Chem.-Biol. Interactions, 82 (1992) 135-149

135

Elsevier Scientific Publishers Ireland Ltd.

MELITTIN-INDUCED ALTERATIONS IN DYNAMIC PROPERTIES OF HUMAN RED BLOOD CELL MEMBRANES

CEZARY WATALA and KRZYSZTOF GWO~DZII~ISKI

Department of Biophysics, University of LddS, ul. Banacha 12/16, 90-237 LSd5 (Poland) (Received September 25th, 1991) (Revision received January 10th, 1992) (Accepted January 12th, 1992)

SUMMARY

The interaction of bee venom melittin with erythrocyte membrane ghosts has been investigated by means of fluorescence quenching of membrane tryptophan residues, fluorescence polarization and ESR spectroscopy. It has been revealed that melittin induces the disorders in lipid-protein matrix both in the hydrophobic core of bilayer and at the polar/non-polar interface of melittin complexed with erythrocyte membranes. The peptide has been found to act most efficiently at the concentration of the order of 10 -l° mol/mg membrane protein. The apparent distance separating the membrane tryptophan and bound 1-anilino-8naphthalenesulphonate (ANS) molecules is decreased upon melittin binding, which results in a significant increase of the maximum energy transfer efficiency. Significant changes in the fluorescence anisotropy of both 1,6-diphenyl1,3,5-hexatriene and 1-anilino-8-naphthalenesulphonate bound to erythrocyte ghosts, which have been observed in the presence of melittin and crude venom, indicate membrane lipid bilayer rigidization. The effect of crude honey bee venom has been found to be of similar magnitude as the effect of pure melittin at the concentration of 10-10 mol/mg membrane protein. Using two lipophilic spin labels, methyl 5-doxylpalmitate and 16-doxylstearic acid, we found that melittin at its increasing concentrations induces a well marked rigidization in the deeper regions of lipid bilayer, whereas the effect of rigidization near the membrane surface maximizes at the melittin concentration of 10-lO moYmg (10 -4 tool melittin per mole of membrane phospholipid). The decrease in the ratio hw/hs of maleimide and the rise in relative rotational correlation time (re) of iodoacetamid spin label, indicate that melittin effectively immobilizes membrane proteins in the plane of the lipid bilayer. We conclude that melittin-induced rigidization of the lipid bilayer may induce a reorganization of lipid assemblies Carrespondene,e to: C. Watala, Department of Biophysics, University of LSd~., ul. Banacha 12/16, 90-237 L6dt, Poland. 0009-2797/92/$05.00 © 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland

136

as well as the rearrangements in membrane protein pattern and consequently the alterations in lipid-protein interactions. Thus, the interaction of melittin with erythrocyte membranes is supposed to produce local conformational changes in membranes, which are discussed in the connection with their significance during the synergistic action of melittin and phospholipase of bee venom on red blood cells.

Key words: Melittin -- Lipid-protein interaction -- Erythrocyte membrane Spin labelling -- Fluorescence polarization -- Fluorescence quenching

INTRODUCTION

Melittin, the main component of bee venom, is a biologically active hydrophobic peptide which can be inserted spontaneously into natural and synthetic membranes to destruct cells in micromolar concentrations. Melittin sufficiently destabilizes the plasma membrane of cultured cells to allow cell disruption in the absence of detergents with minimal homogenization [1]. On the other hand, it has been shown that melittin acts as an inhibitor of Hil-phase formation and as a stabilizer of the bilayer organization [2,3]. The remarkable variety of effects of melittin on the organization of different membrane phospholipids can be understood in a relatively simple model, based on the shape-structure concept of lipid polymorphism and taking into account the position of the peptide molecule with respect to the lipids [4]. The evidence gathered hitherto suggests that melittin has the ability to stimulate the release of membrane fragments out of the cells and this may bring about the perforation of molecules of small size, leading to a colloid-osmotic haemolysis [5]. The proposed action mechanism of melittin in red blood cells attributes the predominant accumulation of the peptide molecules in the outer half of the bilayer to the deforming of the erythrocyte cell into crenature. It is believed that a large accumulation makes the membrane structure unstable, resulting in the release of membrane fragments and the simultaneous enhancement of permeability [1,6,7,8]. Although numerous reports deal with melittin interaction with liposomes of artificial lipid vesicles, there are only a few papers which concern the effects of melittin on natural biological membranes. The present paper is concerned with the analysis of the obtained results on the effects of melittin on the dynamic properties of human red blood cell membranes, discussed in connection with the enhanced susceptibility of erythrocytes during the synergistic action of melittin and phospholipase [9,10]. MATERIALS AND METHODS

Preparation of red blood cell membranes Red blood cells washed four times with phosphate-buffered saline (pH 7.4)

137

were subjected to moderate haemolysis in Tris-HC1/EDTANa2 buffer (pH 7.0) according to Marchesi and Palade [11]. The isolated erythrocyte membranes were resuspended in ice-cold phosphate-buffered saline, used within a few hours and kept at 4°C until used. The protein content in erythrocyte membrane suspensions was measured according to Lowry et al. [12]. Incubation with melittin

Commercially available melittin (Sigma Co., USA; approx. 70% by HPLC) was purified by HPLC. The stock solution (10 -4 tool/l) was prepared in PBS, and the aliquots were subsequently added to membrane samples (2.5 mg/ml protein; approx. 2.3 mmol/1 membrane phospholipid [13]). Independently of whether the membrane samples were labelled with fluorescent or spin labels the ratio of melittin to membrane protein was maintained at 2 x 10 -11 mol/mg, 2 x 10 -1° mol/mg and 2 x 10-9 tool/rag, respectively. The above ratios corresponded to the amounts of melittin per mole of erythrocyte membrane phospholipid equal to 2.15 x 10 -5 mol, 2.15 x 10-4 mol and 2.15 x 10 .3 mol, respectively; the absolute melittin concentrations were, respectively, 0.05 ~mol/1, 0.5 #mol/1 and 5 tLmol/1. Crude venom extracts were prepared from honey bee venom sacs as described earlier [14], diluted in physiological saline (0.5 mg/ml) and added to membrane samples at the ratio of crude venom protein to membrane protein equal 2.5 tLg/mg (approx. 2.62 t~g/mol membrane phospholipid). Fluorescence measurements

The polarization of fluorescence reflects the mobility of the fluorescent molecules and the transfer of excitation energy between them. In the case of rigidly arranged, randomly oriented molecules, the value of the degree of polarization for linearly polarized excitation light depends only on the angle between the absorption and emission oscillators. As the extent of depolarization depends on the extent of the rotation, reflecting the mobility of the molecules surrounding the fluorescent probe, the fluorescence polarization technique can be used to determine the microviscosity of biological membranes [15,16]. In the present study two fluorescent probes were employed in order to determine erythrocyte membrane lipid fluidity: 1,6-diphenyl-l,3,5-hexatriene (DPH) and 1-anilino-8-naphthalenesulphonate (ANS) (Molecular Probes, Eugene, OR, USA). DPH solution (0.5 mmol/1 in tetrahydrofuran) was diluted 1:100 in phosphate-buffered saline (pH 7.4) and vigorously mixed immediately before use. One volume of this diluted DPH dispersion was added to 1 volume of the erythrocyte membrane suspension containing 200 #g/ml protein and the mixture was incubated at 37°C for 35 min. The final protein concentration was 100 #g/ml. Magnesium 1-anilino-8-naphthalenesulphonate (5 mmol/l in PBS) was added to a suspension of erythrocyte ghosts (100 ~g/ml protein) to a final ANS concentration of 25 tLmol/1. The binding of ANS with erythrocyte membranes was analyzed in the system containing membrane samples corresponding to 150 tLg/ml protein and increasing concentrations of ANS (10-80 t~moi/1) [17]. Steady-state fluorescence polarization was measured at 37°C with a PerkinElmer LS-5B spectrofluorometer equipped with polarizers in the excitation and emission beams. The excitation (2.5 nm slit) and emission (5 nm slit) wavelengths

138

were, respectively, 360 and 430 nm for DPH and 360 nm and 470 nm for ANS. Fluorescence polarization was determined using a standard formula from emission intensities that were polarized parallel and perpendicular to the direction of the polarized excitation [18]. The grating transmission factor of Chen and Bowman [19] was used to correct for the depolarization effect of grating monochromators [20]. A fluorescence intensity value for a non-labelled blank was subtracted as a correction for scattered light. Tryptophan fluorescence was excited at 295 nm and recorded at 333 nm. All fluorescence measurements were performed at a temperature of 23 ± 1 °C.

Spin label synthesis Methyl 5-doxylpalmitate was synthesized according to Hubbell and McConnell [21] and 16-doxylstearic acid was purchased from Sigma. Maleimide spin label (4-maleimido-2,2,6,6-tetramethylpiperidine-l-oxyl; MSL) was prepared according to Misharin and Polianovskij [22] and Gaffney [23]; iodoacetamide spin label (ISL) according to McConnell et al. [24].

Spin labelling of erythrocyte membranes Erythrocyte membranes were labelled by the introduction of the 100 mmol/i ethanol solution of spin labelled fatty acid esters to give a final spin label concentration of 5 × 10 -5 mol/1 and then the suspensions were incubated for 30 min at room temperature. The final ethanol concentration did not exceed 0.05% (v/v). In the case of maleimide spin label (2 ~l of 100 mmolfl MSL solution per 1 ml of membrane suspension approx. 3 mg/ml protein) the labelled samples were incubated for 1 h, whereas the incubation with iodoacetamide spin label (2 ul of 100 mmol/l ISL solution per I ml of membrane suspension approx. 3 mg/ml) was performed overnight. The unbound spin label was removed by four washings with PBS. It was checked that no ESR signal could be detected in the supernatant after the last wash. In all the ESR spectra, the ordinate was represented as the amplitude of an ESR signal and expressed in arbitrary units. ESR scannings were routinely recorded as the first derivatives of absorption spectra. The estimated ratios were calculated from the ESR graphs taking into account the relevant amplitudes measured as the heights of the respective peaks (expressed in metric units). ESR measurements were performed at ambient temperature (23 ± 1°C) in a SE/X-20 X-band spectrometer (Wroclaw Technical University, Poland).

Statistical analysis The normal distribution of data was checked out by means of Shapiro-Wilk's test. The data were elaborated by means of analysis of variance and Tukey's test assigned for paired comparisons [25]. RESULTS

Figure 1 shows the double-reciprocal plot of the decrease in the fluorescence intensity of the tryptophan vs. ANS concentration. The intercept of the ordinate

139

fluorescence

~o

"

c

(arbitrary

"

/ To

20

30

ao

6o

units)

,z__z2_

/

/

60

70

/

:7

80

14. I O 14. O tl.

0 *

0

|

.02

*

l

i

.04

|

*

.06

II(ANS)

i

.08

*

i

*

I

.I 0

[II#mo1 ]

Fig. 1. Doublereciprocalplot of fluorescenceintensityvs. ANS concentration.Fo and F denote the tryptophan fluorescencein the absence and presence of ANS, respectively.Figure inset: increasing fluorescenceof ANS at 470 nm and decreasingfluorescenceof tryptophanat 333 nm in the absence of melittin(speckledband), at the melittinconcentrationof 2 x 10- n tool/rag(diagonal-linedband) and 2 x 10-9 tool/rag(solid band) vs. ANS concentration.

gives the reciprocal maximum transfer efficiency which corresponds to a state of complete occupation of all ANS binding sites in the membrane [26]. The average values of the maximum transfer efficiency calculated for various melittin concentrations are presented in Table I. Melittin induced a significant increase of the maximum transfer efficiency, as revealed by analysis of variance and Tukey's test. The apparent interchromophore separation, R, which means the distance separating the membrane tryptophan and bound ANS molecules was determined from transfer efficiency (E) values [26]. The relative alterations of R with respect to control samples are given in Table I. During the action of melittin R diminishes, reaching its minimum at the melittin concentration of 2 × 10- io moYmg. The inset of Fig. 1 shows the descending fluorescence of tryptophan in the presence of increasing amounts of ANS. There was no significant shift of the emission maximum of tryptophan residues and only a little 4 nm shift in the emission maximum of ANS, when the ANS concentration was increased from 10 to 80 ~mol/l (from 466 nm to 470 nm in the absence of melittin and from

140 TABLE I ENERGY TRANSFER EFFICIENCY (E) AND THE RELATIVE APPARENT INTERCHROMOPHORE SEPARATION (R') BETWEEN ERYTHROCYTE MEMBRANE TRYPTOPHAN RESIDUES AND 1-ANILINO-8-NAPHTHALENESULPHONATE MOLECULES IN THE ABSENCE AND PRESENCE OF MELITTIN Mean ± S.D.; n = 6. The significance of differences is given as calculated by analysis of variance (aANOVA) and by Tukey's test (aT). Melittin concentration Parameter

~o (control)

P1 (2 x 10 -11 tool/rag)

P2 (2 x 10 -10 mol/mg)

P3 (2 × 10 -9 mol/mg)

Energy transfer efficiency (E) s T < 0.001 ahNOVA < 0.001 Relative interchromophore separation (R')

0.51 ± 0.04

0.88 ± 0.08

0.91 ~: 0.09

0.84 ± 0.06

~0

1.0

~

Pl

0.726

=

P2

0.681

=

P'3

0.765

464 nm to 468 nm in its presence). In the absence of ANS, fluorescence was emitted by tryptophan residues with maximum at 332 nm (hexc = 295 nm). When adding the increasing amounts of ANS the tryptophyl fluorescence band was quenched and a second peak appeared with a maximum at 466 nm. These data would indicate the alterations concerning ANS localization in membrane lipid bilayer, its interactions with membrane proteins and consequently lipid-protein organization, which arise when melittin interacts with erythrocyte membrane. Significant changes in the fluorescence anisotropy of both DPH and ANS bound to erythrocyte ghosts were observed in the presence of melittin and crude venom. Whereas a progressive decrease of rotational fluidity was revealed in the case of DPH molecules vs. melittin concentration, the ANS fluorescence polarization fluctuated in a specific way. It increased up to a concentration of 10-10 mol/mg, where the most pronounced rigidizing effect was recorded and fell again at the concentration of 10-9 mol/mg (Fig. 1). Although the reduction in the ANS fluorescence polarization was significant at 10 -9 mol/mg, it was remarkably smaller than at 10-l0 mol/mg. As the effect of crude honey bee venom is considered, it was of similar magnitude than the effect of pure melittin at the concentration of the order of 10-lo mol/mg. On the assumption that melittin constitutes approximately 50% of crude venom the aliquot of 2.5 ~g/mg of crude venom extract corresponds to 3.4 x 10-10 mol/mg of pure melittin. These results indicate a decreased rotational mobility of fluorescent probes employed in the present study upon the action of either pure melittin or crude bee venom extract.

141

The results obtained from the spectra of lipophilic spin labels are very compatible with those for fluorescent labels. In the case of 16-DS and Met 5-DP, the experimental parameter h+l/h0, where h+ 1 and h0 are the heights of low-field line and middle-line of the spectra (Fig. 2), respectively, was used as a semiquantitative measure of the carbon chain mobility [27]. These two spin labels are commonly used to monitor 'lipid fluidity' at two different depths in the hydrocarbon region of the lipid bilayer. The lower values of h~ l/h0 parameter indicate the decreased spin label mobility and correspond to the lower membrane 'lipid fluidity'. The progression of changes in 16-DS h . l/h0 is consistent with the changes monitored by DPH, whereas the pattern of changes revealed for 5-DP is very much alike the alterations found for ANS. Since different fluorescent and spin probes monitor different regions of lipid bilayer, one may argue that melittin induces the disorders in lipid-protein matrix both in hydrophobic core of bilayer and at the polar/non-polar interface of melittin complexed with erythrocyte membranes. It seems reasonably certain that melittin at its increasing concentrations induces a well marked rigidization in the deeper regions of lipid bilayer, whereas the effect of rigidization near the membrane surface maximizes at the melittin concentration of 10-10 mol/mg. In order to monitor the possible alterations in membrane protein conformation two protein spin labels, MSL and ISL, were employed. As shown in Fig. 3 the spectrum of MSL attached to erythrocyte membranes consists of two dominant

a

/

~

I

/Tll lho

f

h.i

Fig. 2. ESR spectra of methyl 5-doxylpalmitate (a) and 16-doxyl stearic acid (b) labelled human erythrocytes; h+ l and ho are the heights of low and middle field lines, respectively.

142

Fig. 3. ESR spectrum of 4-maleimido-2,2,6,6-tetramethylpiperidine-l-oxyl (MSL) attached to erythrocyte membrane ghosts; hw and hs correspond to weakly and strongly immobilized membrane MSL residues.

classes of spin label. The broad anisotropic spectrum arose from the subpopulation of label molecules embedded into strongly immobilized membrane environment, the narrower one in turn represents the subpopulation of weakly immobilized or more mobile residues of MSL. The ratio of amplitudes of low-field peaks of weakly {hw) to strongly (hs) immobilized fraction of MSL is regarded an indicator of the physical state of membrane proteins and, moreover, a sensitive measure of conformational state of proteins [28]. This ratio was reduced upon melittin action and this effect deepened as the concentration increased, reaching its maximum for the melittin concentration of 2 × 10 - 10 mol/mg. This suggests that melittin rigidizes membrane protein conformation in a concentrationdependent manner. This implies the immobilization of membrane protein moieties in lipid bilayer as confirmed by the rise in relative rotational correlation time (~¢) of ISL. Relative rotational correlation times ~c were calculated from the heights and widths of spin label hyperfine lines according to the equation [29]: re = 6.5 x 10 -1° x Wo [(ho/h-1) 1/2 - 11 where W0, ho and h_l are mid-field line width, mid-field line height and highfield line height, respectively (Fig. 4). Iodoacetamide spin label binds to the surface - SH groups of membrane proteins and is a sensitive probe of protein structure and conformation [28]. The higher values rc, the slower rotational mobility

143

ho h-1

I

Fig. 4. ESR spectrum of iodoacetamide spin label (ISL) attached to - SH groups of erythrocyte membrane proteins; W0, ho and h_ 1 are mid-field line width, mid-field line height and high-field line height, respectively.

of ISL, which implies the stronger immobilization of protein moieties. The significant elevation of rc was revealed at the melittin concentration of 2 x 10-10 tool/rag and this effect did not deepen while melittin concentration increased. The incubation of erythrocyte membrane ghosts with melittin induced the successive increase in the release of erythrocyte membrane phospholipids (from about 0.085% at the lowest up to approximately 4.9% at the highest peptide concentration). DISCUSSION

Melittin interaction with the structure of the lipid bilayer may lead to changes in the order and dynamics of lipids [30,31], formation of channels [32], membrane fusion [2,33], or alterations in membrane curvature [4]. Several models have been proposed for melittin arrangement in a lipid bilayer, which differed in the postulated depths of melittin penetration in the hydrophobic core. The evidence has been accomplished for the insertion of the melittin helix

144

perpendicular to the bilayer plane [34,35], with the possibility of traversing the bilayer by the peptide [34] and also parallel orientation has been suggested [35]. Although the applicability of any structural model may well depend on the lipid species concerned, the physical state of lipids or the protein-to lipid ratio, all the models are based on the assumption that melittin penetrates into the bilayer or, at least, interacts closely with the lipid interior of the bilayer [1]. The action of melittin can considerably influence membrane lipid order and membrane protein arrangement. It has an ability to stimulate the release of membrane phospholipid outside of cells resulting in an enhancement of permeability [5,8]. Our results indicate that melittin acts as a rigidizer on the membrane hydrocarbon core and effectively immobilizes membrane proteins in the plane of lipid bilayer. In fact, they remain consistent with the data reported hitherto and concerning mainly artificial lipid vesicles. Deuterium magnetic resonance revealed a conformational change of the phosphatidylcholine head group upon melittin binding [36]. The 1,6-diphenyl-l,3,5-hexatriene emission anisotropy results, which are found to be consistent with the wobble-in-cone model, show that melittin induces an increase in the order parameter, S, of the acyl chains of unilamellar vesicles below, at, and above their phase transition temperature [31]. At high protein-to-lipid ratios the data revealed the formation of small particles of melittin and dipalmitoylphosphatidylcholineand the disruption of membrane order in lipid bilayers [37]. The large increase in the bilayer area caused by the melittin inserted into erythrocyte membrane could result in local expansion of the outer leaflet of the membrane. In doing so, the peptide becomes intercalated between the phospholipid side chains. This supposedly results in an increase in the order parameter and the subsequent immobilization of membrane proteins, even at low melittin concentrations (Tables I -III). With the further increasing concentration of melittin the effect is markedly abolished. This remains compatible with the conclusions outlined by Katsu et al. [8]. As they reported, after membrane fragments have been expelled outside cells, a pore might be formed initially, which, at the increasing melittin concentration, might be subsequently annealed by the lateral diffusion of remaining lipid molecules [5,8]. This would result in a significant retardation of cell lysis (slow phase) at the high peptide concentrations [71. In the absence of melittin the maximal transfer efficiency from tryptophan to ANS molecules is low; consequently the calculated distance between them is high. As melittin interacts with the membrane this distance diminishes, resulting in the increase of transfer efficiency. Furthermore, ANS molecules may bind to membranes more efficiently and subsequently occupy the sites which are further away from the membrane proteins, namely membrane lipids. Thus, the data indicate that upon melittin binding the membrane tryptophan residues are exposed to the solvent molecules to the greater extent or that the quenchable fraction of membrane tryptophan becomes greater than in the absence of melittin. The spectrum of ANS is characterized by a little shift towards longer wavelengths, which can be accounted for by an increase in the polarity of the environment, suggesting a diminished contact with hydrophobic phospholipid fatty acyl chains. This would imply that membrane-bound ANS molecules may emerge

145 T A B L E II S T E A D Y - S T A T E F L U O R E S C E N C E P O L A R I Z A T I O N OF 1 - A N I L I N O - 8 - N A P H T H A L E N E S U L P H O N A T E A N D 1 , 6 - D I P H E N Y L - 1 , 3 , 5 - H E X A T R I E N E IN C O N T R O L A N D M E L I T T I N OR CRUDE VENOM-TREATED ERYTHROCYTE MEMBRANES Mean ± S.D.; n = 7. The significance of differences is given as calculated by analysis of variance (aANOV^) and by T u k e y ' s t e s t (aT).

Melittin

Po

Pl

/~2

/~3

D4

concentration Fluorescent probe

(control)

(2 x 10 -10 mol/mg)

(2 x 10 -10 mol/mg)

(2 x 10 -9 tool/rag)

crude venom

DPH

0.330 ± 0.010

0.347 ± 0.009

0.359 :e 0.006

0.359 • 0.005

0.355 ± 0.011

a T < 0.05 a T < 0.005

~0 # Pl P'0 # P4

#I ;~ P2

a T < 0.001

g0 # P2

~

ANS

0.307 ± 0.012

0.312 ± 0.044

0.366 ± 0.036

0.359 ± 0.019

0.359 ± 0.009

a w < 0.05 a T < 0.025

P0 # D3 /L1 # /v2

Pl # P3

a T < 0.01

~0 # /~2

;e #3

aANOV A < 0.0001

aANOVA < 0.001

towards the external environment upon melittin binding, which supposedly results in more efficient energy transfer between ANS and membrane protein tryptophan residues. These tryptophan residues might be originally screened from ANS by phospholipid molecules. Since melittin alters the order in the bilayer lipids as it penetrates the membrane, some of the tryptophan residues may become exposed or, at least, less hindered by phospholipid molecules. Such an exposure of selected membrane tryptophan residues might be responsible for the enhanced energy transfer efficiency during melittin action. Although the depth of melittin penetration into membranes is still questionable [38], our study seems to indicate that its penetration is deep enough to influence considerably the structure of membrane proteins. Both - S H spin labels employed in this study revealed a significant immobilization of membrane proteins upon melittin binding. In fact, based on MSL and ISL spectra, one might presume that the behaviour of both spin labels resembled very much the process of protein inactivation. Clague and Cherry reported decreased mobility of band 3 in erythrocyte ghosts measured subsequent to the addition of melittin. This retardation has been interpreted to reflect aggregation or 'clustering' of the protein in the plane of the membrane [39]. Otherwise, the very recent brief report on the interaction of melittin with dimyristoylphosphatidylcholine unilamellar vesicles has furnished the evidence for relatively shallow penetration of melittin into lipid bilayer [40]. Since that report concerned model liposome vesicles, whereas ours deals with cell membranes, they do not necessarily exclude each

0.818 ± 0.028

16-DS h , l / h o (7)

aANOVA < 0.002

aANOVA < 0.012 Met 5-DS h+lPao (9) aANOV A < 0.002 hw/hs (8) aANOVA < 0.008 re (6)

(control)

0.426 ± 0.019 2.14 ± 0.26 11.17 ± 0.21

0.410 + 0.010

2.42 ± 0.26

11.36 ± 0.19

0.807 ~- 0.014

#I (2 x 10 - l l tool/rag)

DO

Melittin concentration Parameter

by Tukey's test (aT).

11.87 ± 0.32

1.94 ± 0.37

0.410 ± 0.015

0.797 ± 0.025

(2 x 10 - m tool/rag)

~2

11.96 ± 0.32

1.99 ± 0.27

0.387 ± 0.021

0.773 ± 0.022

#3 (2 x 10 -9 mol/mg)

a T < 0.001

a T < 0.025 a T < 0.005

D1 # ~'3

DO # ~2 DO # ~3

DO ;~ u2

~1 a T < 0.025

#O ;~ P'3

a w < 0.001

~1 ~ ~3 DO # ~a a w < 0.025

aT < 0.05 a T < 0.01

~-2 # $¢3

DO ;~ ~3

~ ~3

/z2 ;~ ~3

Mean + S.D. N u m b e r of e x p e r i m e n t s given in p a r e n t h e s e s . T h e significance of differences is given as calculated by analysis of variance (aANOVA) and

T H E M O B I L I T I E S O F M E T 5-DP A N D 16-DS ( E X P R E S S E D A S h , l/ho RATIO), ISL R O T A T I O N A L C O R R E L A T I O N T I M E (%) A N D IMMOBILIZATION OF M S L ( E X P R E S S E D AS hwfhs RATIO) IN C O N T R O L A N D M E L I T T I N - T R E A T E D E R Y T H R O C Y T E M E M B R A N E S

T A B L E III

0'#

147

other. Featuring the possible reason for that difference, one should point out that membrane proteins present in erythrocyte membrane ghosts may potentiate melittin penetration into phospholipid bilayer. Whether membrane protein immobilization by melittin is lipid-mediated remains to be established. However, it seems reasonably certain that melittin-induced rigidization of the lipid bilayer may render some membrane integral proteins more exposed to the external environment. It is tempting to argue whether such a displacement may not account for the enhanced energy transfer efficiency between membrane tryptophan residues and ANS molecules upon melittin binding. The subsequent conclusion may be that melittin significantly reorganizes the lipid bilayer and possibly also rearranges membrane protein pattern. If so, such an argument might be also extrapolated for other proteins and peptides which are able to interact with cell membranes. Herein the phospholipase is of the utmost interest, as it is a natural component of aculeate venoms. It was suggested that melittin facilitates phospholipase action considerably, and both these components of bee venom were postulated to act synergistically in causing red blood cell lysis [1,10]. Such a facilitation could occur by either exposing membrane phospholipids to phospholipase active sites, or by enabling phospholipase transport across melittin-induced membrane channels [1,2]. Subsequent activation of phospbolipase A2 by melittin further increases the destruction of outer membrane leaflet and melittin is supposed, also, to enter the cell [6,7]. The lipidmediated action of melittin on membrane protein may well underlie the mechanism of the pronounced efficiency of melittin and phospholipase acting together under conditions, in which neither alone would be effective. It may be concluded that melittin induces a reorganization of lipid assemblies which can imply the alterations in lipid-protein interactions. Probably the interaction of erythrocyte membranes with melittin produces local conformational changes in membrane proteins, due to which tryptophan residues may be cooperatively modified to a different microenvironment. Depending on experimental design either vesicularization, fusion of small lipid vesicles or fragmentation into micelles may occur [41]. Such processes, when discussed in connection with the mechanism of the action of melittin, involving the lysis of biological membranes and the synergism between melittin and phospholipases, might shed new light on the altered lipid-protein interaction in the membrane, and thus contribute to the explanation of the enhanced susceptibility of erythrocytes to lysis during the action of crude aculeate venoms. ACKNOWLEDGEMENTS

The authors are greatly indebted to M. Mastalerz for supplying the honeybee workers for the isolation of crude venom extract used in this study. REFERENCES 1 A.W. Bernheimer and B. Rudy, Interactions between membranes and cytolytic peptides, Biochim. Biophys. Acta, 864 (1986) 123.

148 2 3 4 5

6 7 8

9 10

11 12 13 14 15 16

17

18 19 20 21 22

23 24 25 26

A.M.Batenburg, M. van Esch and B. de B. Kruijff, Melittin-induced changes of the macroscopic structure of phosphatidylethanolamines, Biochemistry, 27 (1988) 2324. A.M. Batenburg, J.C. HibDeln, A.J. Verldeij and B. de Kruijff, Melittin induces H n phase formation in cardiolipin model membranes, Biochim. Biophys. Acta, 903 (1987) 142. A.M. Batenburg and B. de Kruijff, Modulation of membrane surface curvature by peptide-lipid interactions, Biosci. Rep., 8 (1988) 299. T. Katsu, C. Ninomiya, M. Kuroko, H. Kobayashi, T. Hirota and Y. Fujita, Action mechanism of amphipathic peptides gramicidin S and melittin on erythrocyte membrane, Biochim. Biophys. Acta, 939 (1988) 57. W.F. Dawson, A.F. Drake, J. Helliwell and R.C. Hider, The interaction of Dee melittin with lipid bilayer membranes, Biochim. Biophys. Acta, 510 (1978) 75. W.F. Degrado, G.F. Musso, M. LieDer, E.T. Kaiser and F.T. Kezdy, Kinetics and mechanism of haemolysis induced by melittin and a synthetic melittin analogue, Biophys. J., 37 (1982) 329. T. Katsu, M. Kuroko, T. Morikawa, K. Sanchika, Y. Fujita, H. Yamamura and M. Uda, Mechanism of membrane damage induced by the amphipathic peptides gramicidin S and melittin, Biochim. Biophys. Acta, 983 (1989) 135. C. Mollay, B. Kreil and H. Berger, Action of phospholipase A2 on the cytoplasmic membrane of E. coli; stimulation by melittin, Biochim. Biophys. Acta, 426 (1976) 317. W. Vogt, P. Patzor, L. Lege, H.D. Oldigs and G. Willie, Synergism between phospholipase A2 and various peptides and SH-reagents in causing hemolysis, Nauyn-Schmiedebergs. Arch. Pharmak., 265 (1970) 442. V.T. Marchesi and G.E. Palade, The localization of Mg-Na-K-activated adenosine triphosphatase on red blood cell membranes, J. Cell Biol., 35 (1967) 385. O.H. Lowry, N.J. Rosenbrough, A.L. Farr and R.J. Randall, Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265. B.N. Ames, Assay of inorganic phosphate, total phosphate and phosphates, Meth. Enzymol., 8 (1966) 115-118. C. Watala and J.K. Kowalczyk, Hemolytic potency and phospholipase activity of some bee and wasp venoms, Comp. Biochem. Physiol., 97C (1990) 187. J. Slavik, Anilinonaphthalenesulfonate as a probe of membrane composition and function, Biochim. Biophys. Acta, 694 (1982) 1. W.J. van Blitterswijk, R.P. Hoeven and B.W. van der Meer, Lipid structural order parameters reciprocal of fluidity in biomembranes derived from steady-state fluorescence polarization measurements, Biochim. Biophys. Acta, 644 (1981) 323. C. Watala and Z. J6~wiak, 1-Anilino-8-naphthalenesulphonate binding parameters in red cell membranes. Does diabetes mellitus affect cell membrane dynamics, J. Clin. Chem. Clin. Biochem., 27 (1989) 839. M. Shinitzky and Y. Barenholz, Fluidity parameters of lipid regions determined by fluorescence polarization, Biochim. Biophys. Acta, 515 (1978) 367. R.F. Chen and R.L. Bowman, Fluorescence polarization measurement with ultra-polarizing filters in a spectrophotofluorometer, Science, 147 (1965) 729. T. Azumi and S.P. McGlynn, Polarization of the luminescence of phenanthrene, J. Chem. Phys., 37 (1962) 2413. W.L. Hubbell and H.M. McConnell, Molecular motion in spin labeled phospholipids and membranes, J. Am. Chem. Soc., 93 (1971) 314. A. Misharin and O. Polianovskij, Synthesis and use of spin labels: 4-maleimido-2,2,6,6-tetramethylpiperidine-l-oxyl and 4-isomaleimido-2,2,6,6-tetramethylpiperidine-l-oxyl, Izv Acad Nauk SSSR, Ser Khim, 11 (1974) 2526. B. Gaffney, The chemistry of spin labels, in: L.J. Berliner (Ed.), Spin Labeling, Academic Press, New York, 1976, p. 183. H.M. McConnell, W. Deal and R.T. Ogata, Spin-labeled hemoglobin derivatives in solution, polycrystalline suspensions and single crystals, Biochemistry, 8 (1969) 2580. J. Zar, Biostatistical Analysis, Prentice-Hall, Englewood Cliffs, NJ, 1984. U.P. Andley and B. Chakrabarti, Interaction of 8-anilino-l-naphthalenesulfonatewith rod outer segment membrane, Biochemistry, 20 (1981) 1687.

149 27 J.D. Morrisett, H.J. Pownall, R.T. Plumlee, L.C. Smith, Z.E. Zahner, M. Esfahani and S.J. Wakil, Multiple thermotropic phase transition in Escherichia coli membranes and membrane lipids, J. Biol. Chem., 250 (1975) 6969. 28 L.J. Berliner, The spin label approach to labeling membrane protein sulfhydryl groups, Ann. N. Y. Acad. Sci., 414 (1983) 153. 29 A. Keith, G. Bulfield and W. Snipes, Spin labeled Neuro,~pora mitochondria, Biophys. J., 10 (1970) 618. 30 T.D. Bradrick, J.L. Dasseux, M. Abdalla, A. Aminzadeh and S. Georghiou, Effects of bee venom melittin on the order and dynamics of dimyristoylphosphatidylcholine unilamellar and multilamellar vesicles, Biochim. Biophys. Acta, 900 (1987) 17. 31 E.J. Dufourc, I.C. Smith and J. Dufourcq, Molecular details of melittin-indueed lysis of phospholipid membranes as revealed by deuterium and phosphorus NMR, Biochemistry, 25 (1986) 6448. 32 M.T. Tosteson, J.J. Levy, L.H. Caporale, M. Rosenblatt and D.C. Tosteson, Solid-phase synthesis of melittin: purification and functional characterization, Biochemistry, 26 (1987) 6627. 33 T.D. Bradrick, E. Freire and S. Georghiou, A high-sensitivity differential scanning calorimetric study of the interaction of melittin with dipalmitoylphosphatidylcholine fused unilamellar vesicles, Biochim. Biophys. Acta, 982 (1989) 94. 34 H. Vogel and F. Jahnig, The structure of melittin in membranes, Biophys. J., 50 (1986) 573. 35 H. Vogel, Comparison of the conformation and orientation of alamethicin and melittin in lipid membranes, Biochemistry, 26 (1987) 4562. 36 E. Kuchinka and J. Seelig, Interaction of melittin with phosphatidylcholine membranes. Binding isotherm and lipid headgroup conformation, Biochemistry, 28 (1989) 4216. 37 J.F. Faucon and J.R. Lakowicz, Anisotropy decay of diphenylhexatriene in melittinphospholipid complexes by multifrequency phase-modulation fluorometry, Arch. Biochem. Biophys., 252 (1987) 245. 38 A.M. Batenburg, J.C. Hibbeln and B. de Kruijff, Lipid specific penetration of melittin into phospholipid model membranes, Biochim. Biophys. Acta, 903 (1987) 155. 39 M.J. Clague and R.J. Cherry, A comparative study of band 3 aggregation in erythrocyte membranes by melittin and other cationic agents, Biochim. Biophys. Acta, 980 (1989) 93. 40 P. Kaszyeki and Z. Wasylewski, Fluorescence-quenching-resolved spectra of melittin in lipid bilayers, Biochim. Biophys. Acta, 1040 (1990) 337. 41 J. Dufourcq, J.F. Faucon, G. Fourche, J.L. Dasseux, M. Le Maire and T. Gulik-Krzywieki,Morphological changes of phosphatidylcholine bilayers induced by melittin: vesicularization, fusion, discoidal particles, Biocbim. Biophys. Acta, 859 (1986) 33.