Effect of lipid physical state on the rate of peroxidation of liposomes

Free Radical Biology & Medicine, Vol. 12, pp. 113-119, 1992 Printed in the USA. All fights reserved.

0891-5849/92 $5.00 + .00 Copyright © 1992 Pergamon Press plc

Original Contribution EFFECT OF LIPID PHYSICAL STATE ON THE RATE OF PEROXIDATION OF LIPOSOMES LARRY R . M C L E A N * and KAREN A. HAGAMAN Marion Merrell Dow Research Institute, 2110 East Galbraith Road, Cincinnati, OH 45215, U.S.A. (Received 1 May 1991; Revised 6 June 1991; Accepted 8 October 1991 ) Abstract--The effect of cholesterol on the rate ofperoxidation ofarachidonic acid and 1-palmitoyl-2-arachidonoyl phosphatidylcholine (PAPC) in dimyristoylphosphatidylcholine (DMPC) liposomes was examined above and below the phase transition temperature (Tin) of the lipid. The rate of peroxidation of arachidonic acid was more rapid below the phase transition temperature of the host lipid. At a temperature below the Tm (4°C), increasing concentrations of cholesterol reduced the rate of peroxidation of arachidonic acid as judged by the production of thiobarbituric acid reactive substances. Above Tm (37°C), cholesterol increased the rate of peroxidation of the fatty acid. Similarly, PAPC was peroxidized more rapidly at 4°C than at 37°C. However, cholesterol had little effect on the rate of peroxidation of PAPC at 4°C. The rate of peroxidation of arachidonic acid was related to the lipid bilayer fluidity as judged by fluorescence anisotropy measurements of diphenylhexatriene. The rate of peroxidation increased slowly with increasing rigidity of the probe environment when the bilayer was relatively fluid and more rapidly as the environment became more rigid. The increase in the rate ofperoxidation ofarachidonic acid in the less fluid host lipid was unrelated to differences in iron binding or to transfer of arachidonic acid to the aqueous phase. Decreasing the concentration of arachidonic acid in DMPC to <2 mole% dramatically decreased the rate of peroxidation at 4°C, suggesting that formation of clusters of fatty acids at 4°C is required for rapid peroxidation. These data support the hypothesis that an increase in the packing density of the acyl chains of peroxidizable lipids promotes lipid peroxidation in less fluid environments which favor phase separation. Keywords--Fluorescence polarization, Arachidonic acid, Phase transitions, Vesicles, Fluidity, Free radicals

the rate of peroxidation of lipids and, to our knowledge, no data are available on the effects of membrane proteins on lipid peroxidation. In a paper examining the influence of EDTA on lipid peroxidation, Tien and co-workers 1° showed that detergent-dispersed phospholipids were more susceptible to oxidation than were phospholipid liposomes. This suggests that dispersion of the lipid allows better access to the site of radical generation. Phosphatidylethanolamines in 1-palmitoyl-2-arachidonyl phosphatidylcholine (PAPC) liposomes decrease the rate of peroxidation of the PAPC in proportion to the mole fraction ofphosphatidylethanolamine in the liposomes.~t Low concentrations of lysophosphatidylcholine have relatively less effect than phosphatidylethanolamines and sphingomyelin inhibits peroxidation only over a narrow range of concentrations in PAPC.11 Cholesterol has been reported to increase s or decrease t 1,z2the rate ofperoxidation ofbilayer lipids depending on the composition of the liposome. For PAPC/distearoyl phosphatidylcholine liposomes at

INTRODUCTION

The composition and physical properties of cell membranes are subject to alteration in several disease states in which membrane peroxidative damage is thought to be a causative factor.~-5 In addition, peroxidative damage to membranes may be modified by lipophilic drugs which have membrane-stabilizing effects. 6-8 L u c y 9 has also proposed that vitamin E may have important membrane-stabilizing effects which supplement its antioxidant activity. These data suggest that an improved understanding of the relationship between membrane structure and the rate of lipid peroxidation may give some insight into possible new treatment modalities for such diseases and may shed light on molecular mechanisms underlying certain pathological processes. Relatively limited information is available on the effect of lipid environment on

* Author to whom correspondence should be addressed. 113

L.R. McLEAN and K. A. HAGAMAN

114

temperatures below the phase transition temperature of the host lipid or pure PAPC liposomes at 4°C, cholesterol at high concentrations (33-50 mole%) decreases the rate of oxidation of the arachidonyl chains. 12By contrast, in soy PC liposomes, cholesterol concentrations in excess of 1 mole% increase the rate of lipid peroxidation initiated by 6°Co gamma radiation. 8 Corresponding to the increased rate of peroxidation was a decrease in bilayer fluidity based on an observed increase in the order parameter of doxylstearic acid in the bilayer. A relationship between membrane fluidity and the rate of peroxidation of bilayer lipids has also been suggested by Mowri and coworkers 12 who showed that peroxidation of PAPC in dimyristoylphosphatidylcholine (DMPC) liposomes proceeded more rapidly at temperatures below the phase transition temperature (Tm) of the lipid and by Cestaro and c o - w o r k e r s 13'14 who observed a similar increase in the rate of peroxidation of arachidonic acid in dipalmitoylphosphatidylcholine liposomes below Tm which was related to a decrease in the fluidity of the bilayers. These data suggest that lipid bilayers in which membrane probes are relatively less mobile are more prone to peroxidation than are more fluid bilayers. This potential coupling between lipid bilayer fluidity and the rate of peroxidation of lipids suggests that the physical state of the lipids may be related to the effects of membrane composition on peroxidation. It is of considerable interest to examine the general applicability of these physical effects. One complication of experiments employing single lipid species for such experiments is that the physical state of the lipid depends on the extent of lipid peroxidation. ~5,16For this reason it is preferable to study host lipid effects with liposomes doped with minimal quantities of peroxidizable lipids, such as arachidonic acid, which do not dramatically alter the physical state of the host lipid upon peroxidation. Arachidonic acid in DMPC is ideally suited for such studies because it has little effect of its own on bilayer fluidity and phase separation in mixtures of fatty acids and saturated phosphatidylcholine liposomes is fairly well understood. In the present report, the relationship between the physical state of the lipids and the rate of peroxidation was examined by measuring the effects of temperature and cholesterol on the rate of peroxidation of fatty acids and phospholipids in saturated phosphatidylcholine liposomes. MATERIALS A N D M E T H O D S

Phosphatidylcholines were purchased from Avanti Polar Lipids (Birmingham, AL), cholesterol from

Calbiochem (San Diego, CA), arachidonic acid from NuChek Prep (Elysian, MN) and fluorescent probes from Molecular Probes (Eugene, OR). Liposomes were prepared by swelling dry lipids in 50-mM NaC1, 50-mM Tris-HC1, pH 7.0 above their phase transition temperature.~7 For some samples, the liposomes were sonicated under argon for 20 min with a microtip in a Branson 350 sonifier operating at setting 7 with a 50% duty cycle. Fluorescence anisotropy measurements with 1,6diphenyl-1,3,5-hexatriene (DPH) were performed essentially as described previously, ~8 except that measurements were made on an SLM 4800 Spectrofluorimeter in the T-format. The ratio of the parallel to horizontal emission signals was measured with the excitation polarizer oriented vertically (R v) and horizontally (RH). The anisotropy is given by (Rv/RI4 - 1)/ (Rv/RI-I + 2). Temperatures were regulated with an external water bath to _0.2°C. For lipid peroxidation experiments, liposomes were diluted with 50-mM NaC1, 50-mM Tris-HCl, pH 7.0 so that the final concentration after addition of peroxidation initiators would be 0.5 mg/ml. The samples were pre-equilibrated for 1-2 h at the measurement temperature. A sample was taken for initial measurement of thiobarbituric acid reactive substances. After addition of peroxidation initiators, aliquots of 0.5 ml taken at intervals were added to 0. l ml ofbutylated hydroxytoluene (2%). Then, 1.5 ml each of 20% trichloroacetic acid and 0.67% thiobarbituric acid/ 0.05-N NaOH were added and the mixture heated at 100°C for 30 min. The tubes were then cooled, centrifuged for 15 min at 3000 rpm and transferred to glass cuvettes. The difference in absorbance at 532 and 700 nm (to correct for light scattering) was measured in a Beckman DU-7 spectrophotometer. The thiobarbituric acid reactive substances were calculated in units of malondialdehyde equivalents using a molar extinction coefficient of 1.56 × l05 M -1 cm -~. In most experiments, an Fe-ascorbate oxidation system was used. For this, a solution of ascorbic acid (10-mg/ml water) was added to the liposomes to attain a concentration of 0.5 mM. Then, a freshly prepared solution of Fe(NH4)2(SO4) 2 (1-mg/ml water) was added to attain a final concentration of 0.05 mM. For Fe-ADP oxidations, a solution of 1-mM ADP plus 0.2-mM Fe(NH4)2(SO4)2 was prepared in 50-mM NaC1; 50-mM Tris-HC1, pH 7.0 and incubated for 15 min at 37°C prior to mixing with liposomes in the same buffer. The final concentrations were 0.1-mM Fe(NH4)2(SO4) 2 and 0.5-mM ADP. For FeE+-Fe3+/ EDTA oxidations, a solution of 10-mM FeC13 plus 1 l-raM EDTA in water (pH ~ 7 ) was added to the liposomes to give a 25 laM solution of FeC13, then a

Lip©some peroxidation

<

80

O

E

~ 60 -6 40 E

~ 20 m 0

i

0

i

i

10

20 30 40 50 60 IncubationTime,min Fig. 1. Comparison of methods of preparation and concentrations ofarachidonic acid in DMPC liposomes on the rate ofperoxidation as measured by thiobarbituric acid reactive substances. Peroxidation ofliposomes (0.5 mg/ml) was initiated with Fe-ascorbate (0.05mM Fe(NH4)2(SO4)2 and 0.5-mM ascorbic acid) at either 4°C (open symbols) or 37°C (closed symbols). (O, O) 10% arachidonic acid in unsonicated liposomes; (n, D) 10% arachidonic acid in sonicated liposomes; (0, O) 5% arachidonic acid in unsonicated liposomes. AA--arachidonic acid.

fresh solution ©fEeS© 4 (1-mg/ml water) was added to attain a 25 IxM concentration. The amount of iron bound to liposomes (0.5 mg/ e l ) was quantitated following incubation for 2-3 rain with 0.05-mM Fe(NH4)2(SO4)2 and 0.5-mM ascorbic acid. The liposomes were separated from the buffer by filtration through a 0.2 ~t Acro-Disk filter previously cooled or heated to the temperature of the experiment. Another sample of buffer without liposomes was filtered to show that iron did not bind to the filter. The iron concentration was measured 19 on a 0.2-ml aliquot of the filtrate by adding 0.1 ml ofascorbic acid (1 mg/ml) and 0.1 ml of a 10-mg/ml solution of 3-(2pyridyl)-5,6-bis(4-phenylsulfonic acid)- 1,2,4-triazine (Ferrozine, Sigma) and 0.6 ml of water. The difference in absorbance at 562 and 700 nm (to correct for light scattering) was measured and the iron concentration calculated from measurements on standard solutions of FeSO4.

115

amount of thiobarbituric acid reactive substances produced was proportional to the concentration of arachidonic acid in the liposomes at 5 and 10 mole% as shown by the relative superposition of the kinetic curves in Fig. 1 in which the data are normalized for initial arachidonic acid concentration. A 10 mole% concentration in unsonicated liposomes was chosen to attain a high sensitivity with minimal perturbation of the lip©some structure based on the values of DPH polarization (data not shown). Similarly, the rate of peroxidation of linolenic acid in DMPC liposomes is much faster at 4°C than at 37°C (Fig. 2). By contrast, the rate ofperoxidation of soy PC liposomes proceeds more rapidly at 37°C than at 4°C (data not shown). Thus, the decreased rate of pero xidation at 37 °C with arachidonic acid in DMPC is related to the phase transition of the host lipid and not a decrease in oxygen solubility. The effect of cholesterol on the rate of production of thiobarbituric acid reactive substances in liposomes containing arachidonic acid is shown in Fig. 3 at 4 and 37°C. At the higher temperature, almost no thiobarbituric acid reactive substances are formed. At 4°C, increasing concentrations of cholesterol decrease the rate of peroxidation of arachidonic acid in the liposomes. Conversely, at 37°C, cholesterol increases the rate of peroxidation of arachidonic acid. Cholesterol broadens the phase transition of DMPC as measured by diphenylhexatriene (data not shown) corresponding to a fluidizing effect at temperatures below the phase transition temperature (Tm) and a rigidifying effect above Tm.20At 4°C, only a minimal change in anisotropy is observed, while at 37°C, cholesterol markedly increases the anisotropy (decreases the fluidity) of the acyl chain region of the bilayer. The concentrations ofthiobarbituric acid reactive sub-

25

~

o-

o

2o

RESULTS

In order to establish the best conditions for examining the effects of bilayer fluidity on the rate of peroxidation of molecules within a bilayer, several concentrations of arachidonic acid and methods of lip©some preparation were examined (Fig. 1). At 4°C the rate of peroxidation of arachidonic acid was considerably faster than at 37 °C, regardless of the method of preparation or the concentration of arachidonic acid in the host lipid. The 20 mole% concentration of arachidonic acid used by Cestaro and co-workers ]3 was not necessary to elicit a rapid rate of peroxidation. The

~-C < m 1

0

20

0

5

40

60 80 Time,min

1oo 120

Fig. 2. Effect of temperature on the rate of peroxidation of cis-linolenic acid in DMPC unsonicated liposomes. Peroxidation of liposomes containing 10 mole% cis-linolenic acid was measured at 37°C (O) and 4°C (©) as described under Fig. 1. LA--linolenic acid.

116

L.R. MCLEAN and K. A. HAGAMAN 80

o

40

60

3o 40

< ch

,< -6 E

20

,.-.,

0

E

20 c J

60

i

,

37"C

0

50

0

rr" ,,( IE) 40 )--

10

20

30

40

50

60

Incubation Time, rain

Fig. 5. Peroxidation of PAPC in DMPC unsonicated liposomes containing various concentrations of cholesterol below and above Tin. The incubation temperatures of the experiments were 4°C (open symbols) and 37°C (closed symbols). Peroxidation w a s i n i t i a t e d with Fe-ascorbate as described under Fig. I. DMPC liposomes contained ( L ©) no cholesterol; (A, A) 5% cholesterol; (i, D) 10% cholesterol; (0, 0) 20% cholesterol.

30 20

0 0

20

40

60

Incubation Time, min

Fig. 3. Peroxidation of arachidonic acid in DMPC unsonicated liposomes containing various concentrations of cholesterol below and above Tin. Peroxidation w a s i n i t i a t e d w i t h F e - a s c o r b a t e a s described under Fig. l at either 4°C (upper panel) or 37°C (lower panel) in DMPC liposomes containing (©) no cholesterol; (A) 5% cholesterol; (I) 10% cholesterol; ([3) 20% cholesterol.

stances produced over the first 5 and 15 min as a function of the fluorescence anisotropy of diphenylhexatriene in the liposomes is shown in Fig. 4. The most dramatic increases in the rate of peroxidation occur when the fluorescence anisotropy begins to reach a value in excess of 0.3. The increase in the rate of peroxidation at high anisotropies is striking. The effects of temperature and cholesterol concentration on the rate of oxidation of PAPC in DMPC liposomes were also examined. The rate of peroxidation of PAPC is faster at 4°C than at 37°C (Fig. 5), as

was observed for arachidonic acid oxidation in equivalent environments. Increasing cholesterol concentrations in PAPC-containing liposomes has no significant effect on the rate ofperoxidation either above or below the phase transition temperature of the host lipid. However, a dramatic temperature-dependent increase in peroxidation is observed for PAPC in DMPC liposomes commencing at temperatures below ~ 2 2 ° C (Fig. 6), closely corresponding to the phase transition temperature of the host lipid. The possibility that at temperatures below T,,, phase separation of the arachidonic acid results in increased iron binding that influences the kinetics of peroxidation was investigated by measuring the rates of peroxidation of arachidonic acid in DMPC under several conditions. Addition of EDTA, in concentra-

o o_

7o ,( <

60

-~

5o

40

Q_

~

3o

D "~c

20

~

40

-~

c m" rr < en J--

to

3o

rn

20 0

10

0 i

i

i

i

i

i

0.10

0.15

0.20

0.25

0.30

0.35

Fluorescence Anisotropy Fig. 4. P r o d u c t i o n o f T B A reactive substances as a f u n c t i o n o f fluorescence anisotropy of diphenylhexatriene. The concentrations of

thiobarbituric acid reactive substances (as nmol MDA) after 5 min (I) and 15 min (O) of incubation are plotted as a function of t h e fluorescence anisotropy measured prior to peroxidation.

i 10

i 20

i 30

40

Temperature, °C

Fig. 6. Effect of temperature on the rate ofperoxidation of PAPC in DMPC unsonicated liposomes. Peroxidation w a s i n i t i a t e d w i t h Feascorbate as described under Fig. l with 10 mole% PAPC in DMPC liposomes. The concentrations of TBA-reactive materials were measured over the course of 60 min. The temperature-dependence of the TBARS prior to incubation (©) and after 60 min of incubation (e) are shown.

Liposome peroxidation Table 1. Effect of Incubation Conditions on Arachidonic Acid Peroxidation TBARS (nmol MDA) Peroxidation Condition

4°C

37°C

Fe-ascorbate Fe-ascorbate/EDTAi" Fe-ascorbate/Ca2÷~t Fe-ADP Fe2+/Fe3+-EDTA

4.45 0.64 3.93 7.82 4.80

1.03 0.92 1.72 4.20 1.96

Samples of arachidonic acid (10 mole%) in DMPC liposomes were incubated for 60 min at 4 or 37°C with the indicated peroxidation initiators as described under Materials and Methods. Then, TBARS were measured as described under Materials and Methods. t Incubations were conducted in 50-mM NaC1, 50-mM trisHCL, pH 7.0 containing 1-mM EDTA. Incubations were conducted in 50-mM NaCl, 50-mM trisHCL, pH 7.0 containing 0.05-mM CaCl 2.

tions greater than the iron, prevents peroxidation at low temperatures as has been observed at 37°C by Tien et al. 21 However, addition of ADP or calcium does not change the more rapid relative rate ofperoxidation at 4°C vs 37°C (Table 1). A similar enhancement in peroxidation at 4°C is also evident with FEE+/ Fe3÷-EDTA as an initiator. The amount of iron bound to the liposomes was determined by measurement of the decrease in iron concentration in the buffer after addition of liposomes. The liposomes were separated from the buffer by filtration. For liposomes at a concentration of 1.47-mM DMPC (147~tM arachidonic acid) and 1004tM iron, at 4°C, 12% of the iron is bound and at 37°C, 15% is bound. Thus, the concentration of iron bound does not differ dramatically at the two temperatures. To test whether at 4°C the arachidonic acid is displaced from the DMPC vesicles into the aqueous phase where it is rapidly oxidized, two experiments were performed. In the first, the rate of peroxidation of arachidonic acid in the absence of DMPC was measured. In the second, the DMPC liposomes containing arachidonic acid were incubated at 4°C for 1 h. Then, the liposomes were separated from the aqueous phase by centrifugation and the pellet and the supernatant were oxidized separately. Although relatively rapid peroxidation of arachidonic acid was observed in the aqueous phase in the absence of DMPC, little or no peroxidation was observed in the supernatant separated from the DMPC liposomes in the second experiment (Fig. 7). The rate of peroxidation of the pellet which was resuspended in buffer nearly matched that observed in previous experiments in which the liposomes were not separated from the supernatant. In a related experiment, the liposomes were separated from the aqueous phase after peroxidation. In this

117

case, the supernatant at 4 or 37°C contained virtually all of the oxidized lipid. Thus, arachidonic acid does not appear to transfer to the aqueous phase at 4°C to any appreciable extent and the products of peroxidation are relatively water soluble. The fluorescence polarization data do not differentiate between decreased acyl chain fluidity and phase separation of the fatty acids. 22 At concentrations greater than about 5 mole%, fatty acids are phaseseparated into clusters. 23 Thus, dilution of arachidonic acid below 5 mole% should result in a more random distribution of the fatty acids in the lipid bilayer. The resulting loss of close packing of the fatty acid hydrocarbon chains would decrease the probability of peroxidative chain reactions provided that the more rapid rate of peroxidation of arachidonic acid below the phase transition temperature of the host lipid is due to formation of fatty acid clusters phaseseparated from the host lipid. On the other hand, if the effect were simply due to increased acyl chain rigidity, disruption of close packing would not affect the increased rate of peroxidation observed below the transition temperature. Dilutions were made by increasing the concentration of DMPC at constant arachidonic acid concentrations (Fig. 8A) and by decreasing the concentration of arachidonic acid at constant DMPC concentrations (Fig. 8B). Regardless of which method was chosen for reducing the concentration of arachidonic acid in DMPC, decreasing the concentration below 2 mole% dramatically inhibits the rate of peroxidation at 4°C.

< <

IOO

'

'

o

'

'

' ~

ao ~

60

~

40

~

o 0

10

,,

?

•

•

?

20

30

40

50

60

Incubation Time, min

Fig. 7. Relative rates of production of TBA-reactive substances in the bilayer and aqueous phases in arachidonic acid-containing unsonicated liposomes. DMPC liposomes (1 mg/ml) containing 10 mole% arachidonic acid were incubated at 4°C for 1 h, then the liposomes were separated from the aqueous phase by centrifugation at 40,000 rpm for 30 min. The liposomes (resuspended at a concentration of I mg/ml) and the supernatant were incubated separately for up to 60 min in the presence of the Fe-ascorbate oxidizing system as described under Fig. 1. At intervals, samples were removed and the TBARS were measured in the supernatant (©) and the pellet (e).

118

L.R. MCLEANand K. A. HAGAMAN 200

A

150

100 < E

g E

50

0 200

15O

/

100

/C~°/

/ jo J-

5O

o ~

0

2

4

6

8

10

12

14

16

Incubation Time, min

Fig. 8. Effectof arachidonic acid concentration in DMPC unsonicared liposomeson the rate ofperoxidation.Peroxidationwas initiated with Fe-ascorbateas describedunder Fig. 1 at constant concentrations of arachidonic acid ( 11/ag/ml) and increasing concentrations of DMPC (Panel A) or at constant concentrations of DMPC (2.5 mg/ml) with decreasingconcentrations of arachidonic acid. In both panelsthe concentrationsof arachidonic acid relative to DMPCare 1 mole%(IS]),2 mole%(O) and 5 mole%(O).

DISCUSSION The present set of experiments show that with arachidonic and linolenic acids, the rate of peroxidation is faster below the phase transition temperature of the host lipid than above. In relatively rigid bilayers, a minimal decrease in bilayer fluidity as judged by an increase in diphenylhexatriene fluorescence anisotropy results in large increases in the rate of peroxidation. In relatively fluid bilayers, increasing acyl chain rigidity has much less effect on the rate of peroxidation. The increase in rate ofperoxidation below T,, is not due to either an increase in iron binding or partitioning of arachidonic acid into the aqueous phase. Cholesterol, which increases the fluidity ofbilayers at temperatures below Tin, increases the rate of peroxidation above and decreases it below the phase transition temperature of the lipid. Thus, there is a correspondence between the effects of cholesterol on bilayer awl chain fluidity and the rate of lipid peroxidation. Similarly, the rate of peroxidation of the phospholipid palmitoyl-arachidonoyl phosphatidylcholine increases dramatically below the transition temperature of the host lipid. However, cholesterol has only a minimal effect on the rate of peroxidation of this mixed chain phospholipid. For arachidonic acid in lipid bilayers the increase in rate ofperoxidation below Tm may be attributed to

either an decrease in fluidity of the acyl chains of the lipid or to phase separation of the peroxidizable lipid. The dramatic increase in rate of peroxidation of arachidonic acid at high anisotropies (Fig. 4) suggests a phase boundary beyond which local regions of rapid peroxidation exist in the bilayer. However, it is not possible from the fluorescence polarization data to differentiate between lipid phase separation and a decrease in bilayer fluidity. A separation of a fluidity effect from that of fatty acid clustering was achieved by decreasing the mole fraction of arachidonic acid either at constant DMPC concentrations or by increasing DMPC concentrations at constant arachidonic acid concentration. Both approaches resulted in dramatic decreases in the rate of peroxidation of arachidonic acid at concentrations below about 2 mole%. This concentration corresponds to that at which phase separation is observed. 23 These data support the notion that the increase in rate of peroxidation of arachidonic acid at 4°C is due to phase separation of a cluster of fatty acids. This phase separation correlates with a measured decrease in bilayer fluidity and presumably promotes propagation of the free radical process. The observed similarity in the effects of cholesterol, temperature and dilution suggests that a general relationship between membrane fluidity and phase separation and the rate of lipid peroxidation may exist. A special packing of the acyl chains of the phaseseparated lipid may be promoted at lower temperatures and in the presence of cholesterol with both arachidonic acid and the arachidonoyl chains of PAPC. With PAPC only the clustering of arachidonyl chains would be effective in promoting peroxidation. Thus, the size of clusters of arachidonyl chains is limited by the packing constraints of the palmitoyl chain in the sn-1 position. In biological lipids, with a variety of acyl chains, only limited clustering may be possible. However, local regions in which packing is favorable may be promoted by the presence of other lipids or by preferential head group interactions as have been observed with phosphatidylserine in the presence of calcium. 24-26 In summary, the data support the hypothesis that the increase in rate of lipid peroxidation is the result of a phase separation of the fatty acids which increases the local concentration of peroxidizable chains and promotes propagation of oxidation. In cell membranes, lipid domains 27 in which the local acyl chain fluidity is decreased may be, according to this model, regions prone to lipid peroxidation. In addition, the data suggest that long-range effects of proteins on lipid dynamics which may induce domain formation 2s may promote peroxidation of membrane pro-

Liposome peroxidation

teins. Recent data on the effects of ATP depletion in hepatocytes indicate that significant increases in the order parameter of plasma membrane blebs occur early in hypoxic cell i n j u r y . 29 Such changes in membrane fluidity may influence membrane structure in a way that predisposes such cells to peroxidative injury, such as that observed following reoxygenation after ischemia.30 These notions support the intriguing possibility that antioxidants, such as vitamin E and probucol, which are not only effective as antioxidants, but also modulate membrane bilayer fluidity, ~7'27may be particularly effective in preventing peroxidative damage in vivo by preventing the formation of lipid acyl chain clusters which may promote long radical chain events. Acknowledgements - - We thank Dr. Craig E. Thomas for valuable suggestions and his critical reading of the manuscript.

REFERENCES

1. Yasuda, Y.; Uyesaka, N.; Shio, H.; Akiguchi, 1.; Kameyama, M. Electron spin resonance studies of erythrocyte membrane in spinocerebellar degeneration. J. Neurol. Sci. 90:281-290; 1989. 2. Zubenko, G.S.; Malinakova, I.; Cohen, B.M. Temperature dependence of the molecular dynamics of platelet membranes in Alzheimer's disease. Biol. Psychiatry 22:987-994; 1987. 3. Hitzemann, R.J.; Hirschowitz, J.; Garver, D.L. On the physical properties of red cell ghost membranes in the affective disorders and psychoses. A fluorescencepolarization study. J. Affect. Dis. 10:227-232; 1986. 4. Patel, J.M.; Block, E.R. Nitrogen dioxide-induced changes in cell membrane fluidity and function. Am. Rev. Respir. Dis. 134:1196-1202; 1986. 5. Patel, J.M.; Block, E.R. The effect of oxidant gases on membrane fluidity and function in pulmonary endothelial cells. Free Radical Biol. Med. 4:121-134; 1988. 6. Janero, D.R.; Burghardt, B. Prevention of oxidative injury to cardiac phospholipid by membrane-active "stabilizing agents". Res. Commun. Chem. Path. Pharm. 63:163-173; 1989. 7. Hillard, C.J.; Harris, R.A.; Bloom, A.S. Effects of the cannabinoids on physical properties of brain membranes and phospholipid vesicles:fluorescence studies. J. Pharm. Exp. Therap. 232:579-588; 1985. 8. Nakazawa, T.; Nagatsuka, S.; Yukawa, O. Effects of membrane stabilizing agents and radiation on liposomal membranes. Drugs Exp. Clin. Res. 12:831-835; 1986. 9. Lucy, J.A. 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. 203:4-11; 1972. 10. Tien, M.; Morehouse, L.A.; Bucher, J.R.; Aust, S.D. The multiple effects of ethylenediaminetetraacetate in several model lipid peroxidation systems. Arch. Biochem. Biophys. 218:450458; 1982. 11. Montfoort, A.; Bezstarosti, K.; Groh, M.M.; Metsfi-Ketelh, T.J. The influence of the lipid composition on the degree of

119

lipid-peroxidation of liposomes. Biochem. Int. 15:525-543; 1987. 12. Mowri, H.; Nojima, S.; Inoue, K. Effect of lipid composition of liposomes on their sensitivity to peroxidation. J. Biochem. (Tokyo) 95:551-558; 1984. 13. Cervato, G.; Viani, P.; Masserini, M.; Di Iorio, C.; Cestaro, B. Studies on peroxidation of arachidonic acid in different liposomes below and above phase transition temperature. Chem. Phys. Lipids 49:135-139; 1988. 14. Viani, P.; Cervato, G.; Fiorilli, A.; Rigamonti, E.; Cestaro, B. Studies on peroxidation processes of model membranes and synaptosomes: Role of phosphatidic acid. Chem. Phys. Lipids 52:49-55; 1990. 15. Keough, K.M.W.; Parsons, C.S. Differentialscanning calorimetry of dispersions of products of oxidation of l-stearoyl-2-1inoleoyl-sn-glycero-3-phosphocholine. Biochem. Cell Biol. 68:300-307; 1990. 16. Coolbear, K.P.; Keough, K.M.W. Lipid oxidation and gel to liquid-crystalline transition temperatures of synthetic polyunsaturated mixed-acid phosphatidylcholines. Biochim. Biophys. Acta 732:531-540; 1983. 17. McLean, L.R.; Hagaman, K.A. Probucol reduces the rate of association of apolipoprotein C-Ill with dimyristoylphosphatidylcholine. Biochim. Biophys. Acta 959:201-205; 1988. 18. McLean, L.R.; Hagaman, K.A. Antioxidant activity ofprobucol and its effects on phase transitions in phosphatidylcholine liposomes. Biochim. Biophys. Acta 1029:161-166; 1990. 19. Boyer, J.L.; Hepler, J.R.; Harden, T.K. Hormone and growth factor receptor-mediated regulation of phospholipase C activity. Trends Pharmacol. Sci. 10:360-364; 1989. 20. Phillips, M.C. Physical state of phospholipids and cholesterol in monolayers, bilayers, and membranes. Progr. Surface Membrane Sci. 5:139-221; 1972. 21. Tien, M.; Svingen, B. A.; Aust, S.D. An investigation into the role of hydroxyl radicals in xanthine oxidase-dependent lipid peroxidation Arch. Biochem. Biophys. 216:142-151; 1982. 22. Sklar, L.A.; Miljanich, G.P.; Dratz, E,A. Phospholipid lateral phase separation and the partition of cis-parinaric acid among aqueous, solid lipid, and fluid lipid phases. Biochemistry 18:1707-1716; 1979. 23. Hauser, H.; Guyer, W.; Howell, K. Lateral distribution of negatively charged lipids in lecithin membranes. Clustering of fatty acids. Biochemistry 18:3285-3291; 1979. 24. Jacobson, K.; Papahadjopoulos, D. Phase transitions and phase separations in phospholipid membranes induced by changes in temperature, pH, and concentration of bivalent cations. Biochemistry 14:152-161; 1975. 25. Feigenson, G.W. On the nature of calcium ion binding between phosphatidylserine lamellae. Biochemistry 25:58195825; 1986. 26. Hui, S.W.; Boni, L.T.; Stewart, T.P.; lsac, T. Identification of phosphatidylserine and phosphatidylcholine in calcium-induced phase separated domains. Biochemistry 22:3511-3516; 1983. 27. Klausner, R.D.; Kleinfeld, A.M.; Hoover, R.L.; Karnovsky, M.J. Lipid domains in membranes. Evidence derived from structural perturbations induced by free fatty acids and lifetime heterogeneity analysis. J. Biol. Chem. 255:1286-1295; 1980. 28. Jain, M.K.; White, III, H.B. Long-range order in biomembranes. Adv. Lipid Res. 15:1-60; 1977. 29. Florine-Casteel, K.; Lemasters, J.J.; Herman, B. Lipid order in hepatocyte plasma membrane blebs during ATP depletion measured by digitized video fluorescence polarization microscopy. FASEB J. 5:2078-2084; 1991. 30. Gutteridge, J.M.C.;Halliwell, B. Reoxygenationinjuryandantioxidant protection: A tale of two paradoxes. Arch. Biochem. Biophys. 283:223-226; 1990.