Farnesol and geranylgeraniol modulate the structural properties of phosphatidylethanolamine model membranes

Abstract

The biological activity of farnesol (FN) and geranylgeraniol (GG) and their isoprenyl groups is related to membrane-associated processes. We have studied the interactions of FN and GG with 1,2-dielaidoyl-sn-glycero-3-phosphoethanolamine (DEPE) membranes using DSC and X-ray diffraction. Storage of samples at low temperature for a long time favors a multidomain system formed by a lamellar crystalline (Lc) phase and isoprenoids (ISPs) aggregates. We demonstrate that ISPs alter the thermotropic behavior of DEPE, thereby promoting a H_II growth in a lamellar Lc phase with a reduced degree of hydration. The H_II phase occurs with the same repeat distance (d_HII=5.4 nm) as the Lc phase and upon heating it expands considerably (δd/δT≈0.22 nm/°C). The dimensional stabilization of this H_II phase coincides with the transition temperature of the Lc to L_α phase. Thereafter, the system DEPE/ISP will progress by increasing the nonlamellar-forming propensity and reaching a single H_II phase at high temperature. The cooling scan followed a similar structural path, except that the system went into a stable gel phase L_β with a repeat distance, d_Lβ=6.5 nm, in co-existence with a H_II phase. The formation of ISP microdomains in model PE membranes substantiates the importance of the isoprenyl group in the binding of isoprenylated proteins to membranes and in lipid–lipid interactions through modulation of the membrane structure.

• Isoprenoids
• gel lamellar phase
• lamellar-crystalline phase
• liquid-crystalline lamellar phase
• inverted hexagonal phase
• phase transition
• domain structures

Abbreviations
FN	=	farnesol
GG	=	geranylgeraniol
ISP	=	isoprenoid
PL	=	phospholipid
PC	=	phosphocholine
PE	=	phosphoethanolamine
DEPE	=	1,2-dielaidoyl-sn-glycero-3-phosphoethanolamine
Lβ	=	gel lamellar phase
Lc	=	lamellar-crystalline phase
Lα	=	liquid-crystalline lamellar phase
HII	=	inverted hexagonal phase
Tm	=	gel-to-liquid lamellar phase transition temperature
TH	=	lamellar-to-hexagonal phase transition temperatures

Introduction

The isoprenoid (ISP) lipids, farnesol (FN) and geranylgeraniol (GG), are hydrophobic compounds with one double bond and a methyl side chain every four carbon atoms along the main chain and a single hydroxide group at the headgroup (Figure 1). The biological activity of FN and GG is well-recognized [1–6]. One interesting aspect is that both compounds are implicated in numerous membrane-related biological processes [1], [2], [4]. On the other hand, the prenyl groups, farnesyl (C₁₅) and geranylgeranyl (C₂₀) are attached to proteins at C-terminal cysteine-rich signal sequences, in a post-translational protein modification process [7]. Protein isoprenylation occurs on certain cytoplasmic proteins such as G proteins [8], Rab proteins [9] and Ras proteins [10]. Such proteins are membrane-associated and they act as signal transducers participating in cell regulatory functions [11], [12]. The isoprenyl group of these proteins is known to be crucial for their function [9], [12–16]. The hydrophobicity of the isoprenyl group attached to the protein makes it a good candidate for insertion into the cellular membrane when protein membrane attachment is necessary. It has been reported that prenylated peptides can interact with model phospholipid bilayers [14], [17–19]. Isoprenylation seems to be crucial for the membrane binding of certain proteins [16], [20], [21]. Some studies support the idea that protein-bound prenyl groups act as a membrane anchor in cells and intercalate directly into the cellular membrane [22]. Interestingly, it has been demonstrated that the heterotrimeric G_i protein (Gαβγ) and the Gβγ dimer can bind to PC and PE model membranes. Indeed, the Gβγ dimer drives the interaction of G_i proteins with nonlamellar (hexagonal H_II) membrane structures [16].

Figure 1. Structure of the isoprenoids farnesol (FN) and geranylgeraniol (GG).

Although one important mode of the biological activity of FN and GG and their isoprenyl groups is through the membrane-associated events, there have been few studies of how ISPs interact with the lipid bilayer. Insertion of ISPs into PC bilayers has been demonstrated, and also the corresponding effect on the biophysical properties [23–26]. One interesting conclusion is that FN stabilizes the fluid phase and promotes fluid-gel phase co-existence [25], [26]. However, a main question, not yet resolved concerns the ISP role on the lipid polymorphism. We focus our study on the capacity of FN and GG to modulate the thermotropic phase behavior of model PE membranes. By means of differential scanning calorimetry and X-ray scattering analysis, we demonstrate the formation of structural microdomains rich in ISP in a PE surrounding.

Materials and methods

Materials

Farnesol (all-trans-3,7,11-trimethyl-2,6,10-dodecatrien-1-ol) (FN), geranylgeraniol (all-trans-3,7,11,15-tetramethyl-2,6,10,14-hexadecatetraen-1-ol) (GG) and N-(2-hydroxy ethyl) piperazine-N′-(2-ethanesulfonic acid) sodium salt (Hepes) were obtained from Sigma Chemicals Co. (Madrid, Spain). 1,2-dielaidoyl-sn-glycero-3-phosphoethanolamine (DEPE) was purchased from Avanti Polar Lipids, Inc. (Alabaster, USA). Lipids were stored under argon at −80°C, until use. Differential scanning calorimetry (DSC) of multilamellar liposomes was used to evaluate the phospholipid purity.

Model membrane preparation

Multilamellar lipid vesicles (MLV; 15% w/w) were prepared in 10 mM Hepes, 100 mM NaCl, 1 mM EDTA, pH 7.4 (Hepes buffer). For X-ray scattering, the lipid mixtures were hydrated in the presence or absence of farnesol (FN) or geranylgeraniol (GG) at the indicated molar ratios, homogenized with a pestle-type mini-homogenizer (Sigma), and vortexed until a suspension was obtained. The suspensions were then submitted to five heating and cooling cycles (heating to 70°C and cooling to 4°C) as previously reported [27] and stored at −20°C until use. For DSC measurements, the lipid mixtures (5 mg) were dissolved in chloroform, evaporated under argon and vacuum-dried. A known weight of the samples was transferred to an aluminium pan and hydrated by adding the Hepes buffer. The pans were sealed and submitted to ten heating and cooling cycles (to 75°C and to −20°C), to ensure complete homogenization before measurements.

Differential scanning calorimetry (DSC)

Experiments were run in a DSC 2920 scanning calorimeter (TA instruments). The samples were run at a scan rate of 5, 2 and 1°C/min after their preparation. To ensure the reproducibility of the results, three scan cycles were recorded for each sample. Since each scan yielded a similar thermogram, but with the transitions in the cooling curves shifted to lower temperatures, the second heating scan was used to evaluate the thermotropic transitions. The lipid phase transition temperatures (T_m and T_H) were determined from the maximum of the excess heat flow vs. temperature curves and the transition enthalpies were obtained from the area under the peaks. To check the reproducibility of DSC measurements, two to three samples of each DEPE:ISP molar ratio were prepared. The deconvolution analysis of the calorimetric peaks was done using mathematical fitting curves.

X-ray scattering

Small and Wide-Angle (SAXS and WAXS) synchrotron radiation X-ray scattering data were collected simultaneously on the Soft Condensed Matter beamline A2 of Hasylab at the storage ring DORIS III of the Deutsches Elektronen Synchrotron (DESY). Data collection was performed as described previously [27]. Samples were heated from 25 to 75°C at a scan rate of 1°C/min; afterwards they were kept at the highest temperature for 5 min and finally cooled down to the lowest temperature at the same scan rate. Positions of the observed peaks were converted into distances, d, after calibration with the standards rat tendon tail and poly-(ethylene therephtalate) for the SAXS and WAXS regions respectively. Interplanar distances, d_hkl, were calculated according to Equation 1:Eq. 1 where s is the scattering vector, 2θ is the scattering angle, λ (0.154 nm) is the X-ray wavelength and hkl are the Miller Indexes of the scattering planes.

Results

Differential scanning calorimetry (DSC)

Figure 2 illustrates the DSC thermograms of DEPE alone and in the presence of FN and GG. In the heating mode, DEPE showed two endothermic phase transitions with high cooperativity (Figures 2a and 2c), in agreement with earlier reports [28], [29]. The first one corresponds to the gel to liquid-crystalline (L_β–L_α) phase transition centered at 37.1°C with an enthalpy of 7.2 kcal/mol, while the second one, corresponds to the lamellar to inverted hexagonal (L_α–H_II) phase transition centered around 65°C with an enthalpy of 0.6 kcal/mol. An important effect of FN or GG on the thermotropic behavior of DEPE is a concentration-dependent decrease in the T_H value and a broadening of the L_α–H_II transition peak (a reduction of the thermal transition cooperativity) (Figures 2c and 2d). Indeed, the gel-to-liquid crystalline phase transition splits into two main peaks in presence of ISPs (Figures 2a, 2b and 2e). The shift in the maximum excess heat capacity and the relative area of the two peaks was also dependent on the DEPE:ISP molar ratio (Table I). However, the transition enthalpy of L_β–L_α and L_α–H_II curves was scarcely affected by the presence of ISP in the range of the molar ratio studied (up to 10:1, DEPE:ISP, mol:mol) (Table I). DSC scans of DEPE:ISP mixtures proved to be sensitive to the thermal history of the sample. A sample pre-incubated at −20°C for a long time, showed a transition composed of two peaks with a higher combined enthalpy (11.9 Kcal/mol for DEPE:FN, 20:1 molar ratio) (data not shown). However, when this sample was re-scanned without removing it from the calorimeter, the calorimetric curve also exhibited two peaks but the transition enthalpy decreased (7.2 Kcal/mol). The later scan was identical to the calorimetric profile of DEPE:ISP samples run immediately after their preparation without storage at −20°C (Figure 2e). DSC experiments were performed at different scan rates between 1 and 5°C/min (data not shown). The dependency of the transition temperature on the scan rate was small (≅1°C for both the heating and cooling scans) compared to the shift in the L_α-to-H_II phase transition induced by FN and GG (Figures 2c and Figure 2d).

Figure 2. DSC heating (a, c) and cooling (b, d) thermograms of aqueous dispersions of 1,2-dielaidoyl-sn-glycero-3-phosphoethanolamine (DEPE) in absence and in presence of farnesol (FN) and geranygeraniol (GG) at the molar ratio indicated. The arrows (panel c) indicate the endothermic peak of the L_α-to-H_II phase transition. The scan rate was 2°C/min. (e) Deconvolution of the calorimetric peak of DEPE:FN (20:1, mol:mol) corresponding to the L_β-to-L_α transition (heating scan): experimental data (solid line), fitted overall (broken line) and individual components (dotted lines).

Table I.  Calorimetric data of the effect of FN and GN in DEPE mixtures.

Samples	Molar ratio	Tm(1) (°C)	Tm(2) (°C)	Hm(1 + 2) (kcal/mol)	Hm(1) (%)	Hm(2) (%)	Th (°C)	Hh (kcal/mol)
DEPE	1:0	–	37.1	7.2	–	100.0	65.2	0.6
DEPE:FN	40:1	33.7	36.0	7.0	19.0	73.6	57.2	0.6
DEPE:FN	20:1	34.1	35.7	7.2	15.0	77.9	56.4	0.6
DEPE:FN	10:1	34.2	35.1	7.2	16.0	78.5	51.9	0.6
DEPE:GG	40:1	34.3	36.2	7.3	11.2	84.6	58.0	0.6
DEPE:GG	20:1	34.6	35.8	7.2	7.8	87.4	54.4	0.6

The enthalpy value of the L_β-to-L_α and L_α-to-H_II transitions correspond to the whole area of the curve. The transition temperature represent the maximum of the excess heat flow vs. temperature curves. T_m1 and T_m2 are the transition temperature of the two peaks obtained in the deconvolution of the L_β-to-L_α transition curve. H_m1 and H_m2 are the relative area of each peak. The scan rate was 2°C/min.

X-ray scattering studies

DSC data gave limited information about the gel-to-liquid lamellar and lamellar-to-nonlamellar phase transitions. Therefore, we also studied the structure of DEPE:ISP samples by X-ray scattering (Figure 3 and Table II). DEPE:ISP mixtures showed a particularly interesting structural behavior, depending on the time and storage conditions to which they were submitted (Figures 3–5). We will focus the analysis mainly on the thermotropic phase behaviour of DEPE:FN at the highest FN concentration studied with this technique (20:1, mol:mol) (Figure 4). At the beginning of the scan, DEPE:FN (20:1, mol:mol) mixtures stored at −20°C for a long time (days), showed scattering patterns associated with a lamellar-crystalline Lc1 phase (Figures 3c, Figure 4 and Figure 5) and a lattice parameter (d_Lc=5.4 nm) invariant with the temperature and mixture composition. Although the WAXS peaks were not sharp, looking at the scan and comparing it with other phases, one can see the differences and identify the Lc phase. Furthermore, the SAXS patterns also showed high scattering intensity at very small angles which decreased as the temperature increased (up to ≈39°C) (Figure 4). These properties are characteristics of micellar scattering and suggest a two-phase system. We note a very weak signal at smaller s-values that the peak associated with the Lc phase. It matches the s-value measured for the L_β phase seen on cooling at the end of the temperature scan. However, no measurable signal from L_β phase can be seen on the WAXS range of the scattering pattern (Figures 3c, Figure 3d and Figure 4). Therefore, even though there is evidence of L_β phase co-existing with Lc at low temperatures, its proportion is so small that its presence can be neglected. More interestingly, we also observed the appearance (at ca. 34°C) of weak signals (Figures 3c, Figure 4 and Figure 5) that were unequivocally associated with a H_II phase (d_HII=5.4 nm and δd/δT=0.22 nm/°C). The system went into a Lc + H_II two-phase region during a short range temperature (ΔT_Lc–HII=34–38°C). The Lc phase completely melted at 38°C, going into a L_α phase (d_Lα=5.5 nm), still co-existing with an H_II phase (ΔT_Lα–HII=39–59°C). Finally, a single H_II phase was reached at 60°C (d_HII=6.3 nm at 72°C and δd/δT= − 0.024 nm/°C). It is worth noting that the co-existence of Lc and H_II phases are not commonly seen in phospholipids, however phase transitions from Lc-to-H_II have been observed in mixtures of saturated short-chain phosphatidylcholines and the respective fatty acids [30], glycoglycerolipids [31–33] and DSPE dispersed in glycerol/water (>70wt% glycerol) [34].

Figure 3. Sequence of SAXS (left) and WAXS (right) scattering patterns of (a, b) DEPE, (c, d) DEPE:FN (20:1, mol:mol) and (e, f) DEPE:GG (20:1, mol:mol). The scan rate was 1°C/min. Successive diffraction patterns were collected for 15 s every minute. The thermal sequence of the transition from (Lc + H_II) to L_α and L_α-to-H_II phases was clearly observed in the heating scan. A L_β phase occur at the end of the cooling scan. The Lc and L_β phases were identified by the reflections in the WAXS region. The L_α phase was identified by a single peak of reflection in the SAXS region with a very good signal-to-noise ratio. The H_II phase was characterised by the appearance of three diffraction orders in the SAXS region with a d-spacing ratio of 1:1/√3:1/√4.

Figure 4. Sequence of scattering patterns from DEPE:FN (20:1, mol:mol). (a) In the heating scan, the transition from Lc to L_α is seen on the WAXS part of the scattering. Also the transition from Lc into H_II (associated peaks marked by arrows) and L_α (sharp peaks) is evidenced. (b) In the cooling sequence, the transition from H_II+Lc to L_α and the re-entrant H_II co-existing with L_β is shown.

Figure 5. Dependence of the interplanar repeat distance (d) on temperature for (a) DEPE heating run, (b) DEPE:FN (40:1, mol:mol) heating run, (c) DEPE:FN (20:1, mol:mol) heating and (d) cooling run, (e) DEPE:GG (40:1, mol:mol) heating run and (f) DEPE:GG (20:1, mol:mol) heating run. The different phases represented are: Lc, L_β, L_α and H_II. The co-existence of the Lc + H_II and L_α+H_II phases is indicated by the interval temperature range covered by the vertical lines. The spacing corresponding to the first order of the re-entrant H_II phase is not plotted for simplicity of the figure.

Table II.  Summary of the structural properties.

Mixture		DEPE:FN	DEPE:FN	DEPE:FN	DEPE:GG	DEPE:GG	DEPE:GG
Molar ratio	DEPE	20:1^a	40:1	60:1	20:1^b	40:1	60:1
dLc (28°C)	–	5.39	5.40	5.39	5.34	5.38	5.40
ΔTLc–HII	–	34–38	35–40	37–38	36–42	37–42	36–42
δd/δT (HII)	–	0.22	0.22	0.22	0.27	0.28	0.27
dLα (45°C)	5.52	5.50	5.49	5.49	5.50	5.51	5.57
ΔTLα–HII	66–69	39–59	60–66	60–65	51–62	56–64	67–69
δd/δT (HII)^c	−0.021	−0.025	−0.022	−0.022	−0.022	−0.025	−0.023
dHII (72°C)	6.59	6.46	6.43	6.48	6.31	6.39	6.44
dLβ (28°C)	6.63	6.54	6.57	6.53	6.55	6.52	6.66

^aThis sample was kept at low temperature for several days prior to the X-ray measurements, while the others were pre-heated.

^bThe co-existence of Lc and H_II phases has been ignored due to the extremely low contribution of the H_II phase to the scattering patterns (signals above background level). d_Lc (28°C): interbilayer distance for the lamellar crystalline Lc phase in the heating scan. ΔT_Lc–HII: temperature range of Lc and H_II phase co-existence. δd/δT (H_II): thermal expansion co-efficient of the H_II phase upon initial heating.

^cThermal expansion coefficient of the second H_II phase upon heating. d_Lα (45°C): interbilayer distance of the L_α phase. ΔT_Lα–HII: temperature range of the L_α and H_II phase co-existence. d_HII (72°C): interplanar distance of the H_II phase. d_Lβ (28°C): interbilayer distance for the L_β phase in the cooling scan.

The cooling process also followed an interesting structural path (Figures 3d, Figure 4 and Figure 5d). First, we observed the L_α phase occurring as a single phase between 42 and 35°C and then, a re-entrant H_II phase. Second, the system reached a stable gel L_β phase (d_Lβ=6.5 nm) instead of the crystalline Lc phase co-existing with the H_II phase. The gel phase was stable over a period of days. Only a weak signal associated with the Lc phase could be seen at the end of the cooling scan. Taking the relative proportions of the scattering peaks as indicative of the amount of each phase present, we could estimate that there were only traces of the Lc phase at the end of the scan. However, DEPE:FN mixtures pre-heated to 45°C and then cooled to 25°C (the starting point of the temperature scan) showed only the L_β phase in the heating scan (data not shown).

DEPE:GG mixtures showed similar diffraction patterns and sequence of phases (Figures 3e, Figure 3f, Figure 5e and Figure 5f). Comparatively, GG increased the temperature for the range of co-existence between the Lc + H_II phases (36–42°C). Nevertheless, the main differential effect was on the transition temperature of the L_α-to-H_II phase (Table II). GG was more efficient to lower both the onset and the end of the transition temperature (ΔT_Lα–HII=51–62°C for DEPE:GG at 20:1, mol:mol).

Discussion

The isoprenoids, FN and GG, affected the thermotropic behavior of the nonlamellar-prone phospholipid DEPE by altering the gel-to-liquid lamellar phase transition and promoting the nonllamelar phase, H_II (Figures 2–5). DEPE:ISP mixtures showed scattering patterns that allowed characterization of the Lc, L_β, L_α and H_II phases. Interestingly, the two ISP molecules exhibited similar trends and promoted an H_II phase in co-existence with the Lc or the L_β phase (Figures 4, Figure 5 and Table II).

DSC profiles show asymmetric peaks in the (Lc + L_β)-to-L_α phase transition, indicating that DEPE:ISP mixtures exhibit non-ideal mixing (Figure 2). These data suggest that ISP induced the appearance of domains in the (Lc + L_β)-to-L_α phase transition. However, the main argument comes from the X-ray data that demonstrate unequivocally the presence of co-existing domains (Figures 3–5). Considering the X-ray scattering pattern of samples previously heated (Figure 5d), we can conclude: (i) the first calorimetric peak (T_m1) is associated with a co-existence of Lc/L_β and H_II phases and (ii) the second peak (T_m2) matches with the range temperature of co-existence of L_β and L_α phases. In addition, the L_α phase alone appears on cooling at a temperature (≈42°C) and was scarcely affected by the presence of the ISP molecules.

It should be noted that the form of the DEPE:ISP thermograms differs from the ones previously reported with the lamellar-prone phospholipid DMPC:FN mixtures [19], [25], [26]. From DSC and ²H-NMR, it was concluded that FN promotes the L_β–L_α phase co-existence and induces a marked reduction in the T_m value and the cooperativity of the phase transition [19], [25], [26]. These studies support a model in which FN partitions preferentially into the liquid L_α phase, without affecting the lipid chain order. However, our data showed that ISPs induce a H_II phase in co-existence with either Lc or L_β phases at physiological temperature (i.e., near 37°C). Indeed, the transition of DEPE:ISP to the L_α phase does not show the common characteristics of a DEPE temperature-induced L_α phase formation (compare Figures 5a and Figure 5c). These differences between DEPE:ISP and DMPC:FN systems should be due to the differences in the molecular geometry of PE compared to PC phospholipids, which facilitate the aggregation of the ISP in the former class of lipids.

The thermotropic behavior of DEPE:ISP samples in relation to the formation of the Lc or L_β phase at a low temperature (below 34°C) (Figures 4, Figure 5), should be interpreted in terms of a balance between the thermodynamic and the kinetic parameters to reach each phase formation. Thus, the more stable crystalline Lc phase requires a longer time (days) for its formation at low temperatures. Considering intuitively, that the transition from Lc-to-L_α phases increases the entropy of the system and that the transition enthalpy is similar to the L_β-to-L_α phase transition, we conclude that the energy levels of the phases Lc and L_β should be similar. Moreover, because only one phase, crystalline or gel, was seen at once in considerable amounts, at the beginning of the heating or at the end of the cooling scans, we suggest that the activation energy of the Lc/L_β transition should be significantly high. This is not surprising, because the Lc and L_β structures have different molecular arrangements, making the transition between them difficult to occur, and therefore slow.

Thermotropic transitions from multicomponent systems were observed in lipid–lipid and lipid–peptide mixtures [35–38]. The molecular shape of the lipids and the packing characteristics of hydrocarbon side chains were reported as the main cause to induce lateral heterogeneity. However, we propose the following model to interpret the particular structural behaviour of DEPE:ISP mixtures: Storage of samples at a low temperature for a long time favours a system formed by a layered crystalline Lc phase with a rather reduced hydration. Due to the low miscibility between ISPs and lipid molecules, ISPs would aggregate to form ISP-rich domains in the Lc or the L_β phase of DEPE. ISP lipids are characterized by a small headgroup and a large hydrocarbon cross-sectional area due to the methyl branches along the chain (inverted molecular geometry). Therefore, ISP aggregates will strongly promote the formation of H_II phases in water-poor conditions. The H_II phase expands until 38°C, when the Lc-to-L_α phase transition temperature is reached. Then, the increased fluidity enhances diffusion and miscibility of ISP into the DEPE matrix forming ISP-poor domains into the fluid phase. Thereafter, the DEPE:ISP system will progress by increasing the nonlamellar-forming propensity and the formation of a H_II phase, as observed by calorimetry and X-ray diffraction. The L_α-to-H_II phase transition follows a similar trend as the systems formed by DEPE and unsaturated fatty acids previously reported [27], [29].

This model is supported by the observation that (1) the initial H_II phase occurs with the same repeat distance (d_HII=5.4 nm) as the Lc phase, i.e., it fits well within the Lc bilayer, and upon heating it expands considerably (δd/δT≈0.22 nm/°C); and (2) the lattice parameter stabilization of the H_II phase coincides with the transition temperature of the Lc-to-L_α phase. Further support for such organization of DEPE:ISP components into Lc/H_II domains could be found by estimating the molecule sizes and the lattice parameters of the phases involved. A simple estimation can be done by comparing the lattice parameter of the Lc phase with the molecular length of DEPE. A PE bilayer containing 18 all-trans carbons on each side would be 5.6 nm thick (assuming 0.5 nm for the lipid headgroup on each side). This value agrees with the experimental lattice parameter for the Lc phase (d_Lc=5.4 nm) and the initial d-value of the H_II phase. Although the experimental data do not give unequivocal evidence for such model, they are in line with the molecular mechanism for the lamellar/inverse hexagonal phase transition described by Rappolt et al. [39]. Interestingly, it has been shown that POPE undergoes a L_β-to-H_II phase transition in a water-poor system [40]. Thus, DEPE:ISP systems constitute a new example where an H_II domain grows in a multilamellar matrix with a reduced degree of hydration.

Biological significance

The reported data are related to the context of (i) lipid microdomains in a membrane-rich in PE lipids and (ii) the role that isoprenyl groups could play in the isoprenyl protein membrane interactions. Our experimental concentrations of ISP used in vitro are high, compared to those of ISP or isoprenyl groups protein-bound in a cellular system. However, the presence of structural microdomains rich in ISP, observed within the temperature range of the gel/fluid phases in DEPE membranes, could substantiate the role of lipid-driven protein targeting to cellular membranes. If this is the case, it is likely that the isoprenyl group could be a protein anchoring to enhance the clustering of prenylated proteins in a PE surrounding and through it modulate the membrane structure. We have recently demonstrated that the presence or propensity to form inverted hexagonal (H_II) phases favors the membrane association of Gαβγ heterotrimer and Gβγ dimer proteins, whereas Gα monomers prefer ordered lamellar structures [16]. In this context, the isoprenyl group on Gγ subunits could favor recruitment of G protein heterotrimers to nonlamellar-prone phases and facilitate the subsequent exit of activated Gα monomers to lamellar-prone regions, e.g., membrane rafts. This possibility represents a new point to be considered in biophysical studies of the cellular signaling process, in addition to the importance of rafts. This paper was first published online on prEview on 29 June 2005.

This work was supported by grants SAF2001-0839, SAF2003-00232, SAF2004-05249 (Ministerio de Educación y Ciencia, Spain), PRDIB-2002GC2-11 (Govern Balear), and I-02-066EC from Deutsches Elektronen-Synchrotron DESY, HASYLAB (Hamburg, Germany) from the IHP-Contract HPRI-CT-2001-00140 of the European Commission. We are grateful to Dr J. Cifre for his laboratory assistance.

Notes

¹A similar phase was analysed by DSC and ³¹P NMR and it was called “subgel” phase (Lewis R, Mannock DA, McElhaney RN 1989). Effect of fatty acyl chain length and structure on the lamellar gel to liquid-crystalline and lamellar to reversed hexagonal phase transitions of aqueous phosphatidylethanolamine dispersions. Biochemistry 28: 541–548).