- Email: [email protected]
Evaluation of amiodarone free radical toxicity in rat hepatocytes
Toxicolt~g~ Ler tefs, 56 ( 1991) 1I 7-t 26 @ 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0378-4274~91~$3.50
117
ADONlS037842749100056X
TOXLET 02539
Evaluation of amiodarone free radical toxicity in rat hepatocytes
Randall J. Ruchl, Subhankar Bandyopadhyayz, and James E. Klaunigl”
Pitambar Somani2
Departments of 1Pathology and 2Pharmaco/ogy, Medical College of Ohio, Toledo, OH (U.S.A. j
(Received 19 January 1990) (Accepted 31 October 1990) Ke,y words: Amiodarone; Free radicals; Hepatocytes _-._ SUMMARY The possible rofes of free radicals and lipid peroxidation in the m~hanism of toxicity of amiodarone (AD) [2-butyl-3-(3’,S-diiodo-4’tr-diethylaminoethoxybenzoyl)~nzofuran] and its principle metabolite, desethylamiodarone (DE), were examined in primary cultured Sprague-Dawley male rat hepatocytes. AD (20 and 40 ,ug/ml) and DE (10 and 25 pg/ml) killed hepatocytes in concentration- and time-dependent fashions, Several antioxidants [Cu,Zn-superoxide dismutase (200 U/ml), catalase (200 U/ml), N,hr-diphenylphenylenediamine (DPPD; 25 PM), butylated hydroxytoluene (0.1 mM), and N-acetylcysteine (5 mM)] were incapable of preventing AD and DE hepatocyte toxicity. Only vitamin E (VE, d,l-a-tocopherol acetate; 20-200 /IM) prevented AD and DE toxicity. No correlation between the onset of hepatocyte death by AD and DE and hepatocyte lipid peroxidation was seen. Both drugs inhibited NADPH-dependent rat liver microsomal superoxide production. These results, excluding the preventive effects of VE, do not support a free radical/lipid peroxidation mechanism of hepatocyte toxicity by AD and DE. VE may have prevented hepatocyte toxicity through non-antioxidant effects.
INTRODUCTION
Amiodarone (AD) [2-butyl-3-(3’,5’-diiodo-4’cc-diethylaminoethoxybenzoyl)benzofuran] is a class III antiarrhythmic drug that is effective in the treatment of resistant ventricular and supraventricular arrhythmia [ 1,2]. Several side-effects of AD therapy have been reported which include cornea1 deposits, altered thyroid function, muscle weakness and paresthesias, phototoxicity, pulmonary fibrosis, and hepatotoxicity [3].
Addressf#r eorre~po~dence; James E. Klaunig, Ph.D., Department of Pathology, Medical College of Ohio,
3000 Arlington Ave., Toledo, OH 43614, U.S.A.
118
The cellular
mechanism
of AD toxicity
is unknown.
Free radicals
are generated
upon exposure of AD and DE to ultraviolet light and this effect has been shown to be important in the phototoxicity of these drugs [4-61. Kennedy et al [7] have recently proposed that AD also induces lung toxicity through free radical generation. They showed that in perfused rabbit lungs in the absence of ultraviolet light exposure, the toxicity of AD was inhibited by the antioxidants, vitamin E, butylated hydroxyanisole, and N-acetylcysteine, but not by superoxide dismutase or catalase. They also demonstrated that lung chemiluminescence, superoxide production, and tissue levels of glutathione disulfide were increased by AD. Since another important side-effect of AD therapy is hepatotoxicity [3], we have examined whether AD hepatotoxicity can also be attributed to free radical generation. MATERIALS
AND METHODS
Animals Male Sprague-Dawley rats (1 O&l 50 g) were purchased from Charles River Laboratories (Wilmington, MA), housed in polycarbonate cages, and fed Purina Rodent Chow (Ralston Purina Co., St. Louis, MO) and water ad libitum. Chemicals Amiodarone
(AD) and desethylamiodarone
(DE) were generous
gifts of Sanofi
Labaz Company (Montpellier, France). Copper, zinc-superoxide dismutase (SOD), catalase (CAT), N&V’-diphenyl-p-phenylenediamine (DPPD), vitamin E (d&cctocopherol acetate) (VE), N-acetylcysteine (NAC), butylated hydroxytoluene (BHT) and paraquat were obtained from Sigma Chemical Co. (St. Louis, MO). All other reagents and tissue culture supplies were purchased from sources as previously indicated [S]. Hepatocyte isolation andprimary culture Rat hepatocytes were isolated by two-stage collagenase perfusion of the liver [9]. Hepatocyte viability after isolation was determined by trypan blue dye exclusion and was greater than 90%. Hepatocytes were plated onto 60 mm plastic tissue culture dishes at one million viable cells per dish in 3 ml medium. Hepatocyte culture medium consisted of Leibovitz’s L-15 medium supplemented with 1 mg/ml a-D-glucose, 10% fetal bovine serum, 1 ,uM dexamethasone and 50 pg/ml gentamicin sulfate [lo]. The cultures were incubated in a humidified 100% air incubator at 37°C. After 2 h culture to allow for cell attachment, the cultures were refed with 5 ml supplemented L- 15 medium per dish and used for experiments. Hepatocyte killing by AD and DE and prevention by antioxidants After attachment and refeeding, the 2-h-old hepatocyte cultures were treated with AD (20 or 40 pg/ml) or DE (10 or 25 pg/ml) dissolved in DMSO. Control cultures
119
were treated with DMSO only (2 $/ml). The cells were incubated with the compounds in the dark in the tissue culture incubator for 2,4, 8 and 24 h. Aliquots (0.1 ml) of the culture media were then removed and analyzed for lactate dehydrogenase (LDH) activity as an indication of hepatocyte killing. LDH activity was determined spectrophotometrically as we have described [8,1 I]. Hepatocyte toxicity (LDH release) was expressed as a percentage of the total LDH activity in the cells determined in non-treated cultures lysed with 0.01% Triton X-100. We also evaluated the effects of several antioxidants on AD- and DE-induced hepatocyte toxicity. Hepatocyte cultures were first treated with one of several antioxidants [SOD (200 U/ml), CAT (200 U/ml), DPPD (25 PM), VE (20, 100 or 200 PM), NAC (5 mM) or BHT (0.1 mM)], then treated immediately thereafter with AD or DE. LDH release was analyzed after 2,4,8 and 24 h of exposure to the compounds. Hepatacyte iipidperoxidation
Lipid peroxidation products in primary cultured rat hepatocytes treated with AD or DE were determined as thiobarbituric-acid-reactive substances (TRS) as we have previously described [I 11.Cell killing was also determined in these cultures by analyzing LDH release into the culture medium. Microsomal superoxide production
Superoxide radical production in rat liver microsomes treated with AD or DE was determined by the cytochrome c reduction assay as we have described [8]. As a positive control, we also monitored superoxide generation in paraquat-treated microsomes. Statistics
Statistical differences in LDH release, lipid peroxidation, and superoxide production between control and treated hepatocyte cultures or microsomes were compared by analysis of variance and Dunnett’s post-hoc t-test [ 121. RESULTS
AD (20 or 40 lug/ml) and DE (10 or 25 pg/ml) were toxic to rat hepatocytes in dose- and time-dependent manners (Fig. I). DE was more toxic than AD as we have previously noted [13,14]. To evaluate whether AD and DE induced hepatocyte toxicity through free radical generation, hepatocyte cultures were treated with AD (40 pg/ml) or DE (25 @g/ml) in the presence of one of several antioxidants: VE (200 PM), DPPD (25 ,uM), SOD (200 U/ml), CAT (200 U/ml), NAC (0.5 mM), and BHT (0.1 mM). Of these 6 antioxidants, only VE decreased the toxicity of AD and DE (Figs. 2 and 3). DPPD enhanced AD and DE toxicity (Figs. 2 and 3) and the other antioxidants had no effect (data not shown). The preventive effects of VE were dependent on VE concentration
120 o-0
Q
DMSO 0.2%
A--A
AD 20 ug/ml
n
A--A
AD 40 ug/ml
0-O
4
12
8 Treatment
1. Concentration
toxicity
(LDH
per datum
DE 10 ug/ml DE 25 ug/ml
loo-
0
Fig.
-m
and temporal
release) in primary
relationships cultured
16
Duration
of amiodarone
rat hepatocytes
point; mean k SD. *PC 0.05 vs. the DMSO
20
24
(hr)
(AD) and desethylamiodarone
over 24-h treatment
solvent (0.2%) control
duration.
group
(DE) cyn = 3 cultures
at the same sampling
time.
and were different for AD versus DE (Figs. 4 and 5). VE at 100 and 200 ,BM produced nearly complete protection from AD-induced toxicity over the 24-h treatment period (Fig. 4), while VE at 100 and 200 PM only partially reduced the toxicity of DE (Fig. 5). None of the antioxidants by themselves affected hepatocyte LDH release (data not shown). 8-8
DMSO
B-8
AD + DPPD
A-A
AD alone
l
AD + VE
Treatment
Fig. 2. Effects of NJ’-diphenylphenylenediamine cytotoxicity treatment
(LDH duration.
release) of amiodarone n=3
--•*
Duration
(DPPD; (AD; 40 pg/ml)
(hr)
25 pM) and vitamin in primary
cultured
E (VE; 200 PM) on the rat hepatocytes
cultures per point; mean&SD. *P
over 24-h
The toxicity
of
121
u
loo-
8
go-
,$
80-
g -1 io z t%
70-
g
30-
0
o-o
DMSO
D-•
DE + DPPD
A-A
DE alone
+--0
DE + VE
B f 1
*
i
6050*
40-
*
/
r.
0
4
8
12
Treatment
16
Duration
20
24
(hr)
Fig. 3. Effects of NJ’-diphenylphenylenediamine (DPPD; 25 FM) and vitamin E (VE; 200 PM) on the cytotoxicity (LDH release) of desethylamiodarone (DE; 25 pg/ml) in primary cultured rat hepatocytes over 24-h treatment duration. n = 3 cultures per point; mean + SD. *PC 0.05 vs. DE alone group. The toxicity of the DE solvent, DMSO (0.2%), is also shown.
The ability of VE to prevent AD and DE hepatotoxicity suggested that these drugs kill hepatocytes by a mechanism involving free radical generation and lipid peroxidation. This is because VE is the major membrane-soluble antioxidant and inhibitor of lipid peroxidation in cells [15]. Therefore, we tested whether AD and DE coufd stimulate microsomal superoxide radical production and hepatocyte lipid peroxidaO-O
0
DMSO
A--A
AD alone
0-O
AD + VE 100 uM
n
AD + VE 20 uM
l
AD + VE 200
-m
4
8 Treatment
12 Duration
---+
16
20
uM
24
(hr)
Fig. 4. Dose-responsive protective effects of vitamin E (VE) on amiodarone (AD; 40 pg/ml) cytotoxicity (LDH release) in primary cultured rat hepatocytes over 24-h treatment duration. n=3 cultures per point; mean + SD. *P < 0.05vs.AD-only group. Toxicity in DMSO (0.2%) solvent control cultures is also shown.
122 DMSO
0-o
m
A-A
DE alone
H--B
DE + VE 20
uM
4
12
8 Treatment
Fig. 5. Dose-responsive (LDH
per point;
DE+VElOOuM DE + VE 200
uM
loo-
0
toxicity
O-0 +-+
protective
release) in primary
*P
mean*SD.
Duration
effects of vitamin cultured
16
20
24
(hr)
E (VE) on desethylamiodarone
rat hepatocytes group.
(DE; 25 pg/mI) cyto-
over 24-h treatment
Toxicity
in DMSO
duration.
(0.2%) solvent
n = 3 cultures control
cultures
is also shown.
tion at cytotoxic doses. Surprisingly, both compounds inhibited microsomal superoxide generation in a dose-dependent fashion. As a positive control for microsomal superoxide production, paraquat was utilized. Paraquat has been shown to undergo redox-cycling via NADPH cytochrome P-450 reductase to generate superoxide and TABLE
I
RATES
OF
TREATED
SUPEROXIDE
FREE
WITH AMIODARONE,
Treatment
RADICAL
PRODUCTION
DESETHYLAMIODARONE
IN RAT
LIVER
Superoxide
production
(nmol:minimg DMSO 0.2% (control)
0.96+0.05”
Amiodarone
0.96+0.07
0.4 fig/ml
0.72+0.06*
4 /lg/rnl
0.57&0.03’
40 pg/mI Desethylamiodarone
0.25 pg/mI
0.97 f 0.02
2.5 pg/ml
0.81 +0.02* 0*
25 pg/ml No treatment kiK%qUat
0.05
(control) mM
0.5 mM 5mM * Mean + SD; n = 3 trials per treatment *PC 0.05 vs. control group.
MICROSOMES
OR PARAQUAT
1.04+0.06 3.38*0.21* 10.29+0.27* 24.95 + 1.07*
protein)
123 Cytotoxicity 0
DMSO
m
AD
Lipid Peroxidation FZi
DE
AD
ffl
DE
100 90
90
0
c?
80
80
;
5
70 60
7o ‘2 60 h
100
50
50
z
40
40
& 0 7
30
30
E
20
20
2;
kl E
10
10
E
0 4
8
Fig. 6. Relationship
of amiodarone
rat hepatocytes
4 Duration
8
24
(hr)
(AD; 40 pg/ml) and desethylamiodarone
release) and lipid peroxidation
OJ co
0
24 Treatment
cultured
m
: P
a
(LDH
DMSO
-0
0 2
toxicity
0
(DE; 25 fig/ml) induced
(TRS; thiobarbituric-acid-reactive
after 4, 8, and 24 h of treatment.
n = 3 cultures
substances)
cyto-
in primary
per point; mean k SD. *P < 0.05
vs.DMSO (0.2%) solvent control group.
other oxygen radicals [ 161. Paraquat stimulated microsomal superoxide production in a concentration-dependent manner (Table I). AD and DE also inhibited hepatocyte lipid peroxidation at sampling times prior to the development of 100% hepatocyte death (i.e. 100% LDH release) (Fig. 6). After 24 h of treatment with AD or DE, however, LDH release was approximately 100% and TRS were significantly elevated in both groups. Thus, elevated lipid peroxidation products (TRS) in AD- and DEtreated cultures were not present until 100% hepatocyte toxicity had occurred. Prior to this stage, both compounds
inhibited
lipid peroxidation.
DISCUSSION
The mechanism
of AD toxicity
is unknown.
AD has been shown to inhibit
lysoso-
ma1 phospholipases [ 17-201 and to induce osmophillic granules and myeloid inclusions in cells [2&22]. AD is also capable of disrupting membrane phospholipid fluidity and protein function due to its cationic amphiphilic nature [23]. AD is metabolized to desethylamiodarone (DE) by cytochrome P-450 mixed-function oxidases [24,25] and can inactivate these enzymes following its metabolism [26,27]. AD has also been recently shown to be a potent inhibitor of calmodulin activity [28]. Free radicals have also been invoked in the mechanism of AD toxicity. This is especially true for the phototoxicity that can result with AD therapy [3]. Hasan et al. [4] have demonstrated that AD and DE can kill UV-light-exposed red blood cells and lymphocytes by a phototoxic mechanism that is inhibitable by antioxidants.
124
Other
investigators
have shown
that oxygen
radicals
and organic
radicals
are gen-
erated by the photolysis of AD during exposure to UV light [5,6]. In the absence of UV light exposure, however, it is unclear whether AD can induce toxicity through a free radical mechanism. Kennedy et al. [7] have demonstrated that AD-induced toxicity in perfused rabbit lungs (evidenced by increased pulmonary arterial pressure and edema) was decreased by pretreatment of the rabbits with butylated hydroxyanisole or VE or by the addition of NAC to the perfusate. They also showed that AD stimulated superoxide production, lipid peroxidation, and glutathione oxidation in the lung preparations. Lung toxicity was not dependent on ultraviolet light exposure, so that free radicals could not have been generated through photolysis. In the present study, VE prevented the hepatocyte toxicity of AD and DE in the absence of UV light exposure, suggesting that these compounds kill hepatocytes through free radical generation and lipid peroxidation. However, other results in this study argue against a free radical/lipid peroxidation mechanism of toxicity. First, several other antioxidants of diverse nature were unable to prevent AD and DE toxicity. Secondly, no stimulation of superoxide production in AD- and DE-treated liver microsomes was seen, but instead these compounds inhibited production of the radical. Microsomal superoxide production can be attributed to cytochrome P-450 mixed-function oxidase reduction of dioxygen. Since AD and DE are capable of inhibiting cytochrome P-450 activity [26,27], the inhibition of superoxide generation by these compounds may have been due to the loss of cytochrome P-450 activity. Thirdly, no evidence that lipid peroxidation contributed to the hepatotoxicity of AD or DE was observed. At exposure durations resulting in submaximal cell killing, both compounds inhibited hepatocyte lipid peroxidation. Only when 100% cell death had occurred was there an elevation in lipid peroxidation. This increase in lipid peroxidation, occurring after AD- and DE-induced hepatocyte death, may have been due to the auto-oxidation of dead cells. Therefore, although these data do not conclusively rule out a free radical mechanism of AD and DE hepatoxicity, the majority of these studies do not support such a mechanism. If AD and DE do not kill hepatocytes through free radical generation/ lipid peroxidation, how could VE have prevented the toxicity of these compounds‘? In addition to its antioxidant capacity, VE has been shown to alter drug uptake and metabolism [29], membrane fluidity [30,31], and phospholipase activity [32]. These other effects have all been implicated in the mechanism of AD and DE toxicity [17-231 and it is possible that VE prevented hepatotoxicity through these non-antioxidant actions. ACKNOWLEDGEMENTS
This work was supported 05700 and S07-RR-05700.
by NIH
Biomedical
Research
Support
Grants
S07-PR-
125
REFERENCES 1 Leak,
D. and Eydt,
Intern.
J.N. (1979) Control
of refractory
cardiac
arrhythmias
with amiodarone.
Arch.
Med. 139,425-428.
2 Heger,
J.J., Prystowsky,
recurrent
ventricular
3 Green, J.L., Graham,
of cardiac
T., Kochevar,
and lymphocytes. 5 Li, A.S.W.
arrhythmias.
I.E. and Abduiah,
Photochem.
and Chignell,
A spin trapping
and ventricular
fibrillation.
efficacy of amiodarone
J. Am. Coil. Cardioi.
D. (1984) Amiodarone
Photobiol.
of
2, lI14-I
effects of amioda-
128.
phototoxicity
to human
erythrocytes
40,715-719.
C.F. (1987) Spectroscopic
study of the photolysis
in treatment
Am. Heart J. 106, 887-894.
E.L., Muller, U. and Ladin, H.P. (1979) Toxic and therapeutic
rone in the treatment 4 Hasan,
E.N. and Zipes. D.P. (1983) Clinical
tachycardia
studies of cutaneous
of amiodarone
photosensitizing
and desethylamiodarone.
agents.
Photochem.
IX.
Photo-
biol. 45, 191-197. 6 Paillous,
N. and Vernier,
Photobiol.
M. (1988) Photolysis
7 Kennedy,
T.P., Gordon,
G.B., Paky, A., McShane,
man, W., Sciuto, A.M. and Gurtner, tilated and perfused 8 Ruth.
rabbit
paraquat-generated 9 Klaunig,
J.E., Goldblatt,
IO Klaunig,
P.J., Hinton,
1. Hepatocyte
J.E., Goldblatt, II. Primary hepatocyte
Caldwell, NJ. 13 Bandyopadhyay, amiodarone
toxicity.
12,23336. intercellular
communication
by
94,427-436.
M.M., Chacko,
J. and Trump,
B.F. (1981) Mouse
In Vitro 17,913.-925. D.E., Lipsky,
M.M. and Trump,
J.E. (1989) Effects of culture
Toxicol.
S., Klaunig,
(DA) induced
K., Jack-
lung injury in ven-
B.F. (1981) Mouse
liver cell
In Vitro 17.926934. Appl. Pharmacol.
12 Gad, S. and Weil, C.S. (1986) Statistics
14 Gross,
Photochem.
N.F., Peters, S.P., Friday,
Appl. Pharmacol.
D.E., Lipsky,
I I Ruth, R.J., Crist. K.A. and Klaunig,
(Suppl.)
drug.
causes acute oxidant
of mouse hepatocyte
Toxicol.
isolation.
P.J., Hinton,
culture.
A., Adkinson,
Pharmacol.
J.E. (1988) Inhibition
oxygen free radicals.
liver ceil culture.
induced
an antiarrhythmic
G.H. (1988) Amiodarone
lungs. J. Cardiovasc.
R.J. and Klaunig,
culture.
of amiodarone,
47, 337-343.
and Experimental
inclusion
on hydrogen
Design for Toxicologists.
J.E., Gross, S.A. and Somani. multilamellar
duration
peroxide-
100, 451464. The Telford
P. (1987) Amiodarone
bodies
in rat hepatocytes
Press,
(AM) and desethyl-
in culture.
Circulation
76. 212. S., Bandyopadhyay.
rone toxicity
I.5 Freeman.
in isolated
S., Klaunig, hepatocytes
B.A. and Crapo,
J.E. and Somani,
in culture.
P. (1989) Amiodarone
and desethylamioda-
Prof. Sot. Exp. Bioi. Med. (NY) 190, 163-169.
J.D. (1982) Biology of disease: free radicals
and tissue injury.
Lab. Invest.
47.412426. 16 Bus, J. and Gibson,
J. (1984) Paraquat:
model for oxidant
initiated
toxicity,
Environ.
Health
Perspect.
55, 3746. 17 Downar,
E., Shaikh,
Am. Coll. Cardiol. 18 Hostetler,
N. and Butany,
K.Y., Reasor,
lipase A inhibition 19 Hostetler.
K.Y.,
M.J., Walker,
in amiodarone Giordano,
II, W.J.. Kachel,
in amiodarone
pulmonary
E.R., Yazaki,
pulmonary
T., Simon,
city. Ultrastruct. 23 Maher,
G.T.,
Pathol.
molecules.
Biochim.
a potent
phospholipase
inhibitor.
J.
J. Pharmacol.
Biochim.
Acta 875,40@405. of lysosomal
Biophys.
phospho-
Acta 959, 316321.
V. (1989) Mechanism
of phospholipidosis
Exp. Ther. 251,272-278.
H.L. and Josephson,
Kay, J.M., Pugsley,
B.W. (1986) Role of phospho-
Biophys.
E.J. (1988) In vitro inhibition
D.L., Vilen, T. and Natarajan, toxicity.
M.E. (1982) Amiodarone
S.O. and Eydt, J. (1984) Amiodarone
pulmonary
pulmonary
toxi-
6, 1999207.
P. and Singer, S.J. (1984) Structural
amphipathic
-
P.J. and Frazere,
and desethylamiodarone.
21 Marchiinski, F.E., Gansler. T.S., Waxman, toxicity. Ann. Intern. Med. 97, 839-845. 22 Colgan,
toxicity.
J.R. and Jellison,
lipase A, of rat lung by amiodarone 20 Martin,
J. (1984) Amiodarone
3.604.
Biochemistry
changes
23,232--240.
in membranes
produced
by the binding
of small
126
24 Young, R.A. and Mehendaie, H.M. (1986) In vitro metabolism of amiodarone by rabbit and rat liver and small intestine. Drug Metab. Disposit. 14.423429. 25 Young, R.A. and Mehendale, H.M. (1987) Effect ofcytochrome P450 and flavin-containing monooxygenase modifying factors on the in vitro metabolism of amiodarone by rat and rabbit. Drug Metab. Disposit. IS, 511-517. 26 Grech-Belanger, 0. (1984) Depressive effect of amiodarone on hepatic drug metabolism in the rat. Res. Commun. Chem. Pathol. Pharmacol. 44, i S-30. 27 Larrey, D.. Tinel, M., Letteron, P., Geneve, J., Descatoire, V. and Pessayre, D. (1986) Formation of inactive cytochrome P-450Fe(II)metabolite complex after administration of amiodarone in rats, mice and hamsters. Photochem. Photobiol. 35. 221332220. 28 Nokin, P., Blondiaux, J.P., Schaeffer, P., Jungbluth, L. and Lugnier, C. (1989) Amiodarone is a potent calmodulin antagonist. Naunyn-Schmiedeberg’s Arch. Pharmacol. 339,367-373. 29 Statham, C.N., Minchin, R.F., Sasame, H.A., Kim, S.H. and Boyd, M.R. (1985) Effects of vitamin E on the distribution and metabolism of nitrofu~ntoin in rats. Drug Metab. Disposit. 13, 532 534. 30 Lucy, J.A. (1972) Functional and structural aspects of biological membranes: a suggested structural role for vitamin E in the control of membrane permeability and stability. Ann. NY Acad. Sci. 203, J--II. 3 I Urano, S., Yano, K. and Matsuo. M. (1988) Membrane-stabilizing effect of vitamin E: effect of alphatocopherol and its model compounds on fluidity of lecithin liposomes. Biochem. Biophys. Res. Commun. 150,46Y-475. 32 Tran, K. and C’han. A.C. (1988) Effect of vitamin E enrichment on arachidonic acid release and cellular phospholipids in cultured human endothelial cells. Biochim. Biophys. Acta 963,468 475.
117
ADONlS037842749100056X
TOXLET 02539
Evaluation of amiodarone free radical toxicity in rat hepatocytes
Randall J. Ruchl, Subhankar Bandyopadhyayz, and James E. Klaunigl”
Pitambar Somani2
Departments of 1Pathology and 2Pharmaco/ogy, Medical College of Ohio, Toledo, OH (U.S.A. j
(Received 19 January 1990) (Accepted 31 October 1990) Ke,y words: Amiodarone; Free radicals; Hepatocytes _-._ SUMMARY The possible rofes of free radicals and lipid peroxidation in the m~hanism of toxicity of amiodarone (AD) [2-butyl-3-(3’,S-diiodo-4’tr-diethylaminoethoxybenzoyl)~nzofuran] and its principle metabolite, desethylamiodarone (DE), were examined in primary cultured Sprague-Dawley male rat hepatocytes. AD (20 and 40 ,ug/ml) and DE (10 and 25 pg/ml) killed hepatocytes in concentration- and time-dependent fashions, Several antioxidants [Cu,Zn-superoxide dismutase (200 U/ml), catalase (200 U/ml), N,hr-diphenylphenylenediamine (DPPD; 25 PM), butylated hydroxytoluene (0.1 mM), and N-acetylcysteine (5 mM)] were incapable of preventing AD and DE hepatocyte toxicity. Only vitamin E (VE, d,l-a-tocopherol acetate; 20-200 /IM) prevented AD and DE toxicity. No correlation between the onset of hepatocyte death by AD and DE and hepatocyte lipid peroxidation was seen. Both drugs inhibited NADPH-dependent rat liver microsomal superoxide production. These results, excluding the preventive effects of VE, do not support a free radical/lipid peroxidation mechanism of hepatocyte toxicity by AD and DE. VE may have prevented hepatocyte toxicity through non-antioxidant effects.
INTRODUCTION
Amiodarone (AD) [2-butyl-3-(3’,5’-diiodo-4’cc-diethylaminoethoxybenzoyl)benzofuran] is a class III antiarrhythmic drug that is effective in the treatment of resistant ventricular and supraventricular arrhythmia [ 1,2]. Several side-effects of AD therapy have been reported which include cornea1 deposits, altered thyroid function, muscle weakness and paresthesias, phototoxicity, pulmonary fibrosis, and hepatotoxicity [3].
Addressf#r eorre~po~dence; James E. Klaunig, Ph.D., Department of Pathology, Medical College of Ohio,
3000 Arlington Ave., Toledo, OH 43614, U.S.A.
118
The cellular
mechanism
of AD toxicity
is unknown.
Free radicals
are generated
upon exposure of AD and DE to ultraviolet light and this effect has been shown to be important in the phototoxicity of these drugs [4-61. Kennedy et al [7] have recently proposed that AD also induces lung toxicity through free radical generation. They showed that in perfused rabbit lungs in the absence of ultraviolet light exposure, the toxicity of AD was inhibited by the antioxidants, vitamin E, butylated hydroxyanisole, and N-acetylcysteine, but not by superoxide dismutase or catalase. They also demonstrated that lung chemiluminescence, superoxide production, and tissue levels of glutathione disulfide were increased by AD. Since another important side-effect of AD therapy is hepatotoxicity [3], we have examined whether AD hepatotoxicity can also be attributed to free radical generation. MATERIALS
AND METHODS
Animals Male Sprague-Dawley rats (1 O&l 50 g) were purchased from Charles River Laboratories (Wilmington, MA), housed in polycarbonate cages, and fed Purina Rodent Chow (Ralston Purina Co., St. Louis, MO) and water ad libitum. Chemicals Amiodarone
(AD) and desethylamiodarone
(DE) were generous
gifts of Sanofi
Labaz Company (Montpellier, France). Copper, zinc-superoxide dismutase (SOD), catalase (CAT), N&V’-diphenyl-p-phenylenediamine (DPPD), vitamin E (d&cctocopherol acetate) (VE), N-acetylcysteine (NAC), butylated hydroxytoluene (BHT) and paraquat were obtained from Sigma Chemical Co. (St. Louis, MO). All other reagents and tissue culture supplies were purchased from sources as previously indicated [S]. Hepatocyte isolation andprimary culture Rat hepatocytes were isolated by two-stage collagenase perfusion of the liver [9]. Hepatocyte viability after isolation was determined by trypan blue dye exclusion and was greater than 90%. Hepatocytes were plated onto 60 mm plastic tissue culture dishes at one million viable cells per dish in 3 ml medium. Hepatocyte culture medium consisted of Leibovitz’s L-15 medium supplemented with 1 mg/ml a-D-glucose, 10% fetal bovine serum, 1 ,uM dexamethasone and 50 pg/ml gentamicin sulfate [lo]. The cultures were incubated in a humidified 100% air incubator at 37°C. After 2 h culture to allow for cell attachment, the cultures were refed with 5 ml supplemented L- 15 medium per dish and used for experiments. Hepatocyte killing by AD and DE and prevention by antioxidants After attachment and refeeding, the 2-h-old hepatocyte cultures were treated with AD (20 or 40 pg/ml) or DE (10 or 25 pg/ml) dissolved in DMSO. Control cultures
119
were treated with DMSO only (2 $/ml). The cells were incubated with the compounds in the dark in the tissue culture incubator for 2,4, 8 and 24 h. Aliquots (0.1 ml) of the culture media were then removed and analyzed for lactate dehydrogenase (LDH) activity as an indication of hepatocyte killing. LDH activity was determined spectrophotometrically as we have described [8,1 I]. Hepatocyte toxicity (LDH release) was expressed as a percentage of the total LDH activity in the cells determined in non-treated cultures lysed with 0.01% Triton X-100. We also evaluated the effects of several antioxidants on AD- and DE-induced hepatocyte toxicity. Hepatocyte cultures were first treated with one of several antioxidants [SOD (200 U/ml), CAT (200 U/ml), DPPD (25 PM), VE (20, 100 or 200 PM), NAC (5 mM) or BHT (0.1 mM)], then treated immediately thereafter with AD or DE. LDH release was analyzed after 2,4,8 and 24 h of exposure to the compounds. Hepatacyte iipidperoxidation
Lipid peroxidation products in primary cultured rat hepatocytes treated with AD or DE were determined as thiobarbituric-acid-reactive substances (TRS) as we have previously described [I 11.Cell killing was also determined in these cultures by analyzing LDH release into the culture medium. Microsomal superoxide production
Superoxide radical production in rat liver microsomes treated with AD or DE was determined by the cytochrome c reduction assay as we have described [8]. As a positive control, we also monitored superoxide generation in paraquat-treated microsomes. Statistics
Statistical differences in LDH release, lipid peroxidation, and superoxide production between control and treated hepatocyte cultures or microsomes were compared by analysis of variance and Dunnett’s post-hoc t-test [ 121. RESULTS
AD (20 or 40 lug/ml) and DE (10 or 25 pg/ml) were toxic to rat hepatocytes in dose- and time-dependent manners (Fig. I). DE was more toxic than AD as we have previously noted [13,14]. To evaluate whether AD and DE induced hepatocyte toxicity through free radical generation, hepatocyte cultures were treated with AD (40 pg/ml) or DE (25 @g/ml) in the presence of one of several antioxidants: VE (200 PM), DPPD (25 ,uM), SOD (200 U/ml), CAT (200 U/ml), NAC (0.5 mM), and BHT (0.1 mM). Of these 6 antioxidants, only VE decreased the toxicity of AD and DE (Figs. 2 and 3). DPPD enhanced AD and DE toxicity (Figs. 2 and 3) and the other antioxidants had no effect (data not shown). The preventive effects of VE were dependent on VE concentration
120 o-0
Q
DMSO 0.2%
A--A
AD 20 ug/ml
n
A--A
AD 40 ug/ml
0-O
4
12
8 Treatment
1. Concentration
toxicity
(LDH
per datum
DE 10 ug/ml DE 25 ug/ml
loo-
0
Fig.
-m
and temporal
release) in primary
relationships cultured
16
Duration
of amiodarone
rat hepatocytes
point; mean k SD. *PC 0.05 vs. the DMSO
20
24
(hr)
(AD) and desethylamiodarone
over 24-h treatment
solvent (0.2%) control
duration.
group
(DE) cyn = 3 cultures
at the same sampling
time.
and were different for AD versus DE (Figs. 4 and 5). VE at 100 and 200 ,BM produced nearly complete protection from AD-induced toxicity over the 24-h treatment period (Fig. 4), while VE at 100 and 200 PM only partially reduced the toxicity of DE (Fig. 5). None of the antioxidants by themselves affected hepatocyte LDH release (data not shown). 8-8
DMSO
B-8
AD + DPPD
A-A
AD alone
l
AD + VE
Treatment
Fig. 2. Effects of NJ’-diphenylphenylenediamine cytotoxicity treatment
(LDH duration.
release) of amiodarone n=3
--•*
Duration
(DPPD; (AD; 40 pg/ml)
(hr)
25 pM) and vitamin in primary
cultured
E (VE; 200 PM) on the rat hepatocytes
cultures per point; mean&SD. *P
over 24-h
The toxicity
of
121
u
loo-
8
go-
,$
80-
g -1 io z t%
70-
g
30-
0
o-o
DMSO
D-•
DE + DPPD
A-A
DE alone
+--0
DE + VE
B f 1
*
i
6050*
40-
*
/
r.
0
4
8
12
Treatment
16
Duration
20
24
(hr)
Fig. 3. Effects of NJ’-diphenylphenylenediamine (DPPD; 25 FM) and vitamin E (VE; 200 PM) on the cytotoxicity (LDH release) of desethylamiodarone (DE; 25 pg/ml) in primary cultured rat hepatocytes over 24-h treatment duration. n = 3 cultures per point; mean + SD. *PC 0.05 vs. DE alone group. The toxicity of the DE solvent, DMSO (0.2%), is also shown.
The ability of VE to prevent AD and DE hepatotoxicity suggested that these drugs kill hepatocytes by a mechanism involving free radical generation and lipid peroxidation. This is because VE is the major membrane-soluble antioxidant and inhibitor of lipid peroxidation in cells [15]. Therefore, we tested whether AD and DE coufd stimulate microsomal superoxide radical production and hepatocyte lipid peroxidaO-O
0
DMSO
A--A
AD alone
0-O
AD + VE 100 uM
n
AD + VE 20 uM
l
AD + VE 200
-m
4
8 Treatment
12 Duration
---+
16
20
uM
24
(hr)
Fig. 4. Dose-responsive protective effects of vitamin E (VE) on amiodarone (AD; 40 pg/ml) cytotoxicity (LDH release) in primary cultured rat hepatocytes over 24-h treatment duration. n=3 cultures per point; mean + SD. *P < 0.05vs.AD-only group. Toxicity in DMSO (0.2%) solvent control cultures is also shown.
122 DMSO
0-o
m
A-A
DE alone
H--B
DE + VE 20
uM
4
12
8 Treatment
Fig. 5. Dose-responsive (LDH
per point;
DE+VElOOuM DE + VE 200
uM
loo-
0
toxicity
O-0 +-+
protective
release) in primary
*P
mean*SD.
Duration
effects of vitamin cultured
16
20
24
(hr)
E (VE) on desethylamiodarone
rat hepatocytes group.
(DE; 25 pg/mI) cyto-
over 24-h treatment
Toxicity
in DMSO
duration.
(0.2%) solvent
n = 3 cultures control
cultures
is also shown.
tion at cytotoxic doses. Surprisingly, both compounds inhibited microsomal superoxide generation in a dose-dependent fashion. As a positive control for microsomal superoxide production, paraquat was utilized. Paraquat has been shown to undergo redox-cycling via NADPH cytochrome P-450 reductase to generate superoxide and TABLE
I
RATES
OF
TREATED
SUPEROXIDE
FREE
WITH AMIODARONE,
Treatment
RADICAL
PRODUCTION
DESETHYLAMIODARONE
IN RAT
LIVER
Superoxide
production
(nmol:minimg DMSO 0.2% (control)
0.96+0.05”
Amiodarone
0.96+0.07
0.4 fig/ml
0.72+0.06*
4 /lg/rnl
0.57&0.03’
40 pg/mI Desethylamiodarone
0.25 pg/mI
0.97 f 0.02
2.5 pg/ml
0.81 +0.02* 0*
25 pg/ml No treatment kiK%qUat
0.05
(control) mM
0.5 mM 5mM * Mean + SD; n = 3 trials per treatment *PC 0.05 vs. control group.
MICROSOMES
OR PARAQUAT
1.04+0.06 3.38*0.21* 10.29+0.27* 24.95 + 1.07*
protein)
123 Cytotoxicity 0
DMSO
m
AD
Lipid Peroxidation FZi
DE
AD
ffl
DE
100 90
90
0
c?
80
80
;
5
70 60
7o ‘2 60 h
100
50
50
z
40
40
& 0 7
30
30
E
20
20
2;
kl E
10
10
E
0 4
8
Fig. 6. Relationship
of amiodarone
rat hepatocytes
4 Duration
8
24
(hr)
(AD; 40 pg/ml) and desethylamiodarone
release) and lipid peroxidation
OJ co
0
24 Treatment
cultured
m
: P
a
(LDH
DMSO
-0
0 2
toxicity
0
(DE; 25 fig/ml) induced
(TRS; thiobarbituric-acid-reactive
after 4, 8, and 24 h of treatment.
n = 3 cultures
substances)
cyto-
in primary
per point; mean k SD. *P < 0.05
vs.DMSO (0.2%) solvent control group.
other oxygen radicals [ 161. Paraquat stimulated microsomal superoxide production in a concentration-dependent manner (Table I). AD and DE also inhibited hepatocyte lipid peroxidation at sampling times prior to the development of 100% hepatocyte death (i.e. 100% LDH release) (Fig. 6). After 24 h of treatment with AD or DE, however, LDH release was approximately 100% and TRS were significantly elevated in both groups. Thus, elevated lipid peroxidation products (TRS) in AD- and DEtreated cultures were not present until 100% hepatocyte toxicity had occurred. Prior to this stage, both compounds
inhibited
lipid peroxidation.
DISCUSSION
The mechanism
of AD toxicity
is unknown.
AD has been shown to inhibit
lysoso-
ma1 phospholipases [ 17-201 and to induce osmophillic granules and myeloid inclusions in cells [2&22]. AD is also capable of disrupting membrane phospholipid fluidity and protein function due to its cationic amphiphilic nature [23]. AD is metabolized to desethylamiodarone (DE) by cytochrome P-450 mixed-function oxidases [24,25] and can inactivate these enzymes following its metabolism [26,27]. AD has also been recently shown to be a potent inhibitor of calmodulin activity [28]. Free radicals have also been invoked in the mechanism of AD toxicity. This is especially true for the phototoxicity that can result with AD therapy [3]. Hasan et al. [4] have demonstrated that AD and DE can kill UV-light-exposed red blood cells and lymphocytes by a phototoxic mechanism that is inhibitable by antioxidants.
124
Other
investigators
have shown
that oxygen
radicals
and organic
radicals
are gen-
erated by the photolysis of AD during exposure to UV light [5,6]. In the absence of UV light exposure, however, it is unclear whether AD can induce toxicity through a free radical mechanism. Kennedy et al. [7] have demonstrated that AD-induced toxicity in perfused rabbit lungs (evidenced by increased pulmonary arterial pressure and edema) was decreased by pretreatment of the rabbits with butylated hydroxyanisole or VE or by the addition of NAC to the perfusate. They also showed that AD stimulated superoxide production, lipid peroxidation, and glutathione oxidation in the lung preparations. Lung toxicity was not dependent on ultraviolet light exposure, so that free radicals could not have been generated through photolysis. In the present study, VE prevented the hepatocyte toxicity of AD and DE in the absence of UV light exposure, suggesting that these compounds kill hepatocytes through free radical generation and lipid peroxidation. However, other results in this study argue against a free radical/lipid peroxidation mechanism of toxicity. First, several other antioxidants of diverse nature were unable to prevent AD and DE toxicity. Secondly, no stimulation of superoxide production in AD- and DE-treated liver microsomes was seen, but instead these compounds inhibited production of the radical. Microsomal superoxide production can be attributed to cytochrome P-450 mixed-function oxidase reduction of dioxygen. Since AD and DE are capable of inhibiting cytochrome P-450 activity [26,27], the inhibition of superoxide generation by these compounds may have been due to the loss of cytochrome P-450 activity. Thirdly, no evidence that lipid peroxidation contributed to the hepatotoxicity of AD or DE was observed. At exposure durations resulting in submaximal cell killing, both compounds inhibited hepatocyte lipid peroxidation. Only when 100% cell death had occurred was there an elevation in lipid peroxidation. This increase in lipid peroxidation, occurring after AD- and DE-induced hepatocyte death, may have been due to the auto-oxidation of dead cells. Therefore, although these data do not conclusively rule out a free radical mechanism of AD and DE hepatoxicity, the majority of these studies do not support such a mechanism. If AD and DE do not kill hepatocytes through free radical generation/ lipid peroxidation, how could VE have prevented the toxicity of these compounds‘? In addition to its antioxidant capacity, VE has been shown to alter drug uptake and metabolism [29], membrane fluidity [30,31], and phospholipase activity [32]. These other effects have all been implicated in the mechanism of AD and DE toxicity [17-231 and it is possible that VE prevented hepatotoxicity through these non-antioxidant actions. ACKNOWLEDGEMENTS
This work was supported 05700 and S07-RR-05700.
by NIH
Biomedical
Research
Support
Grants
S07-PR-
125
REFERENCES 1 Leak,
D. and Eydt,
Intern.
J.N. (1979) Control
of refractory
cardiac
arrhythmias
with amiodarone.
Arch.
Med. 139,425-428.
2 Heger,
J.J., Prystowsky,
recurrent
ventricular
3 Green, J.L., Graham,
of cardiac
T., Kochevar,
and lymphocytes. 5 Li, A.S.W.
arrhythmias.
I.E. and Abduiah,
Photochem.
and Chignell,
A spin trapping
and ventricular
fibrillation.
efficacy of amiodarone
J. Am. Coil. Cardioi.
D. (1984) Amiodarone
Photobiol.
of
2, lI14-I
effects of amioda-
128.
phototoxicity
to human
erythrocytes
40,715-719.
C.F. (1987) Spectroscopic
study of the photolysis
in treatment
Am. Heart J. 106, 887-894.
E.L., Muller, U. and Ladin, H.P. (1979) Toxic and therapeutic
rone in the treatment 4 Hasan,
E.N. and Zipes. D.P. (1983) Clinical
tachycardia
studies of cutaneous
of amiodarone
photosensitizing
and desethylamiodarone.
agents.
Photochem.
IX.
Photo-
biol. 45, 191-197. 6 Paillous,
N. and Vernier,
Photobiol.
M. (1988) Photolysis
7 Kennedy,
T.P., Gordon,
G.B., Paky, A., McShane,
man, W., Sciuto, A.M. and Gurtner, tilated and perfused 8 Ruth.
rabbit
paraquat-generated 9 Klaunig,
J.E., Goldblatt,
IO Klaunig,
P.J., Hinton,
1. Hepatocyte
J.E., Goldblatt, II. Primary hepatocyte
Caldwell, NJ. 13 Bandyopadhyay, amiodarone
toxicity.
12,23336. intercellular
communication
by
94,427-436.
M.M., Chacko,
J. and Trump,
B.F. (1981) Mouse
In Vitro 17,913.-925. D.E., Lipsky,
M.M. and Trump,
J.E. (1989) Effects of culture
Toxicol.
S., Klaunig,
(DA) induced
K., Jack-
lung injury in ven-
B.F. (1981) Mouse
liver cell
In Vitro 17.926934. Appl. Pharmacol.
12 Gad, S. and Weil, C.S. (1986) Statistics
14 Gross,
Photochem.
N.F., Peters, S.P., Friday,
Appl. Pharmacol.
D.E., Lipsky,
I I Ruth, R.J., Crist. K.A. and Klaunig,
(Suppl.)
drug.
causes acute oxidant
of mouse hepatocyte
Toxicol.
isolation.
P.J., Hinton,
culture.
A., Adkinson,
Pharmacol.
J.E. (1988) Inhibition
oxygen free radicals.
liver ceil culture.
induced
an antiarrhythmic
G.H. (1988) Amiodarone
lungs. J. Cardiovasc.
R.J. and Klaunig,
culture.
of amiodarone,
47, 337-343.
and Experimental
inclusion
on hydrogen
Design for Toxicologists.
J.E., Gross, S.A. and Somani. multilamellar
duration
peroxide-
100, 451464. The Telford
P. (1987) Amiodarone
bodies
in rat hepatocytes
Press,
(AM) and desethyl-
in culture.
Circulation
76. 212. S., Bandyopadhyay.
rone toxicity
I.5 Freeman.
in isolated
S., Klaunig, hepatocytes
B.A. and Crapo,
J.E. and Somani,
in culture.
P. (1989) Amiodarone
and desethylamioda-
Prof. Sot. Exp. Bioi. Med. (NY) 190, 163-169.
J.D. (1982) Biology of disease: free radicals
and tissue injury.
Lab. Invest.
47.412426. 16 Bus, J. and Gibson,
J. (1984) Paraquat:
model for oxidant
initiated
toxicity,
Environ.
Health
Perspect.
55, 3746. 17 Downar,
E., Shaikh,
Am. Coll. Cardiol. 18 Hostetler,
N. and Butany,
K.Y., Reasor,
lipase A inhibition 19 Hostetler.
K.Y.,
M.J., Walker,
in amiodarone Giordano,
II, W.J.. Kachel,
in amiodarone
pulmonary
E.R., Yazaki,
pulmonary
T., Simon,
city. Ultrastruct. 23 Maher,
G.T.,
Pathol.
molecules.
Biochim.
a potent
phospholipase
inhibitor.
J.
J. Pharmacol.
Biochim.
Acta 875,40@405. of lysosomal
Biophys.
phospho-
Acta 959, 316321.
V. (1989) Mechanism
of phospholipidosis
Exp. Ther. 251,272-278.
H.L. and Josephson,
Kay, J.M., Pugsley,
B.W. (1986) Role of phospho-
Biophys.
E.J. (1988) In vitro inhibition
D.L., Vilen, T. and Natarajan, toxicity.
M.E. (1982) Amiodarone
S.O. and Eydt, J. (1984) Amiodarone
pulmonary
pulmonary
toxi-
6, 1999207.
P. and Singer, S.J. (1984) Structural
amphipathic
-
P.J. and Frazere,
and desethylamiodarone.
21 Marchiinski, F.E., Gansler. T.S., Waxman, toxicity. Ann. Intern. Med. 97, 839-845. 22 Colgan,
toxicity.
J.R. and Jellison,
lipase A, of rat lung by amiodarone 20 Martin,
J. (1984) Amiodarone
3.604.
Biochemistry
changes
23,232--240.
in membranes
produced
by the binding
of small
126
24 Young, R.A. and Mehendaie, H.M. (1986) In vitro metabolism of amiodarone by rabbit and rat liver and small intestine. Drug Metab. Disposit. 14.423429. 25 Young, R.A. and Mehendale, H.M. (1987) Effect ofcytochrome P450 and flavin-containing monooxygenase modifying factors on the in vitro metabolism of amiodarone by rat and rabbit. Drug Metab. Disposit. IS, 511-517. 26 Grech-Belanger, 0. (1984) Depressive effect of amiodarone on hepatic drug metabolism in the rat. Res. Commun. Chem. Pathol. Pharmacol. 44, i S-30. 27 Larrey, D.. Tinel, M., Letteron, P., Geneve, J., Descatoire, V. and Pessayre, D. (1986) Formation of inactive cytochrome P-450Fe(II)metabolite complex after administration of amiodarone in rats, mice and hamsters. Photochem. Photobiol. 35. 221332220. 28 Nokin, P., Blondiaux, J.P., Schaeffer, P., Jungbluth, L. and Lugnier, C. (1989) Amiodarone is a potent calmodulin antagonist. Naunyn-Schmiedeberg’s Arch. Pharmacol. 339,367-373. 29 Statham, C.N., Minchin, R.F., Sasame, H.A., Kim, S.H. and Boyd, M.R. (1985) Effects of vitamin E on the distribution and metabolism of nitrofu~ntoin in rats. Drug Metab. Disposit. 13, 532 534. 30 Lucy, J.A. (1972) Functional and structural aspects of biological membranes: a suggested structural role for vitamin E in the control of membrane permeability and stability. Ann. NY Acad. Sci. 203, J--II. 3 I Urano, S., Yano, K. and Matsuo. M. (1988) Membrane-stabilizing effect of vitamin E: effect of alphatocopherol and its model compounds on fluidity of lecithin liposomes. Biochem. Biophys. Res. Commun. 150,46Y-475. 32 Tran, K. and C’han. A.C. (1988) Effect of vitamin E enrichment on arachidonic acid release and cellular phospholipids in cultured human endothelial cells. Biochim. Biophys. Acta 963,468 475.