mol. Physiol. 8.5 15-524. UITTI, R. J.. RAJPIJT. A. H.. ASHENHIJRST. E. M.. AUL) ROZDILSKY, B. ( 1985). Cyanide-induced parkinsonism: A clinicopathological report. Nuzrvo/o~r35.92 l-925. VAN REEMPTS. J. (1984). The hypoxic brain: Histological and ultrastructural aspects. Behav. Brain Rev 14, 99-108.
