G protein-membrane interactions II: Effect of G Protein-linked Lipids on Membrane Structure and G Protein-membrane Interactions

1 G protein-membrane interactions II: Effect of G Protein-linked Lipids on Membrane Structure and G Protein-membrane Interactions Jesús Casasa, Maitane Ibargurena,b, Rafael Álvareza, Silvia Terésa,b, Victoria Lladóa, Stefano P. Piottoc, Simona Conciliod, Xavier Busquetsa,b, David J. Lópeza,b,* and Pablo V. Escribáa,b aLaboratory of Molecular Cell Biomedicine, Department of Biology, IUNICS, University of the Balearic Islands, E-07122 Palma de Mallorca, Spain. bLipopharma Therapeutics, S.L., ParcBit, 07121 Palma de Mallorca, Spain. cDepartment of Pharmacy, University of Salerno, Via Ponte don Melillo, 84084 Fisciano (SA), Italy. dDepartment of Industrial Engineering, University of Salerno, Via Ponte don Melillo, 84084 Fisciano (SA), Italy. *To whom correspondence should be addressed: David J. López, PhD, Laboratory of Molecular Cell Biomedicine, Department of Biology, University of the Balearic Islands, Crta. Valldemossa km. 7.5, 07122, Palma (Spain). Tel.: +34-97117 33 31; Fax +34-971 17 31 84; E-mail: [email protected] Abbreviations used: DiI, 1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate; DSC, differential scanning calorimetry; GG, geranylgeraniol; GPCR, G protein-coupled receptor; GUV, giant unilamellar vesicle; HII; inverted hexagonal; LUV, large unilamellar vesicle; MA, myristic acid; MOH, myristic alcohol; PA, palmitic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; POH, palmitic alcohol; POPC, 1-palmitoyl-2-oleoyl-sn-glycero3-phosphocholine; POPE, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine; TH, lamellar-to-inverted hexagonal phase transition temperature. This is the accepted manuscript of the article that appeared in final form in Biochimica et Biophysica Acta - Biomembranes 1859(9) : 1526-1535 (2017), which has been published in final form at https://doi.org/10.1016/j.bbamem.2017.04.005. © 2017 Elsevier under CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 2 ABSTRACT G proteins often bear myristoyl, palmitoyl and isoprenyl moieties, which favor their association with the membrane and their accumulation in G Protein Coupled Receptor-rich microdomains. These lipids influence the biophysical properties of membranes and thereby modulate G protein binding to bilayers. In this context, we showed here that geranylgeraniol, but neither myristate nor palmitate, increased the inverted hexagonal (HII) phase propensity of phosphatidylethanolamine-containing membranes. While myristate and palmitate preferentially associated with phosphatidylcholine membranes, geranylgeraniol favored nonlamellar-prone membranes. In addition, Gi1 monomers had a higher affinity for lamellar phases, while G and G showed a marked preference for nonlamellar prone membranes. Moreover, geranylgeraniol enhanced the binding of G protein dimers and trimers to phosphatidylethanolamine-containing membranes, yet it decreased that of monomers. By contrast, both myristate and palmitate increased the Gi1 preference for lamellar membranes. Palmitoylation reinforced the binding of the monomer to PC membranes and myristoylation decreased its binding to PE-enriched bilayer. Finally, binding of dimers and trimers to lamellar-prone membranes was decreased by palmitate and myristate, but it was increased in nonlamellar-prone bilayers. These results demonstrate that co/post-translational G protein lipid modifications regulate the membrane lipid structure and that they influence the physico-chemical properties of membranes, which in part explains why G protein subunits sort to different plasma membrane domains. Keywords: G-proteins, Membranes, Palmitoylation, Myristoylation, Cell Signaling, Isoprenoids. 3 1. INTRODUCTION Upon agonist-mediated activation, G protein-coupled receptor (GPCR)-mediated cell signaling is amplified through the larger number of G protein molecules present at the plasma membrane compared to the number of receptors [1]. Indeed, one agonist-activated GPCR can activate dozens and even hundreds of G molecules [2]. Therefore, many thousands of G proteins can be found in membrane regions where there is a high density of GPCRs. In the plasma membrane, protein-lipid and lipid-lipid interactions define the membrane lipid structure, which in turn influences the type of proteins found in a given membrane region, as well as the activity of GPCRs and related signaling proteins [3, 4]. When a G protein is activated by a GPCR, the G subunit dissociates from the G dimer. The released dimer remains in the vicinity of the receptor and it recruits GPCR kinases, which inactivate GPCRs and regulate other signaling proteins [5]. By contrast, the G monomer regulates the activity of effector proteins (e.g., adenylyl cyclase, phospholipase C and ion channels) often located in different membrane domains, such as lipid rafts [6-8]. The mobilization of each subunit to the correct membrane environment largely depends on their preference for certain lipids or lipid structures. However, the molecular mechanisms underlying G protein interactions with membranes, their mobilization to different domains and their influence on membrane lipid structure remain largely unknown. In the present study we used different approaches to investigate the effect of the coand post-translational lipid modifications of G proteins on membrane lipid structure and protein-lipid interactions. The transmembrane domains of the GPCR, such as those of α2-adrenergic receptor, increase the HII phase propensity of the membrane [9]. G and G proteins are also located preferentially in this nonlamellar-prone environment [4], which may partly explain why G proteins accumulate near GPCRs. On the other hand, certain G monomers prefer lamellar-prone regions, such as lipid rafts, explaining how they may be mobilized from the receptor to effector rich membrane domains [3, 4, 6]. Therefore, membrane lipid structure plays an important role in propagating GPCR-mediated signals. GPCRs frequently cluster in defined membrane regions, where Gαβγ proteins co-localize in molar excess [10]. In these regions, G proteins interact with the cytosolic leaflet of the plasma membrane, aided by the myristoyl and palmitoyl moieties that are associated with the G subunit, and the isoprenyl moieties associated with the G subunit [11]. In addition to facilitating 4 G protein binding to membranes, these lipid anchors may also modify the lipid bilayer environment and the G protein-membrane interactions. The effect of lipid moieties of G proteins on membrane structure and protein-lipid interactions has received little attention to date. Thus, here we have investigated the role of certain lipids on the structural properties of membranes in further detail. For this purpose, we have used model membranes that contain the lamellar-prone phospholipid phosphatidylcholine (PC) and the nonlamellar-prone phospholipid phosphatidylethanolamine (PE), testing their interactions with purified Gi1, G or G proteins in the presence or absence of palmitic acid (PA), myristic acid (MA) or geranylgeraniol (GG). In contrast to other studies in which point mutants were analyzed with or without G protein-anchored lipids [12], this approach enabled us to determine the effect of these lipids on wild type G protein-membrane interactions. Accordingly, we found that these lipid moieties had different effects on membrane lipid structure, and on the interactions of the G proteins with lamellarand nonlamellar-prone membranes. In summary, these results show the role of these lipid modifications in the complex interactions between G proteins and membranes and the possible implications in human health are discussed. 2. EXPERIMENTAL 2.1. Materials Egg yolk PC and bovine liver PE were purchased from Avanti Polar Lipids (Alabaster, AL). MA, PA and GG were obtained from Sigma-Aldrich (Saint Louis, MO). The purified G proteins (myristoylated Gi1, G and Gi) were from Calbiochem (Darmstadt, Germany), the monoclonal anti-Gi1 antiserum was purchased to Santa Cruz Biotechnology (Santa Cruz, CA) and the monoclonal antibody anti-G was from BD Biosciences (Franklin Lakes, NJ). 1,1’- dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate (DiI) and Alexa Fluor 488 C5 maleimide were procured from Invitrogen (Eugene, OR). ECL Western blot detection system and Hyperfilm were from GE Healthcare (Pittsburgh, PA). 2.2. Differential Scanning Calorimetry DSC measurements were made with a Microcal MC-2 microcalorimeter (MicroCal Inc., Northampton, MA, USA), as described elsewhere [13]. Briefly, phospholipids were dissolved in chloroform: methanol (2:1, by vol) and dried under an argon flux. Solvent traces were removed 5 under vacuum for at least 3 h at room temperature before hydration. Multilamellar vesicles were formed by resuspending the lipid film in 10 mM HEPES, 100 mM KCl, 1 mM EDTA, pH 7.4, following vortexing at 42ºC. The mixture was degassed for 5 min and the DSC measurements were then carried out from 10 to 50ºC at a scan rate of 1ºC/min. All samples were subjected to three consecutive scans and calorimetric transitions were found to be reversible. The transition enthalpy and temperature values shown here corresponded to the means of three independent experiments and they were obtained using the software provided by the manufacturer (Microcal Origin). 2.3. 31P-Nuclear Magnetic Resonance Multilamellar vesicles were prepared by mixing 56 mg of bovine liver PE with deionized deuterated water (D2O, 15% w/w) in the presence or absence of 5 mol% MA, PA or GG. Lipid suspensions were hydrated and homogenized with a pestle-type minihomogenizer (Sigma), and vortexed to homogeneity. The suspensions were then subjected to 10 cycles of heating (60°C) and freezing (-80°C), and then equilibrated before data acquisition, as reported previously [14]. 31P-NMR measurements were made in 5 mm tubes on an Advance-300 multinuclear NMR spectrometer (Bruker Instruments). Data were acquired every 5ºC between 5 and 55ºC, equilibrating the temperature for 15 min before each measurement. The accumulated 31P-NMR free induction decay was obtained for 128 transients using a 4.4 s 90° radio-frequency pulse, a 24.3 kHz sweep width and 65,000 data points. The delay between the transients was 2 s and spectra were obtained by scanning from lower to higher temperatures. 2.4. Molecular dynamics Two all-atom lipid bilayers were used for symmetric membrane models containing 1palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine: 1-palmitoyl-2-oleoyl-sn-glycero-3phosphatidylethanolamine (POPC:POPE) (6:4; mol ratio) and POPC. The POPC membrane was made with 98 POPC molecules, 33 Na+ ions and 33 Clcounter ions, and 11,301 water molecules. The POPC:POPE membrane was made with 58 POPC and 40 POPE molecules, 28 Na+ ions and 28 Clcounter ions, and 9,825 water molecules. In both cases, the water density was of 0.997 g/mL. Simulations were performed using the YASARA program [15] at 310 K and 1 atm under a NPT ensemble, coupling the system to a Berendsen thermostat and barostat [16] combined with a control of solvent density as implemented in the software Yasara. The AMBER03 force field was used and the geometry of the molecules was optimized by the semi- 6 empirical AM1 method, using the COSMO solvation model [17]. Partial atomic charges were calculated using the same level of theory as the Mulliken point charge approach [18]. Electrostatic interactions were calculated with a 10.48 Å cut-off, and the long-range electrostatic interactions were handled by the particle mesh Ewald (PME) algorithm [19] using a sixth-order B-spline interpolation and a grid spacing of 1 Å. The leapfrog algorithm was used in all simulations with a 1.25 fs step time for intramolecular forces and a 2.5 fs step time for intermolecular forces. Five types of molecules (MOL_SET) were included to both membrane systems: GG, MA, PA, myristic alcohol (MOH) and palmitic alcohol (POH). The lipid bilayers were assembled and relaxed, reducing the box dimension until the Van der Waals energy of the system started to increase, and the structural parameters of the membranes were then compared using the experimental data [20]. To avoid atom-atom bumps and abnormal non-covalent interactions, the size of the MOL_SET molecules was initially reduced to 20% of the original size and the noncovalent interactions to 10% of their normal value [21]. The MOL_SET molecules were then placed at 160 different positions across the membrane using a software specially designed for this purpose [22]. The size and energy constants were then gradually increased until they reached normal values through cycles of steepest descent minimization, and a cycle of annealing was undertaken until the speed of the fastest atom dropped below 500 m/s. For each membrane/lipid system, the minimum in potential energy was then heated to 310 K and equilibration dynamics of 50 ns were completed. The lateral pressure profile was calculated as described [23-25] and the result are expressed as the average of 5 snapshots of the last 5 ns of simulations. Free energy of insertion was estimated by means of metadynamics calculations [26] using the Desmond program [27] under periodic boundary conditions. Simulations were performed in an isothermal-isobaric ensemble (1 atm, 310 K) with Langevin barostat and thermostat. The metadynamics simulations were carried out after equilibration. The free energy profile G(z) associated with lipid molecule translocation through the lipid bilayer was calculated along the zcomponent of the distance vector joining the membrane and the lipid molecule center of mass (zdist). For this study, the distance between one headgroup oxygen of each molecule of MOL_SET and the center of the membrane was chosen as collective variable. A second collective variable was the distance between the terminal carbon of each molecule of MOL_SET and the center of the membrane. The time interval between the addition of two Gaussian functions, τ, as well as 7 the Gaussian height, w, and Gaussian width, δ, were tuned to optimize the ratio between accuracy and computational cost. We used: τ = 100 fs, w = 0.2 kJ/mol, δ = 0.5 Å. The free energy of the insertion of lipid molecules in the membrane was calculated as the difference between the minimum energy of the molecule inserted in membrane and the free energy in water. 2.5. G protein binding to large unilamellar vesicles (LUVs) LUVs containing different molar ratios of PC:PE were prepared in the presence or absence of 5 mol% PA, MA or GG. The lipids were dissolved in chloroform/methanol (2:1) and mixed at the appropriate volumes. The solvent was evaporated under argon flux and solvent traces were removed under vacuum for at least 3 h. Lipid films were hydrated in 10 mM HEPES, 100 mM KCl, 0.1 mM EDTA, pH 7.4 at 42°C for 1 h, with vigorous vortexing every 15 min. The lipid suspension was submitted to five freeze/thaw cycles and was sonicated in a probe-type sonicator from Branson (Danbury, CT) for 10 s at 15 W. The LUVs were then incubated for 1 h at 37°C with purified Gαβγ (300 ng), G dimers (100 ng) or Gi1 monomers (150 ng) in a total volume of 200 l. The binding of Gαβγ to membranes was carried out in the presence of GDPS (50 M) and that of the Gi1 monomers in the presence of GTPS (50 M). Unbound G proteins were then separated from membrane-bound G proteins by centrifugation for 1 h at 25°C at 100,000 × g. Finally, the membrane pellets were resuspended in electrophoresis loading buffer (84 mM Tris-HCl, pH 6.8, 4% SDS, 1% 2-mercaptoethanol, 5% glycerol, 0.01% bromophenol blue) and boiled for 5 min. In all experiments, myristoylated G subunits and geranylgeranylated G subunits were used. Immunoblot and quantification of bound G proteinsImmunoblotting was performed as described elsewhere [28]. Briefly, samples from the binding experiments were resolved on 1020% gradient SDS-polyacrylamide gels and the proteins were then transferred to nitrocellulose membranes. The membranes were blocked with PBS containing 5% non-fat dry milk, 0.5% bovine serum albumin and 0.02% Tween-20 (blocking solution), and they were then incubated with anti-Gi1 (1:1000 dilution in fresh blocking solution) to detect Gi1 and G, or anti-G (1:1000) to detect G. Antobody binding was detected with a horseradish peroxidase-linked anti-mouse IgG (1:2000) in fresh blocking solution, which was visualized by ECL. The immunoreactive bands on the films were quantified by image analysis and the binding of G proteins to pure PC liposomes in the absence of other lipids was considered as the control value (100%). For each ratio of PE, the relative effect of PA, MA and GG on G protein binding was 8 compared to the same membrane without any lipid moiety and indicated in parentheses in Table 1. 2.6. Confocal microscopy Giant unilamellar vesicles (GUVs) were prepared using the electroformation method [29, 30]. For this purpose, lipid solutions containing 0.3 mM total lipid supplemented with 0.4 mol% DiI were prepared in chloroform: methanol (2:1; v:v). Three µl of the lipid mixture were added to the surface of platinum electrodes and solvent traces were removed under vacuum for 60 minutes. Platinum electrodes were covered with 400 µl of 25 mM HEPES, pH 7.4, previously heated at 50ºC. The platinum wires were connected to an electric wave generator at 50ºC under the following AC field conditions: 500 Hz, 0.22 V for 5 min; 500 Hz, 1.9 V for 20 min and finally 500 Hz, 5.3 V for 90 min. After GUV formation, the chamber was placed on a Leica TCS SPE inverted confocal fluorescence microscope (Barcelona, Spain). The GVMD software from the Beckman Institute (University of Illinois) was used to localize free cysteine residues on the surface of Gi1, G and Gi in order to be used for protein labeling (see Fig. S1) [31]. In brief, free cysteine residues were labeled with Alexa 488 C5maleimide by mixing 10 µl of 0.8 µg/ml protein with 0.5 µl Alexa Fluor 488 (10 µg/ml stock solution) for 10 min at RT. Fluorescently-labeled G proteins were added to GUVs at a final concentration of 15 ng/ml. The binding of Gαβγ to membranes was carried out in the presence of GDPS (50 M) and that of the Gi1 monomers in the presence of GTPS (50 M). The excitation wavelength for DiI was 532 nm and the emission was collected at 555-750 nm; the excitation for Alexa Fluor 488 was 488 nm and the emission 500-533 nm. The binding of fluorescently labeled G protein subunits to the GUVs was measured using the software provided with the microscope. The fluorescence signal sorrounding the lipid membrane was used as background. 2.7. Data analysis The data shown correspond to mean ± SEM values from the number of experiments indicated. One-way ANOVA followed by a Bonferroni test or two-tailed t-test was used for statistical evaluation. Differences were considered statistically significant at p< 0.05. 3. RESULTS 3.1. Effects of PA, MA and GG on membrane lipid structure 9 In the temperature range studied (15 - 45ºC), DSC showed a lamellar-to-inverted hexagonal (HII) phase transition peak for bovine liver PE at 22.4ºC (Fig. 1A). The presence of MA and PA increased the lamellar-to-inverted hexagonal phase transition temperature (TH) value to 31.2ºC and 28.5ºC, respectively. By contrast, when GG is added no transition is observed in the studied range of temperatures, although 31P-NMR experiments indicated that this lipid decreased the TH value, suggesting that it favored the occurrence of nonlamellar phases (Fig. 1B) [32]. When assessed by 31P-NMR, PE organized into lamellar phases at temperatures below 20°C; between 20ºC and 25ºC, both lamellar and HII phases co-existed, and at higher temperatures (≥ 30°C), PE molecules adopted HII phases. 31P-NMR scans also showed that MA and PA increased the TH, whereas GG decreased it about 10ºC (Fig. 1B). The binding free energies of GG, PA, POH, MA and MOH to POPC:POPE and POPC membranes were calculated by computational analysis (Fig. 2 and S2). MOH and POH were studied to isolate the effect of the acyl chain and to compare their effect on lipid membranes with GG, which shares the same alcohol headgroup. MA and PA, which are negatively charged at pH 7.4, exhibited greater binding energies than GG, POH and MOH, which lack of net charge. On the other hand, the binding of GG to POPC:POPE membranes was ~6 kcal/mol higher than that to POPC bilayers, indicating a preference of GG moieties for HII-prone domains. This difference is due to the balance between the hydrophobic match of the isoprenyl chain in the membrane and the hydrogen bonds between the hydroxyl group and water molecules. The differences in free energies for MA binding to POPC and POPC:POPE membranes were larger than those of GG, although the former preferred lamellar (POPC) membranes. In addition, MA showed an even higher propensity to associate with lamellar membranes than PA. These results suggest that GG rapidly segregates to membrane domains rich in the nonlamellar prone phospholipid PE, while PA and MA prefer lamellar-prone membrane domains. In addition, this binding behavior also contributes to explain the membrane microdomain preference of the G-containing G proteins (G and G complexes with an isoprenyl moiety) and G monomers containing MA and/or PA. The presence of each type of lipid has an important effect on the lipid membrane organization. Indeed, POPC membranes displayed an altered stress profile after the addition of any of G protein lipids, MA, PA and GG (Fig. 3A), indicating that regions rich in G proteins may undergo structural lipid regulations that could contribute to control the localization and activity of certain 16 lamellar-prone regions, and G and G to nonlamellar-prone microdomains, where these proteins can participate in productive interactions with specific signaling effectors. Finally, both MA and PA regulate membrane fluidity, which could modulate the activity of GPCRs and other membrane proteins [49]. It has been shown recently that membrane lipid composition and its structural regulation influences physiological processes such as blood pressure, platelet aggregation, cell proliferation and apoptosis, as well as underlying the mechanism of action of certain drugs [50-52]. The regulatory effect of membrane lipid composition on the localization and activity of peripheral and integral proteins can be partly explained by changes in the lipid bilayer lateral pressure [53] or fluidity [54]. Treatment with lipids or lipid-interacting molecules can regulate the composition and structure of membranes, reversing important pathological alterations such as cancer, hypertension or obesity [32, 50]. This novel therapeutic strategy, called “membrane-lipid therapy”, is based on the regulation of the activity of important signaling proteins by modulating the reorganization of membrane microdomains [55] and the subsequent protein-lipid interactions [50, 56]. By contrast to the general opinion that interventions on membranes could affect a large number of processes, this approach has been shown to be highly specific [52, 57], further demonstrating that the structure-function relationships of membrane lipids can be finely regulated. Thus, the present study sheds further light on the molecular mechanisms governing pharmaceutical and nutraceutical therapies targeting membrane lipids. ACKNOWLEDGMENTS - FUNDING This study was supported by the Spanish Ministerio de Economía y Competitividad grant BIO2010-21132, BIO2013-49006-C2-1-R, RTC-2015-3542-1 and RTC-2015-4094-1, cofinanced by FEDER funds from the EU (“Una manera de hacer Europa”), by the Govern de les Illes Balears (Grups competitius and Research Excellent Grant) and the Marathon Foundation. JC and RA were supported by predoctoral fellowships from the Ministerio de Ciencia e Innovación and from the Ministerio de Educación, Cultura y Deporte, respectively. ST, MI and DJL hold a Torres-Quevedo contract from the Spanish Ministerio de Economía y Competitividad. VL is supported by a postdoctoral contract from the Asociación Española Contra el Cáncer. 17 18 REFERENCES [1] G.D. Kim, E.J. Adie, G. Milligan, Quantitative stoichiometry of the proteins of the stimulatory arm of the adenylyl cyclase cascade in neuroblastoma x glioma hybrid, NG108-15 cells, Eur J Biochem, 219 (1994) 135-143. [2] L. Lagnado, Signal amplification: let's turn down the lights, Curr Biol, 12 (2002) R215-217. [3] V. Rudajev, J. Novotny, L. Hejnova, G. Milligan, P. Svoboda, Dominant portion of thyrotropin-releasing hormone receptor is excluded from lipid domains. Detergent-resistant and detergent-sensitive pools of TRH receptor and Gqalpha/G11alpha protein, J Biochem, 138 (2005) 111-125. [4] O. Vögler, J. Casas, D. Capo, T. Nagy, G. Borchert, G. Martorell, P.V. Escribá, The Gbetagamma dimer drives the interaction of heterotrimeric Gi proteins with nonlamellar membrane structures, J Biol Chem, 279 (2004) 36540-36545. [5] R.J. Lefkowitz, Seven transmembrane receptors: a brief personal retrospective, Biochim Biophys Acta, 1768 (2007) 748-755. [6] K.S. Nair, N. Balasubramanian, V.Z. Slepak, Signal-dependent translocation of transducin, RGS9-1-Gbeta5L complex, and arrestin to detergent-resistant membrane rafts in photoreceptors, Curr Biol, 12 (2002) 421-425. [7] E.M. Hur, Y.S. Park, B.D. Lee, I.H. Jang, H.S. Kim, T.D. Kim, P.G. Suh, S.H. Ryu, K.T. Kim, Sensitization of epidermal growth factor-induced signaling by bradykinin is mediated by cSrc. Implications for a role of lipid microdomains, J Biol Chem, 279 (2004) 5852-5860. [8] R. Álvarez, D.J. López, J. Casas, V. Lladó, M. Higuera, T. Nagy, M. Barceló, X. Busquets, P.V. Escribá, G protein-membrane interactions I: Gαi1 myristoyl and palmitoyl modifications in protein-lipid interactions and its implications in membrane microdomain localization, Biochim Biophys Acta, (2015). [9] J. Prades, J.A. Encinar, S.S. Funari, J.M. González-Ros, P.V. Escribá, F. Barceló, Interaction of transmembrane-spanning segments of the alpha2-adrenergic receptor with model membranes, Mol Membr Biol, 26 (2009) 265-278. [10] T. Drmota, J. Novotny, G.W. Gould, P. Svoboda, G. Milligan, Visualization of distinct patterns of subcellular redistribution of the thyrotropin-releasing hormone receptor-1 and gqalpha /G11alpha induced by agonist stimulation, Biochem J, 340 ( Pt 2) (1999) 529-538. [11] P.V. Escribá, P.B. Wedegaertner, F.M. Goñi, O. Vögler, Lipid-protein interactions in GPCR-associated signaling, Biochim Biophys Acta, 1768 (2007) 836-852. [12] S. Moffett, D.A. Brown, M.E. Linder, Lipid-dependent targeting of G proteins into rafts, J Biol Chem, 275 (2000) 2191-2198. [13] P.V. Escribá, M. Sastre, J.A. García-Sevilla, Disruption of cellular signaling pathways by daunomycin through destabilization of nonlamellar membrane structures, Proc Natl Acad Sci, 92 (1995) 7595-7599. [14] S.S. Funari, F. Barceló, P.V. Escribá, Effects of oleic acid and its congeners, elaidic and stearic acids, on the structural properties of phosphatidylethanolamine membranes, J Lipid Res, 44 (2003) 567-575. [15] E. Krieger, T. Darden, S. Nabuurs, A. Finkelstein, G. Vriend, Making optimal use of empirical energy functions: force-field parameterization in crystal space., Proteins, 57 (2004) 678-683. [16] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. Di Nola, J.R. Haak, Molecular dynamics with coupling to an external bath., J. Chem. Phys. , 81 (1984) 3684-3698. 19 [17] A. Klamt, Conductor-like Screening Model for Real Solvents. A new Approach to the Quantitative Calculation of Solvatation Phenomena. , J. Phys. Chem., 99 (1995) 2224-2235. [18] J.J. Stewart, MOPAC: a semiempirical molecular orbital program, J Comput Aided Mol Des, 4 (1990) 1-105. [19] U. Essmann, L. Perera, M.L. Berkowitz, T. Darden, H. Lee, L.G. Pedersen, A sooth particle mesh Ewald method., J. Chem. Phys., 103 (1995) 193-202. [20] N. Kucerka, S. Tristram-Nagle, J.F. Nagle, Structure of fully hydrated fluid phase lipid bilayers with monounsaturated chains, J Membr Biol, 208 (2005) 193-202. [21] T. Crul, N. Toth, S. Piotto, P. Literati-Nagy, K. Tory, P. Haldimann, B. Kalmar, L. Greensmith, Z. Torok, G. Balogh, I. Gombos, F. Campana, S. Concilio, F. Gallyas, G. Nagy, Z. Berente, B. Gungor, M. Peter, A. Glatz, A. Hunya, Z. Literati-Nagy, L. Vigh, Jr., F. HoogstraBerends, A. Heeres, I. Kuipers, L. Loen, J.P. Seerden, D. Zhang, R.A. Meijering, R.H. Henning, B.J. Brundel, H.H. Kampinga, L. Koranyi, Z. Szilvassy, J. Mandl, B. Sumegi, M.A. Febbraio, I. Horvath, P.L. Hooper, L. Vigh, Hydroximic acid derivatives: pleiotropic hsp co-inducers restoring homeostasis and robustness, Curr Pharm Des, 19 (2013) 309-346. [22] S. Piotto, L. Di Biasi, S. Concilio, A. Castiglione, G. Cattaneo, GRIMD: distributed computing for chemists and biologists, (2014). [23] R.S. Cantor, Lateral pressures in cell membranes: a mechanism for modulation of protein function., J. Phys. Chem. , 101 (1997) 1723-1725. [24] S. Piotto, S. Concilio, E. Bianchino, P. Iannelli, D.J. López, S. Terés, M. Ibarguren, G. Barceló-Coblijn, M.L. Martin, F. Guardiola-Serrano, M. Alonso-Sande, S.S. Funari, X. Busquets, P.V. Escribá, Differential effect of 2-hydroxyoleic acid enantiomers on protein (sphingomyelin synthase) and lipid (membrane) targets, Biochim Biophys Acta, 1838 (2014) 1628-1637. [25] S. Piotto, A. Trapani, E. Bianchino, M. Ibarguren, D.J. López, X. Busquets, S. Concilio, The effect of hydroxylated fatty acid-containing phospholipids in the remodeling of lipid membranes, Biochim Biophys Acta, 1838 (2014) 1509-1517. [26] A. Laio, M. Parrinello, Escaping free-energy minima, Proceedings of the National Academy of Sciences, 99 (2002) 12562-12566. [27] K.J. Bowers, E. Chow, H. Xu, R.O. Dror, M.P. Eastwood, B.A. Gregersen, J.L. Klepeis, I. Kolossvary, M.A. Moraes, F.D. Sacerdoti, Scalable algorithms for molecular dynamics simulations on commodity clusters, in: SC 2006 Conference, Proceedings of the ACM/IEEE, IEEE, 2006, pp. 43-43. [28] P.V. Escribá, M. Sastre, J.A. García-Sevilla, Increased density of guanine nucleotidebinding proteins in the postmortem brains of heroin addicts, Arch Gen Psychiatry, 51 (1994) 494-501. [29] I.M. Angelova, D.S. Dimitrov, Liposome electroformation, Faraday Discuss Chem Soc 81: 303-311., (1986). [30] M. Ibarguren, D.J. López, L.R. Montes, J. Sot, A.I. Vasil, M.L. Vasil, F.M. Goñi, A. Alonso, Imaging the early stages of phospholipase C/sphingomyelinase activity on vesicles containing coexisting ordered-disordered and gel-fluid domains, J Lipid Res, 52 (2011) 635-645. [31] Y. Kim, S.O. Ho, N.R. Gassman, Y. Korlann, E.V. Landorf, F.R. Collart, S. Weiss, Efficient site-specific labeling of proteins via cysteines, Bioconjug Chem, 19 (2008) 786-791. [32] J. Martínez, O. Vögler, J. Casas, F. Barceló, R. Alemany, J. Prades, T. Nagy, C. Baamonde, P.G. Kasprzyk, S. Terés, C. Saus, P.V. Escribá, Membrane structure modulation, protein kinase C alpha activation, and anticancer activity of minerval, Mol Pharmacol, 67 (2005) 531-540. 20 [33] A.E. Cardenas, R. Shrestha, L.J. Webb, R. Elber, Membrane permeation of a peptide: it is better to be positive, J Phys Chem B, 119 (2015) 6412-6420. [34] G. van Meer, D.R. Voelker, G.W. Feigenson, Membrane lipids: where they are and how they behave, Nat Rev Mol Cell Biol, 9 (2008) 112-124. [35] J.F. McCallum, A. Wise, M.A. Grassie, A.I. Magee, F. Guzzi, M. Parenti, G. Milligan, The role of palmitoylation of the guanine nucleotide binding protein G11 alpha in defining interaction with the plasma membrane, Biochem J, 310 ( Pt 3) (1995) 1021-1027. [36] S. Clarke, Protein isoprenylation and methylation at carboxyl-terminal cysteine residues, Annu Rev Biochem, 61 (1992) 355-386. [37] L.F. Agnati, D. Guidolin, G. Leo, K. Fuxe, A boolean network modelling of receptor mosaics relevance of topology and cooperativity, J Neural Transm, 114 (2007) 77-92. [38] C. Huang, J.R. Hepler, L.T. Chen, A.G. Gilman, R.G. Anderson, S.M. Mumby, Organization of G proteins and adenylyl cyclase at the plasma membrane, Mol Biol Cell, 8 (1997) 2365-2378. [39] R.R. Neubig, Membrane organization in G-protein mechanisms, FASEB J, 8 (1994) 939946. [40] P.V. Escribá, A. Ozaita, C. Ribas, A. Miralles, E. Fodor, T. Farkas, J.A. García-Sevilla, Role of lipid polymorphism in G protein-membrane interactions: nonlamellar-prone phospholipids and peripheral protein binding to membranes, Proc Natl Acad Sci U S A, 94 (1997) 11375-11380. [41] J. Novotny, D. Durchankova, R.J. Ward, J.J. Carrillo, P. Svoboda, G. Milligan, Functional interactions between the alpha1b-adrenoceptor and Galpha11 are compromised by depalmitoylation of the G protein but not of the receptor., Cell Signal, 18 (2006) 1244-1251. [42] R.M. Epand, R.F. Epand, N. Ahmed, R. Chen, Promotion of hexagonal phase formation and lipid mixing by fatty acids with varying degrees of unsaturation, Chem Phys Lipids, 57 (1991) 75-80. [43] P. Oh, J.E. Schnitzer, Segregation of heterotrimeric G proteins in cell surface microdomains. G(q) binds caveolin to concentrate in caveolae, whereas G(i) and G(s) target lipid rafts by default, Mol Biol Cell, 12 (2001) 685-698. [44] A. González-Horta, D. Andreu, M.R. Morrow, J. Pérez-Gil, Effects of palmitoylation on dynamics and phospholipid-bilayer-perturbing properties of the N-terminal segment of pulmonary surfactant protein SP-C as shown by 2H-NMR, Biophys J, 95 (2008) 2308-2317. [45] K.A. Melkonian, A.G. Ostermeyer, J.Z. Chen, M.G. Roth, D.A. Brown, Role of lipid modifications in targeting proteins to detergent-resistant membrane rafts. Many raft proteins are acylated, while few are prenylated, J Biol Chem, 274 (1999) 3910-3917. [46] S.S. Funari, J. Prades, P.V. Escribá, F. Barceló, Farnesol and geranylgeraniol modulate the structural properties of phosphatidylethanolamine model membranes, Mol Membr Biol, 22 (2005) 303-311. [47] F. Barceló, J. Prades, J.A. Encinar, S.S. Funari, O. Vögler, J.M. González-Ros, P.V. Escribá, Interaction of the C-terminal region of the Ggamma protein with model membranes, Biophys J, 93 (2007) 2530-2541. [48] W.F. Simonds, J.E. Butrynski, N. Gautam, C.G. Unson, A.M. Spiegel, G-protein beta gamma dimers. Membrane targeting requires subunit coexpression and intact gamma C-A-A-X domain, J Biol Chem, 266 (1991) 5363-5366. [49] T.M. Fong, M.G. McNamee, Correlation between acetylcholine receptor function and structural properties of membranes., Biochemistry, 25 (1986) 830-840. 21 [50] P.V. Escribá, Membrane-lipid therapy: a new approach in molecular medicine, Trends Mol Med, 12 (2006) 34-43. [51] P.V. Escribá, J.M. González-Ros, F.M. Goñi, P.K. Kinnunen, L. Vigh, L. SánchezMagraner, A.M. Fernández, X. Busquets, I. Horvath, G. Barceló-Coblijn, Membranes: a meeting point for lipids, proteins and therapies, J Cell Mol Med, 12 (2008) 829-875. [52] L. Vigh, P.N. Literati, I. Horvath, Z. Torok, G. Balogh, A. Glatz, E. Kovacs, I. Boros, P. Ferdinandy, B. Farkas, L. Jaszlits, A. Jednakovits, L. Koranyi, B. Maresca, Bimoclomol: a nontoxic, hydroxylamine derivative with stress protein-inducing activity and cytoprotective effects, Nat Med, 3 (1997) 1150-1154. [53] D.J. López, M. Egido-Gabas, I. López-Montero, Busto, J. V., J. Casas, M. Garnier, F. Monroy, B. Larijani, F.M. Goñi, A. Alonso, Accumulated Bending Energy Elicits Neutral Sphingomyelinase Activity in Human Red Blood Cells. , Biophys J, 102(9) (2012) 2077-2085. [54] G. Lenaz, Lipid fluidity and membrane protein dynamics, Biosci Rep, 7 (1987) 823-837. [55] M. Ibarguren, D.J. López, J.A. Encinar, J.M. González-Ros, X. Busquets, P.V. Escribá, Partitioning of liquid-ordered/liquid-disordered membrane microdomains induced by the fluidifying effect of 2-hydroxylated fatty acid derivatives, Biochim Biophys Acta, 1828 (2013) 2553-2563. [56] P.V. Escribá, X. Busquets, J.I. Inokuchi, G. Balogh, Z. Torok, I. Horvath, J.L. Harwood, L. Vigh, Membrane lipid therapy: Modulation of the cell membrane composition and structure as a molecular base for drug discovery and new disease treatment, Prog Lipid Res, (2015). [57] V. Lladó, S. Terés, M. Higuera, R. Álvarez, M.A. Noguera-Salva, J.E. Halver, P.V. Escribá, X. Busquets, Pivotal role of dihydrofolate reductase knockdown in the anticancer activity of 2hydroxyoleic acid, Proc Natl Acad Sci, 106 (2009) 13754-13758. 22 Table 1 Binding of G proteins to LUVs in the presence or absence of PA, MA and GG Binding of Gi, G and G to LUVs of PC and PE at various molar ratios, in the presence or absence of 5 mol% PA, MA or GG. The binding of these G proteins to PC:PE (10:0, mol:mol) membranes in the absence of PA, MA or GG was considered 100%. Results are expressed as average values ± SD of 3 experiments. t-tests were used to determine statistical significance * p < 0.05, ** p < 0.01, *** p <0.001. The data in parentheses are the percent change (positive or negative) with respect to the corresponding control model membrane (same PC:PE ratio), considering the binding of each G protein to each membrane composition without PA, MA or GG as 0% change. Gi1 PC:PE Control PA MA GG 10:0 100.0± 2.6 (0) 123.6 ± 7.4* (23.6) 109.3 ± 7.1 (9.3) 86.8 ± 7 (-13.2) 8:2 89.5 ± 3.9 (0) 98.4 ± 11.4 (9.9) 64.8 ± 8.1* (-27.6) 74.3 ± 12.8 (-17) 6:4 68.7 ± 2.7 (0) 66.1 ± 9.7 (-3.8) 54.0 ± 7.4* (-21.4) 52.8 ± 7.1* (-23.2) 4:6 36.6 ± 5.5 (0) 41.3 ± 7.5 (12.8) 48.0 ± 6* (31.1) 28.5 ± 6.6 (-22.2) 2:8 22.9 ± 3.2 (0) 23.8 ± 3.6 (3.9) 32.1 ± 3.9* (41.5) 20.1 ± 6.7 (-12.3) G PC:PE Control PA MA GG 10:0 100.0± 1.9 (0) 86.0 ± 12.7 (-14.0) 104.5 ± 16.6 (4.5) 125.0 ± 12.2* (25.0) 8:2 134.9 ± 15.8 (0) 164.5 ± 5.5* (21.9) 170.1 ± 12.3* (26.1) 169.6 ± 5.4* (25.7) 6:4 207.9 ± 16.7 (0) 250.3 ± 12.2* (20.4) 210.3 ± 24.5 (1.1) 266.6 ± 5.9** (28.2) 4:6 289.3 ± 24.8 (0) 287.5 ± 47.2 (-0.6) 309.7 ± 51.0 (7.5) 298.4 ± 29.6 (9.1) 2:8 334.7 ± 9.8 (0) 277.3 ± 33.9 (-17.2) 374.5 ± 13.5 (11.9) 339.5 ± 40.3 (1.4) Gi1 PC:PE Control PA MA GG 10:0 100.0± 3.4 (0) 118.1 ± 18.5 (18.1) 95.2 ± 18.2 (-4.8) 108.8 ± 8.8 (8.8) 8:2 140.1 ± 5.4 (0) 175.3 ± 20.9* (25.1) 141.3 ± 26.8 (0.8) 179.0 ± 6.7* (27.8) 6:4 166.3 ± 14.5 (0) 228.8 ± 28.7** (37.6) 172.2 ± 23.9 (3.5) 288.8 ± 23.0** (73.7) 4:6 225.7 ± 17.9 (0) 233.2 ± 35.7 (3.3) 224.3 ± 30.9 (-0.6) 331.2 ± 32.1* (46.7) 2:8 272.5 ± 13.0 (0) 285.7 ± 30.1 (4.8) 267.7 ± 36.8 (-1.8) 371.1 ± 51.3* (36.2) 23 Table 2 Binding of G proteins to GUVs assessed by confocal microscopy. Gi, G and Gi were fluorescently labeled with Alexa Fluor 488 and the binding to GUVs composed of PC:PE was determined from the amount of fluorescence in the membrane vesicle (a.u., arbitrary units), using the fluorescence of surrounding areas as background. Data are expressed as average values ± S.D of at least 5 vesicles per condition. t-tests were used to determine statistical significance * p < 0.05, ** p< 0.01, *** p< 0.001. PC:PE Gi1 G Gi1 10:0 11.18 ± 1.23 0.53 ± 0.23 1.07 ± 0.76 8:2 3.84 ± 0.88*** 1.18 ± 0.18 2.94 ± 0.98 6:4 0.41 ± 0.3*** 3.6 ± 0.56* 3.35 ± 1.11* 4:6 No binding 6.04 ± 0.66** 7.47 ± 0.37*** 2:8 No binding 14.82 ± 1.08*** 10.51 ± 0.59*** 24 FIGURE LEGENDS Figure 1 The effect of G protein lipids on membrane lipid structure. (A) DSC thermograms of bovine liver PE membranes in the presence or absence of 5 mol% PA, MA or GG. The peaks correspond to the lamellar-to-hexagonal phase transition. (B) 31P-NMR of bovine liver PE membranes in the presence or absence of 5 mol% PA, MA or GG. NMR scans were recorded at the temperatures indicated on the left of the panel. Figure 2 Binding energies of GG, MA and PA to model membranes. Binding energies of the G protein lipids to POPC (filled bars) and POPC:POPE (6:4, mol:mol) (grey bars) membranes. The bars correspond to the mean binding energy values ± S.D. of three simulation experiments. Figure 3 Effects of GG, MA and PA on the bilayer lateral pressure. The figure shows the lateral pressure in (A) POPC and (B) POPC:POPE (6:4, mol:mol) membranes in the absence (black line) or presence of GG (red line), MA (green line) and PA (blue line). The X-axis indicates the distance from the center of the membrane. Figure 4 Effects of G protein lipids on the binding of G proteins to lipid bilayers. Model membranes composed of PC:PE (10:0) or PC:PE (6:4; molar ratio), in the presence (gray bars) or absence (filled bars) of 5 mol% PA, MA or GG. Bars correspond to the mean ± SEM of five independent experiments for the binding of Gαi1, G and G to lipid bilayers. Representative immunoblots of each graph are shown above each histogram. Anti-Gi1 was used to detect Gi1 and G, or anti-G to detect G. * indicates statistical significance (p < 0.05) compared to the same membrane composition in the absence of PA, MA or GG. Figure 5 Confocal images of G protein binding to GUVs. Gαi1, Gβγ and Gαβγ were fluorescently labeled with Alexa Fluor 488 and incubated with GUVs composed of different mol ratios of PC:PE (10:0, 8:2, 6:4, 4:6, 2:8). Lipid membranes were stained with DiI and are shown in yellow, while the fluorescently labeled G proteins appear in red. Binding of G protein subunits to lipid vesicles was assessed by quantification of the Alexa 488 fluorescence signal associated to membranes. The fluorescence surrounding the vesicle was used as background. Bar=10 µm. 25 Figure 1 32 G protein-membrane interactions II: Effect of G Protein-linked Lipids on Membrane Structure and G Protein-membrane Interactions Jesús Casasa, Maitane Ibargurena,b,*, Rafael Álvareza, David J. Lópeza,b, Silvia Terésa,b, Victoria Lladóa, Stefano P. Piottoc, Simona Conciliod, Xavier Busquetsa,b and Pablo V. Escribáa,b,* Highlights Myristoyl, palmitoyl and isoprenyl groups influence biophysical properties of membranes. Myristoyl, palmitoyl and isoprenyl moieties of G proteins modulate their binding to membranes. The isoprenyl moieties favor the accumulation of heterotrimeric G proteins in phosphatidylethanolamine-enriched domains. Changes in the lipid composition of G proteins might modulate cell signaling.