Chemistry and Physics of

LIPIDS ELSEVIER SCIENTIFIC PUBLISHERS IRELAND Chemistry and Physics of Lipids 66 (1993) 123-134

Influence of sphingosine on the thermal phase behaviour of neutral and acidic phospholipid liposomes

A n u Kc3iv, Pekka Mustonen, Paavo K.J. Kinnunen*

Department of Medical Chemistry, University of Helsinki, Siltavuorenpenger 10, 00170 Helsinki, Finland

(Received 18 January 1993; revision received ; accepted 30 April 1993)

Abstract

The physical state of lipids is known to have pronounced effects on membrane functions. We studied the influence of sphingosine, a modulator of diverse cellular processes on the thermal phase behaviour and molecular packing of neutral and acidic phospholipids. Differential scanning calorimetry of multilamellar liposomes as well as the monolayer technique were employed. Inclusion of sphingosine in diacylphosphatidylcholine liposomes increased their pre- transition temperature Tp until at about 10 tool% sphingosine this transition was abolished. For these liposomes a gradual increase in both the temperature T m and enthalpy AH m of the main transition caused by sphingosine was observed. In contrast to diacylphosphatidylcholines, the Tp for dihexadecylphosphatidylcholine was lowered by sphingosine, demonstrating that the latter destabilizes the interdigitated gel phase. Inclusion of sphingosine in dimyristoylphosphatidic acid and dipalmitoylphosphatidylserine liposomes first elevated the Tm without significant changes in AH m, while at sphingosine contents > 50 mol% the appearance of complex melting profiles was evident. The transition temperature for the egg yolk phosphatidic a6id was shifted from below 0 to 29°C when mixed with sphin- gosine in a molar ratio of 1:1. Sphingosine also condensed the eggPA monolayers residing on an air-buffer interface. Accordingly, besides introducing a positive surface charge allowing the binding or activation of some proteins, sphin- gosine could influence membrane-mediated cellular processes by altering the organization and state of membrane lipids.

Key words: Sphingosine; Phospholipid; Thermal phase behaviour; DSC

* Corresponding author. Abbreviations: DHPC, dihexadecyiphosphatidylcholine; DMPA, dimyristoylphosphatidic acid; DMPC, dimyristoyl- phosphatidylcholine; DPPC, dipalmitoylphosphatidylcholine; DSC, differential scanning calorimetry; egg PA, phosphatidic acid from egg yolk; PA, phosphatidic acid; PC, phosphatidyl- choline; PKC, protein kinase C; PS, phosphatidyiserine; T m, main phase transition temperature; Tp, pre-transition tempera- ture; AH m, main phase transition enthalpy; AHp, pre-transition enthalpy;

1. Introduction

Sphingosine, a single-chain amphiphilic sphingolipid metabolite, has been shown to modulate diverse cellular functions, including growth and differentiation, activation of host de- fence systems, receptor function and oncogenesis [1,21. Free sphingosine is present in many cell

0009-3084/93/$06.00 © 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. SSDI 0009-3084(93)02182-Q

124 A. Kdiv et al./ Chem. Phys. Lipids 66 (1993) 123-134

types [3-5], and its content can vary either through breakdown of the complex lipids or by de novo synthesis.

Interest in the physiological significance and biochemical mechanisms of action of sphingosine has been growing since the discovery that this lipid is a potent inhibitor of protein kinase C (PKC), a central enzyme in signal transduction [6]. Besides blocking PKC activity, sphingosine has been shown to inhibit several PKC-mediated processes in vivo, including platelet secretion and aggrega- tion [6-8], oxidative burst in neutrophils [9-11] and attachment of murine Lewis lung carcinoma cells to extracellular matrix proteins [12]. In recent years also numerous PKC-independent cellular effects of sphingosine have been reported. Ex- amples of such phenomena include the stimulation of quiescent Swiss 3T3 fibroblast proliferation and potentiation of the mitogenic responses to other growth factors [13], regulation of epidermal growth factor receptor phosphorylation in A431 human epidermoid carcinoma cells [14] and stimu- lation of Ca 2÷ influx and the release of Ca 2÷ from agonist-sensitive intracellular pools in rat parotid acinar cells [15]. These findings clearly show that the biochemical processes underlying the influence of sphingosine on cellular functions are not limited to PKC inhibition, and other possible targets of its action remain to be elucidated.

Interestingly, the efficiency of sphingosine as an in vitro inhibitor of PKC was dependent on the relative amount of this lipid present in liposomes or micelles, and the inhibition could be reversed by sufficient amounts of acidic phospholipids [6]. Accordingly, sphingosine appears to exert its action in the lipid membrane. Since PKC requires acidic phospholipids for activity, the inhibitory action of sphingosine could involve neutralization of the negative charge of the acidic phospholipids and thus abolish their ability to bind and/or to activate the enzyme [16]. Little is known about the physical state of sphingosine in membranes. In particular, as long as the pK a of sphingosine in cell membranes or in liposomes cannot be directly determined, no strictly quantitative measure can be given of the role of charge neutralization caused by sphingosine.

Besides the requirement for the presence of cer-

tain lipids for activity, several membrane proteins have been shown to be sensitive to the phase state and organization of their lipid microenvironment [17,18]. Notably, modulation of PKC activity caused by alterations of the bulk biophysical pro- perties of the lipid membrane has been suggested [19]. Coupling between the organization and func- tion of biological membranes is an emerging con- cept, providing a framework for the development of understanding of the significance of the phase behaviour of lipids [20,21]. Accordingly, it was of interest to study the influence of sphingosine on the thermal phase behaviour of liposomes compos- ed of either neutral or acidic phospholipids as well as of their mixtures.

2. Experimental procedures

2.1. Materials HEPES, Tris, EDTA, DMPC, DPPC, DMPA,

eggPA, DPPS and D-sphingosine (from bovine brain sphingomyelin) were from Sigma. DHPC was from Calbiochem (San Diego, USA). The aci- dic phospholipids were calcium-free sodium salts. The purity of these lipids was checked by thin layer chromatography on plates coated with silicic acid (Merck, Darmstadt), using a chloroform/ methanol/water (65:25:4 v/v) solvent system. In each case only one homogeneous spot was detected after staining with iodine vapour. Water was freshly deionized in a Milli RO/Milli Q (Millipore) filtering system.

2.2 Differential scanning calorimetry Multilamellar liposomes were prepared by mix-

ing appropriate amounts of the lipid stock solu- tions in chloroform. Unless otherwise indicated, the amount of phospholipid(s) was usually kept constant, while the amount of sphingosine and, correspondingly, that of total lipid varied. The lipid mixtures were first evaporated to dryness under a stream of nitrogen, and traces of solvent were then removed by evacuating under reduced pressure for 2 h. The dry residue was hydrated above the main transition temperature o f the lipid components with 20 mM HEPES buffer, pH 7.4, containing 0.1 mM EDTA. For some DMPA/ sphingosine and DPPS/sphingosine mixtures a

A. K6iv et al./ Chem. Phys. Lipids 66 (1993) !-23-134 125

brief ( - 6 0 s) irradiation in a bath-type sonicator (Branson) was used. The samples were then main- tained on ice for at least 15 h before the heat capa- city scans were recorded, using a high-sensitivity adiabatic differential scanning calorimeter DASM-4 (Biopribor, Pushchino, Russia) at a heating rate of 0.5°C/min. Transition enthalpies were determined by planimetry of the peaks, using the internal electrical power calibration signal as a reference, and they are expressed in kilocalories per mole of phospholipid; i.e. the contribution due to the varying amounts of sphingosine was includ- ed. Phospholipid concentrations were determined by phosphorus assay [22]. Sphingosine concentra- tion was determined by dry weight. The critical micellar concentration (CMC) of sphingosine in 20 mM HEPES buffer, pH 7.4, containing 0.1 mM EDTA was determined by measuring the increase at the CMC in fluorescence intensity of 0.5 /zM diphenylhexatriene included in samples containing different concentrations of sphingosine.

2.3. Monolayer studies Compression isotherms were determined using a

rectangular trough drilled in Teflon. Surface pres- sure was measured using a Pt-Wilhelmy plate con- nected to an electrobalance. Appropriate amounts of the lipid stock solutions in chloroform were mixed to obtain the desired composition. Lipid monolayers were spread on the air-buffer (20 mM Tris-HC1, pH 7.4, containing 0.1 mM EDTA) interface from chloroform with a microsyringe. Isotherms were recorded at ambient temperature (~ 24°C) using a constant compression rate of 7.9 A2/chain • min. Both eggPA and sphingosine formed stable monolayers, with collapse pressures of 44 and 46 mN/m, respectively.

3. Results

3.1. Phase behaviour of sphingosine/phosphatidyl- choline liposomes

Sphingosine (2-3 mM) hydrated into 20 mM HEPES/0.1 mM EDTA buffer (pH 7.4) showed on DSC a single-phase transition at about 39°C with an enthalpy of AH = 3.4 kcal/mol (Fig. 1). At [sphingosine] < 2 mM both the transition temper- ature and its enthalpy were found to vary with the

>.. _P

,< Q. ,,¢ (J

I-- ,< U,I -I-

U') U') UJ L ) X U,I

I

30 40

TEMPERATURE, °C

i

5O

Fig. 1. DSC thermogram for neat 3 mM sphingosine in 20 mM HEPES buffer, pH 7.4, containing 0.1 mM EDTA.

concentration of the lipid (Table 1). As revealed by diphenylhexatriene fluorescence, the CMC of sphingosine was - 2 5 #M. Therefore this concen- tration dependence of the thermal phase behaviour does not reflect micellarization of this lipid. The nature of the process causing the heat absorption remains uncertain.

Increasing amounts of sphingosine had a mark- ed effect on the phase behaviour of the disaturated diacylphosphatidylcholines DMPC and DPPC. As shown in Fig. 2A, 3 mol% of sphingosine in DMPC liposomes increased their pre-transition and main transition temperatures Tp and Tm by 0.4 and 0.5°C, respectively. Concomitant broaden- ing of the pre-transition peak and a decrease in

Table 1 Phase transition temperatures and enthalpies determined from measurements on samples of different sphingosine concen- tration

[Sphingosine] Ttransition AH (mM) (°C) (kcal/mol)

0.5 40.0 1.6 1.0 ~ .1 2.5 2.0 39.4 3.4 3.0 39.3 3.4

126 A. K6iv et a l . / Chem. Phys. Lipids 66 (1993) 123-134

>.

(J ,,¢ a.

0

I,,,-

II1 "1"

¢il tu 0 X tu

®

2b :~o ,~ O

TEMPERATURE, C

>-

a.

i.- ¢¢ LLI

W 0 X LLI

0

2C) 30 4 0 50 60

TEMPERATURE,°C

@

2b 30 '/0 gO o

TEMPERATURE, C

Fig. 2. Influence of sphingosine on the thermal phase behaviour of phosphatidylcholine liposomes. A: DSC thermograms for 0.9 mM DMPC in the presence of (a) 0, (b) 3, (c) 7, (d) 12, (e) 30 and (f) 54 tool% sphingosine. B: DSC thermograms for 0.9 mM DPPC in the presence of(a) 0, (b) 4, (c) 9, (d) 23, (e) 33 and (f) 54 mol% sphingosine. C: DSC thermograms for 0.8 mM DHPC in the presence

of(a) 0, (b) 2, (c) 4, (d) 6, (e) 13, (f) 31 and (g) 56 mol% of sphingosine.

AHp were observed. At sphingosine concentra- tions > 12 mol% the pre-transition could no longer be detected. When the sphingosine content was in- creased, T m continued to shift upwards, with a linear dependence on the mol% of sphingosine (Fig. 3A). At a [sphingosine]/[DMPC] molar ratio of about 1:1 the T m was 36°C, i.e. 11.7°C above

that for pure DMPC. Simultaneously a nearly 2- fold elevation in z~Hm, from 5.6 to 10.2 kcal/mol phospholipid, was observed, corresponding approximately to the sum of Z~Hm (oMPC) + AHp (oMPC) + AH(sph) (Fig. 3B). This behaviour was accompanied by an increase in the half-width of the peak.

A. K6iv et al . / Chem. Phys. Lipids 66 (1993) 123-134 127

~ 50

-2

30

C o 20

c

p-

/ ®

"~ 15 E

5

T -

, , , ~ , t . = 0 , , T , , t~

0 10 20 30 40 50 60 i- 0 10 20 30 40 50 I -

[Sphingosine], mol % [Sphingosine], mol % 60

Fig. 3. The dependence of the transition temperatures and enthalpies on the concentration of sphingosine present in phosphatidyl- choline liposomes. A: T m (0) and Tp (O) for DMPC; T m (A) and Tp (V) for DPPC; T m (1) and Tp (El) of DHPC. B: AHrn (0) and

AHp (O) for DMPC; AH m (A) and AHp (V) for DPPC, AH m (ll) and AHp (El) for DHPC.

As the Tm of DMPC is well below the melting temperature of sphingosine, it was of interest to study the effects of this single-chain amphiphile on DPPC having a T m at 41.7°C, about 3°C above the transition of neat sphingosine. Interestingly, the influence of sphingosine on the phase behav- iour of DPPC was essentially similar to that on DMPC (Fig. 2B). The values for Tp were increas- ed with the addition of sphingosine until the pre- transition disappeared at sphingosine contents exceeding 9 mol%. The subtransition present at 16.4°C for pure DPPC disappeared when 2 mol% of sphingosine was included. The values for T m increased gradually to 48°C with an increasing content of sphingosine up to a [sphingosine]/ [DPPC] molar ratio of about 1:1 (Fig. 3A). Notably, for all sphingosine/DPPC mixtures tested the observed Tm was higher than that of either of the pure components. A considerable broadening of the transition peak, together with an increase in the AH m corresponding approximately to AHtSph), were also observed.

DHPC is a close structural analogue of DPPC, differing from the latter only in the type of linkage of the hydrocarbon chains to the glycerol backbone. However, at temperatures below the pre-transition, DHPC forms the interdigitated gel

phase instead of the La phase observed for DPPC [23]. Therefore we studied the influence of sphin- gosine on DHPC to compare its interactions with the interdigitated and non-interdigitated gel phases. As depicted in Fig. 2C, the phase behav- iour of the diether lipid DHPC was even more sen- sitive to the presence of sphingosine than those of DMPC and DPPC. Interestingly, in contrast to the diacyl PCs, the Tp of DHPC was decreased by sphingosine. As little as 1 mol% of sphingosine lowered the Tp for DHPC by 3.4°C, while 4 mol% of sphingosine abolished the pre-transition. The Tm of DHPC was gradually increased from 43.7 to 49.4°C with the addition of sphingosine up to 56 mol%, thus again being higher than the melting temperatures of the pure components. Beginning from 6 mol% sphingosine, a shoulder was present on the low-temperature side of the transition peak. Also, an increase in the AHm similar to that for DPPC was observed (Fig. 3).

Multilamellar liposomes composed of 50 mol% DPPC and 50 mol% DHPC (total lipid concentra- tion 1 mM) showed a pre-transition at Tp = 25.5°C and a main transition with a Tm = 42.4°C, in good agreement with previously published data [24,25]. The inclusion of 1-10 mol% of sphingo- sine first resulted in an increase in the Tp followed

128 A. Kdiv et al./ Chem. Phys. Lipids 66 (1993) 123-134

by disappearance of the pre-transition peak. Similarly to pure DPPC and DHPC, the values for T m were increased in the presence of sphingosine. Beginning from 5 tool% sphingosine, a shoulder was present on the low-temperature side of the peak. No clear indications for sphingosine- induced phase separation phenomena were observ- ed (data not shown).

3.2. Influence of sphingosine on the phase behaviour of acidic phospholipids

Incorporation of increasing amounts of sphin- gosine in liposomes of dimyristoylphospatidic acid (DMPA) caused a significant increase in their T m with a linear dependence on the concentration of sphingosine. At [sphingosine]/[DMPA] exceeding 1:1 the main transition peak separated into two peaks, at 61.3 and 62.5°C. As shown in Fig. 4, a further increase in sphingosine concentration first led to a downward shift in the temperatures for these peaks, followed by their flattening into a broad heat absorption profile lacking fine struc- ture. At sphingosine contents > 75 mol% a new peak appeared, at -40°C, i.e. close to the transi- tion observed for neat sphingosine. The transition

enthalpy of 5.7 kcal/mol for pure DMPA was in- significantly altered by the addition of up to 50 mol% sphingosine. At higher sphingosine concen- trations accurate determination of transition en- thalpies could not be carried out, because of extensive sample aggregation. Preparation of the [DMPA]/[sphingosine] l:l liposomes in the buffer containing 1 M NaCl did not induce phase separ- ation, although the value for T m decreased by 3.5°C.

Increasing the amounts of sphingosine in dipalmitoylphosphatidylserine (DPPS) liposomes caused similar thermal phase behaviour alterations as observed for DMPA. A rise in Tm could first be observed up to a [sphingosine]/[DPPS] molar ratio of about 1:1, after which separation into two peaks occurred. At high sphingosine contents a low tem- perature phase separation with the appearance of a new peak at 36°C could be observed (Fig. 5).

Sphingosine also considerably elevated the tran- sition temperature for the phosphatidic acid derived from egg yolk (eggPA). Because of the unsaturation of its fatty acid chains, this lipid nor- mally exists in the liquid crystalline phase at temperatures above 0°C. However, in the presence of 50 mol% sphingosine two small and broad tran- sition peaks at about 29 and 32°C were observed.

> - I,-,-

o ,< D. ,< o

l.--

-i-

t/) t/) I.U O X I,U

a

b C

-

g

h

3b 5b 7b TEMPERATURE,

Fig. 4. Influence of sphingosine on the thermal phase behaviour of acidic liposomes. DSC traces for I mM DM P A in the ab- sence (a) and presence of 20 (b), 39 (c), 50 (d), 60 (e), 67 (f), 75

(g) and 85 (h) mol% of sphingosine.

>- I -

,< i l l . ,< O I.-- ~t u l "!"

t / )

W o x UJ

3'o

a • ~ b

C

S ~ - - - d ~ - - - e

g

h

6o TEMPERATURE,°C

Fig. 5. DSC thermograms for l mM DPPS in absence (a) and presence of l0 (b), 30 (c), 60 (d), 55 (e), 67 (f), 75 (g) and 85 (h)

mol% of sphingosine.

A. K&iv et al . / Chem. Phys. Lipids 66 (1993) 123-134 129

> . I - ra

<c Q.

I -

uJ 3:

u~ (I) uJ ( J x LLI

a

_ b

lb 2'o 3'o s'o TEMPERATURE, °C

Fig. 6. DSC recordings for liposomes composed of eggPA and sphingosine in the molar ratio of (a) 1:1 and (b) 1:2.

The mixture of sphingosine and eggPA at a 2:1 molar ratio also exhibited two very broad transi- tions, with maximal heat capacities around 19 and 31°C (Fig. 6).

We also investigated the influence of sphingo- sine on eggPA monolayers residing on an air- buffer interface. Compression isotherms were

recorded for the pure lipids and for their mixtures containing different concentrations of sphingo- sine. At each surface pressure value up to 40 mN/m, monolayers of neat sphingosine were more tightly packed than those of PA (Fig. 7). Also il- lustrated is the area occupied per acyl/alkyl chain versus the mol% of sphingosine present in the monolayer at different surface pressures. At sur- face pressure values < 30 mN/m the dependence clearly deviated from linear in a manner indicating an attractive interaction between PA and sphingo- sine. No similar behaviour was observed when DMPC was used instead of egg PA (data not shown).

3.3. Phase behaviour of phosphatidylcholine/phos- phatidic acid/sphingosine tertiary mixtures

Addition of DPPC to a 1:1 mixture of DMPA and sphingosine exhibiting a single transition at 61.3°C (Fig. 4) resulted in a decrease of the transi- tion temperature (Fig. 8) with a concomitant increase in the transition enthalpy and broadening of the peak. The DMPA/sphingosine mixture could mix with fairly large amounts of phosphati- dylcholine, as even for a DPPC/DMPA/sphingo-

e-

04 °<:

80

201 0 20 40 60 80 100

[Sphingosine], mol %

Fig. 7. Influence of sphingosine content on the area occupied per acyl/alkyl chain in eggPA-sphingosine monolayers formed at an air-buffer interface (for details see Experimental Pro- cedures) at surface pressures of 2.5 (O), 5.0 (A), 10.0 (T), 20.0

(O), 30.0 (11) and 40.0 02) mN/m.

70

T m

5O

40 I i I I

0 10 20 30 40 50

[PC], mol % of total lipid

Fig. 8. Influence of inclusion of increasing amounts of DPPC (O) or DMPC (1~ on the T m of liposomes composed of DMPA/sphingosine at molar ratios of 1:1 and 1:0.8, respec-

tively.

130 A. K~iv et al./ Chem. Phys. Lipids 66 (1993) 123-134

sine ratio of 1:1:1 no evidence for phase separation could be detected. Similar behaviour was observed when increasing amounts of DMPC were added to a 1:0.8 mixture of DMPA and sphingosine (Fig. 8).

We further studied the influence of sphingosine on the phase behaviour of liposomes composed of 90 mol% DMPC and 10 mol% eggPA and thus resembling the ratio of zwitterionic and acidic phospholipids in natural membranes. As shown in Fig. 9, such liposomes exhibited a single com- paratively broad transition at 21.8°C. With the addition of increasing amounts of sphingosine, this transition shifted upwards, and a shoulder appeared on the high temperature side together with further broadening of the peak. At sphingo- sine concentrations of > 15 mol%, heat absorption profiles with two or three maxima was observed.

When DMPA was used instead of eggPA the main transition was observed at 26°C and a pre- transition at 20°C. Addition of 5 mol% of sphingo- sine abolished the pre-transition. Consequently the main transition peak diminished and a broad shoulder appeared on the high-temperature side.

>. l--

In <

I-- <C uJ -r"

or} ¢/) uJ 0 X uJ

a

20 ~ 40

O T E M P E R A T U R E , C

Fig. 9. DSC thermograms for the [DMPC]/[eggPA] 9:1 lipo- somes in the absence (a) and presence of 5 (b), 10 (c), 15 (d),

20 (e), 25 (f) and 30 (g) mol% of sphingosine.

>.

t,.)

o.

uJ -l-

t,/) u) w (.1 x i i i

q

. _J

J

~ a

b c

3'0

T E M P E R A T U R E ,

Fig. 10. Heat capacity scans for the DMPC-DMPA 9:1 lipo- somes in the absence (a) and presence of 5 (b), 10 (c), 20 (e), 30

(d) and 50 (f) mol% of sphingosine.

At sphingosine contents > 10 mol% this peak disappeared, and very broad and featureless heat absorption was observed instead (Fig. 10).

4. Discussion

The molecular mechanisms of action of sphin- gosine, a lipid regulating a variety of cellular pro- cesses [1,2], have remained largely unresolved. What is perhaps the currently best understood property of sphingosine, its ability to inhibit PKC, has been shown to depend on the presence of both the long alkyl chain and the positive charge [26,27]. Structural analogues of sphingosine and long-chain aliphatic amines [26] mimic the influ- ence of sphingosine on PKC. The addition of suffi- cient amounts of acidic phospholipids reverses the inhibitory effect. Accordingly it has been sug- gested that the inhibition of PKC by sphingosine occurs because of electrostatic neutralization of the negative charge of phosphatidylserine required for the activation of this enzyme [16]. Sphingosine has been shown to detach PKC from membranes [28]. Also, positively charged polyamines that bind to acidic phospholipids [29,30] interfere with the membrane attachment of this enzyme [31]. In ad- dition, sphingosine prevents the association of the

A. K6iv et ai. / Chem. Phys. Lipida 66 (1993) 123-134 131

positively charged substrate histone with lipid- Triton mixed micelles [16]. Similarly to PKC, the inhibition of insulin receptor tyrosine kinase [32] and the activation of casein kinase [33] by sphin- gosine could be reversed by the addition of phos- phatidylserine, suggesting that electrostatic phenomena may be involved. The creation of a positive surface charge has been suggested to ac- count for the inhibition of CTP:phosphocholine cytidyltranferase by sphingosine [34]. Quantitative characterization of these perhaps at least partly electrostatically controlled processes has, how- ever, remained elusive, as the protonation level of sphingosine in lipid membranes is not known. Interestingly, sphingosine has been shown to influ- ence the cellular metabolism of acidic phospholipids. Enhanced phosphatidylserine syn- thesis by LA-N-2 cells [35] and increased levels of phosphatidic acid in Swiss 3T3 cells [36] following inclusion of sphingosine have been reported. Ad- ditionally, the inhibition of phosphatidate phosphohydrolase [37,38] as well as the activation of phospholipase D [39,40] and phospholipase C [41], L-serine base exchange enzyme [42] and the 80 kDa diacylglycerol kinase [43] by sphingosine lead to the accumulation of acidic phospholipids in cells.

Numerous studies thus suggest that some of the effects of sphingosine are mediated via the mem- brane rather than by direct interactions with pro- teins. However, little attention has been paid to the influence of sphingosine on the organization and thermotropic behaviour of lipids and the possible significance of such alterations for the regulation of enzyme functions. A body of evidence has accumulated to indicate the existence of composi- tionally as well as functionally specific micro- domains in biological membranes [21]. Notably, several recent studies have suggested the formation of a domain enriched with acidic phospholipids, particularly PS, underneath PKC on its membrane attachment [44,45]. Analogously, although the amounts of sphingosine required to evoke cellular responses are small, enrichment of this lipid into distinct domains may cause high local concentra- tions to be present. In such membrane regions sphingosine may effectively modulate the fluidity and phase state of the bilayer.

Our present study demonstrates pronounced effects by sphingosine on the phase behaviour of both neutral and acidic phospholipids at physi- ological pH. The presence of about 10 mol% of sphingosine was enough to abolish the pre- transition of the diester phosphatidylcholines DMPC and DPPC and only 4 tool% to abolish that of the diether phosphatidylcholine DHPC. Sphingosine seems to have a different influence on the stability of the L~ gel phase of DMPC and DPPC than on the interdigitated gel phase exhibited by DHPC. For the diacyl lipids Tp was increased until merging of the pre-transition into the main transition occurred, showing a preference of sphingosine for the gel phase. For DHPC a decrease in Tp caused by sphingosine was seen, in- dicating that the presence of sphingosine did not favour interdigitation. Addition of sphingosine elevated the values of Tm for DMPC, DPPC and DHPC, the increase being largest for DMPC. For DPPC and DHPC the values of their Tm were higher than those for the pure components at each sphingosine concentration, revealing strongly non- ideal mixing. The influence of cationic amphi- philes on phosphatidylcholine bilayers is known to depend on the chain length of the amphiphile. Alkyl trimethylammonium bromides with short alkyl chains up to fourteen methylene segments decrease the orientational order [46] and lower the phase transition temperature of phosphatidyl- cholines [47,48], while those containing 14 to 18 carbon alkyis cause only a very slight increase in Tm [47,49]. However, mixtures of some double- chain cationic amphiphiles with DPPC give sharp, narrow phase transitions at temperatures higher than the T m for either of the pure components [50]. It has been suggested that the latter phenome- non is the result of reorientation of the PC phos- phocholine moiety in the presence of the positive surface charge [51], thus decreasing the cross- sectional area of the headgroups. Electrostatic interactions between the positive charges and PC headgroup dipoles probably also favour tighter intramolecular packing [50]. Similar mechanisms might be involved in the effects of sphingosine on phosphatidylcholine liposomes. However, in this case the transition widths were considerably broadened by sphingosine, revealing a weakening

132 A. K6iv et al. / Chem. Phys. Lipids 66 (1993) 123-134

of the transition co-operativity. In addition to the amino group, sphingosine contains two hydroxyls in its polar headgroup. Alcohols with chain lengths of 12-18 carbons increase the main transi- tion temperature of DPPC and also slightly increase the transition enthalpy [47,52] similarly to sphingosine. Thus the mechanism underlying the changes in the phase behaviour of PCs might in- volve - - in addition to electrostatic interaction be- tween the at least partially protonated amino group of sphingosine [26,27] and the lipid phos- phate - - hydrogen bonding between sphingosine hydroxyls and the lipid phosphate. Alterations in the membranes' hydration because of sphingosine are also possible.

Compared with the pure compounds, mixtures of acidic phospholipids DMPA or DPPS with sphingosine also exhibited elevated transition temperatures. Similar phenomena have been reported earlier for acidic lipids and some cationic detergents [50,53] and are probably attributable to the condensing effect of the strong electrostatic interactions between the positively and negatively charged compounds. For these acidic phospho- lipids the transition peaks remained comparatively narrow in the presence of moderate concentrations of sphingosine, showing little effect on transition cooperativity. Interestingly, transition enthalpies at moderate sphingosine concentrations <50 mol% were insignificantly altered. Maximum values for Tm were evident close to a [sphingo- sine]/[acidic lipid] ratio of 1:1, referring to the tightest packing and presumably complete neutralization of either DPPS or DMPA by sphin- gosine. DPPS should carry one and DMPA slight- ly more than one negative charge at physiological pH [54]. At sphingosine contents > 50 mol% the liposomes exhibited complex melting profiles with low temperature phase separation. However, no direct quantitative conclusion on the stoichio- metry of the sphingosine-PA/PS electrostatic neutralization can be drawn on the basis of this data, because of the uncertainty of the pKa value for sphingosine. There is also a lack of knowledge about how these lipids influence each other's pro- tonation level in liposomes of varying composi- tion. So far the pKa values for sphingosine have been measured in mixed micelles with Triton X-

100 or octyl glucoside, and the results obtained are at variance [26,27]. As demonstrated for eggPA liposomes, the presence of sphingosine can rigidify fluid lipid bilayers to such an extent that the tran- sition from the gel to the liquid crystalline phase, normally occurring below 0°C, is shifted close to the physiological temperature range. Additional support for the condensing effect of sphingosine was obtained from studies on monolayers. Mix- tures of sphingosine and eggPA films turned out to occupy smaller areas per acyl chain than could be expected by simple arithmetic averages.

The introduction of sphingosine into mixed liposomes of DMPC and 10 mol% of either DMPA or eggPA (having an acidic lipid content resembling that of natural membranes) led to com- plex phase behaviour. Generally, the presence of sphingosine resulted on the one hand in an increase of the phase transition temperature, reflecting rigidification of the membrane, and on the other hand broadening of the transitions, in- dicating low co-operativity and enlargement of the temperature range where domains of crystalline and liquid crystalline phases can coexist. The shape of the transitions was strongly dependent on the acidic phospholipid present in the liposome.

A correlation between the influence of mem- brane additives on PKC activity and their ability to promote or inhibit the transition of phospha- tidylethanolamine bilayers to the hexagonal phase has been reported [55,56]. On the basis of this finding, the idea of a possible modulation of PKC function by altering the physical state of the lipid bilayer was put forward [19]. The authors suggest that hexagonal phase promoting agents cause in- creased separation of phospholipid headgroups, thus destabilizing the bilayer and facilitating the penetration of PKC into the membrane interior, leading to the activation of this enzyme. The pre- sent results further support the concept of lipid physical state mediated regulation, showing that the presence of the PKC inhibitor sphingosine leads to tighter lipid packing in both neutral and acidic phospolipid bilayers. Formation of mem- brane domains containing sufficiently high local concentrations of sphingosine would, accordingly, prevent the bilayer penetration of PKC and result in its inhibition.

A. K6iv et al . / Chem. Phys. Lipids 66 (1993) 123-134 133

In summary, we have demonstrated that the presence of sphingosine exerts a significant influ- ence on the thermotropic behaviour and organiza- tion of membrane lipids, offering another possible mechanism for affecting cellular processes medi- ated via the membrane besides surface charge- controlled binding or activation of proteins.

5. Acknowledgements

We wish to thank Mrs. Birgitta Rantala for skilful technical assistance. This study was sup- ported by the Finnish State Medical Research Council.

6. References

1 Y.A. Hannun and R.M. Bell (1989) Science 243, 500-507. 2 A.H. Merrill and V.L. Stevens (1989) Biochim. Biophys.

Acta 1010, 131-139. 3 E. Wilson, E. Wang, R.E. Mullins, D.C. Liotta, J.D.

Lambeth and A.H. Merrill, Jr. (1988) J. Biol. Chem. 263, 9304-9309.

4 A.H. Merrill, Jr., E. Wang, R.E. Mullins, W.C.L. Jamison, S. Nimkar and D.C. Liotta (1988) Anal. Biochem. 171, 373-381.

5 T. Kobayashi, K. Mitsuo and 1. Goto (1988) Eur. J. Biochem. 171, 747-752.

6 Y.A. Hannun, C.R. Loomis, A.H. Merrill, Jr. and R.M. Bell (1986) J. Biol. Chem. 261, 12604-12609.

7 Y.A. Hannun, C.S. Greenberg and R.M. Bell (1987) J. Biol. Chem. 262, 13620-13626.

8 T. Wiedmer, B. Ando and P.J. Sims (1987) J. Biol. Chem. 262, 13674-13681.

9 E. Wilson, M.C. Olcott, R.M. Bell, A.H. Merrill, Jr. and J.D. Lambeth (1986) J. Biol. Chem. 261, 12616-12623.

10 E. Wilson, W.G. Rice, R.M. Kinkade, Jr., A.H. Merrill, Jr., R. Arnold and J.D. Lambeth (1987) Arch. Biochem. Biophys. 259, 204-214.

11 J.A. Badwey, J.M. Robinson, W. Horn, R.J. Soberman, M.J. Karnovsky and M.L. Karnovsky (1988) J. Biol. Chem. 263, 2779-2786.

12 J. lnokuchi, Y. Kumamoto, M. Jimbo, H. Shimeno and A. Nagamatsu (1991) FEBS Lett. 286, 39-43.

13 H. Zhang, N.E. Buckley, K. Gibson and S. Spiegel (1990) J. Biol. Chem. 265, 76-81.

14 M. Faucher, N. Girones, Y.A. Hannun, R.M. Bell and R.J. Davis (1988) J. Biol. Chem. 263, 5319-5327.

15 H. Sugiya and S. Furuyama (1991) FEBS Lett. 286, 113-116.

16 M.D. Bazzi and G.L. Nelsestuen (1987) Biochem. Biophys. Res. Commun. 146, 203-207.

17 R.D. Klausner, N. Kumar, J.N. Weinstein, R. Blumenthal and M. Flavin (1981) J. Biol. Chem. 256, 5879-5885.

18 L. Fesus, A. Horvath and J. Harsfalvi (1983) FEBS Lett. 155, 1-5.

19 R.M. Epand and D.S. Lester (1990) TIPS II, 317-320. 20 E. Sackmann (1984) in: Biological Membranes, vol. 5,

Academic Press, London, pp. 105-143. 21 P.K.J. Kinnunen (1991) Chem. Phys. Lipids 57, 375-399.

• 22 D.R. Bartlett (1959) J. Biol. Chem. 234, 466-468. 23 M.J. Ruocco, D.J. Siminovitch and R.G. Griffin 0985)

Biochemistry 24, 2406-2411. 24 K. Lohner, A. Schuster, G. Degovics, K. Muller and P.

Laggner (1987) Chem. Phys. Lipids 44, 61-70. 25 J.T. Kim, J. Mattai and G.G. Shipley (1987) Biochemistry

26, 6599-6603. 26 A.H. Merrill, S. Nimkar, D. Menaldino, Y.A. Hannun, C.

Loomis, R.M. Bell, S.R. Tyagi, J.D. Lambeth, V.L. Stevens, R. Hunter and D.C. Liotta (1989) Biochemistry 28, 3138-3145.

27 R. Bottega, R. Epand and E.H. Ball (1989) Biochem. Biophys. Res. Commun. 164, 102-107.

28 R.N. Kolesnick and S. Clegg (1988) J. Biol. Chem. 263, 6534-6537.

29 F. Schuber, K. Hong, N. Diizgiines and D. Papahad- jopoulos (1983) Biochemistry 22, 6134-6140.

30 K.K. Eklund and P.K.J. Kinnunen (1986) Chem. Phys. Lipids 39, 109-117.

31 M. Moruzzi, B. Barbiroli, M.G. Monti, B. Tadolini, G. Hakim and G. Mezzetti (1987) Biochemistry J. 247, 175-180.

32 R.S. Arnold and A.C. Newton (1991) Biochemistry 30, 7747-7754.

33 O.B. McDonald, Y.A. Hannun, C.H. Reynolds and N. Sahyoun (1991) J. Biol. Chem. 266, 21773-21776.

34 P.S. Sohal and R.B. Cornell (1990) J. Biol. Chem. 265, 11746- 11750.

35 I.N. Singh, G. Sorrentino, R. Massarelli and J.N. Kanfer 0992) FEBS Lett. 296, 166-168.

36 H. Zhang, N.N. Desai, J.M. Murphey and S. Spiegel (1990) J. Biol. Chem. 265, 21309-21316.

37 Y. Lavie, O. Piterman and M. Liscovitch (1990) FEBS Lett. 277, 7-10.

38 T.J. Mullmann, M.I. Siegel, R.W. Egan and M.M. Billah (1991) J. Biol. Chem. 266, 2013-2016.

39 Z. Kiss and W.B. Anderson (1990) J. Biol. Chem. 265, 7345-7350.

40 Y. Lavie and M. Liscovitch (1990) J. Biol. Chem. 265, 3868-3872.

41 T. Hashizume, T. Sato and T. Fujii (1992) Biochem. J. 282, 243-247.

42 J.N. Kanfer and D. McCartney (1991) FEBS Lett. 291, 63-66.

43 F. Sakane, K. Yamada and H. Kanoh (1989) FEBS Len. 255,409-413.

44 A.C. Newton and D.E. Koshland, Jr. (1989) J. Biol. Chem. 264, 14909-14915.

45 M. Bazzi and G.L. Nelsestuen (1991) Biochemistry 30, 7961-7969.

46 J. Gallova, F. Devinsky and P. Balgavy 0990) Chem. Phys. Lipids 53, 231-241.

134

47 A.W. Eliasz, D. Chapman and D.F. Ewing (1976) Biochim. Biophys. Acta 448, 220-230.

48 T. Inoue, K. Miyakawa and R. Shimozawa (1986) Chem. Phys. Lipids 42, 261-270.

49 T. lnoue, K. Fukushima and R. Shimozawa, R. (1990) Chem. Phys. Lipids 52, 157-161.

50 J.R. Silvius (1991) Biochim. Biophys. Acta 1070, 51-59. 51 P.G. Seherer and J. Seelig (1989) Biochemistry 28,

7720-7728.

A. K6iv et al./ Chem. Phys. Lipids 66 (1993) 123-134

52 J.M. Boggs, G. Rangaraj and K.M. Koshy (1986) Chem. Phys. Lipids 40, 23-34.

53 T. Inoue, Y. Suezaki, K. Fufushima and R. Shimosawa (1990) Chem. Phys. Lipids 55, 145-154.

54 H. TrOuble and H.J. Eibl (1974) Proc. Natl. Acad. Sci. USA 71,214-219.

55 R.M. Epand (1987) Chem.-Biol. Inter. 63, 239-247. 56 R.M. Epand, A.R. Stafford, J.C. Cheetham, R. Bottega

and E.H. Ball (1988) Bioscience Reports 8, 49-54.