v

Archives of Biochemistry and BiophysicsVol. 389, No. 1, May 1, pp. 57–67, 2001doi:10.1006/abbi.2001.2314, available online at http://www.idealibrary.com on

Melittin Exerts Multiple Effects on the Release of FreeFatty Acids from L1210 Cells: Lack of Selective Activationof Phospholipase A2 by Melittin

Sang Yoon Lee, Heung Soon Park, Soo Jae Lee, and Myung-Un Choi1

School of Chemistry and Molecular Engineering and Center for Molecular Catalysis, Seoul National University,Seoul 151-747, South Korea

Received September 28, 2000; published online April 5, 2001

t

nn

Melittin is known as a phospholipase A2 (PLA2) acti-ator, but the selectivity of its effect on PLA2 is uncer-

tain. We examined the selectivity of melittin effect onthe release of free fatty acids (FFAs) from L1210 cellsusing various inhibitors. A systemic lipid analysis byHPLC and GLC revealed that melittin induced releaseof various FFAs including saturated, monounsatu-rated, and polyunsaturated FFAs. Various PLA2 inhib-itors examined exerted only minimal effects on themelittin-induced arachidonic acid (AA) and palmiticacid (PAL) releases. Specific inhibitors of phosphati-dylinositol-phospholipase C (U73122) and diacylglyc-erol lipase (RHC80267) exerted significant inhibitoryeffects on both AA and PAL releases. These resultssuggest that melittin-induced FFA release is mostlikely due to multiple participations of various typesof lipases. Since BAPTA/AM, an intracellular Ca21 che-lator, did not influence the FFA release, the Ca21 in-fluxed by melittin appeared not to be a key factor forthe FFA release. The mimicking of the melittin-in-duced FFA release by digitonin, a membrane-perme-abilizing agent, implies that the membrane-perturb-ing action of melittin is likely the cause of the FFArelease. Melittin also induced release of multiple FFAsfrom other cell lines including P388D1 and HL60. Therapid melittin-stimulated phospholipase D (PLD) ob-served in L1210 cells appeared not directly related tothe steady release of FFA, as indicated by the fact thatthe PLD was not blocked by RHC80267. In view ofmelittin’s multiple effects on the composition of cellu-lar lipids, we conclude that melittin does neither ex-clusively release any single FFA nor selectively acti-vate PLA2 in L1210 cells. The problem of using melittinas a PLA2 activator is discussed. © 2001 Academic Press

1 To whom correspondence should be addressed. Fax: 182-2-871-

4927. E-mail: muchoi@snu.ac.kr.

0003-9861/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

Key Words: melittin; free fatty acids; phospholipaseA2; membrane perturbation; cellular lipid composi-ion; L1210 cells.

Melittin, an amphiphilic peptide (26 amino acid res-idues) isolated from honeybee Apis mellifera (1), isknown to exert a variety of membrane-perturbing ef-fects such as hemolytic and antimicrobial activity (2).Melittin also induces structural alterations of mem-branes including pore formation, fusion, and vesicula-tion (3–5). These morphological changes of membranesbrought about by melittin could be attributed to induc-tion of hormone secretion (6), aggregation of membraneproteins (7), and change of membrane potential (8).Furthermore, melittin stimulates various enzymes in-cluding G-protein (9), protein kinase C (10), adenylatecyclase (11), phospholipase C (PLC)2 (12, 13), and phos-pholipase D (PLD) (14–16). These diverse effects re-ported implicate that melittin exerts multiple effectson cellular functions.

2 Abbreviations used: AA, arachidonic acid; AACOCF3, arachido-yl trifluoromethyl ketone; BAPTA/AM, 1,2-bis(o-aminophe-oxy)ethane-N,N,N9,N9-tetraacetic acid tetra(acetoxymethyl) ester;

BEL, E-6-(bromomethylene)tetrahydro-3-(1-naphthalenyl)-2H-py-ran-2-one; CE, cholesterol ester; Cer, ceramide; Chol, cholesterol;CL, cardiolipin; DAG, diacylglycerol; EtBr, ethidium bromide;FAME, fatty acid methyl ester; FBS, fetal bovine serum; FFA, freefatty acid; LDH, lactate dehydrogenase; LPC, lysophosphatidylcho-line; LPE, lysophosphatidylethanolamine; MAG, monoacylglycerol;OA, oleic acid; PA, phosphatidic acid; PACOCF3, palmitoyl triflu-oromethyl ketone; PAL, palmitic acid; PC, phosphatidylcholine; PE,phosphatidylethanolamine; PEt, phosphatidylethanol; PG, phos-phatidylglycerol; PI, phosphatidylinositol; PLA2, phospholipase A2;PLC, phospholipase C; PLD, phospholipase D; PS, phosphatidylser-

ine; SM, sphingomyelin; TAG, triacylglycerol.

57

lP

soa

[

adcfP

kc1w

w2

m

w1r

ipw

iiJwctsO2

d

58 LEE ET AL.

In spite of diverse effects of melittin, this peptide hasbeen widely accepted as an activator for phospholipaseA2 (PLA2) in various intact cells. There are reportssuggesting that melittin induces release of arachidonicacid (AA) from Swiss mouse 3T3-4a fibroblasts (17),human fibroblasts (18), and ras-transformed NIH3T3cells (19) by stimulating PLA2 activity. Some experi-ments even used melittin to specifically stimulate en-dogenous PLA2. For example, melittin-induced AA re-lease was correlated with calpain activation in necroticrat hepatocytes (20), and the role of AA in mediatingthe effect of interferon-g on sphingomyelin (SM) hydro-lysis was mimicked by melittin treatment in HL60cells (21). However, there is also a significant numberof reports that cast some doubts on the selectivity ofmelittin as a PLA2 activator: apparent failure to corre-late melittin-induced AA released from AA-labeled ratislets with lysophosphatidylcholine (LPC) that waspresumably produced by PLA2 along with the AA re-ease (22); lack of antagonizing effect of quinacrine, aLA2 inhibitor, on the melittin-induced AA release in

pancreatic acinar cells (23). The action target for melit-tin appeared to be not the endogenous PLA2 but theendogenous PLC and triglyceride lipase in skeletalmuscle cells (24). Furthermore, a recent report sug-gested that AA produced in human leukemia U937cells by melittin might be due to other types of acylhy-drolases instead of PLA2 (16). All these discrepanciesurrounding the melittin effect on AA release in vari-us cell lines complicate the usage of melittin as anctivator for PLA2.In a series of experiments we observed that masto-

paran 7, another amphiphilic peptide similar to melit-tin, activated PLD and induced an extensive alterationof lipid composition in L1210 cells (25, 26). One keyfinding was that the free fatty acid (FFA) release in-duced by mastoparan 7 was not restricted to AA butincluded various types of FFAs (26). In view of themultiple releases of FFAs we extended the investiga-tion of the FFAs release to melittin, the most exten-sively characterized amphiphilic peptide. In particularour efforts had been focused on some biochemical as-pects of enzymes that might be involved in the FFAsrelease by melittin. We examined how the inhibitors ofvarious lipases, including PLA2 inhibitors, affect themelittin-induced FFAs release and carried out sys-temic analysis of the cellular lipids of L1210 cells be-fore and after melittin treatment. In this paper wedemonstrate the multiple effects of melittin on FFAsrelease and the lack of specific stimulation of PLA2 inL1210 cells. A broad effect on the composition of cellu-lar lipids brought about by melittin is also presented.

MATERIALS AND METHODS

Materials. [5,6,8,9,11,12,14,15-3H(N)]Arachidonic acid ([3H]AA),

9,10-3H(N)]palmitic acid ([3H]PAL), [1-14C]oleic acid ([14C]OA), and t

[4-14C]cholesterol ([14C]Chol) were purchased from Dupont NEN(Boston, MA). Synthesized melittin was obtained from Peptron (Tae-jon, Korea). RPMI 1640 and fetal bovine serum (FBS) were fromGibcoBRL (Gaithersburg, MD). Arachidonyl trifluoromethyl ketone(AACOCF3), palmitoyl trifluoromethyl ketone (PACOCF3), E-6-(bromomethylene)tetrahydro-3-(1-naphthalenyl)-2H-pyran-2-one(BEL), U73122 (1-[6-((17b-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl]-1H-pyrrole-2,5-dione), and 1,2-bis(o-aminophe-noxy)ethane-N,N,N9,N9-tetraacetic acid tetra(acetoxymethyl) ester(BAPTA/AM) were obtained from Calbiochem (San Diego, CA).RHC80267 (1,6-bis-(cyclohexyloximinocarbonylamino)hexane) wasfrom ICN (Costa Mesa, CA), and precoated TLC (silica gel 60F254)nd digitonin were from Merck (Darmstadt, Germany). Phosphati-ylethanol (PEt) was prepared from phosphatidylcholine (PC) usingabbage PLD according to the previous method (27). Standards foratty acid methyl esters (FAMEs) were from Supelco (Bellefonte,A). All other chemicals were from Sigma (St. Louis, MO).Cell culture and isotope labeling. L1210 (mouse lymphocytic leu-

emia), P388D1 (mouse lymphoma), and HL60 (human promyelo-ytic leukemia) cells were grown in Hepes (20 mM)-buffered RPMI640 medium supplemented with 10% (v/v) FBS. The confluent cellsere labeled with [3H]AA (0.5 mCi/ml), [3H]PAL (1.5 mCi/ml), or

[14C]OA (0.5 mCi/ml) for 3 h according to the previous method (14).Assay of FFA release. The isotope-labeled cells harvested wereashed three times with phosphate-buffered saline (137 mM NaCl,.7 mM KCl, 1.8 mM KH2PO4, and 10 mM Na2HPO4, pH 7.4) as

described previously (25) and were resuspended in an assay mediumcontaining 20 mM Hepes (pH 7.2), 137 mM NaCl, 2.7 mM KCl, 3 mMMgCl2, 2 mM CaCl2, 2 mM EGTA, and 1 mg/ml bovine serum albu-

in. The final assay mixtures (200 ml, 2 3 107 cells/ml) containing 5mM melittin were incubated for 20 min at 37°C. For inhibitor studies,inhibitors were added to the cells for 15 min before melittin wasadded to the assay mixture. At the end of the reactions, total lipidswere extracted by addition of 1 ml of chloroform/methanol (2/1) andthe FFA released was separated on a TLC plate using a heptane/diethylether/acetic acid (60/40/2) solvent system. The radioactivity oftotal lipids esterified by [3H]AA, [3H]PAL, or [14C]OA per 106 cells

as 64,000 6 2400 dpm, 120,000 6 20,000 dpm, and 122,000 62,000 dpm, respectively. The data were presented as percentage ofadioactivity in FFA with respect to the total lipid radioactivity.Assay of PLD activity. PLD activity was determined by measur-

ng PEt, the unique transphosphatidylation product of PLD in theresence of ethanol, as described previously (25). Cells were labeledith [3H]PAL and resuspended in the assay medium supplemented

with 1.5% ethanol. After melittin treatment for 20 min at 37°C,[3H]PEt produced was separated on a TLC plate using ethylacetate/isooctane/acetic acid/water (13/2/3/10).

Measurement of membrane permeability. Membrane permeabil-ity was determined by means of ethidium bromide (EtBr) influx andlactate dehydrogenase (LDH) release. The influx of EtBr was mea-sured as described previously (25). An aliquot of L1210 cells (5 3 106

cells) was resuspended in 2 ml of the assay medium supplementedwith 30 mM EtBr. In the presence of melittin the time-dependentnflux of EtBr was determined by measuring changes in fluorescencentensity at 37°C with a FP777 spectrofluorometer (Jasco, Tokyo,apan). The excitation wavelength was 360 nm and the emissionavelength was 580 nm. To measure the LDH released by melittin

ell-free supernatant of the melittin-treated L1210 cells was ob-ained by microcentrifugation (3000 rpm, 5 min). LDH activity of theupernatant was assayed according to the published method (28).ne milliliter of an assay solution containing 81.3 mM Tris (pH 7.2),03.3 mM NaCl, 187 mM NADH, and 1.5 mM pyruvate was added to

the supernatant (70 ml). After incubation for 10 min at 30°C, theecrease in absorbance at 340 nm, due to NADH, was determined as

he LDH release.

[

whmiPc

fcai

matL

59MULTIPLE RELEASES OF FREE FATTY ACIDS BY MELITTIN

HPLC analysis of cellular lipids. HPLC analysis was performedaccording to the method described previously (26). L1210 cells (2 3107 cells/ml) incubated with melittin for 20 min were quenched byaddition of 30 ml of chloroform/methanol (1/2) containing 200 mg ofdipalmitoyl phosphatidyl-N,N-dimethylethanolamine, 0.24 mCi of

3H]PAL, and 0.08 mCi of [14C]Chol as internal standards. Cellularlipids of L1210 cells were extracted according to the method of Blighand Dyer (29) and the extracted lipids were analyzed using Young-Lin M930 HPLC system (Young-Lin Instrument, Anyang, Korea)equipped with Waters Nova-Pak silica column (3.9 3 300 mm) andAlltech 500 evaporative light scattering detector (Alltech Associates,Deerfield, IL). The cellular lipids were sequentially separated bythree-step chromatography. Major phospholipids were separated inthe first step with a linear gradient from chloroform/methanol/trif-luoroacetic acid (80/20/0.05, v/v) to chloroform/methanol/water/trif-luoroacetic acid (60/34/6/0.05, v/v). Acidic lipids were analyzed in thesecond step using a gradient system from chloroform/methanol/am-monium hydroxide (90/10/0.5, v/v) to chloroform/methanol/water/am-monium hydroxide (75/22/2.7/0.5, v/v). The third step using a gradi-ent of 0–4% ethanol in n-hexane/chloroform (90/10, v/v) allowedseparation of neutral lipids. The mass of each lipid class was deter-mined from the peak area of the signals of evaporative light scatter-ing detector on the basis of a calibration curve obtained by parallelanalysis of appropriate authentic lipid standards.

GLC analysis of free fatty acids. FFAs analysis was performedaccording to the method described previously (26). FFAs were iso-lated by HPLC from melittin-treated or untreated L1210 cells asdescribed above and analyzed by GLC after esterification with BF3-methanol. The FFAs dissolved in 100 ml of n-hexane were treated

ith 1 ml of 10% (w/w) BF3-methanol. After the reaction mixture waseated at 60°C for 20 min, the reaction was stopped by addition of 0.5l of n-hexane and 0.5 ml of water. The resulting FAMEs extracted

nto the upper organic phase were analyzed by GLC using a Hewlett-ackard 6890 gas chromatograph equipped with a Supelco SP-2560apillary column (100 m 3 0.25 mm 3 0.2 mm film thickness) and a

flame ionization detector. Carrier gas was nitrogen and injector anddetector temperatures were set at 260°C. The oven temperatureincreased from 150 to 225°C at a 1.5°C/min rate. The composition ofFFAs in the sample was determined on the basis of a parallel anal-ysis of FAME standards.

RESULTS

Perturbation of the permeability of L1210 cells. Ef-ect of melittin on membrane permeability of L1210ells was investigated by measuring the influx of EtBr,fluorescent dye (Fig. 1A). The EtBr fluorescence rap-

dly increased when 5 mM melittin was added to aL1210 cell suspension. Half the maximum level of flu-orescence was reached within 250 s. No change influorescence was observed in the absence of melittin.Melittin also enhanced the release of LDH, a cytoplas-mic enzyme, more than threefold (Fig. 1B). This mag-nitude of LDH release was comparable with that ofdigitonin-drived LDH release, suggesting that melittinhad cytotoxic influence on L1210 cells by disruptingthe cell membrane integrity. The melittin-perturbedpermeability increase was frequently observed in var-ious intact cells. For example, melittin induced EtBrinflux in HL60 cells (30) and LDH release in primaryrat neurons (31).

Release of free fatty acids. Since melittin is known

as a PLA2 activator, the release of AA from L1210 cells

was examined using isotope-labeled cells with [3H]AA(Fig. 2A). The release of AA induced by melittin in-creased linearly during the 20-min incubation. Whencells were labeled with [3H]PAL, PAL was also releasedby melittin. In the absence of melittin, neither AA norPAL release was observed throughout the 20-min in-cubation. For further evaluation of the content of FFAreleased by melittin, HPLC-GLC analysis of all the

FIG. 1. Effect of melittin on membrane permeability. (A) EtBr (30mM) influx was measured in the presence and absence of 5 mM

elittin. Fluorescence was determined at 360 nm for excitation andt 580 nm for emission. (B) LDH released to extracellular superna-ant was measured as decrease of NADH absorbance at 340 nm.1210 cells were incubated for 20 min with or without 5 mM melittin.

As a positive control, the digitonin effect on LDH release was carriedout. The amount of digitonin used was 30 mg/ml.

FFA released was carried out without isotope labeling

imc

(tm

mudT

60 LEE ET AL.

(Fig. 2B). The composition of FAME derived from FFArevealed that melittin induced the release of various

FIG. 2. Effect of melittin on FFA release. (A) [3H]AA-labeledsquare) and [3H]PAL-labeled (circle) L1210 cells were incubated inhe presence (closed symbols) and absence (open symbols) of 5 mMelittin. The released [3H]AA and [3H]PAL were separated by TLC

and expressed as percentage of total lipid radioactivity. The datapresented are the means 6 SD for 3–4 measurements performed induplicate. (B) HPLC-GLC determination of FFAs released by melit-tin. L1210 cells were incubated for 20 min with or without 5 mMmelittin and the FFAs released were isolated from total cellularlipids by HPLC. After the isolated FFAs were esterified by BF3-

ethanol, the resulting FAMEs were analyzed by GLC as describednder Materials and Methods. The absolute amount of each FFA wasetermined on the basis of a parallel analysis of FAME standards.he data were normalized to nmol/107 cells and presented as

means 6 range of duplicate analysis.

types of FFAs. The total amount of FFAs increased

about 3-fold to a total of 12.1 nmol/107 cells. The largestincrease in the amount of FFA was 3.0 nmol of PAL per107 cells followed by 2.7 nmol of OA. It is noteworthythat a significant amount of saturated FFAs was re-leased by melittin (4.2 nmol/107 cells). The relativeincrease in FFA was highest for palmitoleic acid (6.8-fold), followed by myristic acid (5.6-fold), OA (3.7-fold),AA (3.2-fold), PAL and linoleic acid (2.8-fold), andstearic acid (1.8-fold). These multiple alterations ofFFA components observed strongly support that melit-tin induces release of not only AA but also variousother kinds of FFAs.

Partial inhibition of AA and PAL release by PLA2

inhibitors. In order to examine whether the endoge-nous PLA2 activity in the cells was responsible for theFFA release, the cells were preincubated with variousPLA2 inhibitors (50 mM each) prior to the addition of 5mM melittin (Fig. 3). Surprisingly, the melittin-in-duced FFA release was only partially blocked by all thePLA2 inhibitors examined. In AA-labeled cells (Fig. 3A)PACOCF3 (a known Ca21-independent PLA2 inhibitor),AACOCF3 (a known Ca21-dependent PLA2 inhibitor),and quinacrine (a general PLA2 inhibitor) exerted onlysmall inhibitory effects (less than 20%). Another PLA2

inhibitor, BEL, known as a Ca21-independent PLA2

inhibitor, showed a relatively significant inhibition ofthe AA release (approximately 40%). Similarly in PAL-labeled cells (Fig. 3B) only small inhibitory effects wereobserved with AACOCF3, PACOCF3, and quinacrinebut slightly more inhibitory effects with BEL. Al-though there were subtle differences in inhibitory ef-fects of the PLA2 inhibitors between AA and PAL re-leases, the ineffectiveness of PLA2 inhibitors for block-ng both FFAs releases implies that other lipases

ight be involved in the release of FFAs from L1210ells when the cells were stimulated by melittin.Involvement of other lipases besides PLA2 in melittin-

induced FFA release. As shown in Figs. 3A and 3B,the release of AA and PAL by melittin was not fullyblocked by PLA2 inhibitors. Therefore it was necessaryto search for other possible pathways involved in theFFA release. To test this assumption, some lipase in-hibitors such as U73122 (a phosphatidylinositol (PI)-PLC inhibitor), RHC80267 (a diacylglycerol (DAG)lipase inhibitor), and BEL were added before melittin(Fig. 4). Both AA (Fig. 4A) and PAL (Fig. 4B) releasesinduced by melittin were significantly inhibited byU73122 (approximately 30–40%) and RHC80267 (ap-proximately 50%). Three combinations of any two in-hibitors showed more inhibitory effects (approximately70–80%) on the FFA release, and combined pretreat-ment with all three inhibitors appeared to block theFFA release completely. These results suggest thatseveral lipases including PLA2 participate in the FFA

release by melittin in L1210 cells.

ea[

w

pm

61MULTIPLE RELEASES OF FREE FATTY ACIDS BY MELITTIN

Effect of Ca21 on melittin-induced AA and PAL re-lease. Melittin has been known to evoke large influxof extracellular Ca21, leading to an elevation of intra-cellular Ca21 (19). To investigate the effect of Ca21 onFFA release by melittin, the melittin-induced FFAsreleases were carried out under conditions of differentCa21 concentrations (Fig. 5). The concentration of freextracellular Ca21 in the standard assay medium

FIG. 3. Effects of various PLA2 inhibitors on the melittin-inducedFFA release. [3H]AA-labeled (A) and [3H]PAL-labeled (B) L1210 cells

ere pretreated for 15 min with various PLA2 inhibitors at a con-centration of 50 mM each. Then the cells were incubated with 5 mMmelittin for 20 min. The FFAs released were separated by TLC. Themeasured radioactivities of [3H]AA and [3H]PAL were expressed as

ercentage of total lipid radioactivity. The data presented are theeans 6 SD for three to four measurements performed in duplicate.

which contains 2 mM Ca21 along with 2 mM EGTA was s

estimated to be approximately 19.3 mM by a computerprogramming method (32). In the presence of extracel-lular Ca21 there was a fourfold increase in AA releaseby melittin. However, in the absence of extracellularCa21 the melittin effect on AA release was found to beminimal (50% increase) (Fig. 5A). On the other hand,PAL was released by melittin even in the assay me-dium containing EGTA without Ca21 (Fig. 5B). Unlike

FIG. 4. Effects of various lipase inhibitors on FFA release by melit-tin. [3H]AA-labeled (A) and [3H]PAL-labeled (B) L1210 cells werepretreated for 15 min with U73122, RHC80267, and BEL at a con-centration of 50 mM each. After addition of 5 mM melittin the finalssay mixtures were incubated for 20 min. The radioactivities of

3H]AA and [3H]PAL released were measured and presented as de-

cribed in the legend of Fig. 3.

pmc

Fe

62 LEE ET AL.

the AA release, the PAL released by melittin did notseem to be dependent on the presence of free extracel-lular Ca21. To distinguish probable elevation of intra-

FIG. 5. Effect of extracellular Ca21 on FFA release by melittin.[3H]AA-labeled (A) and [3H]PAL-labeled (B) L1210 cells were sus-

ended in an assay medium containing 2 mM Ca21 with or without 2M EGTA. The cells were also suspended in an assay medium

ontaining no Ca21 but with 2 mM EGTA. Each suspension wasincubated for 20 min in the presence and absence of 5 mM melittin.For comparison, a complete set of parallel experiment was carriedout with 10 mM A23187, a Ca21 ionophore. To examine the intracel-lular Ca21 effect the cells in the standard medium (containing 2 mMCa21 and 2 mM EGTA) were pretreated for 15 min with 50 mMBAPTA/AM, an intracellular Ca21 chelator. The radioactivity of each

FA released was measured and presented as described in the leg-nd of Fig. 3.

cellular Ca21 concentration by influx from the extracel-

lular Ca21, the cells were preincubated with BAPTA/AM, an intracellular Ca21 chelator, at a concentrationof 50 mM in the standard assay medium (Figs. 5A and5B). Neither AA nor PAL release by melittin was af-fected by BAPTA/AM. For comparison the effect ofA23187 (10 mM), a Ca21 ionophore, was examined withregard to Ca21 dependence of the FFA release. Anincreased release of AA (4.5-fold) and PAL (2.4-fold) byA23187 could be observed in the assay medium con-taining 2 mM Ca21 without EGTA (Figs. 5A and 5B).However, neither AA nor PAL released by A23187 wasobserved under any other conditions including thestandard assay medium. The requirement of 2 mMCa21 for the A23187-dependent AA and PAL releasewas distinct from the melittin-induced FFA releaseobserved in the standard assay medium containingonly a small amount (19.3 mM) of Ca21. To confirm theeffects of the PLA2 inhibitors on the melittin-inducedAA and PAL release, the inhibition study was repeatedusing an assay medium containing 2 mM Ca21 withoutEGTA. The PLA2 inhibitors exerted only partial block-ing of the AA and PAL release even in the presence of2 mM Ca21 (data not shown) like in the standard assaymedium (Fig. 3).

Melittin-induced release of PAL, OA, and AA in othercell lines. Effect of melittin on FFA release was fur-ther investigated in other cell lines including P388D1and HL60 cells (Fig. 6). For this purpose each cell linewas labeled with [3H]PAL, [14C]OA, or [3H]AA andtreated with 5 mM melittin. As expected, releases ofcorresponding FFAs by melittin were found in each cellline. Similar to L1210 cells, P388D1 and HL60 cellsshowed two- to threefold enhancement of respectiveFFA release upon melittin treatment. These observa-tions imply that the melittin-induced release of variousFFAs is common among leukocytes.

Effect of digitonin on L1210 cell membranes. Digi-tonin, a steroid glycoside, has been used as a mem-brane-permeabilizing agent to introduce nonperme-able molecules into cells. Upon digitonin treatmentcells usually take up EtBr and release various cyto-plasmic contents including LDH (33). We confirmedthe large increase in EtBr permeability (25) and LDHrelease (Fig. 1B) in digitonin-treated L1210 cells. Thedigitonin effects on EtBr influx and LDH release wereanalogous to the melittin effects observed in Fig. 1.When effect of digitonin on FFA release was examined,it was found that digitonin increased AA and PALrelease as expected (Fig. 7). The increase in AA andPAL was 5.4-fold and 3.5-fold, respectively. The DAGlipase inhibitor, RHC80267, also inhibited the digito-nin-induced FFAs releases (approximately 50%). Thesimilarity of digitonin and melittin effects on FFA re-lease suggests that the membrane-perturbing action of

melittin is similar to that of digitonin.

o

m C.

63MULTIPLE RELEASES OF FREE FATTY ACIDS BY MELITTIN

Lack of link between FFA release and the melittin-stimulated PLD activity. The PLD in L1210 cells wasreported to be activated when melittin or unsaturatedFFAs such as OA were exogenously added (14). Be-cause melittin induced the release of unsaturatedFFAs including OA (Figs. 2B and 6), a possible rela-tionship between unsaturated FFA release and melit-tin-stimulated PLD was investigated. The PLD activ-ity, ascertained by formation of PEt in the presence ofethanol (34), was found to be strongly stimulatedwithin 5 min (more than sevenfold increase over thecontrol) by melittin (Fig. 8A). This rapid formation ofPEt is quite different from the steady increase of AAand PAL release during the 20-min incubation periodexamined (Fig. 2A). When the effect of various PLA2

inhibitors on melittin-stimulated PLD activity was ex-amined, no major inhibitory effect was observed (Fig.8B). Inhibitory effects found were 24, 12, and 7% forAACOCF3, PACOCF3, and quinacrine, respectively. In-terestingly, RHC80267, which significantly blocked therelease of FFAs (Fig. 4), enhanced the melittin-inducedPEt formation. These results strongly suggest that themelittin-stimulated PLD activity is hardly affected bythe FFA release induced by melittin.

Broad alteration of lipid components of L1210 cellsby melittin. Multiple effect of melittin on cellular lip-ids of L1210 cells was further confirmed by determin-ing the amount of each lipid component. The mass ofeach lipid in the absence and presence of melittin wasanalyzed by HPLC and listed in Table I. The content ofPC was the highest among the lipid components incontrol cells (44.6%, w/w total) followed by phosphati-dylethanolamine (PE, 16.1%) and PI (9.2%). The con-tents of Chol, cholesterol ester (CE), and triacylglycerol

FIG. 6. Effect of melittin on FFA release in L1210, P388D1, and HLrespectively, and resuspended in the assay medium. After 20 min in

elittin the radioactivity of each FFA released was separated by TL

(TAG) were nearly 6.0 to 7.0% and those of cardiolipin/

phosphatidylglycerol (CL/PG) and phosphatidylserine(PS) were approximately 3.0%. The proportions of lipidmetabolites such as lysophosphatidylethanolamine(LPE), FFA, and DAG were approximately 1.0% each.Ceramide (Cer), monoacylglycerol (MAG), LPC, phos-phatidic acid (PA), and PEt were minor lipid compo-nents. After melittin treatment, an extensive alter-ation of the lipid composition was observed. In partic-ular, the increase in FFA and PEt presented a strikingcontrast to the decrease of PC and PE. The increase inFFA and PEt upon melittin treatment was 2.20 and2.25 mg, respectively, per 107 cells. In contrast 5.04 mgf PC (9.5% of control PC) and 1.04 mg of PE (5.4% of

control PE) were decreased by melittin per 107 cells.Slight increases in PA, CL/PG, DAG, MAG, LPE, andCer were also observed with minor decreases in Chol,CE, PS, and TAG. Changes in LPC and SM were neg-ligible. The melittin-induced FFA release was not af-fected by the presence of 1.5% ethanol added to mea-sure the PLD activity. The alteration of lipid composi-tion brought about by melittin is very similar to that bymastoparan 7 reported previously (26).

DISCUSSION

The observation of melittin-induced EtBr influx (Fig.1A) and LDH release (Fig. 1B) confirms that melittinperturbs the L1210 cellular membranes. The mem-brane-perturbing effect of melittin has been observedin various mammalian cells as hemolysis of humanerythrocytes (4), EtBr influx in HL60 cells (30), andLDH release in primary cortical neuronal cultures (31).Under the membrane-perturbing conditions (Fig. 1),melittin induced release of not only AA but also various

cells. Each cell line was labeled with [3H]PAL, [14C]OA, and [3H]AA,bation in the presence (filled bar) and absence (empty bar) of 5 mMThe data are presented as described in the legend of Fig. 3.

60cu

types of FFAs including saturated FFAs (myristic acid,

Eu1

w

64 LEE ET AL.

PAL, and stearic acid), monounsaturated FFAs (palmi-toleic acid and OA) as well as other polyunsaturatedFFA (linoleic acid) (Fig. 2B). The portion of releasedPAL to the total amount of released FFAs was thehighest (38.5%), whereas that of AA was the lowest(1.7%), though the relative increases of AA (3.2-fold)and PAL (2.8-fold) over the control were almost equalbetween them. The combined amount of saturatedFFAs released including PAL, stearic acid, and myris-tic acid was 53.8% of the total FFAs released. The

FIG. 7. Effect of digitonin on FFA release. [3H]AA-labeled (A) and[3H]PAL-labeled (B) L1210 cells were incubated for 20 min in thepresence of 30 mg/ml digitonin. For inhibition study the labeled cells

ere pretreated with 50 mM RHC80267 for 15 min prior to theaddition of digitonin. The radioactivities of [3H]AA and [3H]PALwere measured and presented as described in the legend of Fig. 3.

multiple release observed strongly implies that melit-

tin exerts multiple effects on the FFA release fromL1210 cells. The releases of PAL, OA, and AA obtainedfrom labeled P388D1 and HL60 cells also support themultiple effects of melittin on the FFA release (Fig. 6).Recently, release of various FFAs by melittin in ratcerebral cortex (35) and in trophocytes of Periplaneta

FIG. 8. Effect of melittin on PLD activity in L1210 cells. (A) PEtformation was measured with or without 5 mM melittin in the pres-ence of 1.5% ethanol. The cells were labeled with [3H]PAL for 3 h. (B)

ffects of various PLA2 inhibitors and RHC80267 on melittin-stim-lated PLD activity. The [3H]PAL-labeled cells were pretreated for5 min with various inhibitors at a concentration of 50 mM each.

Then the cells were incubated for 20 min with or without 5 mMmelittin. PEt produced was separated by TLC. The measured radio-activity of [3H]PEt was expressed as percentage of total lipid radio-activity. The data presented are the means 6 SD for three measure-

ments performed in duplicate.

A

PPLPSLS

wdbM

65MULTIPLE RELEASES OF FREE FATTY ACIDS BY MELITTIN

americana (36) was also ascertained by HPLC analy-sis. Thus multiple effect of melittin on FFA releaseappears to occur in broad types of cell lines.

Because melittin is generally presumed to be a PLA2

activator (17–19) the FFAs released in L1210 cells bymelittin could be attributed to the PLA2 activity. How-ever ineffectiveness of PLA2 inhibitors to reduce themelittin-induced AA and PAL release (Fig. 3) castsdoubt on the specificity of melittin effect on activationof PLA2, though slight inhibitions were observed by

ACOCF3 and BEL. On the other hand the significantinhibitions observed with other lipase inhibitors suchas U73122 or RHC80267 implicate that other lipasesbesides PLA2 could be involved in the melittin-inducedFFA release in L1210 cells (Fig. 4). Relatively smallerbut significant increases of DAG and MAG (Table I)further imply a possible participation of PI–PLC–DAGlipase pathway in the FFA release. Similar results ofPI–PLC activations by melittin were reported in ratliver tissue (12) and in human HaCaT keratinocytes(13). The lack of selective stimulation of PLA2 activityby melittin was further supported by the fact that theHPLC analysis revealed only a small amount of lyso-phospholipids, and presumed products of PLA2 activity(Table I). If melittin specifically stimulated PLA2 ac-tivity in L1210 cells, the mass increase of LPC and LPE

TABLE I

Determination of Lipid Components of L1210 Cellsbefore and after Melittin Treatment

Lipids Control (mg/107 cells) Melittin (mg/107 cells)

CE 7.32 6 0.03 7.07 6 0.07TAG 7.04 6 0.59 6.82 6 0.07Chol 8.51 6 0.09 8.25 6 0.11DAG 0.86 6 0.07 1.19 6 0.08Cer 0.27 6 0.05 0.41 6 0.22MAG 0.22 6 0.11 0.38 6 0.05PEt 0.03 6 0.01 2.28 6 0.04FFA 1.23 6 0.09 3.43 6 0.08CL/PG 3.99 6 0.21 4.37 6 0.15PA 0.20 6 0.05 0.62 6 0.01PI 10.95 6 0.12 11.38 6 0.13

S 3.66 6 0.01 3.41 6 0.16E 19.20 6 0.26 18.16 6 0.82PE 1.24 6 0.08 1.38 6 0.06C 53.08 6 0.57 48.04 6 1.61M 0.92 6 0.03 0.84 6 0.01PC 0.21 6 0.02 0.27 6 0.03um 118.93 6 0.58 118.30 6 0.52

Note. L1210 cells (2 3 107/ml) were incubated for 20 min with orwithout 5 mM melittin in the presence of 1.5% ethanol. Cellular lipids

ere separated by HPLC and the mass of each lipid class wasetermined by means of evaporative light scattering detector on theasis of a parallel analysis of lipid standards as described underaterials and Methods. The data are presented as means 6 SD of

triplicate analysis.

(0.20 mg/107 cells) would correspond to that of FFAs

(2.20 mg/107 cells). However, it is still possible that thelysophospholipids generated by PLA2 stimulated bymelittin may not accumulate but may be rapidly hy-drolyzed further by lysophospholipase. There was asimilar observation that melittin caused production ofFFAs and DAG in the absence of lysophospholipid for-mation in primary cultures of skeletal muscle cells(24). Therefore, the inhibition studies in conjunctionwith lipid analysis suggest that the melittin-inducedFFA release is probably due to multiple participationsof various types of lipases.

The insensitivity of BAPTA/AM, an intracellularCa21 chelator, to melittin-induced AA and PAL releaseunder the standard assay medium containing 2 mMCa21 with 2 mM EGTA implies that the Ca21 influxitself cannot influence the FFA release (Fig. 5). Incontrast the A23187-linked FFA release was observedonly in the presence of 2 mM Ca21. However, there wasa subtle difference between the mode of melittin-in-duced AA release and that of PAL release in their Ca21

effects. The AA release was diminished in the absenceof Ca21, whereas the PAL release was still sustainedwithout Ca21. This observation implies that the lipaseinvolved in AA release could be different from the oneinvolved in PAL release. The finding that digitonin, acommonly used membrane-permeabilizing agent,could mimic the melittin-induced FFA release (Fig. 7)supports that the perturbing action of melittin (Fig. 1)is presumably the cause of the FFA release. There is alarge number of reports claiming that FFA release isinduced by membrane-perturbing chemicals in variousmammalian cells, including digitonin-treated pituitarycorticotrope tumor AtT-20 cells (37) and streptolysinO-treated rat submandibular ductal cells (38). Othercytotoxic effects of melittin on cells such as K1 efflux inhuman erythrocytes (4) and release of g-aminobutyricacid in mouse brain synaptosomes (39) have also beenspeculated due to its perturbing effect on cellular mem-branes. Recently the perturbing action of melittin onlipid bilayer membranes has been extensively charac-terized with model vesicles (40, 41).

The FFAs produced by melittin might further affectcellular functions of L1210 cells. Recently PAL wasreported to modulate the regulation of cholesterol syn-thesis in human keratinocytes (42) and the gene ex-pression of insulin in pancreatic beta cells (43), apartfrom well known AA effects such as inductions of reac-tive oxygen species in bovine heart mitochondria (44)and apoptosis in HepG2 cells (45). OA was also knownto promote influenza hemagglutinin-mediated mem-brane fusion (46) and palmitoleic acid behaved as a gapjunction uncoupler (47). Therefore diverse effects ofmelittin on cell physiology of L1210 cells in relationwith the released FFAs remain to be determined. TheHPLC analysis of L1210 cellular lipids clearly showed

that melittin exerted its effect on cellular lipids beyond

1

1

1

1

1

1

1

11

12

2

222

2

2

2

2

2

3

3

66 LEE ET AL.

the release of various types of FFAs (Table I). The lipidmediators produced by melittin other than FFAs suchas PA, Cer, DAG, and MAG are also known to becrucial for cell physiology. Accordingly the conversionof phospholipids to corresponding lipid mediators islikely to be mediated by various lipid-metabolizing en-zymes that were stimulated by melittin. In other wordsthe extensive alteration of lipid composition broughtabout by melittin in L1210 cells could be a multiplemanifestation of membrane perturbation-activatedlipases embedded in cellular membranes. The differenttypes of cellular lipids produced may exert diverseeffects on cell physiology. Actually a great number ofdiverse melittin actions have been observed in variousintact cell lines (6–11, 13, 14).

The fact that melittin may not exclusively releasethe AA or selectively activate PLA2 in various intactcells implies that numerous studies using melittin as aselective PLA2 activator might have misled their re-sults. If they employed only one labeled fatty acid,labeled AA for example, to label the cells of their in-terest they could only observe the AA release. There-fore the melittin-induced AA release should be consid-ered carefully. For example, a participation of PLA2

pathway was suggested based on the melittin treat-ment in reduced 5-HT1B receptor response in CHO cells(48). Recently the release of acetylcholine by melittinfrom rat hippocampus (49) and induction of apoptosisby melittin in human retinoblastoma Y79 cells (50)were also claimed to be the result of selective activationof PLA2. The questions on PLA2 involvement in theseexamples are brought on the basis of the fact thatmelittin exerted multiple effects on cellular lipids in-cluding various types of FFAs (Table I and Fig. 2).Production of various lipid metabolites could lead to abroad alteration of the lipid bilayer structure and con-sequently affect cellular functions in a variety of dif-ferent ways. Therefore it does not seem to be reason-able to correlate the multiple effect of melittin to a fewspecific cellular functions. Considering the complexityof melittin effect, we suggest that the selectivity ofmelittin is considered carefully before it is used as aPLA2 activator. Systemic analysis of cellular lipidscould be useful in resolving the complicated issuesregarding the cellular functions of melittin.

ACKNOWLEDGMENTS

This work was supported by the Brain Korea 21 Program and byKorea Science and Engineering Foundation (KOSEF) through theCenter for Molecular Catalysis at Seoul National University. Wethank Dr. H.-J. Kim for his careful comments on the manuscript.

REFERENCES

1. Eisenberg, D. (1984) Annu. Rev. Biochem. 53, 595–623.

2. Wade, D., Boman, A., Wahlin, B., Drain, C. M., Andreu, D.,

Boman, H. G., and Merrifield, R. B. (1990) Proc. Natl. Acad. Sci.USA 87, 4761–4765.

3. Dempsey, C. E. (1990) Biochim. Biophys. Acta 1031, 143–161.4. Katsu, T., Kuroko, M., Morikawa, T., Sanchika, K., Fujita, Y.,

Yamamura, H., and Uda, M. (1989) Biochim. Biophys. Acta 983,135–141.

5. Ladokhin, A. S., Selsted, M. E., and White, S. H. (1997) Biophys.J. 72, 1762–1766.

6. Kiesel, L., Rabe, T., Hauser, G., Przylipiak, A., Jadali, F., andRunnebaum, B. (1987) Mol. Cell Endocrinol. 51, 1–6.

7. Hui, S. W., Stewart, C. M., and Cherry, R. J. (1990) Biochim.Biophys. Acta 1023, 335–340.

8. Carrasquer, G., Li, M., Yang, S., and Schwartz, M. (1998) Bio-chim. Biophys. Acta 1369, 346–354.

9. Mahady, G. B., Liu, C., and Beecher, C. W. (1998) Phytochemis-try 48, 93–102.

0. Mau, S. E., and Vilhardt, H. (1997) J. Neurochem. 69, 762–772.

1. Knowles, B., and Farndale, R. W. (1988) Biochem. J. 253, 235–241.

2. Haber, M. T., Fukui, T., Lebowitz, M. S., and Lowenstein, J. M.(1991) Arch. Biochem. Biophys. 288, 243–249.

3. Haase, I., Czarnetzki, B. M., and Rosenbach, T. (1996) Exp.Dermatol. 5, 84–88.

4. Lee, S. Y., Yeo, E., and Choi, M.-U. (1998) Biochem. Biophys. Res.Commun. 244, 825–831.

5. Liu, S.-Y., Tappia, P. S., Dai, J., Williams, S. A., and Panagia, V.(1998) J. Mol. Cell Cardiol. 30, 1203–1214.

6. Saini, S. S., Chopra, A. K., and Peterson, J. W. (1999) Toxicon 37,1605–1619.

7. Shier, W. T. (1979) Proc. Natl. Acad. Sci. USA 76, 195–199.8. Argiolas, A., and Pisano, J. J. (1983) J. Biol. Chem. 258, 13697–

13702.9. Sharma, S. V. (1993) Oncogene 8, 939–947.0. Arora, A. S., de Groen, P., Emori, Y., and Gores, G. J. (1996)

Am. J. Physiol. 270, G238–G245.1. Visnjic, D., Batinic, D., and Banfic, H. (1997) Blood 89, 81–

91.2. Metz, S. A. (1986) Biochem. Pharmacol. 35, 3371–3381.3. Heisler, S. (1989) Can. J. Physiol. Pharmacol. 67, 411–416.4. Fletcher, J. E., Jiang, M. S., Gong, Q. H., and Smith, L. A. (1991)

Biochem. Cell Biol. 69, 274–281.5. Lee, S. Y., Park, N. G., and Choi, M.-U. (1998) FEBS Lett. 432,

50–54.6. Park, H. S., Lee, S. Y., Kim, Y. H., Kim, J. Y., Lee, S. J., and Choi,

M.-U. (2000) Biochim. Biophys. Acta 1484, 151–162.7. Jung, K., Koh, E., and Choi, M.-U. (1989) Bull. Korean Chem.

Soc. 10, 585–600.8. Doyle, A., Griffiths, J. B., and Newell, D. G. (1993) in Cell and

Tissue Culture: Laboratory Procedures (Doyle, A., Griffiths,J. B., and Newell, D. G., Eds.), pp. 4B:8.1-4B:8.4, Orbital Press,Letchworth, England.

9. Bligh, E. G., and Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37,911–917.

0. Midoux, P., Mayer, R., and Monsigny, M. (1995) Biochim. Bio-phys. Acta 1239, 249–256.

1. Clapp, L. E., Klette, K. L., DeCoster, M. A., Bernton, E., Petras,J. M., Dave, J. R., Laskosky, M. S., Smallridge, R. C., and

Tortella, F. C. (1995) Brain Res. 693, 101–111.

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

67MULTIPLE RELEASES OF FREE FATTY ACIDS BY MELITTIN

32. Brooks, S. P., and Storey, K. B. (1992) Anal. Biochem. 201,119–126.

33. Vitale, M. L., Castillo, R. D., and Trifaro, J.-M. (1992) J. Neuro-chem. 59, 1717–1728.

34. Liscovitch, M., and Chalifa, V. (1994) in Signal-Activated Phos-pholipases (Liscovitch, M., Ed.), pp. 31–63, CRC Press, Austin,TX.

5. Phillis, J. W., Song, D., and O’Regan, M. H. (1999) Brain Res.847, 270–275.

6. Ali, I., and Steele, J. E. (1997) Insect Biochem. Mol. Biol. 27,681–692.

7. Pompeo, A., Baldassare, M., Luini, A., and Buccione, R. (1999)Life Sci. 64, 837–846.

8. Kabre, E., Chaib, N., Boussard, P., Merino, G., Devleeschouwer,M., and Dehaye, J. P. (1999) Biochim. Biophys. Acta 1436, 616–627.

9. Gonzalez, L., Nekrassov, V., Castell, A., and Sitges, M. (1997)Neurochem. Res. 22, 189–199.

0. Benachir, T., and Lafleur, M. (1995) Biochim. Biophys. Acta

1235, 452–460.

1. Takei, J., Remenyi, A., and Dempsey, C. E. (1999) FEBS Lett.442, 11–14.

2. Siefken, W., Hoppner, H., and Harris, I. R. (2000) Exp. Dermatol.9, 138–145.

3. Jacqueminet, S., Briaud, I., Rouault, C., Reach, G., and Poitout,V. (2000) Metabolism 49, 532–536.

4. Cocco, T., Di Paola, M., Papa, S., and Lorusso, M. (1999) FreeRadical Biol. Med. 27, 51–59.

5. Chen, Q., Galleano, M., and Cederbaum, A. I. (1997) J. Biol.Chem. 272, 14532–14541.

6. Chernomordik, L. V., Leikina, E., Frolov, V., Bronk, P., andZimmerberg, J. (1997) J. Cell Biol. 136, 81–93.

7. Dhein, S., Krusemann, K., and Schaefer, T. (1999) Br. J. Phar-macol. 128, 1375–1384.

8. Berg, K. A., Maayani, S., and Clarke, W. P. (1996) Mol. Phar-macol. 50, 1017–1023.

9. Almeida, T., Cunha, R. A., and Ribeiro, J. A. (1999) Brain Res.826, 104–111.

0. Vento, R., D’Alessandro, N., Giuliano, M., Lauricella, M., Cara-

billo, M., and Tesoriere, G. (2000) Exp. Eye Res. 70, 503–517.