Effects of exogenous phospholipases on brain membrane phospholipid perturbation, (Na+ + K+)-ATPase activity and cellular swelling of brain slices

Neurochem. Int. Vol. 10, No. 3, pp. 303-310, 1987 Printed in Great Britain. All rights reserved

0197-0186/87 $3.00+ 0.00 © 1987 PergamonJournals Ltd

EFFECTS OF EXOGENOUS PHOSPHOLIPASES ON BRAIN MEMBRANE PHOSPHOLIPID PERTURBATION, ( N a + + K÷)-ATPase ACTIVITY AND CELLULAR SWELLING OF BRAIN SLICES PAK HOO CHAN,* SYLVIA CHEN and ROBERT A. FISHMAN Brain Edema Research Center, Department of Neurology, University of California, School of Medicine, San Francisco, CA 94143, U.S.A. (Received 21 August 1986; accepted 14 October 1986)

Abstract--Rat brain membranes were incubated with bee venom phospholipase A 2 (PLA2) or phospholipase C (PLC) from Clostridium perfringens. PLA2 caused a significant increase in free polyunsaturated fatty acids concomitant with membrane phospholipid degradation as monitored by HPLC and by gas chromatography. Equal concentrations of PLC had a much lesser effect than PLA2. Divergent and differential effects were shown on deacylation and incorporation of [3H]arachidonic acid in membrane phospholipids. The incorporation of [3H]arachidonic acid into various phospholipids was greatly reduced by PLA2 (0.018 units/ml) whereas PLC at identical concentration was not effective. PLA2 inhibited (Na + + K +)-ATPase but was not effective on p-nitrophenyl-phosphatase activity whereas PLC stimulated both enzymes. PLA2 induced swelling of cortical brain slices whereas PLC was not effective. Thus, the severity of the perturbation of membrane integrity, and the inhibition of (Na + + K+)-ATPase in brain membranes may play an important role in cellular swelling of brain slices induced by PLA2.

Brain in situ contains a very low concentration of free polyunsaturated fatty acids (PUFAs) like arachidonic acid (20: 4) and docosahexaenoic acid (22: 6) (0.01/~mol/g wet weight) (Lunt and Rowe, 1968; Gardiner et al., 1981; Yoshida et al., 1982). The extremely low concentration of free P U F A s is due to the dynamic control of deacylation and reacylation processes in membrane phospholipids. P U F A s are readily acylated into C-2 position (Yau and Sun, 1974) of the glycerol backbone of membrane phospholipids under normal conditions. In ischemic injured brain such acylation in brain membranes is largely inhibited (Pediconi et al., 1983). Ischemia, cold-injury and other pathological conditions in brain also cause membrane injury and a rapid deacylation of P U F A s from membrane phos-

pholipids (Bazan, 1970; Agardh and Siesj6, 1981; Gardiner et al., 1981; Rehncrona et al., 1982; Tang and Sun, 1982; Yoshida et al., 1982; Chan et al., 1983b). Either PLA 2 or C or both have been proposed to be involved in the deacylation process in brain (Bazan and Turco, 1980; Edgar et al., 1982; Chan et al., 1984b). The localization of both enzymes in brain has been also demonstrated (Van Den Bosch, 1980; Edgar and Freysz, 1982). The exact role of endogneous PLA 2 and PLC on membrane phospholipid breakdown and the alterations of membraneassociated function and the subsequent cellular injury is not clear at present. We have demonstrated that membrane injury and phospholipid degradation are the early and important events following freezing and free radical injuries (Chan et al., 1982, 1983b,d, 1984b). Arachidonic acid derived from membrane injury could be *Address correspondence to: Dr Pak Hoo Chan, Department of Neurology M-794, University of California, metabolized further to form oxygen free radicals, School of Medicine, San Francisco, CA 94143, U.S.A. prostaglandins, thromboxanes and leukotrienes Tel.: (415) 476-2987 (Samuelsson et al., 1979; Wolfe, 1982). These metabAbbreviations: 20: 4, Arachidonic acid; 22: 6, Docosahexa- olites are involved in the development of cellular enoic acid; HPLC, High-performance liquid chromatography; PC, Phosphatidylcholine; PE, Phosphatidy- injury and edema (Chan and Fishman, 1978; Kuehl lethanolamine; PI, Phosphatidylinositol; PS, Phosphati- and Egan, 1980; Samuelsson B, 1983; Chan et al., dylserine; SM, Sphingomyelin; DG, Diacylglycerols; 1983c, 1984a; Chan and Fishman, 1984). Since both PLA2, Phospholipase A2; PLC, Phospholipase C. freezing and free radical injuries affect phos303

304

PAK HOO CHAN et al.

pholipases and membrane phospholipids, the question remains as to what are the comparative effects o f PLA2 and PLC on the degree of membrane perturbation, the level of accumulated arachidonic acid and the development of brain edema and injury. In this study, we compare the effects of bee venom PLA2 and PLC from Clostridium perfringens on both structural and functional alterations of membrane and the development of brain edema. These enzymes have a known lipid specificity and may thus provide some information to correlate the known lipid changes with the possible development of tissue swelling. Bee venom PLA2 causes a release of arachidonic acid and other P U F A at C-2 position from various phospholipids (Van Den Bosch, 1980) whereas bacterial PLC hydrolyzes the phosphodiester bond of PC forming diacylglycerol (DG) and phosphocholine (Sleight and Kent, 1980). The deacylation and the incorporation of [3H]arachidonic acid into the membrane phospholipids affected by these enzymes were monitored by H P L C (Chan et al., 1983d). Arachidonic acid and other fatty acids are further released from D G by D G lipase, a membrane enzyme occurring in brain (Edgar and Freysz, 1982). Other membrane associated functions studied include the activities of phospholipid-dependent (Na ÷ + K ÷ ) ATPase and K+-activated p-nitrophenylphosphatase. Finally, the effects of PLA2 and PLC on the development of cellular swelling and cation levels in brain slices were also studied. EXPERIMENTAL PROCEDURES

Materials PLA 2 (E.C. 3.1.1.4, bee venom, 1500 units/mg protein), PLC (E.C. 3.1.4.3, Clostridium perfringens, 20 units/mg protein), phosphatidylcholine (bovine brain), phosphatidylethanolamine (bovine brain), phosphatidylserine (bovine brain), sphingomyelin (bovine brain), tysophosphatidylethanolamine (bovine brain) were obtained from Sigma, St. Louis, Mo. Phosphatidylinositol (bovine liver) was purchased from Cal Biochem, La Jolla, Calif. Fatty acids and fatty acid methylester standards were obtained from Supelco, Bellafonte, Pa. Hexane and 2-propanol were from Burdick and Jackson laboratories, Muskegon, Mich. H20 used for high-performance liquid chromatography was obtained from distilled water filtered through Milli Q water purification system (Millipore Corp., Bedford, Mass.). [5,6,8,9,11,12,14,15 3H(N)]arachidonic (61.0 Ci/mmol, 98% purity) was obtained from New England Nuclear, Boston, Mass. Separation of free fatty acids and phospholipids by highperformance liquid chromatography The various phospholipids, free fatty acids and neutral lipids of lipid extracts from brain homogenates were separated by high-performance liquid chromatography (HPLC,

1082 B liquid chromatography, Hewlett Packard, Palo Alto, Calif.) using a silica column (Micro-Pak Si-5, 30 cm x 4 mm i.d.), Varian Assoc., Palo Alto, Calif.) according to our previous method (Chan et al., 1983d). The flow rate of the column was maintained at l ml/min. The gradient began with an initial mobile phase of 4% H20 and was increased to 9% H20 at 10 min and maintained at 9% H20 for 10 min, followed by reduction of H20 to 4% at 25 rain. Lipid elution was monitored at 206 nm with a recorder response at 0.512 AU/cm. Fatty acid analysis The free fatty acids eluted from HPLC were pooled and subjected to methylation according to the method previously described (Chan et al., 1983b). The fatty acid methylesters were analyzed by gas chromatography (Hewlett Packard 5830 A) equipped with fused silica capillary column (SP2330, 0.25 mm i.d. x 30 M, Supelco, BeUafonte, Pa). The quantitative procedures were described previously (Chan et al., 1983b). Incorporation of [3H]arachidonic acid into phospholipids and neutral lipids of cortical slice homogenates A pool of six single first cortical rat brain slices was homogenized in 4 ml of Krebs-Ringer buffer with a tissue homogenizer (Kontes, San Leandro, Calif.) for 1 min at 4°C (Chan and Fishman, 1978). The homogenates were centrifuged at 1000g for 20 min and 1.5 ml of crude membrane fraction (St) (2.5mg protein/ml) was incubated with 1.5 # Ci/ml of [5,6,8,9,11,12,14,15 3H(N)]arachidonic acid in Krebs-Ringer buffer containing 2.5 mM ATP and 0.1 mM dithiotreitol and CoA for various times (Majewska and Sun, 1982). The incubation mixtures were washed three times with Krebs-Ringer buffer containing 0.1% BSA to remove the excess free [3H]20:4. The lipids of the incubation mixtures were extracted with chloroform:methanol (2: l, v/v) according to the method of Folch et al. 0957). The lipids extracts were washed with Krebs-Ringer and were dried under N 2. The lipids were resuspended with hexane/2-propranol (3:4, v/v) for HPLC analysis. The neutral lipids obtained from HPLC were further separated by Silica Gel H thin layer chromatography (250pm, Analtech, Newark, Del.) Triacylglyeerols, diacylglycerols and monocylglycerols were separated with a solvent system consisting of heptane/diethylether/formic acid (90: 60: 4 v/v) according to the method of Matsuzawa and Hostetler (1980). The radioactivity of various neutral lipid, free fatty acids and individual phospholipids were determined by scintillation spectrophotometry. (Na + + K +)-A TPase and K +-stimulated p-nitrophenylphosphatase assays (Na + + K+)-ATPase (EC 3.6.1.3) activity was measured as the difference in inorganic phosphate (Pi) formed from ATP in the presence and the absence of 0.25 mM ouabain (Skou and Esmann, 1979). The assay medium contains 3 mM ATP, 80 mM imidazole, I00 mM NaC1, 20 mM KCI, 5 mM MgCI2 pH 7.5 and the assay was carried out at 37°C for 30min. K+-stimulated p-nitrophenylphosphatase (EC 3.1.3.1) was measured by the hydrolysis of colorless pnitrophenylphosphate to form a yellow product which could be measured spectrophotometrically. The assay medium contains 100mM KCI, 20mM Mgel 2, 30mM histidine, 0.2% bovine serum albumin at a pH of 7.6.

Phospholipases and brain swelling NL

Tissue swelling, Na + and K + content o f cortical slices

Sprague-Dawley male rats (purchased from BantinKingman, Fremont, Calif.) weighing 100 + 10 g were decapitated and the brains rapidly removed and placed on ice. The preparation of single first cx~rtiealslices, weighing 40-50 mg with a 0.35 nun thickness, was reported previously (Chan and Fishman, 1978). The initial wet weight of each slicewas measured with a Cahn 25 automatic electrobalance (Cahn Instruments, Cerritos, Calif.) and each slice was then incubated in either 5 ml Krebs-Ringer or Krebs--Ringercontaining different concentrations of either PLA2 or PLC at 37°C for 60 min. The composition of Krebs-Ringer was reported previously (Chan and Fishman, 1978). After each slice was slightly blotted on an acid-washedglass Petri dish to remove excess water adherent to the slice and the final wet weight was measured. The cortical slices were dried at 105°C for 16 h to obtain dry weight. The tissue swellingwas obtained according to the formula published elsewhere (Chan et aL, 1982). Dried sliceswere extracted with 2 M nitric acid (4 ml) for 16 h. The nitric acid extracts were used to measure Na + and K + contents by atomic absorption spectrophotometry (Perkin-Elmer 560) (Chan et al., 1982).

305 P1

°I

J

< A

E e',O O v e u tO

o

"l II'', - ' - - - J \ , , .

es

<

RESULTS

Figure 1 shows the HPLC profile of lipid extracts of brain homogenates. After 10 to 60 min incubation at 37°C with PLA2 at 18 units per ml, the integrated area of phosphatidylethanolamine (PE, plus plasmalogen), phosphatidylinositol (PI), phosphatidylserine (PS), phosphatidylcholine (PC) were decreased significantly. Sphingomyelin (SM) was not affected. The integrated area of free fatty acids were increased significantly at 10 and 60 min respectively. An extra peak which corresponds to lysophosphatidylethanolamine (LPE) was observed. Lysophosphatidylcholine was eluted out at 28 min and was not detected due to its extremely low UV absorbance of its saturated acyl moieties. Other lysophospholipids were also not detected in this HPLC lipid elution profile. Neutral lipids and cholesterol were not affected. When the free fatty acid peak was further analyzed by gas chromatography, it was shown that oleic acid (18: 1), arachidonic acid (20:4) and docosahexaenoic acid (22:6) were increased by 131-, 268, 145-fold respectively at 10m in after the treatment of PLA2 (Table 1). Both palmitic acid (16:0) and stearic acid (18:0) were not affected. 18: 1, 20:4 and 22:6 were increased by 340-, 491-, and 262-fold respectively at 60 rain whereas 16:0, and 18:0 were increased slightly. The low levels of saturated fatty acid findings are quite unexpected since brain lysophospholipases are extremely active and are capable of hydrolyzing lysophospholipids to form saturated fatty acids (Sun and Foudin, 1984). The relatively low activities of brain lysophospholipases observed in the present

JUl I 0

I 5

~ I

I 15

I 20

I 25

I 30

Time (min.) Fig. 1. High-performance liquid chromatography separation of membrane phospholipids and free fatty acids resulting from phospholipase A2 hydrolysis. Brain mere. branous fraction (S0 (2.5 mg protein/ml) was incubated with phospholipase A2 (18 units/ml) at 37°C for 10 and 60rain. The reaction was stopped by adding 10 times volume of chloroform-methanol (2:1, v/v). The extracted lipids were redissolved in hexane: 2-propanol (3:4, v/v) and separated by high-performance liquid chromatography [11]. Upper panel, control; middle panel, 10 min; lower panel, 60 min. Note the complete disappearance of PI, PS and decreased absorbance in PE and PC and the concomitant increase in free fatty acids and LPE in the middle and lower panels. NL, neutral lipids plus cholesterol; FFA, free fatty acids, PE, phosphatidylethanolamine (plus plasmalogen); IS, phosphatidylserine; PC, phosphatidylcholine; SM, sphingomyelin; LPE, lysophosphatidylethanolamine. Lysophosphatidylcholine was not shown in this chromatogram and is eluted out at 28 minutes. studies are not known and require further elucidation. Thus, the increase in UV absorbance of free fatty acids obtained from the HPLC profile is an indication of the increase of polyunsaturated fatty acid.

306

PAK HOO CHAN et al. Table 1. Effects of phospholipase A2 or C on the release of free fatty acids from membrane phospholipids Free Fatty Acid (nmol/mg protein) Incubation time Control 10min 60min

PLA 2 PLC PLA 2 PLC

16:0

18:0

18:1

20:4

22:6

3.0 -+ 0.4 5.2_+2.4 13.2_+1.6 8.9_+5.2 13.9_+0.2

2.7 ± 0.4 2.1 _+0.16 9.7_+3.6 4.7_+0.4 9.6_+4.4

0.13 + 0.05 17.1 _+2.2 10.6_+3.0 44.2_+3.3 13.1 _+2.1

0.06 + 0.01 16.1 _+ 1.9 1.3_+0.3 29.5_+2.7 2.2_+ 1.9

0.13 ± 0.05 18.9_+2.4 2.6_+0.8 24.1_+3.8 3.6_+ 1.3

Brain homogenates (2.5 mg protein/ml), were incubated with bee venom phospholipase A 2 or C (18 units/ml) at 37°C for 10 or 60 min. The free fatty acid peak was separated by HPLC, methylated and determined by gas chromatography using a capillary column. The results are obtained from four different determinations and expressed as a mean _+standard error of means.

pc FFA

°I

Pi

FFA

E ¢,. 0 co r~

<

J

p$

0'

5'

1'0

I '5

2'0

' 25

' 30

Time (min.)

Fig. 2. High performance liquid chromatography separation of membrane and free fatty acids resulted from phospholipase C hydrolysis. The experimental conditions are described in Fig. 1 except that phospholipase C (18 units/ml) was used. Upper panel, control; middle panel, 10 min; lower panel, 60 rain. Note the disappearance of PC and the concomitant increase of neutral lipids and free fatty acids in the middle and lower panels.

The effect of PLC on membrane phospholipids is shown in Fig. 2. The integrated area of PC was decreased 86 and 92% at 10 and 60 min respectively. Other phospholipids were not affected by the treatment of PLC. The integrated areas were increased 4.6- and 4.1-fold for neutral lipids and 4.1- and 5.l-fold for free fatty acids at 10 and 60 min respectively. Free fatty acids, 16:0, 18: 0, 18:1, 20:4, 22:6 were increased by 4.4-, 3.6-, 81.5-, 21.7-, 20.0-fold respectively at 10min and were increased by 4.6, 3.6, 100.8-, 36.7- and 27.7-fold at 60 min (Table 1). The incorporation of [3H]20:4 into phospholipids and neutral lipids in control membranes at various times is shown in Fig. 3. The majority of [3H]20: 4 was preferentially acylated into PI followed by PC, PE, PS and SM respectively (Fig. 3A). In neutral lipid fractions, [3H]20:4 was predominantly acylated into triacylglycerols (TG), followed by diacylglycerols (DG) and monoacylglycerols (MG) respectively. Acylation of [3H]20: 4 into cholesterol and cholesterol ester was minimal (data not shown). After the treatment with PLA 2 at a concentration of 0.018 units/ml for 30 min, the acylation of [3H]20: 4 into PI, PS, PE and PC was reduced by 90, 75, 65 and 32% respectively. The acylation of PI and PC was completely eliminated by high concentrations of PLA 2 (Fig. 4A). On the other hand, PLC (0.018 units ml), reduced slightly only the acylation of PC and PI. PE and PS were not affected (Fig. 4B). However, acylation of PE, PS, PI and PS was reduced by 60, 64, 90 and 92% respectively at higher concentrations of PLC (18 units/ml). Since phospholipases A2 and C have differential effects on deacylation and acylation of 20:4 in membranous fractions, we further examined their effects on membrane-bound (Na ÷ + K + ) - A T P a s e and K+-stimulated p-nitrophenylphosphatase activities (Fig. 5). PLA2 only inhibited ( N a + + K + ) -

Phospholipases and brain swelling 100 C

A

100t~ y~

-6 E

.....¢ /.......; pc

60

40

40 20 "~ <

8 400 _c .-g

pl

,

y_ 80

.u_

<,

0

1oo

|

oc

-g

A

~3M

80

0 Q.

307

60'

~._.._.,~~--~

4o

~•

60

<

40

~'

20

r~ "1-

2O

Pi P$

o

~: -oo -1.75

0

/d." ~

,

,

i

,

,

,

5

I0

15

20

25

30

//

60

Time (min.)

Fig. 3. The incorporation or acylation of [3H]arachidonic acid into phospholipids and neutral lipids of brain membranes. Brain membranes (2.5 mg protein/ml) were incubated with 1.5 #Ci [3H]arachidonic acid in the presence of other co-factors and substrates at 37°C for various times. The reaction was stopped by adding chloroform:methanol (2: I, v/v). The extracted lipids were separated by HPLC. A typical HPLC lipid profile has been described earlier [11]. Neutral lipids were further separated by silica gel H thin layer chromatography [12]. The radioactivity corresponding to each lipid was counted by scintillation counter. A, phospholipids; B, neutral lipids; TG, triacylglycerols; DG, diacylglycerols; MG, monoacylglycerols.

ATPase with prolonged incubation (60 min) whereas it was not effective on p-nitrophenylphosphatase. Since membrane phospholipids are almost completely bydrolyzed by PLA, even at 10 min, these data suggest that cerebral ( N a + + K+)-ATPase may be somewhat insensitive to the treatment of exogenous PLA2. On the other hand, both (Na + + K+)-ATPase and p-nitrophenylpbosphatase activities were significantly stimulated by PLC. Both enzymes reached their maximal stimulation at 10 min following the incubation of PLC. The differential effects of PLA2 and PLC on Na + pump activity and deacylation and acylation of 20: 4 prompted studies on their effects on cellular injury. Table 2 shows that when rat brain cortical slices were incubated with PLA2 (18 units/ml), tissue swelling and sodium content were increased by 248 and 173% respectively. The potassium content was decreased by 71%. PLC at equal concentration was not effective.

-0.75

0.25

1.25

[Log Enzyme units]

Fig. 4. The effects of phospbolipase A 2 and C on acylation of [3H]arachidonic acid. Brain membranes (2.5rag protein/ml) were incubated with phospholipase A2 and C at 0.018, 0.18 and 18 units/ml at 37°C for 30min. The membranes were then incubated with 1.5/zCi [3H]arachidonic acid in the presence of other co-factors and substrates at 37°C for 30 min. Following the incubation, the membranes were washed two times with Krebs-Ringer, one time with 0.1% BSA to remove the excess non-incorporated [3H]arachidonic acid. The labeled phospholipids were separated by HPLC and were collected and counted by scintillation counter.

DISCUSSION The present studies demonstrate the differential effects of exogenous PLA 2 and C on membrane phospholipid metabolism, deacylation and acylation and on cellular swelling. However, caution must be taken when interpreting these data regarding the possible pathophysiological role of exogenous PLA2 on tissue swelling. It is likely that the biochemical properties between the exogenous PLA2 and the endogenous counterpart are quite different. It may require a much lesser degree of stimulation of endogenous PLA2 to elicit the tissue swelling. It is also likely that the endogenous PLA2 may have a different substrate specificity as compared to bee venom PLA2. The latter is known to be non-specific for various brain phospholipids. On the other hand, PLC from Clostridium perfringens has a limited specificity toward PC. Our data thus confirm the PC specificity of PLC from the same source (Sleight and Kent, 1980). The degradation of PC by PLC was concomitant with the increase in D G and free fatty acids. The increase in both saturated and unsaturated free fatty acids was

308

PAK Hoo Ct-IANet al. _..-..

'c

÷

E

o

50-~ 40 30 20 10-1

~

__

6-

a.

0

10

20 30 40 Time (rain.)

50

60

Fig. 5. Effects of phospholipase A 2 and C on (Na + + K+)-ATPase and p-nitrophenylphosphatase activities. Brain membranes (2.5 mg protein/ml) were incubated with phospholipase A2or C (18 units/ml) at 37°C for various times. The reaction was stopped by adding two times volume of cold buffer. The data represented the mean of four different experiments with duplicate assays. A - - A , phospholipase A2; O - - O , phospholipase C.

seen in both HPLC elution profile and from gas chromatographic analysis, indicating the DG lipase, at least in part, is involved in hydrolyzing DG to form MG and free fatty acids. However, it is difficult to discern whether the free fatty acid release due to PLC is an indication of endogenous DG lipase or that some lipase may be present in the bacterial PLC. Nevertheless, the increased level of free 20:4 and other free fatty acids induced by DG iipase was much smaller than the level produced from PLA2, suggesting that DG lipase and the neutral lipids may not be critical for the subsequent development of cellular Table 2. Effect o f phospholipase A 2 and C on cellular swelling, N a + and K + contents o f brain slices Incubation medium

Swelling ( % )

Na + K+ ( p E q . K g dry wet J)

Control (10)

10.33 + 0.88

578 _+21

394 + 20

11.25___1.6 27.91 + 3.9*

555 + 18 960 + 15"

366 + 18 105_+10"

10.77+ 0.67 9.78 + 1.03

563 _+6 569 + 13

357 _+12 366 _ 14

Phospholipase A2

0.18 units/ml(4) 18 units/ml Phospholipase C

0.18 units/ml(4) 18 units/ml

Rat brain slices were incubated with Krebs-Ringer (control) or Krebs-Ringer buffer containing phospholipase A2 or C for various concentrations at 37°C for 60 min. *P < 0.01, compared to control group.

swelling and injury in brain slices following the incubation of exogenous bacterial PLC. Again, the substrate specificity and mode of action of endogenous PLC may be quite different from the bacterial PLC, nevertheless, our data only suggest that the mild perturbation of membrane lipids may not be critical for the subsequent development of tissue swelling. On the other hand, PLA: caused a rapid deacylation of all major phospholipids, indicating this enzyme plays a central role in regulating the release of 20:4 and the subsequent formation of prostaglandins, thromboxanes, oxygen-derived free radicals, and leukotrienes (Samuelsson et al., 1979; Wolfe, 1982). Although these cyclo-oxygenase and lipoxygenase products were not studied, it is well known that these eicosanoid metabolites are involved in inflammatory processes (Kuehl and Egan, 1980; Samuelsson B, 1983; Chan et al., 1984a). Furthermore, we have suggested that the accumulation of 20:4 and other PUFAs may perturb cellular membrane integrity and lead to the development of cellular swelling (Chan and Fishman, 1978; Chan et al., 1983c; Chan and Fishman, 1984; Chan and Fishman, 1985). Besides its overt effects on deacylation of membrane phospholipids, exogenous PLA2 was also much more effective in inhibiting 20:4 acylation. At a concentration as low as 0.018 units per ml, PLA2 reduced the acylation of 20: 4 into PI, PS, PE and PC of BSA-washed membranes by 90, 75, 65 and 32% respectively. PLC at an equal concentration was not effective. Since PC is located on the outside leaflet of plasma membranes, our data suggest that the PLC-induced mild perturbation of membrane phospholipids occurs at the outside leaflet. The data further suggest that deacylation or acylation occurs despite mild degradation of phosphatidylcholine or the accumulation of DG and phosphocholine. On the other hand, the bulk perturbation of inside and outside membrane phospholipids by PLA: might provide a much less suitable environment for acylation. The differential effects of exogenous PLA 2 and PLC on membrane perturbation was further confirmed by their effects on (Na ÷ + K÷)-ATPase and p-nitrophenylphosphatase activities. The inhibition of (Na ÷ +K÷)-ATPase required prolonged incubation of PLA 2 (e.g. 60 min) during which time, the free 20:4 and 22:6 were accumulated by 491- and 262-fold respectively. These data suggest that the accumulation of 20:4 and 22:6 at about 25 nmol/mg protein concentration could cause the inhibition of

Phospholipases and brain swelling (Na + + K+)-ATPase. The present data thus confirm the detrimental effects of the polyunsaturated fatty acids on (Na ÷ + K+)-ATPase activity in both brain slice homogenates and synaptosomal preparations (Chan et al., 1983a). Furthermore, bovine serum albumin, when co-incubated with polyunsaturated fatty acids with a molar ratio of 5 or less (BSA/PUFA) inhibited the edema-inducing effects of polyunsaturated fatty acids (Chan et al., 1980), indicating the detrimental effects of P U F A on tissue swelling in brain slices. However, our data did not exclude the possible detrimental effects of lysophospholipids on (Na + + K ÷ ) - A T P a s e activity. Nevertheless, we have used lysophosphatidylcholine and lysophosphatidylethanolamine that resulted from the PLA 2 (18 units/ml) hydrolysis of commercial phosphatidylcholine and phosphatidylethanolamine and have found that these lysophospholipids were ineffective in inducing brain slice swelling and cation alteration (data not shown). On the other hand, PLC activated both ( N a + + K÷)-ATPase and K ÷activated p-nitrophenylphosphatase significantly. The mechanisms of the PLC-mediated stimulation of the Na+-pump is not clear. Following 60min of incubation, PLC caused an accumulation of 5.8 nmol of polyunsaturated fatty acids (20:4 plus 22:6) per mg protein in brain membranes. This concentration is two orders of magnitude lower than the concentration required to inhibit (Na ÷ + K ÷ ) - A T P a s e in synaptic membranes (Chan et al., 1983a). The activation of (Na ÷ + K+)-ATPase by PLC may be due to the exposure of the active site of the enzyme resulting from limited PC hydrolysis or due to the accumulation of D G and other lipid products. These possibilities warrant further investigation. In summary, our studies present evidence that PLA2 and PLC differentially affect both deacylation and incorporation of 20:4 in brain membrane and membrane-bound ( N a ÷ + K÷)-ATPase activity. We also demonstrated that the severe degradation of membrane phospholipids and the accumulation of high concentrations of polyunsaturated fatty acids might result in cellular injury and the development of edema. Since the mode of action of endogenous phospholipases might behave differently, both biochemically and functionally, as compared to the bee venom and bacterial enzymes, further studies are warranted to elucidate these differences.

Acknowledgements--This investigation was supported by

N.I.H. grant NS-14543. We thank Mrs Sheryl Colwell for editing the manuscript.

309 REFERENCES

Agardh C. D. and Siesj6 B. K. (1981) Hypoglycemic brain injury: phospholipids, free fatty acids and cyclic nucleotides in the cerebellum of the rat after 30 and 60 minutes of severe insulin-induced hypoglycemia. J. Cerebral. Blood Flow Metab. 1, 267-275. Bazan N. G. (1970) Effects of ischemia and electroconvulsive shock on free fatty acid pool in the brain. Biochim. biophys. Acta 218, 1-10. Bazan N. G. and Turco E. B. R. (1980) Membrane lipids in the pathogenesis of brain edema-phospholipids and arachidonic acid, the earliest membrane components changed at the onset of ischemia. Adv. Neurol. 28, 197-205. Chan P. H. and Fishman R. A. (1978) Brain edema: induction in cortical slices by polyunsaturated fatty acids. Science, N.Y. 201, 358-360. Chan P. H., Fishman R. A., Lee J. and Quan S. C. (1980) Arachidonic acid-induced swelling in incubated rat brain cortical slices, effect of bovine serum albumin. Neurochem. Res. 5, 629-640. Chart P. H., Yurko M. and Fishman R. A. (1982) Phospholipid degradation and cellular edema induced by free radicals in brain cortical slices. J. Neurochem. 38, 525-53 I. Chan P. H., Kerlan R. and Fishman R. A. (1983a) Reductions of GABA and glutamate uptake and ( N a + + K+)-ATPase activity in brain slices and synaptosomes by arachidonic acid J. Neurochem. 40, 309-316. Chan P. H., Longar S. and Fishman R. A. (1983b) Phospholipid degradation and edema development in coldinjured rat brain. Brain ICes. 277, 329-337. Chan P. H., Fishman R. A., Caronna J., Schmidley J. W., Prioleau G. and Lee J. (1983c) Induction of brain edema followingintracerebral injection of arachidonic acid. Ann. Neurol. 13, 625~32. Chan P. H., Fishman R. A., Chen S. and Chew S. (1983d) Effects of temperature on arachidonic acid-induced cellular edema and membrane perturbation in rat brain cortical slices. J. Neurochem. 41, 1550-1557. Chan P. H. and Fishman R. A. (1984) The role of arachidonic acid in vasogenic brain edema. Fed. Proc. 43, 210-213. Chan P. H., Schmidley J. W., Fishman R. A. and Longar S. (1984a) Brain injury, edema and vascular permeability changes induced by oxygen-derived free radicals. Neurology, Cleveland 34, 315-320. Chan P. H., Fishman R. A., Schmidley J. W. and Chen S. (1984b) Release of polyunsaturated fatty acids from phospholipids and alteration of brain membrane integrity by oxygen-derived free radicals. J. Neurosci. Res. 12, 595~505. Chan P. H. and Fishman R. A. (1985) Free fatty acids, oxygen free radicals and membrane alterations in brain ischemia and injury. In: Cerebral Diseases, 14th Research (Princeton-Williamsburg) Conference (Plum F. and Pulsinelli W. A. eds), pp. 227-235. Raven Press, New York. Edgar A. D. and Freysz L. (1982) Phospholipase A2 activities of rat brain cytosol. Occurrence of phospholipase C activity with phosphatidylcholine. Biochirn. biophys. Acta 711, 224-228. Edgar A. D., Strosznajdar J. and Horrocks L. A. (1982)

310

PAK Hoo CHAN et al.

Activation of ethanolamine phospholipase A 2 in brain during ischemia. J. Neurochem. 39, 1111-I 116. Folch J., Lees M. and Stanley S. G. H. (1957) A sample method for the isolation and purification of total lipids from animal tissues. J. biol. Chem. 226, 497 509. Gardiner M., Nilsson B., Rehncrona S. and Siesj6 B. K. (1981) Free fatty acids in the rat brain in moderate and severe hypoxia. J. Neurochem. 36, 1500-1505. Kuehl F. A. and Egan R. N. (1980) Prostaglandins, arachidonic acid, and inflammation. Science, N.Y. 210, 978-984. Lunt G. G. and Rowe C. E. (1968) The production of unesterified fatty acid in brain. Biochim. biophys. Acta 152, 681493. Majewska M. D. and Sun G. Y. (1982) Activation of arachidonyl-phosphatidyl-inositol and phosphatidylcholine turnover by K+-evoked stimulation of brain synaptosomes. Neurochem. Int. 4, 427-433. Matsuzawa Y. and Hostetler K. Y. (1980) Inhibition of lysosomal phospholipase A and phospholipase C by chloroquine and 4,4'-Bis(diethylamino-ethoxy)ct,fl-diethyldiphenylethane. J. Biol. Chem. 255, 5190-5194. Pediconi M. F., Rodriguez de Turco E. B. and Bazan N. G. (1983) Effects of postdecapitation ischemia on the metabolism of [t4C] arachidonic acid and [14C] palmitic acid in the mouse brain. Neurochem. Res. 8, 835-845. Rehncrona S., Westerberg E., Akesson B. and Siesj6 B. K. (1982) Brain cortical fatty acids and phospholipids during the following complete and severe incomplete ischemia. J. Neurochem. 38, 84-93. Samuelsson B., Hammarstrom S. and Bogeat P. (1979)

Pathway of arachidonic acid metabolism. Adv. lnflam. Res. 1, 405-411. Samuelsson B. (1983) Leukotrienes: mediators of immediate hypersensitivity reactions and inflammation. Science, N.Y. 220, 568-575. Skou J. C. and Esmann M. (1979) Preparation of membrane bound and of solubilized (Na ÷ + K+)-ATPase from rectal glands of Squalus acanthias. Biochim. biophys. Acta 567, 436-444. Sleight R. and Kent C. (1980) Regulation of phosphatidylcholine biosynthesis in cultured chick embryonic muscle treated with phospholipase C. J. bioL Chem. 255, 10644-10650. Sun G. Y. and Foudin L. L. (1984) On the status of lysolecithin in rat cerebral cortex during ischemia. J. Neurochem. 43, 1081-1086. Tang W. and Sun G. (1982) Factors affecting the free fatty acids in rat brain cortex. Neurochem. Int. 4, 269-273. Van Den Bosch H. (1980) Intracellular phospholipase A 2. Biochim. biophys. Acta 604, 191-246. Wolfe L. S. (1982) Eicosanoids: prostaglandins, thromboxanes, leukotrienes and other derivatives of carbon-20 unsaturated fatty acids. J. Neurochem. 38, 1-14. Yau T. M. and Sun G. Y. (1974) The metabolism of [1J4C]-arachidonic acid in the neutral glycerides and phosphoglycerides of mouse brain. J. Neurochem. 22, 99-104. Yoshida S., Abe K., Busto R., Watson B. D., Kogure K. and Ginsberg M. D. (1982) Influence of transient ischemia on lipid-soluble antioxidants, free fatty acids and energy metabolites in rat brain. Brain Res. 245, 307-316.