Scientific paper Interactions of Divalent Calcium Ions With Head Groups of Zwitterionic Phosphatidylcholine Liposomal Membranes Poornima Budime Santhosh,1 Aljaž Velikonja,2'3 Ekaterina Gongadze,4 Aleš Iglic,4'* Veronika Kralj-Iglic45 and Nataša Poklar Ulrih1'6 1 Department of Food Science and Technology, Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, SI-1000 Ljubljana, Slovenia 2 Laboratory of Biocybernetics, Faculty of Electrical Engineering, University of Ljubljana, Tržaška 25, SI-1000 Ljubljana, Slovenia 3 SMARTEH Research and Development of Electronic Controlling and Regulating Systems, Poljubinj 114, SI-5220 Tolmin, Slovenia 4 Laboratory of Biophysics, Faculty of Electrical Engineering, University of Ljubljana, Tržaška 25, SI-1000 Ljubljana, Slovenia. 5 Laboratory of Clinical Biophysics, Faculty of Health Sciences, University of Ljubljana, Zdravstvena 5, SI-1000 Ljubljana, Slovenia 6 Centre of Excellence for Integrated Approaches in Chemistry and Biology of Proteins (CipKeBiP), Jamova 39, 1000 Ljubljana, Slovenia * Corresponding author: E-mail: ales.iglic@fe.uni-lj.si Tel: (01) 4768 825: Fax: (01) 4768 850 Received: 03-10-2013 Paper based on a presentation at the 4th RSE-SEE 2013 Symposium on Electrochemistry in Ljubljana, Slovenia The interaction of the divalent calcium ions with the zwitterionic lipid membranes was studied by measuring the lipid order parameter which is inversely proportional to the membrane fluidity. Small unilamellar lipid vesicles were prepared from 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine and then treated with different concentrations of divalent calcium ions. An increase in the order parameter and decrease in the fluidity of the liposomal membranes were observed after treatment with the calcium ions. The presence of positively charged iron oxide nanoparticles in the suspension of liposomes negligibly changed the results. The results of experiments were discussed theoretically within modified Lan-gevin-Poisson-Boltzmann (MLPB) model leading to the conclusion that the membrane fluidity and ordering of the membrane lipids are primarily altered by the accumulation of calcium ions in the region of negatively charged phosphate groups within the head groups of the membrane lipids. Keywords: Anisotropy, liposomes, bilayer fluidity, iron oxide nanoparticles, electric double layer theory Abstract Liposomes encapsulating nanoparticles (NPs) with magnetic properties termed as "magnetoliposomes" have attracted a great interest in the recent past due to their potential biomedical applications especially in targeted drug 1. Introduction delivery,1 hyperthermia2 and as contrast agents in magnetic resonance imaging (MRI).3 Among the magnetic NPs, iron oxide (Fe2O3) NPs gains special importance in various clinical applications due to their non-toxicity and bio-degradability.4'5 As the usage of Fe2O3 NPs is tremendously increasing in the medical fields, it becomes impor- tant to study the consequences of interactions of the NPs with the membrane which may alter the physical properties of the bilayer such as fluidity, phase transition temperature, morphology, mechanical stability and permeability. Membrane fusion is an essential process which accounts for various biological events such as fertilization, embryogenesis, neuronal signalling and endocrine hormone secretion. The fusion process includes inter membrane contact, mixing of the membrane lipids and pore formation to mix the inner lipid contents to form large multila-mellar vesicles.6 Membrane fusion is generally induced in vitro and in vivo conditions7, 8 by fusogenic agents called fusogens such as proteins and peptides. Apart from these larger molecules, divalent cations such as Ca2+, Mg2+, Mn2+and Zn2+ have the potential to induce membrane fusion.9 Since calcium is known to be an essential biological component and a key player in inducing membrane fusion, we have chosen divalent calcium ions to study their effect on liposomes which serve as a simple model system to understand the complex fusion process in cells. Membrane fluidity characterizes the viscosity of the lipids in the plasma membrane. The cells maintain the optimal level of fluidity so that the mobility of the lipopro-teins in the membrane is large enough to perform different biological functions.10 Alterations in the fluidity level of the liposomal membranes can be conveniently studied by anisotropy measurements using the fluorescent probes such as 1, 6-diphenyl-1,3,5 hexatriene (DPH) (Fig. 1a).11 Membrane fluidity can be altered by various factors such as temperature, cholesterol, lipid composition including the hydrocarbon chain length, degree of saturation and interaction of NPs.12 Since Ca2+ interacts with the cell membrane, it induces alterations in the ordering of the membrane lipids and therefore affects the bilayer fluidity. Ca2+ may also promote the adhesion and fusion of the adjacent cells/liposomes and therefore mixing of the different lipid components. A lot of work has been done on the study of the interaction of negatively charged membranes mediated by positively charged calcium ions but very less work was carried on with the zwitterionic liposomes. Hence in this a) o Figure 1. Structure of (a) DPH (1,6-diphenyl-1,3,5-hexatriene) and (b) POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine). work we intended to study the effect of Ca2+ on liposomes prepared with a zwitterionic lipid such as 1-palmitoyl-2-oleoyl-OT-glycero-3-phosphocholine (POPC) (Fig. 1b). In order to study the effect of electrostatic interactions of Ca2+ with the NPs in detail, we have used in this work the positively charged Fe2O3 NPs functionalized with amino groups. Through mathematical modeling using the modified Langevin-Poisson-Boltzmann (MLPB) model of electric double layer, we aim to discuss the theory and the obtained experimental results. The scope of the present study is to gain more knowledge for better understanding of the role of Ca2+ (also in the presence of Fe2O3 NPs) in altering the membrane lipid bilayer fluidity which is important to consider prior to their biomedical applications. 2. Materials and Methods POPC (1 -palmitoyl-2-oleoyl-i«-glycero-3-phosp-hocholine) was purchased from Avanti Polar Lipids Inc., USA. DPH (1,6-diphenyl-1,3,5-hexatriene) and anhydrous calcium chloride were obtained from Sigma-Aldrich Chemie GmbH, Steinheim, Germany. All the chemicals obtained have high purity (> 99%) and used without any further purification. Iron oxide amine (Fe2O3-NH2) NPs were obtained from J. Stefan Institute, Ljubljana, Slovenia. 2. 1. Preparation of Small Unilamellar Vesicles Small unilamellar vesicles (SUVs) were prepared by the thin film method using the rotary evaporator. The POPC lipid (2 mg) was dissolved in 1 ml of chloroform in a round bottomed flask. The argon gas was bubbled through the lipid-chloroform mixture in the flask to avoid the oxidation of lipids. The organic solvent in the flask was slowly evaporated by the rotary evaporator under reduced pressure (1.7 kPa) to obtain a thin lipid film. The control liposomes containing pure POPC lipid were prepared by hydrating the obtained lipid film with double distilled water at pH 5.5. A thin lipid film was prepared in a similar way and hydrated with the Fe2O3-NH2 NPs suspended in distilled water. The hydrodynamic radius of the Fe2O3-NH2 NPs was 180 nm and their zeta potential value was 4 mV respectively. The lipid suspensions in the flasks were vortexed vigorously with the glass beads for 10 min to obtain multila-mellar vesicles (MLVs). The obtained vesicles were sonicated for 30 min total time with 10 s on-off cycles at 40% amplitude using a Vibracell Ultrasonic Disintegrator VCX 750 (Sonics and Materials, Newtown, USA) to form SUVs. In order to remove the debris formed after sonica-tion, the sample containing SUVs were then centrifuged for 10 min at 212554 g (Eppendorf centrifuge 5415C). The SUVs were incubated with appropriate amount of calcium chloride (CaCl2) solution to reach a final concen- tration of 10 mM and 100 mM in 2.5 ml of 20 mM HEPES buffer and left for 30 min before measuring the anisotropy values. 2. 2. Fluorescence Anisotropy Measurements DPH is one of the widely used fluorescent probes for measuring viscosity, lipid order and polarity of the membrane.13 Being a hydrophobic probe, DPH intercalates between the tail regions of the lipid bilayer and is distributed throughout the membrane. It is almost non-fluorescent in aqueous environment but shows intense fluorescence signals after incorporation into the hydrophobic core of the membrane bilayer.14 Temperature dependent anisotropy values were measured following the incorporation of the fluorescent dye DPH in the POPC liposomes treated with divalent calcium ions. The measurements were carried out in a 10 mm-path-length cuvette using the Cary Eclipse fluorescence spectrophotometer (Varian, Mulgrave, Australia). In the quartz cuvette, 10 pL of DPH was added to 2.5 mL 100 pM solutions of SUVs prepared from POPC to reach a final concentration of 0.5 pM DPH. The anisotropy measurements were performed within the temperature range from 15 °C to 50 °C by gradually increasing the temperature by 5 °C for every measurement, with a time interval of 8 min with constant mixing at pH 7.0. Varian autopola-rizers having the slit widths with a nominal band-pass of 5 nm was used for both the excitation and emission spectra. The fluorescent probe DPH was excited at 358 nm with the excitation polarizer oriented in the vertical position. The emitted polarized light in both the vertical and horizontal planes were recorded by a monochromator and measured at 410 nm. The anisotropy (r> values were measured using the built-in software of the instrument by applying the below formula: <'>=7 A ~ GI± +2GI± (1) where Iy and I± are the parallel and perpendicular emission intensities, respectively. The G-factor value (ratio of the sensitivities of the detection system for vertically [IHV] and horizontally [IHH] polarized light) was determined separately for each sample. The lipid-order parameter S was calculated from the anisotropy values using the following expression:15 S = r Jr) f ' VI r 1 - 2 — + 5 - 1 + — I u J U J 2( -) '0 J (2) cal value of r0 for DPH is 0.4, while experimental values of r0 lie between 0.362 and 0.394.15 In our calculation, the experimental value of r0 was 0.370 for DPH in POPC at 5 °C. To conclude, the anisotropy values obtained using DPH are directly proportional to lipid-order parameter (S) in the membrane, which can be calculated using the formula shown in Equation 2. Lipid-order parameter is inversely proportional to the fluidity and hence from the obtained lipid-order parameter results, bilayer fluidity can be estimated.16 2. 3. Theoretical Model 2. 3. 1. Pure Salt Solution in Contact With the Zwitterionic Lipid Bilayer POPC lipid is a common representative of the zwit-terionic (dipolar) lipids. Dipolar head group consists of a phosphate (negative) and amino (positive) group in the head region. Due to amphiphilic effect, the negative part of the lipid head group bounded to the lipid tails (phosphate group) is in contact with the salt solution forming negatively charged surface at x = 0 (see Fig. 2). The positive part of the lipid head group (amino group) penetrates into the salt solution and can be partially movable inside head group region (see Fig. 2). In this work the contact of the zwitterionic lipid bilayer (e.g. POPC) with the salt solution containing divalent counter-ions and monovalent co-ions (e.g. CaCl2) is where r0 is the fluorescence anisotropy of DPH in the absence of any rotational motion of the probe. The theoreti- Figure 2. Schematic presentation of zwitterionic lipid bilayer in contact with the salt solution containing monovalent co-ions and divalent counter-ions (e.g. Calcium ions). Phosphate groups in the head group region are described by negatively charged surface at x = 0 with negative surface charge density O}. D is the distance between the charges in the single lipid head group, while ffl describes the orientation angle of the single head group. theoretically described by using the modified Langevin-Poisson-Boltzmann (MLPB) model.17-19 The MLPB model takes into account the finite volumes of lipid head groups17, the cavity field in saturation regime, and the electronic polarization of the water dipoles.18-23 The finite volume of ions and water molecules in the solution was not taken into the account. Schematic presentation of the model system can be seen in Fig 2. The Poisson equation describing the system presented in Fig. 2 can be written in the form: (3) where 0(x) is the electric potential, e0 is permittivity of the free space, er(x) is the relative permittivity of the salt solution, pion(x) is the volume charge density of ions in the salt solution and pzw(x) is the volume charge density of the positive charges of zwitterionic lipid head groups. In the model the salt solution is composed of calcium chloride (Ca-Cl2) and water. In the water solution the single CaCl2 molecule dissociates into one divalent counter-ion Ca2+ and two monovalent co-ions Cl-, therefore the volume charge density of ions pion(x) can be written as: P„Jx) = ~eBn_ (x) + 2e„w(-r). (4) where n(x) is the number density of Cl- , m(x) is the number density of Ca2+.and e0 is the elementary charge. Considering the Bolztmann distribution function for co-ions (Cl) and counter-ions (Ca2+): (5) where n0 is bulk concentration of monovalent chloride co-ions, m0 the bulk concentration of divalent calcium counter-ions, f = 1/kT, kT is the thermal energy. The electro neutrality condition: yields Equation 4 can be further rewritten as: (6a) (6b) (7) The volume charge density of the positive charges of zwitterionic lipid head groups pzw(x) can be expressed as:17 (8) where P(x) is the probability density function for angle o, D is the distance between charges in zwitterionic lipid head group (see Fig. 2) and a0 is the area per lipid molecule. Inserting the Equation 7 and 8 into Equation 3 yields the Poisson equation in MLPB model: (9) where the last term in Equation 9 is different from zero only in the region 0 < x < D and the boundary conditions are: —(x = 0) =--- dx e0£r(x = 0) dx dx (10) (11) (12) (13) The surface charge density c1 describes the negative surface charge density of the phosphate groups of lipid head groups at x = 0: o1 = - e0/a0 (see Fig. 2), where a0 is the area per lipid molecule. Equation 9 was solved by using the standard implemented function for the multiboun-dary value problems (bvp4c) in Matlab2012b where the values er(x) and P(x) were calculated in the iteration process outside of bvp4c function. The space dependent permittivity er(x) within MLPB model is17,18: er (jt) = ir + - n0u.po ( 2 + ir "j L(ypt)E(x)P) 3 J E(x) (14) where n is refractive index of water, n0w is bulk concentration of water, p0 is the dipole moment of water molecule, L(u) = (coth(u) - 1/u) is the Langevin function, y = (3/2)((2 + n2)/3) and E(x) is the magnitude (absolute value) of the electric field strength, while the probability density function P(x) (taking into account the finite volumes of lipid head groups) is:17 (15) where a is a model parameter (see also ref. 17) and the value of A is calculated in iterative procedure until the normalization condition is met: I r" — [ P(x)dx = 1. 2. 3. 2. Positively Charged Nanoparticles Added to Salt Solution In the case when in the system are present also positively charged NPs (see Fig. 5) the boundary condition described by Equation 11 is replaced by equation: —(x = D ) =__ dx ' eaer(x = D\' (17) where c2 is the surface charge density of nanoparticle located at the distance x = Dnp from the negatively charged plane in the lipid head group region (see Fig. 5). 2. 3. 3. Osmotic Pressure Osmotic pressure between the zwitterionic lipid bi-layer and nanoparticle can be derived by using the procedure described elsewhere19 by integrating the Poisson equation (Equation 9 in our case) and subtracting the corresponding bulk osmotic pressure value between the lipid surface and nanoparticle surface located at x = Dnp (see Fig. 5). To this end, Equation 9 is first rewritten in the form: d_ dx s„n dx dx + 2e„m0 f - e^™ ) = 0, (18) Dan where Equation 14 was taken into account. Multiplying equation 18 by d$= ty'dx followed by integration leads to: (19) where the integration of the last term in Equation 18 (with non-zero value only in the region 0 < x < D) was omitted since the osmotic pressure was always calculated outside the lipid head group region in the region x > D. The osmotic pressure difference n = ninner- nbulk can be therefore written as: n=~e