Influence of proteins on the reorganization of phospholipid bilayers into large domains

Nanodiscs: A toolkit for membrane protein science

Membrane proteins are involved in numerous vital biological processes, including transport, signal transduction and the enzymes in a variety of metabolic pathways. Integral membrane proteins account for up to 30% of the human proteome and they make up more than half of all currently marketed therapeutic targets. Unfortunately, membrane proteins are inherently recalcitrant to study using the normal toolkit available to scientists, and one is most often left with the challenge of finding inhibitors, activators and specific antibodies using a denatured or detergent solubilized aggregate. The Nanodisc platform circumvents these challenges by providing a self-assembled system that renders typically insoluble, yet biologically and pharmacologically significant, targets such as receptors, transporters, enzymes, and viral antigens soluble in aqueous media in a native-like bilayer environment that maintain a target's functional activity. By providing a bilayer surface of defined composition and structure, Nanodiscs have found great utility in the study of cellular signaling complexes that assemble on a membrane surface. Nanodiscs provide a nanometer scale vehicle for the in vivo delivery of amphipathic drugs, therapeutic lipids, tethered nucleic acids, imaging agents and active protein complexes. This means for generating nanoscale lipid bilayers has spawned the successful use of numerous other polymer and peptide amphipathic systems. This review, in celebration of the Anfinsen Award, summarizes some recent results and provides an inroad into the current and historical literature.

Probing the extended lipid anchorage with cytochrome c and liposomes containing diacylphosphatidylglycerol lipids

Experiments investigating the adsorption and desorption of cytochrome c onto and from liposomes containing 50 mol% 1,2-diacylphosphatidylglycerol lipids [10:0, 12:0, 14:0, 16:0, 18:1(Δ9 cis)] with 1,2-dioleoyl-sn-glycero-3-phosphatidylcholine (DOPC) in pH 7.4 buffered solutions of low to moderate ionic strength are reported. Fluorescence experiments show that cytochrome c has a similar adsorption affinity for the five labeled 50 mol% PG liposome systems investigated. Fluorescence recovery experiments reveal the extent of cytochrome c desorption upon the addition of >10× excess of unlabeled 100% 1,2-dioleoyl-sn-glycero-3-phosphatidylglycerol (DOPG) liposomes is dependent on the lipid's acyl chain length. The extent of desorption is also shown to be independent of temperature, albeit over a narrow range. The differences in the extent of cytochrome c desorption from liposomes containing PG lipids with different acyl chain lengths is attributed to the varying contribution of the binding motif involving the extended lipid anchorage in response to lipid packing stress.

Lateral dynamics of charged lipids and peripheral proteins in spatially heterogeneous membranes: comparison of continuous and Monte Carlo approaches

Biological membranes are complex environments whose physico-chemical properties are of utmost importance for the understanding of many crucial biological processes. Much attention has been given in the literature to the description of membranes along the z-axis perpendicular to the membrane. Here, we instead consider the lateral dynamics of lipids and peripheral proteins due to their electrostatic interaction. Previously, we constructed a Monte Carlo automaton capable of simulating mutual diffusive dynamics of charged lipids and associated positively charged peptides. Here, we derive and numerically analyze a system of Poisson-Boltzmann-Nernst-Planck (PBNP) equations that provide a mean-field approximation compatible with our Monte Carlo model. The thorough comparison between the mean-field PBNP equations and Monte Carlo simulations demonstrates that both the approaches are in a good qualitative agreement in all tested scenarios. We find that the two methods quantitatively deviate when the local charge density is high, presumably because the Poisson-Boltzmann formalism is applicable in the so-called weak coupling limit, whose validity is restricted to low charge densities. Nevertheless, we conclude that the mean-field PBNP approach provides a good approximation for the considerably more detailed Monte Carlo model at only a fraction of the associated computational cost and allows simulation of the membrane lateral dynamics on the space and time scales relevant for the realistic biological problems.

Morphological changes of supported lipid bilayers induced by lysozyme: planar domain formation vs. multilayer stacking

Total internal reflection fluorescence microscopy (TIRFM) has been utilized to explore the effect of cationic protein lysozyme (Lz) on the morphology of solid-supported lipid bilayers (SLBs) comprised of zwitterionic lipid phosphatidylcholine (PC) and its mixture with anionic lipid cardiolipin (CL). Kinetic TIRFM imaging of different systems revealed subtle interplay between lipid lateral segregation accompanied by exchange of neutral and acidic lipids in the protein-lipid interaction zone, and the formation of lipid multilayer stacks. The switch between these states was shown to be controlled by CL content. In weakly charged SLBs containing 5 mol% CL, assembling of CL molecules into planar domains upon Lz adsorption has been observed while at higher content of anionic lipid (25 mol%) in-plane domains tend to transform into multilayer stacks, thereby ensuring the most thermodynamically-favorable membrane conformation.

Cardiolipin clusters and membrane domain formation induced by mitochondrial proteins

We show in this study that mitochondrial creatine kinase promotes segregation and clustering of cardiolipin in mixed membranes, a phenomenon that has been proposed to occur at contact sites in the mitochondria. This property of mitochondrial creatine kinase is dependent on the native octameric structure of the protein and does not occur after heat-denaturation or with the native dimeric form of the protein. Cardiolipin segregation was demonstrated by differential scanning calorimetry using membranes containing cardiolipin and either dipalmitoylphosphatidylethanolamine or 1-palmitoyl-2-oleoylphosphatidylethanolamine. Addition of the ubiquitous form of mitochondrial creatine kinase leads to the formation of a phosphatidylethanolamine-rich domain as a result of the protein binding preferentially to the cardiolipin. Such phase separation does not occur if cardiolipin is replaced with dioleoyl phosphatidylglycerol. Lipid phase separation is observed with other cardiolipin-binding proteins, including cytochrome c and, to a very small extent, with truncated Bid (t-Bid), as well as with the cationic polypeptide poly-L-lysine, but among these proteins the octameric form of mitochondrial creatine kinase is by far the most effective in causing segregation and clustering of cardiolipin. The proteins included in this study are found at mitochondrial contact sites where they are known to associate with cardiolipin. Domains in mitochondria enriched in cardiolipin play an important role in apoptosis and in energy flux processes.

To see or not to see: Lateral organization of biological membranes and fluorescence microscopy

In the last few years several experimental strategies based on epi-, confocal and two photon excitation fluorescence microscopy techniques have been employed to study the lateral structure of membranes using giant vesicles as model systems. This review article discusses the methodological aspects of the aforementioned experimental approaches, particularly stressing the information obtained by the use of UV excited fluorescent probes using two-photon excitation fluorescence microscopy. Additionally, the advantages of utilizing visual information, to correlate the lateral structure of compositionally simple membranes with complex situations, i.e., biological membranes, will be addressed.

Cytochrome C interaction with cardiolipin/phosphatidylcholine model membranes: effect of cardiolipin protonation

Resonance energy transfer between anthrylvinyl-labeled phosphatidylcholine as a donor and heme moiety of cytochrome c (cyt c) as an acceptor has been employed to explore the protein binding to model membranes, composed of phosphatidylcholine and cardiolipin (CL). The existence of two types of protein-lipid complexes has been hypothesized where either deprotonated or partially protonated CL molecules are responsible for cyt c attachment to bilayer surface. To quantitatively describe cyt c membrane binding, the adsorption model based on scaled particle and double layer theories has been employed, with potential-dependent association constants being treated as a function of acidic phospholipid mole fraction, degree of CL protonation, ionic strength, and surface coverage. Multiple arrays of resonance energy transfer data obtained under conditions of varying pH, ionic strength, CL content, and protein/lipid molar ratio have been analyzed in terms of the model of energy transfer in two-dimensional systems combined with the adsorption model allowing for area exclusion and electrostatic effects. The set of recovered model parameters included effective protein charge, intrinsic association constants, and heme distance from the bilayer midplane for both types of protein-lipid complexes. Upon increasing CL mole fraction from 10 to 20 mol % (the value close to that characteristic of the inner mitochondrial membrane), the binding equilibrium dramatically shifted toward cyt c association with partially protonated CL species. The estimates of heme distance from bilayer center suggest shallow bilayer location of cyt c at physiological pH, whereas at pH below 6.0, the protein tends to insert into membrane core.

Coverage-dependent changes of cytochrome c transverse location in phospholipid membranes revealed by FRET

The method of fluorescence resonance energy transfer (FRET) has been employed to monitor cytochrome c interaction with bilayer phospholipid membranes. Liposomes composed of phosphatidylcholine and varying amounts of anionic lipid cardiolipin (CL) were used as model membranes. Trace amount of fluorescent lipid derivative, anthrylvinyl-phosphatidylcholine was incorporated into the membranes to serve energy donor for heme moiety of cytochrome c. Energy transfer efficiency was measured at different lipid and protein concentrations to obtain extensive set of data, which were further analyzed globally in terms of adequate models of protein adsorption and energy transfer on the membrane surface. It has been found that the cytochrome c association with membranes containing 10 mol% CL can be described in terms of equilibrium binding model (yielding dissociation constant Kd = 0.2-0.4 microM and stoichiometry n = 11-13 lipid molecules per protein binding site) combined with FRET model assuming uniform acceptor distribution with the distance of 3.5-3.6 nm between the bilayer midplane and heme moiety of cytochrome c. However, increasing the CL content to 20 or 40 mol% (at low ionic strength) resulted in a different behavior of FRET profiles, inconsistent with the concepts of equilibrium adsorption of cytochrome c at the membrane surface and/or uniform acceptor distribution. To explain this fact, several possibilities are analyzed, including cytochrome c-induced formation of non-bilayer structures and clusters of charged lipids, or changes in the depth of cytochrome c penetration into the bilayer depending on the protein surface density. Additional control experiments have shown that only the latter process can explain the peculiar concentration dependences of FRET at high CL content.

Cytochrome c location in phosphatidylcholine/cardiolipin model membranes: resonance energy transfer study

Resonance energy transfer between lipid-bound fluorescent probe 3-methoxybenzanthrone as a donor and heme group of cytochrome c as an acceptor has been examined to ascertain the protein disposition relative to the surface of model membranes composed of phosphatidylcholine and cardiolipin (10, 50 and 80 mol%). The model of energy transfer in membrane systems has been extended to the case of donors distributed between the two-bilayer leaflets and acceptors located at the outer monolayer taking into account the donor and acceptor orientational behavior. Assuming specific protein orientation relative to the membrane surface and varying lateral distance of the donor-acceptor closest approach in the range from 0 to 3.5 nm the limits for possible heme distances from the bilayer midplane have been found to be 0.8-3 nm (10 mol% CL), 0-2.6 nm (50 mol% CL), and 1.4-3.3 nm (80 mol% CL).

Caveolin scaffolding region and the membrane binding region of SRC form lateral membrane domains

Formation of domains by the membrane binding motifs of caveolin and src were studied in large unilamellar vesicles using fluorescence digital imaging microscopy. Caveolin, a major structural protein of caveolae, contains a scaffolding region (residues 82-101) that contributes to the binding of the protein to the plasma membrane. A caveolin peptide (82-101) corresponding to this scaffolding region induced the formation of membrane domains enriched in the acidic lipids phosphatidylserine and phosphatidylinositol-4,5-bisphosphate. Cholesterol, another predominant component of caveolae, was also enriched in these domains. Caveolae also contain many different signaling molecules including src family tyrosine kinases. Src proteins bind to the plasma membrane via a N-terminal myristate chain and a cluster of basic residues that can interact electrostatically with negatively charged lipids. A peptide corresponding to the src membrane binding motifs (residues myr-2-19) sequestered acidic lipids into lateral membrane domains. Both the src and the caveolin peptides colocalized together with acidic lipids in the domains. Control experiments show the domains are not the result of vesicle aggregation. Two-photon fluorescence correlation spectroscopy experiments suggest diffusion in the domains was slower, but the domains were dynamic. Protein kinase C phosphorylated src in its N-terminal membrane binding region; however, the caveolin scaffolding peptide inhibited this activity. Consequently, protein-induced membrane domains may affect cell signaling by organizing signal transduction components within the membrane and changing reaction rates.

Stabilization of O-pyromellitylgramicidin channels in bilayer lipid membranes through electrostatic interaction with polylysines of different chain lengths

Functioning of membrane proteins, in particular ionic channels, can be modulated by alteration of their arrangement in membranes. We addressed this issue by studying the effect of different chain length polylysines on the kinetics of ionic channels formed in a bilayer lipid membrane (BLM) by O-pyromellitylgramicidin carrying three negative charges at the C-terminus. The method of sensitized photoinactivation was applied to the analysis of the channel association-dissociation kinetics (characterized by the exponential factor of the curve describing the time course of the flash-induced decrease in the transmembrane current, tau). Addition of polylysine to the bathing solutions of BLM led to the deceleration of the photoinactivation kinetics, i.e. to the increase in tau. It was shown here that for a series of polylysines differing in their chain lengths, the value of tau grew as their concentration increased above a threshold level until at a certain concentration of each polylysine tau reached maximum. At higher polylysine concentrations tau began to decrease and finally became close to the control level observed in the absence of polylysine. With lengthening of the polylysine chain the maximum value of tau increased, the concentration dependence became steeper, and the threshold concentration decreased. The increase in the ionic strength of the medium shifted the concentration dependence of tau to higher polylysine concentrations and decreased the maximum value of tau. It was concluded that the increase in tau was caused by the formation of domains of O-pyromellitylgramicidin molecules induced by binding of polylysines. This can be related to functional aspects of polycation-induced sequestering of negatively charged transmembrane peptides in neutral membranes.

A correlation between lipid domain shape and binary phospholipid mixture composition in free standing bilayers: A two-photon fluorescence microscopy study

Giant unilamellar vesicles (GUVs) composed of different phospholipid binary mixtures were studied at different temperatures, by a method combining the sectioning capability of the two-photon excitation fluorescence microscope and the partition and spectral properties of 6-dodecanoyl-2-dimethylamino-naphthalene (Laurdan) and Lissamine rhodamine B 1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (N-Rh-DPPE). We analyzed and compared fluorescence images of GUVs composed of 1,2-dilauroyl-sn-glycero-3-phosphocholine/1, 2-dipalmitoyl-sn-glycero-3-phosphocholine (DLPC/DPPC), 1, 2-dilauroyl-sn-glycero-3-phosphocholine/1, 2-distearoyl-sn-glycero-3-phosphocholine (DLPC/DSPC), 1, 2-dilauroyl-sn-glycero-3-phosphocholine/1, 2-diarachidoyl-sn-glycero-3-phosphocholine (DLPC/DAPC), 1, 2-dimyristoyl-sn-glycero-3-phosphocholine/1, 2-distearoyl-sn-glycero-3-phosphocholine (DMPC/DSPC) (1:1 mol/mol in all cases), and 1,2-dimyristoyl-sn-glycero-3-phosphoethanolamine/1, 2-dimyristoyl-sn-glycero-3-phosphocholine (DMPE/DMPC) (7:3 mol/mol) at temperatures corresponding to the fluid phase and the fluid-solid phase coexistence. In addition, we studied the solid-solid temperature regime for the DMPC/DSPC and DMPE/DMPC mixtures. From the Laurdan intensity images the generalized polarization function (GP) was calculated at different temperatures to characterize the phase state of the lipid domains. We found a homogeneous fluorescence distribution in the GUV images at temperatures corresponding to the fluid region for all of the lipid mixtures. At temperatures corresponding to phase coexistence we observed concurrent fluid and solid domains in the GUVs independent of the lipid mixture. In all cases the lipid solid domains expanded and migrated around the vesicle surface as we decreased the temperature. The migration of the solid domains decreased dramatically at temperatures close to the solid-fluid-->solid phase transition. For the DLPC-containing mixtures, the solid domains showed line, quasicircular, and dendritic shapes as the difference in the hydrophobic chain length between the components of the binary mixture increases. In addition, for the saturated PC-containing mixtures, we found a linear relationship between the GP values for the fluid and solid domains and the difference between the hydrophobic chain length of the binary mixture components. Specifically, at the phase coexistence temperature region the difference in the GP values, associated with the fluid and solid domains, increases as the difference in the chain length of the binary mixture component increases. This last finding suggests that in the solid-phase domains, the local concentration of the low melting temperature phospholipid component increases as the hydrophobic mismatch decreases. At the phase coexistence temperature regime and based on the Laurdan GP data, we observe that when the hydrophobic mismatch is 8 (DLPC/DAPC), the concentration of the low melting temperature phospholipid component in the solid domains is negligible. This last observation extends to the saturated PE/PC mixtures at the phase coexistence temperature range. For the DMPC/DSPC we found that the nonfluorescent solid regions gradually disappear in the solid temperature regime of the phase diagram, suggesting lipid miscibility. This last result is in contrast with that found for DMPE/DMPC mixtures, where the solid domains remain on the GUV surface at temperatures corresponding to that of the solid region. In all cases the solid domains span the inner and outer leaflets of the membrane, suggesting a strong coupling between the inner and outer monolayers of the lipid membrane. This last finding extends previous observations of GUVs composed of DPPE/DPPC and DLPC/DPPC mixtures (, Biophys. J. 78:290-305).

Distribution and stability of membrane proteins in lipid membranes on solid supports

Bacteriorhodopsin and the nicotinic acetylcholine receptor were biotinylated and reconstituted in lipidic membranes on silicon supports by fusion with proteoliposomes. The presence and distribution of the proteins were studied by binding with streptavidin. Radio-labelled streptavidin was employed for quantifying the amounts of protein remaining in the supported membranes after storage in buffer. The proteins within the membranes remained bound to the surface for weeks. The biological activity of reconstituted unlabelled receptor upon storage showed stability in membranes formed on silicon supports and a reduced stability when formed onto lipid monolayer covered supports. Atomic force microscopy studies on preparations in liquid showed bilayer structures but also attached, partly fused liposomes and membrane particles. In air, the surface was smoother and contained less of liposomes and more of stacked lipid layers. Preparations labelled with streptavidin conjugated to colloidal gold and imaged in air showed the proteins individually distributed, with no protein-rich patches or protein aggregates.

Unusual electrostatic effects on binding of C1q to anionic liposomes: role of anionic phospholipid domains and their line tension

The binding of 125I-C1q to anionic liposomes was studied as a function of protein concentration, pH, ionic strength, and anionic lipid composition. The maximum amount of protein bound per micromole of lipid was very sensitive to electrostatic factors, increasing strongly with decreased pH and ionic strength or increased anionic lipid content. The apparent association constant was independent of these electrostatic factors, however, in marked contrast to studies on basic peptide binding to anionic lipid vesicles. Microscopic observations of large unilamellar liposomes containing fluorescently labeled C1q or phosphatidylglycerol demonstrated, under conditions causing strong electrostatic interactions, that C1q and anionic lipids colocalized into domains whose radii of curvature were higher than that of the surrounding lipid. These domains were observed to bud and pinch off into brightly fluorescent vesicles. We propose a model for all of these observations in which the line tension or edge energy at the boundary of the domain resists its increase in circumference as the domain grows by electrostatic effects on binding, eventually resulting in vesiculation. We propose that under favorable electrostatic conditions, as larger domains form the edge energy balances the increases in the electrostatic contribution to binding, resulting in a net binding energy independent of electrostatic factors.

Binding of peripheral proteins to mixed lipid membranes: effect of lipid demixing upon binding

Binding isotherms have been determined for the association of horse heart cytochrome c with dioleoyl phosphatidylglycerol (DOPG)/dioleoyl phosphatidylcholine (DOPC) bilayer membranes over a range of lipid compositions and ionic strengths. In the absence of protein, the DOPG and DOPC lipids mix nearly ideally. The binding isotherms have been analyzed using double layer theory to account for the electrostatics, either the Van der Waals or scaled particle theory equation of state to describe the protein surface distribution, and a statistical thermodynamic formulation consistent with the mass-action law to describe the lipid distribution. Basic parameters governing the electrostatics and intrinsic binding are established from the binding to membranes composed of anionic lipid (DOPG) alone. Both the Van der Waals and scaled particle equations of state can describe the effects of protein distribution on the DOPG binding isotherms equally well, but with different values of the maximum binding stoichiometry (13 lipids/protein for Van der Waals and 8 lipids/protein for scaled particle theory). With these parameters set, it is then possible to derive the association constant, Kr, of DOPG relative to DOPC for surface association with bound cytochrome c by using the binding isotherms obtained with the mixed lipid membranes. A value of Kr (DOPG:DOPC) = 3.3-4.8, depending on the lipid stoichiometry, is determined that consistently describes the binding at different lipid compositions and different ionic strengths. Using the value of Kr obtained it is possible to derive the average in-plane lipid distribution and the enhancement in protein binding induced by lipid redistribution using the statistical thermodynamic theory.

The influence of poly-(L-lysine) and porin on the domain structure of mixed vesicles composed of lipopolysaccharide and phospholipid: an infrared spectroscopic study

Fourier transform infrared (FTIR) spectroscopy has been used to study the thermotropic phase behavior of binary lipid mixtures composed of deuterated phospholipids (PLs) and lipopolysaccharides (LPSs). Furthermore, the influence of an extrinsic high-molecular, polycationic polypeptide (poly-(L-lysine), PLL(500)) and an intrinsic membrane protein (outer membrane protein F, OmpF) on these binary mixtures was investigated by FTIR spectroscopy. "Deep rough" mutant LPS (ReLPS), isolated from Salmonella minnesota R595, and perdeuterated 1,2-dimyristoylphosphatidylethanolamine (DMPEd54) were used as model lipids. Deuteration of one of the lipids permitted the detection of lipid protein interaction with each lipid component separately. For this purpose, the symmetric >CH2 and >CD2 stretching bands were utilized as specific monitors to scrutinize the state of order of the membranes. From the individual phase transition temperatures Tm and the shape of the phase transition profiles, it is established that ReLPS and DMPEd54 are molecularly immiscible. In addition to the two domains of the pure lipid components, a third, domain-like structure is detected that may coexist with these pure domains. This domain-like structure undergoes a gel to liquid-crystalline L1 (beta <--> alpha) phase transition at temperatures distinctly different from that of the respective pure lipid domains. The nature of this type of domain is discussed in terms of a "border region" model that adequately explains the experimentally observed complex phase transition profiles. It is further demonstrated that the extrinsic polycationic polypeptide PLL(500) and the intrinsic, pore-forming protein OmpF isolated from Escherichia coli interact preferentially and highly specifically with the negatively charged ReLPS. Both the synthetic polypeptide and the pore-forming protein increased the tendency of ReLPS and DMPEd54 to segregate into distinct, well-separated domains. Whereas the transition profiles of the ternary system ReLPS/DMPEd54/PLL(500) showed the features of a phase segregation phenomenon not affecting the transition temperatures of the pure lipid components, the ternary system composed of ReLPS/DMPEd54 and OmpF exhibited phase transition curves that were characterized by an unspecific (DMPEd54/OmpF) and a strong and unique (ReLPS/OmpF) type of lipid-protein interaction. Furthermore, semiquantitative estimations supported the supposition that OmpF might be able to induce bilayer asymmetry in preformed symmetrical ReLPS/DMPEd54 vesicles.

Domains in cationic lipid plus polyelectrolyte bilayer membranes: detection and characterization via 2H nuclear magnetic resonance

2H nuclear magnetic resonance (NMR) spectroscopy of choline-deuterolabeled 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC-alpha-d2 and POPC-beta-d2) has been used to detect and quantify domain formation induced in cationic lipid-containing bilayers upon the addition of anionic polyelectrolytes. Three different polyelectrolytes, poly(sodium 4-styrenesulfonate) or PSSS, poly(sodium acrylate) or PACA, and poly(sodium glutamate) or PGLU, were added to POPC lipid bilayers containing 1,2-dioleoyl-3-(dimethylamino)propane (DODAP) as the cationic amphiphile. All three polyelectrolytes produced two-component 2H NMR spectra, consistent with two populations of POPC, one polyelectrolyte-bound and another polyelectrolyte-free. The relative intensities of the two spectral components provided the relative amounts of the two POPC populations. The 2H NMR quadrupolar splitting from either spectral component provided the DODAP content of the particular POPC population. The two POPC populations differed in that the polyelectrolyte-bound population contained a stoichiometric polyelectrolyte anion:DODAP cation ratio leading to enrichment with respect to DODAP, while the polyelectrolyte-free population was depleted of DODAP. Estimates of the size of a polyelectrolyte-defined domain revealed a constant number of bound DODAP but a flexible number of bound POPC, which increased in proportion to the global POPC content. The most compact domains were formed by the most hydrophobic polyelectrolyte, PSSS, while the most expansive domains were formed by the most hydrophilic polyelectrolyte, PGLU.

Spin-label electron spin resonance studies on the interactions of lysine peptides with phospholipid membranes

The interactions of lysine oligopeptides with dimyristoyl phosphatidylglycerol (DMPG) bilayer membranes were studied using spin-labeled lipids and electron spin resonance spectroscopy. Tetralysine and pentalysine were chosen as models for the basic amino acid clusters found in a variety of cytoplasmic membrane-associating proteins, and polylysine was chosen as representative of highly basic peripherally bound proteins. A greater motional restriction of the lipid chains was found with increasing length of the peptide, while the saturation ratio of lipids per peptide was lower for the shorter peptides. In DMPG and dimyristoylphosphatidylserine host membranes, the perturbation of the lipid chain mobility by polylysine was greater for negatively charged spin-labeled lipids than for zwitterionic lipids, but for the shorter lysine peptides these differences were smaller. In mixed bilayers composed of DMPG and dimyristoylphosphatidylcholine, little difference was found in selectivity between spin-labeled phospholipid species on binding pentalysine. Surface binding of the basic lysine peptides strongly reduced the interfacial pK of spin-labeled fatty acid incorporated into the DMPG bilayers, to a greater extent for polylysine than for tetralysine or pentalysine at saturation. The results are consistent with a predominantly electrostatic interaction with the shorter lysine peptides, but with a closer surface association with the longer polylysine peptide.

Formation of membrane domains during the activation of protein kinase C

The lateral membrane organization of phosphatidylserine, diacylglycerol, substrate, and Ca(2+)-dependent protein kinase C in large unilamellar vesicles was investigated by using fluorescence digital imaging microscopy. The formation of phosphatidylserine domains could be induced by either Ca2+, the MARCKS peptide, or protein kinase C. However, only Ca2+ could induce diacylglycerol to partition into the phosphatidylserine domains. In the complete protein kinase C assay mixture, two separate triple-labeling experiments demonstrated the colocalization of phosphatidylserine, protein kinase C, diacylglycerol, and the MARCKS peptide in domains. The amounts of all the labeled components in whole vesicles and in domains were measured at various concentrations of either phosphatidylserine, Ca2+, diacylglycerol, or the MARCKS peptide or with the addition of polylysine. The role of each component in forming membrane domains and in mediating the enzyme activity was analyzed. The results indicated that the inclusion of the MARCKS peptide in the domains, not just the binding of the substrate to vesicles, was especially important for PKC activity. The formation of PKC domains required the presence of DAG and Ca2+ at physiological ionic strength. The PKC activity was proportional to the amounts of PKC and substrate in the domains. The results also showed that the MARCKS peptide left the domains after being phosphorylated. A model for the activation of protein kinase C involving sequestering of the reaction components into membrane domains is proposed. The efficiency of the reaction was greatly increased by concentrating the activators, the enzyme, and the substrate into domains.

Fluorescence signaling of ligand binding and assembly in metal-chelating lipid membranes

BACKGROUND

Chemical information is sometimes transmitted across cell membranes by ligand-induced assembly of receptors. We have previously designed a series of lipids with metal-chelating headgroups that can serve as general receptors for proteins containing accessible histidines. Such lipids can also be derivatized with pyrene, a fluorescent probe that has a different emission maximum when it is aggregated (excimer fluorescence) from that seen for the monomer. We set out to examine whether lipids of this kind would produce a signal in response to ligand binding.

RESULTS

A model ligand, poly-L-histidine (poly(His)), bound specifically to pyrene-labeled Cu(II)-iminodiacetate lipid (Cu-PSIDA) within a membrane matrix. Binding of poly(His) induces the redistribution of Cu-PSIDA, so that it forms pyrene-rich domains that are detectable by the increased ratio of excimer to monomer fluorescence. Using rhodamine-labeled poly(His), we have shown that the receptor lipid domains correspond to poly(His)-rich domains below the lipid interface.

CONCLUSIONS

The Cu-PSIDA receptor signals binding of the macromolecular ligand through its excimer fluorescence and allows the resulting domains formed by ligand assembly to be imaged. Fluorescent Cu-PSIDA can thus serve as an optical reporter of ligand-induced lipid reorganization.

Lateral inhomogeneous lipid membranes: theoretical aspects

The physical concepts underlying the lateral distribution of the components forming a lamellar assembly of amphiphiles are discussed in this review. The role of amphiphiles' molecular structure and/or aqueous environment (ionic strength, water soluble substances) on formation and stability of lateral patterns is investigated. A considerable effort is devoted to the analysis of the properties of patterned structure which can be different from those of randomly mixed multi-component lamellae. Examples include adhesion and fusion among laterally inhomogeneous bilayers, enhanced interfacial adsorption of ions and polymers, enhanced transport across the bilayer, modified mechanical properties, local stabilization of non-planar geometries (pores, edges) and related phenomena (electroporation, budding transition and so on). Furthermore, an analysis of chemical reactivity within or at the water interface of a laterally inhomogeneous bilayer is briefly discussed. A link between these concepts and experimental findings taken from the biological literature is attempted throughout the review.

Lipid dynamics and peripheral interactions of proteins with membrane surfaces

A large body of evidence strongly indicates biomembranes to be organized into compositionally and functionally specialized domains, supramolecular assemblies, existing on different time and length scales. For these domains and intimate coupling between their chemical composition, physical state, organization, and functions has been postulated. One important constituent of biomembranes are peripheral proteins whose activity can be controlled by non-covalent binding to lipids. Importantly, the physical chemistry of the lipid interface allows for a rapid and reversible control of peripheral interactions. In this review examples are provided on how membrane lipid (i) composition (i.e., specific lipid structures), (ii) organization, and (iii) physical state can each regulate peripheral binding of proteins to the lipid surface. In addition, a novel and efficient mechanism for the control of the lipid surface association of peripheral proteins by [Ca2+], lipid composition, and phase state is proposed. The phase state is, in turn, also dependent on factors such as temperature, lateral packing, presence of ions, metabolites and drugs. Confining reactions to interfaces allows for facile and cooperative large scale integration and control of metabolic pathways due to mechanisms which are not possible in bulk systems.

Lipid domains in model and biological membranes

Lipid domains that occur within biological of model membranes encompass a variety of structures with very different lifetimes. The separation of membrane lipids into compositional domains can be due to lateral phase separation, immiscibility within a single phase, or interaction of lipids with integral or peripheral proteins. Lipid domains can affect the extent and rate of reactions in the membrane and provide sites for the activity of specialized proteins. Domains are likely to be involved in the process of lipid sorting to various cellular membranes, as well as in other processes which involve membrane budding or invagination.

Formation of membrane domains by the envelope proteins of vesicular stomatitis virus

The properties of the two envelope-associated proteins of vesicular stomatitis virus, the glycoprotein (G) and the matrix protein (M), were investigated in order to understand the mechanism of virus budding and domain formation in membranes. Fluorescence resonance energy transfer was used to study the interaction between the G protein and specific phospholipids. The protein had the highest affinity for phosphatidic acid among the phospholipids tested. Fluorescence digital imaging microscopy also was used to determine how the protein could alter the lateral distribution of phospholipids in membranes. Large domains enriched in phosphatidic acid were observed when the protein was incorporated into phospholipid vesicles. The G protein colocalized with the phosphatidic acid-enriched domains. Similar experiments carried out with the M protein showed that the M protein induced the formation of domains enriched not only in phosphatidic acid but also in phosphatidylserine. The phosphatidic acid-enriched domains induced by either the G or M proteins were similar in terms of the degree of enrichment of phosphatidic acid and the size of the domains. When the two proteins were reconstituted in vesicles at the same time, the domains were condensed. There was a greater degree of phosphatidic acid enrichment, and the size of the domains was reduced. The formation of domains enriched in the viral proteins and specific phospholipids may mimic the first steps that occur during budding of the virus from the plasma membrane of infected cells.

Pyrularia thionin binding to and the role of tryptophan-8 in the enhancement of phosphatidylserine domains in erythrocyte membranes

Pyrularia thionin is a small, strongly basic peptide which interacts readily with cellular and synthetic membranes. With cells it induces hemolysis, depolarizes the cellular membrane with an accompanying influx of Ca2+, and activates an endogenous phospholipase A2. Evidence points toward a binding site involving phosphatidylserine (PS). This study shows that addition of the peptide to erythrocyte membranes as well as to vesicles formed from phospholipids isolated from erythrocyte membranes causes an enhancement of phospholipid domains which are made visible by the use of fluorescence digital imaging microscopy with fluorescent derivatives of PS (NBD-PS) and phosphatidylcholine (NBD-PC). Addition of thionin caused a large increase in NBD-PS domains, with an accompanying enrichment of NBD-PC in another separate domain. Double-labeling experiments performed with a Texas Red derivative of thionin show that the peptide binds to the domain enriched in NBD-PS. P thionin inactivated by modification of Trp-8 with N-bromosuccinimide lost the ability to enhance PS domains, although it bound to the membrane with the same affinity as native P thionin. This shows that binding to the membrane is not in itself sufficient to cause the NBD-PS and NBD-PC redistribution into domains.

Interaction of glycosphingolipids and glycoproteins: thermotropic properties of model membranes containing GM1 ganglioside and glycophorin

High-sensitivity differential scanning calorimetry (DSC) was used to study the mutual interactions between a glycoprotein (human glycophorin, GPA) and a glycosphingolipid (GM1 ganglioside) embedded in large unilamellar vesicles composed of dimyristoylphosphatidylcholine (DMPC). The DSC thermograms exhibited by DMPC/GM1 vesicles, either in the presence or in the absence of GPA, are resolvable into two components. The relative contribution of the minor component, centered at higher temperature, to the total enthalpy and its transition temperature increase with the concentration of the glycolipid embedded in the vesicles. This minor peak, undetectable in the absence of ganglioside, is indicative of the occurrence of lateral phase separation and suggests that GM1 ganglioside-enriched domains are present within the bilayer. At a given concentration of GM1 embedded in the vesicles, the proportion of the phase-separated peak is higher in the presence of GPA, suggesting that the glycoprotein enhances the tendency of GM1 to segregate. Experiments investigating the thermotropic behavior of GPA show that the temperature of irreversible thermal unfolding of the glycoprotein inserted in DMPC vesicles, centered at 65.9 degrees C in the absence of GM1, is shifted to 57.6 degrees C when GM1 is present in the bilayer. These results indicate that, at least in this experimental system, on the one hand, GPA enhances the tendency of the glycolipid to segregate within the membrane, and on the other hand, the glycolipid clusters affect the protein conformation and oligomerization in the membrane.

Epifluorescence microscopic studies of monolayers containing mixtures of dioleoyl- and dipalmitoylphosphatidylcholines

Monolayers of dipalmitoylphosphatidylcholine (DPPC), dioleoylphosphatidylcholine (DOPC), and some mixtures of these lipids were investigated using an epifluorescence microscopic surface balance. Monolayers were visualized at 23 +/- 1 degree C through the fluorescence of 1 mol% of two different fluorescent probes, 1-palmitoyl-2-(12-[(7-nitro-2-1,3-benzoxadizole-4- yl)amino]dodecanoyl)phosphatidylcholine (NBD-PC), which partitions into the liquid expanded (LE) or disordered lipid phase and 3,3'-dioctadecyloxacarbocyanine perchlorate (DiO-C18), which preferentially associates with the liquid condensed (LC) phase or lipid with ordered chains. LC domains were observed in pure DPPC monolayers at relatively low surface pressures (pi), and these domains grew with increasing surface pressure. Only liquid expanded phase was observed in pure DOPC monolayers up to the point of monolayer collapse. In monolayers containing 29:70:1, 49:50:1, and 69:30:1 (mol/mol/mol) of DPPC:DOPC:probe the domains of LC phase were smaller than those seen in DPPC monolayers at equivalent surface pressures. Quantitative analysis of the visual fields shown by the mixed monolayers showed a distribution of sizes of condensed domains at any given pi. At pi = 30 mN m-1, liquid-expanded, or fluid, regions occupied more than 70% of the total monolayer area in all three mixtures studied, whereas DPPC monolayers were more than 75% condensed or solid at that pressure. For monolayers of DPPC:DOPC:NBD-PC 49:50:1 and 69:30:1 the average domain size and the percentage of the total area covered with LC, or rigid, areas increased to a maximum at pi around 35 mN m-1 followed by a decrease at higher pi. Repetitive compression and expansion of the monolayers containing DPPC:DOPC:NBD-PC 49:50:1 at an initial rate of 3.2 A2 molecule-1 s-1 produced monolayers with visual properties consistent with there being a preferential exclusion of the unsaturated lipid from the monolayer.

Characterization of structural and functional phosphoinositide domains in human erythrocyte membranes

In the erythrocyte membrane, only a fraction (50-60%) of phosphatidylinositol 4,5-bisphosphate (PIP2) and of phosphatidylinositol 4-phosphate (PIP) is rapidly turned over by specific kinases and phosphatases and accessible to hydrolysis by the polyphosphoinositide (PPI)-specific phospholipase C (PLC). To investigate whether the metabolic segregation of PPI resulted from preferential interactions with proteins, we have measured the accessibility of PPI to bee venom phospholipase A2 (PLA2) in native erythrocyte membranes, or after treatments designed to remove peripheral proteins and cytoplasmic domains of integral proteins. In native membranes, PPI, as well as the other major phospholipids, behaved as two distinct fractions (R1 and R2) differing by their sensitivity to PLA2. Such a behavior was not observed in PIP and PIP2 containing artificial vesicles. Evidence was provided that the highly sensitive fraction of PIP and PIP2 (R1) may be identical to the PLC-sensitive and rapidly metabolized pool. Removal of peripheral proteins, followed by proteolysis of the cytoplasmic domain of integral proteins, mainly glycophorins and band 3, led to a reduction of the R1 fraction of PIP and of PIP2. It is proposed that the rapidly metabolized pool of PIP2 and PIP, involved in the regulation of major cellular functions, would be maintained in its functional state through interactions with integral proteins.

Deformations in the cytoplasmic membrane of Escherichia coli direct the synthesis of peptidoglycan. The hernia model

To explain the growth of the Gram-negative envelope and in particular how it could be strengthened where it is weakest, we propose in the hernia model that local weakening of the peptidoglycan sacculus allows turgor pressure to cause the envelope to bulge outwards in a hernia; the consequent local alteration in the radius of curvature of the cytoplasmic membrane causes local alterations in phospholipid structure and composition that determine both the synthesis and hydrolysis of peptidoglycan. This proposal is supported by evidence that phospholipid composition determines the activity of phospho-N-acetylmuramic acid pentapeptide translocase, UDP-N-acetylglucosamine:N-acetylmuramic acid-(pentapeptide)-P-P-bactoprenyl-N-acetylglucosamine transferase, bactoprenyl phosphate phosphokinase, and N-acetylmuramyl-L-alanine amidase. We also propose that the shape of Escherichia coli is maintained by contractile proteins acting at the hernia. Given the universal importance of membranes, these proposals have implications for the determination of shape in eukaryotic cells.

Membrane aging during cell growth ascertained by Laurdan generalized polarization

The sensitivity of the fluorescent probe Laurdan to the phase state of lipids has been utilized to detect modifications in the composition and physical state of cell membranes during cell growth. In phospholipid vesicles, the Laurdan emission spectrum shows a 50-nm red shift by passing from the gel to the liquid-crystalline phase. The Generalized Polarization (GP) value has been used for the data treatment instead of the ratiometric method common in investigations utilizing other fluorescent probes that display spectral sensitivity to medium properties. The GP value can be measured easily and quickly and possesses all the properties of "classical" polarization, including the additivity rule. Once Laurdan limiting GP values have been established for the gel and the liquid-crystalline phase of lipids, the quantitative determination of coexisting phases in natural samples is possible. In the present work the observation of a relevant decrease in the fractional intensity of the liquid-crystalline phase in K562 cell membranes during 5 days of asynchronous growth is reported. A decrease in the "fluidity" of cell membranes in K562 cells kept in culture for several months is also reported. The procedure developed for labeling cell membranes with Laurdan is reported and the influence of cell metabolism on fluorescence parameters is discussed. Also discussed is the influence of cholesterol on Laurdan GP.

Phospholipid domains determine the spatial organization of the Escherichia coli cell cycle: the membrane tectonics model

Escherichia coli normally divides at its equator between segregated nucleoids. Such division is inhibited during perturbations of chromosome replication (even in the absence of inducible division inhibitors); eventually, division resumes at sites which are not at this equator. Escherichia coli will also divide at its poles to generate minicells following overproduction of the FtsZ or MinE proteins. The mechanisms underlying the division inhibition and the positioning of the division sites are unknown. In the membrane tectonics model, I propose that the formation of phospholipid domains within the cytoplasmic membrane positions division sites. The particular phospholipid composition of a domain attracts particular proteins and determines their activity; conversely, particular proteins change the composition of domains. Principally via such proteins, the interaction of the chromosome with the membrane creates a chromosomal domain. The development of chromosomal domains during replication and nucleoid formation contributes to the formation and positioning of a septal domain between them. During septation (cell division), this septal domain matures into a polar domain. Each domain attracts and activates different enzymes. The septal domain attracts and activates enzymes necessary for septation. Preventing the formation of the septal domain by preventing chromosome replication prevents normal division. Altering the composition of the polar domain may allow septation enzymes to function there and generate minicells. A corollary of the model explains how the formation of an origin domain by the attachment of hemi-methylated origin DNA to the membrane may underlie the creation and migration of structures within the envelope, the periseptal annuli.

Fluorescent membrane probes incorporating dipyrrometheneboron difluoride fluorophores

The spectroscopic properties of a new series of fatty acid analogs in which a dipyrrometheneboron difluoride fluorophore forms a segment of the acyl methylene chain are presented and their characteristics as fluorescent membrane probes are examined. When incorporated as a low mole fraction component in model phospholipid membranes, the probes retain the principal characteristics of the parent fluorophore: green fluorescence emission with high quantum yield, extensive spectral overlap, and low environmental sensitivity. The fluorescence quantum yield is typically two to three times that of comparable membrane probes based on the nitrobenzoxadiazole fluorophore. The spectral overlap results in a calculated Förster energy transfer radius (Ro) of about 57 A. Consequently, increasing fluorescence depolarization and quenching are observed as the mole fraction of the probe species incorporated in the membrane is increased. Low environmental sensitivity is manifested by retention of high quantum yield emission in aqueous dispersions of fatty acids. Partition coefficient data derived from fluorescence anisotropy measurements and iodide quenching experiments indicate that in the presence of fluid phase phospholipid bilayers the aqueous fraction of fatty acid is very small. Fluorescence intensity and anisotropy responses to phospholipid phase transitions are examined and found to be indicative of nonrandom fluorophore distribution in the gel phase. It is concluded that the spectroscopic properties of the fatty acid probes and their phospholipid derivatives are particularly suited to applications in fluorescence imaging of cellular lipid distribution and membrane level studies of lateral lipid segregation.

Domains and anomalous adsorption isotherms of dipalmitoylphosphatidylcholine membranes and lipophilic ions: pentachlorophenolate, tetraphenylborate, and dipicrylamine

Dipalmitoylphosphatidylcholine (DPPC) vesicles acquire negative surface charge on adsorption of negatively charged pentachlorophenolate (PCP-), and lipophilic ions tetraphenylborate (TPhB-), and dipicrylamine (DPA-). We have obtained (a) zeta-potential isotherms from the measurements of electrophoretic mobility of DPPC vesicles as a function of concentration of the adsorbing ions at different temperatures (25-42 degrees C), and (b) studied the effect of PCP- on gel-to-fluid phase transition by measuring the temperature dependence of zeta-potential at different PCP- concentrations. The zeta-potential isotherms of PCP- at 25, 32, and 34 degrees C correspond to adsorption to membrane in its gel phase. At 42 degrees C the zeta-potential isotherm corresponds to membrane in its fluid phase. These isotherms are well described by a Langmuir-Stern-Grahame adsorption model proposed by McLaughlin and Harary (1977. Biochemistry. 15:1941-1948). The zeta-potential isotherm at 37 degrees C does not follow the single-phase adsorption model. We have also observed anomalous adsorption isotherms for lipophilic ions TPhB- and DPA- at temperatures as low as 25 degrees C. These isotherms demonstrate a gel-to-fluid phase transition driven by ion adsorption to DPPC membrane during which the membrane changes from weakly to a strongly adsorbing state. The anomalous isotherm of PCP- and the temperature dependence of zeta-potential can be described by a two-phase model based on the combination of (a) Langmuir-Stern-Grahame model for each phase, (b) the coexistence of gel and fluid domains, and (c) depression of gel-to-fluid phase transition temperature by PCP-. Within the anomalous region the magnitude of zeta-potential rapidly increases concentration of adsorbing species, which was characterized in terms of a Esin-Markov coefficient. This effect can be exploited in membrane-based devices. Comments are also made on the possible effect of PCP, as an uncoupler, in energy transducing membranes.

On the principles of functional ordering in biological membranes

Integrating the available data on lipid-protein interactions and ordering in lipid mixtures allows to emanate a refined model for the dynamic organization of biomembranes. An important difference to the fluid mosaic model is that a high degree of spatiotemporal order should prevail also in liquid crystalline, "fluid" membranes and membrane domains. The interactions responsible for ordering the membrane lipids and proteins are hydrophobicity, coulombic forces, van der Waals dispersion, hydrogen bonding, hydration forces and steric elastic strain. Specific lipid-lipid and lipid-protein interactions result in a precisely controlled yet highly dynamic architecture of the membrane components, as well as in its selective modulation by the cell and its environment. Different modes of organization of the compositionally and functionally differentiated domains would correspond to different functional states of the membrane. Major regulators of membrane architecture are proposed to be membrane potential controlled by ion channels, intracellular Ca2+, pH, changes in lipid composition due to the action of phospholipase, cell-cell coupling, as well as coupling of the membrane with the cytoskeleton and the extracellular matrix. Membrane architecture is additionally modulated due to the membrane association of ions, lipo- and amphiphilic hormones, metabolites, drugs, lipid-binding peptide hormones and amphitropic proteins. Intermolecular associations in the membrane and in the membrane-cytoskeleton interface are further selectively controlled by specific phosphorylation and dephosphorylation cascades involving both proteins and lipids, and regulated by the extracellular matrix and the binding of growth factors and hormones to their specific receptor tyrosine kinases. A class of proteins coined architectins is proposed, as a notable example the pp60src kinase. The functional role of architectins would be in causing specific changes in the cytoskeleton-membrane interface, leading to specific configurational changes both in the membrane and cytoskeleton architecture and corresponding to (a) distinct metabolic/differentiation states of the cell, and (b) the formation and maintenance of proper three dimensional membrane structures such as neurites and pseudopods.

Characterization of lipid domains in erythrocyte membranes

Fluorescence digital imaging microscopy was used to study the lateral distribution of the lipid components in erythrocyte membranes. Intact erythrocytes labeled with phospholipids containing a fluorophore attached to one fatty acid chain showed an uneven distribution of the phospholipids in the membrane thereby demonstrating the presence of membrane domains. The enrichment of the lipotropic compound chlor-promazine in domains in intact erythrocytes also suggested that the domains are lipid-enriched regions. Similar membrane domains were present in erythrocyte ghosts. The phospholipid enrichment was increased in the domains by inducing membrane protein aggregation. Double-labeling experiments were done to determine the relative distributions of different phospholipids in the membrane. Vesicles made from extracted lipids did not show the presence of domains consistent with the conclusion that membrane proteins were responsible for creating the domains. Overall, it was found that large domains exist in the red blood cell membrane with unequal enrichment of the different phospholipid species.