Membrane surface charge modulates lipoprotein complex forming capability of peptides derived from the C-terminal domain of apolipoprotein E

Apolipoprotein E (apoE) plays a major role in the transport and metabolism of lipid by acting as a ligand for low density lipoprotein-receptors. The amphipathic helical regions of its C-terminal domain are necessary for the lipoprotein binding and assembly of nascent lipoprotein particles. Lipoproteins in the plasma are known to possess a net negative charge, determined by both its protein and lipid components, which regulates the metabolism of lipoproteins. The role of membrane surface charge on the interaction of apoE has not been studied previously. Also the importance of individual amphipathic helical regions of its C-terminal domain in binding to negatively charged lipid membrane is not addressed. In this study we have compared the interaction of four peptide segments of apoE C-terminal domain (apoE_(202-223), apoE_(223-244), apoE_(245-266), and apoE_(268-289)) with zwitterionic and negatively charged model membranes by employing UV-visible and fluorescence spectroscopy, circular dichroism, and native PAGE analysis. Our results show that the peptide sequence 202-223, 245-266 and 268-289 of apoE has higher affinity towards negatively charged lipid membrane and are independently capable of forming lipoprotein particles of 17 ± 2 nm Stokes diameter. The results suggest that surface charge of lipoprotein regulates its metabolism possibly by modulating the recruitment of apoE on its surface.     

Membrane binding properties of IRSp53-missing in metastasis domain (IMD) protein

The 53-kDa insulin receptor substrate protein (IRSp53) organizes the actin cytoskeleton in response to stimulation of small GTPases, promoting the formation of cell protrusions such as filopodia and lamellipodia. IMD is the N-terminal 250 amino acid domain (IRSp53/MIM Homology Domain) of IRSp53 (also called I-BAR), which can bind to negatively charged lipid molecules. Overexpression of IMD induces filopodia formation in cells and purified IMD assembles finger-like protrusions in reconstituted lipid membranes. IMD was shown by several groups to bundle actin filaments, but other groups showed that it also binds to membranes. IMD binds to negatively charged lipid molecules with preference to clusters of PI(4,5)P₂. Here, we performed a range of different in vitro fluorescence experiments to determine the binding properties of the IMD to phospholipids. We used different constructs of large unilamellar vesicles (LUVETs), containing neutral or negatively charged phospholipids. We found that IMD has a stronger binding interaction with negatively charged PI(4,5)P₂ or PS lipids than PS/PC or neutral PC lipids. The equilibrium dissociation constant for the IMD–lipid interaction falls into the 78–170 μM range for all the lipids tested. The solvent accessibility of the fluorescence labels on the IMD during its binding to lipids is also reduced as the lipids become more negatively charged. Actin affects the IMD–lipid interaction, depending on its polymerization state. Monomeric actin partially disrupts the binding, while filamentous actin can further stabilize the IMD–lipid interaction.     

Lipid modulation of nicotinic acetylcholine receptor function: the role of neutral and negatively charged lipids.

The effects of negatively charged and neutral lipids on the function of the reconstituted nicotinic acetylcholine receptor from Torpedo californica were determined with two assays using acetylcholine receptor-containing vesicles: the ion flux response and the affinity-state transition. The receptor was reconstituted into three different lipid environments, with and without neutral lipids: (1) phosphatidylcholine/phosphatidylserine; (2) phosphatidylcholine/phosphatidic acid; and (3) phosphatidylcholine/cardiolipin. Analysis of the ion flux responses showed that: (1) all three negatively charged lipid environments gave fully functional acetylcholine receptor ion channels, provided neutral lipids were added; (2) in each lipid environment, the neutral lipids tested were functionally equivalent to cholesterol; and (3) the rate of receptor desensitization depends upon the type of neutral lipid and negatively charged phospholipid reconstituted with the receptor. The functional effects of neutral and negatively charged lipids on the acetylcholine receptor are discussed in terms of protein-lipid interactions and stabilization of protein structure by lipids.     

The effect of surface charge density on valinomycin-K+ complex formation in model membranes

The model membrane approach was used to investigate the surface charge effect on the ion-antibiotic complexation process. Mixed monolayers of valinomycin and lipids were spread on subphases containing K⁺ or Na⁺. The surface charge density was modified by spreading ionizable valinomycin analogs on aqueous subphases of different pH or by changing the nature of the lipid (neutral, negatively charged) in the mixed film. Surface pressure and surface potential measurements demonstrated that a neutral lipid (phosphatidylcholine) or positively charged valinomycin analogs didn't enhance the antibiotic complexing capacity. However, a maximal complexation is reached for a critical lipid concentration in the valinomycin-phosphatidylserine mixed film. The role of the surface charge on the valinomycin complexing properties was examined in terms of the Gouy-Chapman theory. As a consequence of the negative charge of the lipid monolayer, the K⁺ concentration near the surface is larger than the bulk concentration, by a Boltzmann factor. A good agreement was observed between the experimental results and the theoretical predictions. Conductance measurements of asymmetric bilayers containing a neutral lipid (egg lecithin) on one side and a negatively charged lipid (phosphatidylserine) on the other, confirm the role of the surface charge. Indeed, addition of K⁺ to the neutral side of the bilayer containing valinomycin had no effect on the conductance whereas addition of K⁺ to the charged side of the bilayer caused a 80-fold conductance increase.     

Correlation between octanol/water and liposome/water distribution coefficients and drug absorption of a set of pharmacologically active compounds

Abstract

Absorption and consequent therapeutic action are key issues in the development of new drugs by the pharmaceutical industry. In this sense, different models can be used to simulate biological membranes to predict the absorption of a drug. This work compared the octanol/water and the liposome/water models. The parameters used to relate the two models were the distribution coefficients between liposomes and water and octanol and water and the fraction of drug orally absorbed. For this study, 66 drugs were collected from literature sources and divided into four groups according to charge and ionization degree: neutral; positively charged; negatively charged; and partially ionized/zwitterionic. The results show a satisfactory linear correlation between the octanol and liposome systems for the neutral (R^2?=?0.9324) and partially ionized compounds (R^2?=?0.9367), contrary to the positive (R^2?=?0.4684) and negatively charged compounds (R^2?=?0.1487). In the case of neutral drugs, results were similar in both models because of the high fraction orally absorbed. However, for the charged drugs (positively, negatively, and partially ionized/zwitterionic), the liposomal model has a more-appropriate correlation with absorption than the octanol model. These results show that the neutral compounds only interact with membranes through hydrophobic bonds, whereas charged drugs favor electrostatic interactions established with the liposomes. With this work, we concluded that liposomes may be a more-appropriate biomembrane model than octanol for charged compounds.     

Examining the contributions of lipid shape and headgroup charge on bilayer behavior

To better understand bilayer property dependency on lipid electrostatics and headgroup size, we use atomistic molecular dynamics simulations to study negatively charged and neutral lipid membranes. We compare the negatively charged phosphatidic acid (PA), which at physiological pH and salt concentration has a negative spontaneous curvature, with the negatively charged phosphatidylglycerol (PG) and neutrally charged phosphatidylcholine (PC), both of which have zero spontaneous curvature. The PA lipids are simulated using two different sets of partial charges for the headgroup and the varied charge distribution between the two PA systems results in significantly different locations for the Na⁺ ions relative to the water/membrane interface. For one PA system, the Na⁺ ions are localized around the phosphate group. In the second PA system, the Na⁺ ions are located near the ester carbonyl atoms, which coincides with the preferred location site for the PG Na⁺ ions. We find that the Na⁺ ion location has a larger effect on bilayer fluidity properties than lipid headgroup size, where the A_lipid and acyl chain order parameter values are more similar between the PA and PG bilayers that have Na⁺ ions located near the ester groups than between the two PA bilayers.     

Transmembrane Lipid Migration in Planar Asymmetric Bilayer Membranes

Planar asymmetric bilayer membranes, formed by apposing a monolayer of the neutral lipid glyceroldioleate (GDO) with one of the negatively charged lipid oleyl acid phosphate (OAP), were used to measure the rate of transmembrane OAP migration. The assay for this lipid flip-flop was the interaction of Ca²⁺ ions with negatively charged lipids which causes membranes to break: when Ca²⁺ is added to the compartment limited initially by the neutral lipid, flip-flop of the charged lipid eventually results in membrane breakdown. At 22 ± 2°C, in the absence of an externally applied electric field, an upper limit to the half time of OAP flip-flop was measured as 18.7 h, with a tentative lower limit of 14.4 h.     

Role of electrostatics in the binding of charged metallophthalocyanines to neutral and charged phospholipid membranes

Binding of the cationic tetra(tributylammoniomethyl)-substituted hydroxoaluminum phthalocyanine (AlPcN₄) to bilayer lipid membranes was studied by fluorescence correlation spectroscopy (FCS) and intramembrane field compensation (IFC) methods. With neutral phosphatidylcholine membranes, AlPcN₄ appeared to bind more effectively than the negatively charged tetrasulfonated aluminum phthalocyanine (AlPcS₄), which was attributed to the enhancement of the coordination interaction of aluminum with the phosphate moiety of phosphatidylcholine by the electric field created by positively charged groups of AlPcN₄. The inhibitory effect of fluoride ions on the membrane binding of both AlPcN₄ and AlPcS₄ supported the essential role of aluminum-phosphate coordination in the interaction of these phthalocyanines with phospholipids. The presence of negative or positive charges on the surface of lipid membranes modulated the binding of AlPcN₄ and AlPcS₄ in accord with the character (attraction or repulsion) of the electrostatic interaction, thus showing the significant contribution of the latter to the phthalocyanine adsorption on lipid bilayers. The data on the photodynamic activity of AlPcN₄ and AlPcS₄ as measured by sensitized photoinactivation of gramicidin channels in bilayer lipid membranes correlated well with the binding data obtained by FCS and IFC techniques. The reduced photodynamic activity of AlPcN₄ with neutral membranes violating this correlation was attributed to the concentration quenching of singlet excited states as proved by the data on the AlPcN₄ fluorescence quenching.     

Electrokinetic Properties of the Sarcoplasmic Reticulum Membrane Obtained from Reconstitution Studies

Electrophoretic mobility data of SR vesicles reconstituted with uncharged and two mixtures of charged and uncharged lipids (Brethes, D., Dulon, D., Johannin, G., Arrio, B., Gulik-Krzywicki, T., Chevallier, J. 1986. Study of the electrokinetic properties of reconstituted sarcoplasmic reticulum vesicles. Arch. Biochem. Biophys. 246:355–356) were analyzed in terms of four models of the membrane-water interface: (I) a smooth, negatively charged surface; (II) a negatively charged surface of lipid bilayer covered with an electrically neutral surface frictional layer; (III) an electrically neutral lipid bilayer covered with a neutral frictional layer containing a sheet of negative charge at some distance above the surface of the bilayer; (IV) an electrically neutral lipid bilayer covered with a homogeneously charged frictional layer. The electrophoretic mobility was predicted from the numerical integration of Poisson-Boltzmann and Navier-Stokes equations. Experimental results were consistent only with predictions based on Model-III with charged sheet about 4 nm above the bilayer and frictional layer about 10 nm thick. Assuming that the charge of the SR membrane is solely due to that on Ca⁺⁺-ATPase pumps, the dominant SR protein, the mobility data of SR and reconstituted SR vesicles are consistent with 12 electron charges/ATPase. This value compares well to the net charge of the cytoplasmic portion of ATPase estimated from the amino acid sequence (-11e). The position of the charged sheet suggests that the charge on the ATPase is concentrated in the middle of the cytoplasmic portion. The frictional layer of SR can be also assigned to the cytoplasmic portion of Ca⁺⁺-ATPase. The layer has been characterized with hydrodynamic shielding length of 1.1 nm. Its thickness is comparable to the height of the cytoplasmic portion of Ca⁺⁺-ATPase. Received: 15 June 1998/Revised: 8 October 1998     

Interaction of the lantibiotic nisin with mixed lipid bilayers: a 31P and 2H NMR study

Nisin is a positively charged antibacterial peptide which binds to the negatively charged membranes of Gram-positive bacteria. The initial interaction of the peptide with model membranes of neutral (phosphatidylcholine) and negatively charged (phosphatidylcholine/phosphatidylglycerol) model lipid membranes was studied using nonperturbing solid state magic angle spinning (MAS) (31)P NMR and (2)H wide-line NMR. In the presence of nisin, the coexistence of two bilayer lipid environments was observed both in charged and in neutral membranes. One lipid environment was found to be associated with lipid directly interacting with nisin and one with noninteracting lipid. Solid state (31)P MAS NMR results show that the acidic membrane lipid component partitions preferentially into the nisin-associated environment. Deuterium NMR ((2)H NMR) of the selectively headgroup-labeled acidic lipid provides further evidence of a strong interaction between the charged lipid component and the peptide. The segregation of acidic lipid into the nisin-bound environment was quantified from (2)H NMR measurements of selectively headgroup-deuterated neutral lipid. It is suggested that the observed lipid partitioning in the presence of nisin is driven, at least initially, by electrostatic interactions. (2)H NMR measurements from chain-perdeuterated neutral lipids indicate that nisin perturbs the hydrophobic region of both charged and neutral bilayers.     

Liposome-Induced Conformational Changes of an Epitopic Peptide and its Palmitoylated Derivative of Influenza Virus Hemagglutinin

The conformation of synthetic HA_317-329-NH₂representing the major B- and T-cell epitopic region of influenza virus hemagglutinin, its palmitoylated derivative (HA_317-329-Thr(Pal)-NH₂), and the intersubunit peptide (HA_317-341-NH₂) comprising also the fusion peptide, were studied in aqueous buffer and in the presence of neutral and negatively charged liposomes. The free peptide is unordered in aqueous solution, even in the presence of liposomes. However, grafting the palmitic acid or the fusion peptide onto the C-terminus of the peptide enables the hydrophilic HA_317-329to adopt folded (turn) and β-strand structure on the surface of neutral and negatively charged liposomes, respectively. The results emphasize the importance of some kind of anchor for achieving a specific conformation of epitopic peptide HA_317-329-NH₂on the surface of liposomes.     

The effect of surface charge density on valinomycin-K+ complex formation in model membranes

The model membrane approach was used to investigate the surface charge effect on the ion-antibiotic complexation process. Mixed monolayers of valinomycin and lipids were spread on subphases containing K+ or Na+. The surface charge density was modified by spreading ionizable valinomycin analogs on aqueous subphases of different pH or by changing the nature of the lipid (neutral, negatively charged) in the mixed film. Surface pressure and surface potential measurements demonstrated that a neutral lipid (phosphatidylcholine) or positively charged valinomycin analogs didn't enhance the anti-biotic complexing capacity. However, a maximal complexation is reached for a critical lipid concentration in the valinomycin-phosphatidylserine mixed film. The role of the surface charge on the valinomycin complexing properties was examined in terms of the Gouy-Chapman theory. As a consequence of the negative charge of the lipid monolayer, the K+ concentration near the surface is larger than the bulk concentration, by a Boltzmann factor. A good agreement was observed between the experimental results and the theoretical predictions. Conductance measurements of asymmetric bilayers containing a neutral lipid (egg lecithin) on one side and a negatively charged lipid (phosphatidyl-serine) on the other, confirm the role of the surface charge. Indeed, addition of K+ to the neutral side of the bilayer containing valinomycin had no effect on the conductance whereas addition of K+ to the charged side of the bilayer caused a 80-fold conductance increase.     

Interaction of IAPP and Insulin with Model Interfaces Studied Using Neutron Reflectometry

The islet amyloid polypeptide (IAPP) and insulin are coproduced by the β-cells of the pancreatic islets of Langerhans. Both peptides can interact with negatively charged lipid membranes. The positively charged islet amyloid polypeptide partially inserts into these membranes and subsequently forms amyloid fibrils. The amyloid fibril formation of insulin is also accelerated by the presence of negatively charged lipids, although insulin has a negative net charge at neutral pH-values. We used water-polymer model interfaces to differentiate between the hydrophobic and electrostatic interactions that can drive these peptides to adsorb at an interface. By applying neutron reflectometry, the scattering-length density profiles of IAPP and insulin, as adsorbed at three different water-polymer interfaces, were determined. The islet amyloid polypeptide most strongly adsorbed at a hydrophobic poly-(styrene) surface, whereas at a hydrophilic, negatively charged poly-(styrene sulfonate) interface, the degree of adsorption was reduced by 50%. Almost no IAPP adsorption was evident at this negatively charged interface when we added 100 mM NaCl. On the other hand, negatively charged insulin was most strongly attracted to a hydrophilic, negatively charged interface. Our results suggest that IAPP is strongly attracted to a hydrophobic surface, whereas the few positive charges of IAPP cannot warrant a permanent immobilization of IAPP at a hydrophilic, negatively charged surface at an ionic strength of 100 mM. Furthermore, the interfacial accumulation of insulin at a hydrophilic, negatively charged surface may represent a favorable precondition for nucleus formation and fibril formation.     

Electrostatically induced change of the conformational order of charged lipid membranes

Raman spectroscopy and X-ray diffraction are used to investigate the influence of surface charges on the structure of ionizable lipid membranes of dimyristoylmethylphosphatidic acid. The membrane surface charge density is regulated by varying the pH of the aqueous phase. Changes of the conformational order of the lipid chains are determined from the intensity of the CC stretch chain vibrations around 1100 cm^?1 in a lipid Raman spectrum. In going from an electrical neutral to a negatively charged membrane, the conformational order is reduced by 5% in the ordered and by 9% in the fluid membrane phase, corresponding to 0.6 and 0.8 CC bonds, respectively, which change from a trans to a gauche conformation. The electrostatically induced conformational change is mainly concentrated at the lipid chain ends as indicated by the spectral variations of the 890 cm^?1 CH₃ rocking band of the chain termini. The X-ray diffraction experiments show that increasing the surface charge density in the ordered membrane phase leads to a lateral expansion of the packing of the lipid polar groups, whereas the packing of the lipid chains in a plane perpendicular to the chain axes remains constant, indicating an increase of the tilt of the lipid chains from δ = 10° (pH 3) to δ = 27° (pH 9).     

Permeability of lipid bilayer to anthracycline derivatives. Role of the bilayer composition and of the temperature

The uptake of three anthracycline derivatives: doxorubicin, daunorubicin and pirarubicin, into large unilamellar vesicles (LUV) in response to a driving force provided by DNA encapsulated inside the LUV has been investigated as a function of the temperature and of the bilayers lipid composition. The kinetics of the decay of the anthracycline fluorescence in the presence of DNA-containing liposome was used to follow the diffusion of the drug through the membrane. For the three drugs, the permeability coefficient of the neutral form of the drug (P⁰) decreases as the amount of negatively charged phospholipid in the bilayers increases. This can be explained by the fact that the kinetics of passive diffusion of the drugs depends on the amount of neutral form embedded in the polar head group region, which decreases as the quantity of negatively charged phospholipids increases. P⁰ also decreases as the amount of cholesterol, that makes the bilayer more rigid, increases. The activation energies, E_a, for the passage of the neutral form of these anthracyclines through the bilayers lie within 100±15 kJ·mol⁻¹, except for pirarubicin and doxorubicin through anionic phospholipid-rich membranes (E_a=57 kJ·mol⁻¹) and cholesterol-rich membranes (E_a=167 kJ·mol⁻¹).     

A molecular dynamics study of the interaction between domain I-BAR of the IRSp53 protein and negatively charged membranes

The methods of computer simulation in all-atom and coarse-grained approximations have been used to study specific interactions of the isolated domain I-BAR of the actin-binding protein IRSp53 with model membranes containing neutral phospholipids and those including negatively charged PI(4,5)P₂ phospholipids. It has been shown that the I-BAR domain does not interact with neutral lipids but induces bending of the synthetic membrane rich in negatively charged phospholipids. Clustering of charged lipids on the surface of the membrane at the sites of its interaction with the protein has been observed. This indicates that the interaction of I-BAR with negatively charged lipids is of electrostatic and hydrophobic nature.     

Enhanced selectivity of a molecularly imprinted polymer toward the target molecule via esterification of non‐specific binding sites with diazomethane

Diazomethane (CH₂N₂) was used to methylate the non‐specific binding sites after molecularly imprinted polymer particles were prepared using methacrylic acid as the functional monomer, ethylene glycol dimethacrylate as the cross‐linker and bisphenol A (BPA) as the template. After diazomethane treatment and subsequent removal of BPA by triethylamine, the treated molecularly imprinted polymer (TMIP) particles were tested for binding selectivity toward BPA and other organic compounds by capillary electrophoresis with ultraviolet detection. Even in the presence of compounds that were positively charged, neutral or negatively charged in the background electrolyte, BPA was selectively bound with the highest efficiency. A significant decrease in the affinity for metformin (MF, a positively charged compound), along with ¹³C nuclear magnetic resonance spectra and electrophoretic mobility data, provided strong evidence for the elimination of non‐specific –COOH binding sites in the TMIP particles. Only 8% of MF and 16% of diclofenac sodium salt (a negatively charged compound) remained as non‐specific bindings because of hydrophobic interactions. Further comparison with poly(methyl methacrylate) revealed the true merits of the TMIP, which exhibited minimal non‐specific bindings while preserving a high level of specific binding owing to molecular recognition. Copyright © 2014 John Wiley & Sons, Ltd.     

Bimodal solubilization of phospholipid-cholesterol vesicles in 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) solutions. Formation of filamentous and helical microstructures depends on negatively charged lipids

Phospholipid/cholesterol vesicles were solu-bilized by 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS). Above 30 mol% cholesterol (Ch) in the lipid vesicles several remarkable changes of the solubilization process were observed. (i) Two modes of solubilization: The effective detergent to lipid ratio R^c(M) for the formation of mixed micelles decreased from R^c(M) = 43 ± 3 at low lipid concentrations, [L]≤ 0.15 mm, to R^c(M) = 2.4 ± 0.3 above [L] = 0.5 mm (40 mol% Ch, T = 20 °C). (ii) At subsolubilizing CHAPS concentrations, filamentous and helical microstructures were formed, similar to those which were observed in native and model bile. (iii) The number of observed fibers was about two orders of magnitude higher in the presence of the negatively charged lipids phosphatidylglycerol (PG) and phosphatidic acid (PA) compared to the zwitterionic phosphatidylcholine (PC). Fiber formation began after 16–18 h using PG and PA compared to 3–4 days in the presence of PC. Screening of the charged lipids by NaCl effectively reduced the formation of fibers. Assuming binding of Na⁺ to the charged lipid aggregates, an intrinsic binding constant K_int = 0.6 M^–1 was determined by applying the Gouy-Chapman theory. After the addition of CHAPS to PG/Ch vesicles, a fast initial solubilization of the vesicles (<1 min) to mixed micelles (r_h = 2.3 ± 0.2 nm) and small vesicles (r_h = 23 ± 1 nm) was observed, followed by an intermediate period of 2 h, after which the formation of fibers occurred (>15 h). The microstructures are visualized by darkfield and electron microscopy. The method of vesicle solubilization is compared to the dilution of concentrated micellar solutions, which is usually applied to model bile systems. Received: 28 May 1996 / Accepted: 26 July 1996     

Changes in surface charge density on liposomes induced by Escherichia coli endotoxin

Changes in surface charge density of liposomes induced by E. coli endotoxin were studied by microelectrophoresis. Endotoxin altered the surface charge of phosphatidylcholine liposomes from neutral to negative. The negative charge of the endotoxin-phosphatidylcholine complex was neutralized electrostatically by binding with Ca²⁺ (2 mM). Phosphatidylcholine liposomes were made positive by addition of the positively charged detergent, hexadecyltrimethylammonium chloride. Endotoxin made the positively charged liposomes less charged. On the other hand, phosphatidylserine liposomes which were negatively charged became less charged in the presence of high concentration of endotoxin (8 mg/ml). The endotoxin effect on phosphatidylserine liposomes was abolished by EDTA (1 mM) but potentiated by CaCl₂ (0.1–2 mM). These results indicate that endotoxin interacts with liposomes both hydrophobically and electrostatically.     

Phase diagram, stability, and overcharging of lamellar cationic lipid-DNA self-assembled complexes.

Cationic lipid-DNA (CL-DNA) complexes comprise a promising new class of synthetic nonviral gene delivery systems. When positively charged, they attach to the anionic cell surface and transfer DNA into the cell cytoplasm. We report a comprehensive x-ray diffraction study of the lamellar CL-DNA self-assemblies as a function of lipid composition and lipid/DNA ratio, aimed at elucidating the interactions determining their structure, charge, and thermodynamic stability. The driving force for the formation of charge-neutral complexes is the release of DNA and lipid counterions. Negatively charged complexes have a higher DNA packing density than isoelectric complexes, whereas positively charged ones have a lower packing density. This indicates that the overcharging of the complex away from its isoelectric point is caused by changes of the bulk structure with absorption of excess DNA or cationic lipid. The degree of overcharging is dependent on the membrane charge density, which is controlled by the ratio of neutral to cationic lipid in the bilayers. Importantly, overcharged complexes are observed to move toward their isoelectric charge-neutral point at higher concentration of salt co-ions, with positively overcharged complexes expelling cationic lipid and negatively overcharged complexes expelling DNA. Our observations should apply universally to the formation and structure of self-assemblies between oppositely charged macromolecules.