Langmuir 2007, 23, 5651-5656
5651
Specific Adsorption of Cytochrome c on Cardiolipin-Glycerophospholipid Monolayers and Bilayers O Å scar Dome`nech,† Lorena Redondo,‡ M. Teresa Montero,‡ and Jordi Herna´ndez-Borrell*,‡ Departament de Quı´mica-Fı´sica, Facultat de Quı´mica, and Departament de Fisicoquı´mica, Facultat de Farmacia, UniVersitat de Barcelona, Barcelona 08028, Spain ReceiVed NoVember 24, 2006. In Final Form: February 9, 2007 In this study, we examined the adsorption of cytochrome c (cyt c) on monolayers and liposomes formed from (i) pure 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), or cardiolipin (CL) and on (ii) the more thermodynamically stable binary mixtures of POPE/CL (0.8:0.2 mol/mol) and POPC/CL (0.6:0.4 mol/mol). Constant surface pressure experiments showed that the maximum and minimum interactions occurred in the pure CL (anionic phospholipid) and the pure POPE (zwitterion) monolayers, respectively. Observation by atomic force microscopy (AFM) of the images of Langmuir-Blodgett (LB) films extracted at 30 mN m-1 suggests that the different interactions of cyt c with POPE/CL and the POPC/CL monolayers could be due to lateral phase separation occurring in the POPE/CL mixture. The competition between 8-anilino-1-naphthalene sulfonate (ANS) and cyt c for the same binding sites in liposomes that have identical nominal compositions with respect to those of the monolayers was used to obtain binding parameters. In agreement with the monolayer experiments, the most binding was observed in POPE/CL liposomes. All of our observations strongly support the existence of selective adsorption of cyt c on CL, which is modulated differently by different neutral phospholipids (POPE and POPC).
1. Introduction Cytochrome c (cyt c) is a peripheral membrane protein that is associated with the inner mitochondrial membrane. It is an essential component of the mitochondrial respiratory chain because it transfers electrons between the CoQH2-cyt c reductase and the cyt c oxidase complexes. In part because cyt c is a soluble protein, it is often considered to be an electron shuttle. Far from this simplistic view, the mechanism by which cyt c diffuses between the reductant and the oxidant protein substrates may involve the translational motion of cyt c in the aqueous intermembrane mitochondrial space or in the 2D plane defined by the bilayer.1 In this regard, it has been established using fluorescence measurements of monolayers2 and solid-state 31P NMR3 that cyt c binds specifically to anionic phospholipids. More recently, the nature of phospholipid-cyt c interactions has been investigated by atomic force microscopy (AFM)4 and combined with force spectroscopy (FS).5 The biological implications of such interactions appear to be related to the transport of electrons themselves6 and to the release of cyt c during the process of apoptosis.7 The involvement of cardiolipin (CL) has been demonstrated in both of these processes. We showed the specific affinity of cyt c for CL by using a combination of fluorescence measurements and AFM observations.8 However, major zwitterionic phospholipids could also influence such * Corresponding author. E-mail:
[email protected]. † Departament de Quı´mica-Fı´sica. ‡ Departament de Fisicoquı´mica. (1) Cramer, W. A.; Knaff, D. B. Energy Transduction in Biological Membranes: A Texbook of Bioenergetics; Springer-Verlag: New York, 1991; p 164. (2) Teissie, J. Biochemistry 1981, 20, 1554-1560. (3) Pinheiro, T. J. T.; Watts, A. Biochemistry 1994, 33, 2451-2458. (4) Lei, C.; Wollenberger, U.; Cheller, F. W. Electroanalysis 1999, 11, 274276. (5) Choi, E. J.; Dimitriadis, E. K. Biophys. J. 2004, 87, 3234-3241. (6) Robinson, N. C. J. Bioenerg. Biomembr. 1993, 25, 153-163. (7) McMillin, J. B.; Dowhan, W. Biochim. Biophys. Acta 2002, 1582, 97107. (8) Dome`nech, O Å .; Sanz, F.; Montero, M. T.; Herna´ndez-Borrell, J. Biochim. Biophys. Acta 2006, 1758, 213-221.
interactions. Such phospholipids are present in significant amounts and constitute the phospholipid matrix’s inner mitochondrial membranes. These phospholipids include phosphoethanolamine (PE) and phosphocholine (PC). In previous studies, we examined the mixing properties of binary systems 1-palmitoy-2-oleoylsn-glycero-3-phosphoethanolamine (POPE), 1-palmitoy-2-oleoyl-sn-glycero-3-phosphocholine (POPC), and cardiolipin monolayers and supported planar bilayers (SPBs).8,9 According to the values of the Gibbs energy of mixing, the most stable monolayers assayed were POPC/CL (0.6:0.4 mol/mol) and POPE/CL (0.8: 0.2 mol/mol). Hence these compositions were used to investigate the topographic characteristics of the SPBs by AFM. Thus, whereas POPC/POPE and POPC/CL yield homogeneous SPBs, lateral phase segregation of the POPE/CL system was observed.10 In the present study, we will investigate the effects of cyt c on pure POPC, POPE, and CL and on the more thermodynamically stable mixtures POPE/CL (0.8:0.2 mol/mol) and POPC/CL (0.6: 0.4 mol/mol) at constant surface pressure. By combining our monolayer results with binding experiments in solution using 8-anilino-1-naphtalene sulfonate (ANS),11 we investigated whether CL is essential for cyt c adsorption onto bilayers. 2. Materials and Methods 2.1. Chemicals. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), and cardiolipin (CL) (specified as 99% pure) were purchased from Avanti Polar Lipids (Alabaster, AL) and used without further purification. The nominal fatty acid composition of CL was 16:0 (1.0%), 16:1 (1.0%), 18:1 (5.5%), 18:2 (90.3%), and 18:3 (1.0%). Horse heart cyt c was purchased from Sigma Chemical Co. (Madrid, Spain). 8-Anilino-1-naphthalene sulfonate (ammonium salt, ANS) was purchased from Molecular Probes (Eugene, OR). The concentration of the spreading solutions of phospholipids was 1 mg mL-1 (9) Dome`nech, O Å .; Torrent-Burgue´s, J.; Merino, S.; Sanz, F.; Montero, M. T.; Herna´ndez-Borrell, J. Colloids Surf., B 2005, 41, 233-238. (10) Dome`nech, O Å .; Montero, M. T.; Herna´ndez-Borrell, J. Biochim. Biophys. Acta 2007, 1768, 100-106. (11) Slavı´k, J. Biochim. Biophys. Acta 1982, 694, 1-25.
10.1021/la0634241 CCC: $37.00 © 2007 American Chemical Society Published on Web 04/10/2007
5652 Langmuir, Vol. 23, No. 10, 2007
Dome` nech et al.
in chloroform/methanol (3:1 v/v). The buffer used was 50 mM TrisHCl buffer (pH 7.40) containing 150 mM NaCl and was prepared in ultrapure water (Milli Q reverse osmosis system, 18.3 MΩ cm resistivity) and filtered with a Kitasato system (450 nm pore diameter) before use. 2.2. Preparation of Lipid Monolayers: Surface PressureArea Isotherms and Constant Surface Pressure Experiments. The surface pressure (π) was measured with a resolution of (0.1 mN m-1 using a 312 DMC Langmuir-Blodgett (LB) trough (NIMA Technology Ltd. Coventry, England). The working surface area of the trough was 137 cm2. The trough was placed on a vibrationisolated table (Newport, Irvine, CA) and enclosed in an environmental chamber. The temperature of the subphase was maintained at 24.0 ( 0.2 °C by an external circulating water bath. Before each experiment, the trough was washed with chloroform and rinsed throughly with purified water. The cleanliness of the trough and subphase formed by the buffer was ensured before each run by cycling the full range of the trough area and aspirating the air-water surface, while at the minimal surface area, to zero surface pressure. Constant surface pressure experiments were performed by spreading the desired lipid onto the subphase free of cyt c and then compressing the lipid film to the desired surface pressure (π ) 30 mN m-1). Appropiate volumes of the spreading solution of lipids were carefully deposited onto the subphase by using a microsyringe (Hamilton, Reno, NV). After 15 min to allow for monolayer stabilization, cyt c was injected beneath the monolayer to a final concentration of 2 µM (equivalent to a lipid/protein molar ratio LPR of 15). The change in surface area per molecule (∆A) with time (t) was then measured. To obtain a quantitative interpretation of these experiments, data were fitted to the following Langmuir-like equation (kit)b ∆A ) ∆Amax 1 + (kit)b
(1)
where ∆Amax is the maximum increase in area reached by the monolayer under steady-state conditions, ki is a rate constant determined by the nature of the substances involved and the experimental conditions, and b is a parameter that is related to the cooperativity of the process. For compression isotherms, the corresponding aliquot of lipid was spread onto the subphase solution. The experiment was started after a 15 min period. The compression barrier speed, calculated at the final surface pressure, was 5 cm2 min-1. Every π-A isotherm was repeated at least three times. The isotherms showed satisfactory reproducibility. Z-type LB films (transfer on Z-type on the upstroke only)12 were then transferred onto freshly cleaved mica. The substrate was lifted at a constant rate of 1 mm min-1. The transfer ratios were evaluated and were close to unity, indicating that the mica was almost totally covered by the monolayer. 2.3. Atomic Force Microscopy. Images were generated with a commercial AFM (Nanoscope III, Digital Instruments, CA) and Si3N4 cantilevers (Olympus, Tokyo, Japan) that had a nominal spring constant of 40 N m-1. The instrument was equipped with a J scanner (120 µm) and was used in tapping mode. Mica squares (0.4 cm2, Asheville-Schoonmaker Mica Co., Newport News, VA) were glued to a steel disc. Images were recorded in air at constant force. The set point was continuously adjusted during the imaging to minimize the force applied. All of the images were processed using Digital Instruments software. 2.4. Preparation of Liposomes and Binding Experiments. Chloroform/methanol (3:1 v/v) stock solutions of the desired lipid compositions were evaporated to dryness in a flask using a rotovapor. The resulting thin lipid film was then kept under reduced pressure overnight to ensure the absence of traces of organic solvent. Multilamellar vesicles were obtained by hydration in buffer to a final concentration of 60 µM. Thereafter, suspensions were filtered through Nucleopore (Costar, Cambridge, MA) polycarbonate filters (12) Petty, M. C. Langmuir-Blodgett Films: An Introduction; Cambridge University Press: New York, 1996; p 39.
(200 nm nominal diameter) using an extruder (Lipex Biomembranes Inc., Canada) device. The size and polydispersity of liposomes were monitored by quasi-elastic light scattering (QLS) using an Autosizer IIc photon correlation spectrophotometer (Malvern Instruments, U.K.). A dominant peak was typically observed indicating the presence of ∼200 ( 20 nm particles. The final lipid concentration was 50 µM as determined by phosphorus quantification. Lipidbound ANS showed intense fluorescence at 480 nm (excitation 380 nm), whereas nonbound ANS exhibited low fluorescence in aqueous solution. All measurements of ANS fluorescence were made in an SLM-Aminco 8100 spectrofluorometer provided with a jacketed cuvette holder. The temperature of the cuvette was kept constant at 24.0 ( 0.1 °C using an external circulating water bath (Haake, Germany). Stock solutions of ANS in ethanol were freshly prepared and protected from light exposure. These experiments consisted of incubating cyt c to a final lipid/protein molar ratio of 15 (LPR ≈ 1, w/w) for 30 min at 24 °C. The samples were then titrated with stock solutions of ANS. The adsorption data for ANS binding on bilayers were fitted to the following Freundlich-like isotherm13 (k[ANS]∞)b [ANS]B ) Cmax 1 + (k[ANS]∞)b
(2)
where k is the binding constant, Cmax is the maximum concentration of ANS bound to liposomes, b is the cooperativity of the binding process, and subscripts B and ∞ refer to the bound and free ANS concentrations, respectively. The value of bound ANS can be calculated by taking into account that it is proportional to the fluorescence. Equation 2 can be used to calculate variations in the electrostatic surface potential according to ∆Ψ )
( )
RT kcyt ln F k0
(3)
where T is the temperature, R is the universal gas constant, F is the Faraday constant, and kcyt and k0 are the binding constants with and without the presence of cyt c, respectively.
3. Results and Discussion First, we have carried out constant surface pressure measurements of monolayers at 30 mN m-1, which is often assimilated to the surface pressure of the biological membranes.14 In addition, high surface pressures have the advantage of ensuring the efficient transfer of monolayers onto solid substrates.15 Upon injection of cyt c beneath the monolayers, changes in the average area per molecule (∆A) with time (t) were followed (Figure 1). As can be seen, maximum and minimum variations in ∆A were observed in pure CL (anionic phospholipid) and pure POPE (zwitterion) monolayers, respectively. For a more detailed comparison, the parameters obtained by fitting the experimental data for all of the monolayers to eq 1 are listed in Table 1. Here, for instance, it can be seen that the ∆Amax for CL is ∼9 times the ∆Amax for POPE. However, the k and b values for the pure CL monolayer were lower than the values for the other phospholipids. In addition, the ∆Amax values were slightly higher for pure POPC and mixed POPC/CL monolayers than for pure POPE. It is worth noting that the ∆Amax reached by the POPE/CL mixture was significantly higher than that of the other monolayers with the exception of that for the pure CL monolayer. Nevertheless, this does not provide an explanation for the different adsorption that we have observed between the two binary compositions reported in this article. Indeed, the difference in (13) Montero, M. T.; Pijoan, M.; Merino-Montero, S.; Hernandez-Borrell, J. Langmuir 2006, 22, 7574-7578. (14) Cevc, G.; Marsh, D. Phospholipid Bilayers: Physical Principles and Models; Wiley-Interscience: New York, 1987. (15) Egusa, S.; Gemma, N.; Azuma, M. J. Phys. Chem. 1990, 94, 2512-2518.
Adsorption of Cytochrome c on Mono- and Bilayers
Langmuir, Vol. 23, No. 10, 2007 5653
Table 1. Parameters Obtained from Constant Surface Pressure Experimentsa
∆Amax (Å2 molecule-1) k (× 10-2) s-1 b
POPC
POPE
CL
POPC/CL (0.6:0.4 mol/mol)
POPE/CL (0.8:0.2 mol/mol)
2.9 ( 0.2
1.49 ( 0.13
12.9 ( 0.5
2.23 ( 0.09
6.4 ( 0.3
2.8 ( 0.5 0.91 ( 0.13
2.5 ( 0.5 1.2 ( 0.3
1.7 ( 0.2 0.67 ( 0.03
2.5 ( 0.2 1.00 ( 0.09
2.9 ( 0.4 0.91 ( 0.11
a ∆Amax is the maximum increase in the area reached by the monolayer under steady-state conditions, ki is the rate constant, and b is the cooperativity of the process.
Figure 1. Increase in surface area as a function of time after the injection of cyt c beneath monolayers of (9) CL, (2) POPE, (O) POPC, (b) POPE/CL (0.8:0.2 mol/mol), and (0) POPC/CL (0.6:0.4 mol/mol).
the behavior of POPC/CL (0.6:0.4 mol/mol) and POPE/CL (0.8: 0.2 mol/mol) could not be a mere indication of the electrostatic nature of the cyt c bilayer. If this were the case and assuming that the main driving force of the interaction would be electrostatic, then the POPC/CL monolayer, with a higher proportion of the anionic phospholipid, should have a ∆A that is higher than that of the POPE/CL monolayer. This was not observed. Our results indicate that the nature of the interaction between cyt c and the phospholipid mixtures is not governed solely by electrostatic interactions.16 Whereas the interaction between cyt c and monolayers of zwitterionic phospholipids POPE and POPC involves a hydrophobic contribution,17 the presence of anionic phospholipids2 (in that particular study, phosphatidylglycerol) has been showed to enhance the penetration of cyt c. Notice that whether the process undergone by the pure CL monolayer is, according k and b values, slower and less cooperative than the rest of the monolayers, it shows the largest ∆Amax values. Taking into account that cyt c is positively charged at this pH,4 the increased penetration observed in the presence of CL could be attributed to the contribution of electrostatic interactions. We know from a previous study, in which we examined the mixing properties of POPC/CL (0.6:0.4 mol/mol) and POPE/CL (0.8:0.2 mol/mol) monolayers,8 that the excess Gibbs energy of mixing (GE) for both compositions is negative. Therefore, the interactions between the molecules are mainly attractive (Supporting Information). We concluded that in the range of the molar fraction studied (χCL ) 0 to 1) either POPE or POPC mix ideally. In a recent study18 based on the existence of two collapse (16) Kostrzewa, A.; Pali, T.; Froncisz, Marsh, D. Biochemistry 2000, 39, 60666074. (17) Quinn, P. J.; Dawson, R. M. C. Biochem. J. 1969, 113, 791-804. (18) Sennato, S.; Bordi, F.; Cametti, C.; Coluzza, C.; Desideri, A.; Rufini, S. J. Phys. Chem. B 2005, 109, 15950-15957.
Figure 2. Surface pressure versus area per molecule isotherms of (9) POPE, (b) POPE/CL (0.95:0.05 mol/mol), (0) POPE/CL (0.90: 0.10 mol/mol), and (O)POPE/CL (0.80:0.20 mol/mol).
pressures in the χCL ) 0.1-0.3 range, it was inferred that phase separation of the components occurs. In addition, the thermodynamic analysis was supported by AFM observations. Therefore, in the search for possible analogies, we extended our study to investigate the surface pressure-area (π-A) isotherms of our systems for χCL < 0.2. Thus, whereas POPC/CL monolayers were miscible in all proportions, even χCL < 0.2 (data not shown), two well-defined surface collapse pressures were observed for the POPE/CL system at χCL < 0.2 (Figure 2). This finding agrees reasonably well with the behavior of the DPPE/CL system.18 According to the Gibbs phase rule, this indicates that POPE and CL are not completely miscible at the molecular level at the air-water interface at χCL < 0.2. Apparently, there are discrepancies between the GE values for POPC/CL and POPE/CL that we have published8 elsewhere and the values reported for DPPC/CL and DPPE/CL monolayers.18 Thus, whereas for the mixtures of CL with DPPC and DPPE the GE values were always positive, the mixtures of CL with POPC and POPE yield negative values of GE in the range of χCL ) 0.2-0.6 and positive values for χCL < 0.2 and χCL > 0.6 (Supporting Information). Therefore, the behavior was similar in both cases, showing a clear tendency toward zero, which indicates ideal mixing, as the CL molar fraction decreases. In any event, our results suggest that attractive interactions between the components prevail, being larger for POPC/CL than for POPE/CL monolayers. In relative terms, that means that the contribution of the repulsive interactions to the net GE values is more important in the POPE/CL than in the POPC/CL system. Theoretically, it is reasonable to assume that this will result in lateral segregation of the molecules in the POPE/CL system at χCL < 0.2. These domains have been observed by AFM in supported planar bilayers of POPE/CL (0.8:0.2 mol/ mol),8 and we found it adequate to use the same technique in LBs of this system and others under investigation in this work.
5654 Langmuir, Vol. 23, No. 10, 2007
Dome` nech et al.
Figure 3. AFM tapping-mode images of (A) POPC, (B) POPE, (C) CL, (D) POPC/CL (0.6:0.4 mol/mol), and (E) POPE/CL (0.8:0.2 mol/mol) monolayers extracted at 30 mN m-1. Amplification of the image (to the left of E).
To perform a topographical characterization of such monolayers, LBs of the systems under investigation were transferred onto a mica substrate at 30 mN m-1 and imaged with AFM in tapping mode (Figure 3). As can be seen, POPC (Figure 3A) and POPC/CL (Figure 3D) (0.6:0.4 mol/mol) appeared to be continuous and smooth with average surface roughnesses of Ra ) 0.10 and 0.06 nm, respectively. However, the LBs of POPE (Figure 3B) and CL (Figure 3C) were not homogeneous. In the monolayer of POPE, with Ra ) 0.14, holes with a diameter of 75 ( 15 nm (n ) 50) were observed. These were randomly distributed, covering ∼6.39% of the total substrate area. The step-height difference between the holes and the monolayer was ∼1.16 nm. In the LB of CL (Figure 3C), structures with no particular shape were seen. The step-height difference between these structures and the monolayer was ∼0.90 nm. This height was lower than that expected for a monolayer; therefore, we scratched the surface to uncover the substrate, which enables us to measure the step height between the top of the monolayer and the uncovered mica (Supporting Information). The height was then established to be ∼1.4 nm. Indeed, we have recently shown that these determinations can be inaccurate in those cases in which the jump-to-contact force exceeds the mechanical stability of the monolayer.19 The LB of the POPE/CL (0.8:0.2 mol/mol)
monolayer also yields a nonhomogeneous structure (Figure 3E). Interestingly, a magnification of two small areas of this LB (see the magnifications in the insets on the left side of Figure 3E) reveals the existence of two different regions resembling the features of the pure POPE (Figure 3B) and CL (Figure 3C). This last observation provides evidence of a lateral segregation of the phospholipids in the monolayer that could be related to or induce the formation of domains observed in supported planar bilayers (SPBs) of the same composition,10 which has been unambiguously demonstrated by differential scanning calorimetry experiments carried out with liposomes of the same composition.20 Although the correlation between monolayer and bilayer properties has been successful in some cases,21 a note of caution should be added because in Figure 3 we are showing the topographies of the acyl chains facing the air and also because the surface pressure of the SPBs is not known. However, as stated elsewhere,22 (19) Garcia-Manyes, S.; Dome`nech, O Å .; Sanz, F.; Montero, M. T.; Herna´ndezBorrell, J. Biochim. Biophys. Acta, in press, 2006. (20) Epand, R. F.; Tokarska-Schlattner, M.; Schlattner, U.; Wallimann, T.; Epand, R. M. J. Mol. Biol. 2007, 365, 968-980. (21) Ross, M.; Steinem, C.; Galla, H. J.; Janshoff, A. Langmuir 2001, 17, 2437-2445. (22) Milhiet, P. E.; Domec, C.; Giocondi, M. C.; Van Mau, N.; Heitz, F.; Le Grimellec, Ch. Biophys. J. 2001, 81, 547-555.
Adsorption of Cytochrome c on Mono- and Bilayers
Langmuir, Vol. 23, No. 10, 2007 5655
Figure 5. Concentration of ANS bound to liposomes formed with (A) POPC/CL (0.6:0.4 mol/mol) and (B) POPE/CL (0.8:0.2 mol/ mol) as a function of the free ANS concentration in the (9) absence and (0) presence of 2 µM cyt c. The total lipid concentration is 50 µM. Each value is the average of three independent experiments.
Figure 4. Concentration of ANS bound to liposomes formed with (A) POPC, (B) POPE, (C) CL as a function of the free ANS concentration, and in the (9) absence and (0) presence of 2 µM cyt c. The total lipid concentration is 50 µM. Each value is the average of three independent experiments.
comparing supported monolayers and bilayers remains in some cases difficult. In addition, the head groups of the phospholipids, particularly the negative charge borne by CL and the hydrogen bonding ability of POPE,9 may play a key role in the formation of the domains. An appealing possibility could be that cyt c is selectively adsorbed on enriched CL domains in monolayers. This would
explain the higher adsorption of cyt c in the POPE/CL (0.8:0.2 mol/mol) system (Figure 1). Conversely, when CL is mixed with POPC, the electrostatic interaction is less favored. Thus, the three to four molecules of CL that are necessary for electron transport through cyt c oxidase23 would be indicative of a stoichiometric requirement for cyt c-CL binding. This requirement could be fulfilled by the POPE/CL (0.8:0.2 mol/mol) system because of lateral segregation. However, it could not be fulfilled by the POPC/CL (0.6:0.4 mol/mol) monolayer. To investigate the favored interaction of cyt c further, binding experiments have been carried out on liposomes. ANS is a fluorescent molecule that has proven to be useful for monitoring changes in the hydrophilic phosphate moiety of phospholipids caused by drugs13,24 and by cyt c itself.25 Typically, fluorescence intensities increase as ANS is added to liposomes until a plateau is reached at high probe concentrations (Figures 4 and 5). In the presence of cyt c, ANS is displaced as a result of the competition between the protein and the label for the same binding sites.10 (23) Robinson, N. C. J. Bioenerg. Biomembr. 1993, 25, 153-163. (24) Va´zquez, J. L.; Berlanga, M.; Merino, S.; Dome`nech, O Å .; Vin˜as, M.; Montero, M. T.; Hernandez-Borrell, J. J. Photochem. Photobiol. 2001, 73, 1419. (25) Teissie, J.; Baudras, A. Biochimie 1977, 59, 693-703.
5656 Langmuir, Vol. 23, No. 10, 2007
Dome` nech et al. Table 2. Parameters Obtained from ANS Binding Dataa
Cmax (µM) b k (µM-1) × 10-2 ∆ψ (mV)
5.05 0.97 3.3
POPC
POPE
+ cyt c
+ cyt c
3.44 1.50 4.2 +2.40 ( 1.50
2.5 1.02 2.5
3.0 1.53 3.6 +3.40 ( 1.80
CL
POPC/CL (0.6:0.4 mol/mol)
+ cyt c 2.92 1.3 3.4
0.47 1.1 5.9 +13.4 ( 1.5
POPE/CL (0.8:0.2 mol/mol)
+ cyt c 7.4 1.22 3.1
4.43 1.49 3.7 +0.20 ( 1.30
+ cyt c 6.4 1.21 2.3
2.38 1.40 5.0 +8.40 ( 1.20
aC max is the fluorescence intensity (related to the maximum concentration bound), b is the cooperativity, k is the binding constant, and ∆ψ is the variation in the surface potential of liposomes.
Consequently, in general, the fluorescence of ANS decreases in the presence of cyt c. The values obtained by fitting the isotherm (eq 2) to the ANS binding data and the calculated surface potential values (eq 3) are shown in Table 2. As can be seen, the values reached at t f ∞ (Cmax) were always lower in the presence of cyt c with the exception of POPE. In the absence of the protein, the highest and lowest values of Cmax were found for the POPC/CL (0.6:0.4 mol/mol) and the CL and POPE bilayers, respectively. In the presence of cyt c, the maximum displacement of ANS (i.e., the highest binding of the protein) was observed in the POPE/CL (0.8:0.2 mol/mol) composition. This is in agreement with the observation that the highest adsorption occurred in the POPE/ CL monolayer (Figure 1). However, the incorporation of cyt c to into liposomes produces a slight increases in k, which means that the probe affinity for bilayers increases in the presence of the protein. With the exception of CL, it can be said that the k and b values are higher in the presence of cyt c in all cases. As can be seen by a simple inspection of the binding curves, liposomes formed from pure phospholipids bind very low amounts of cyt c to POPC (Figure 4A) and almost negligible amounts of cyt c to POPE and CL (Figure 4B,C). However, the highest binding was observed in mixtures of either of POPC/CL (Figure 5A) or POPE/CL (Figure 5B). This behavior is in close agreement with the minimal adsorption of cyt c to POPC and POPE (Figure 1). This suggests, in agreement with the results reported,17 that cyt c cannot penetrate into zwitterionic lipid monolayers. The variations in the electrostatic surface potentials (∆Ψ) calculated according to eq 3 are also shown in Table 2. Because of the high standard deviations of the measurements, only the ∆Ψ values for CL and POPE/CL are significant. The positive values for the POPE/CL and CL systems indicate that cyt c is able to screen the negative charge born by these liposomes. This behavior appears to be related to the high adsorption of cyt c on pure CL and POPE/CL monolayers (Figure 1). It is important to bear in mind that CL is a phospholipid that has two different binding sites for cyt c.26,27 In addition, recent surface force measurements28 have revealed that the absolute surface charge
born by CL liposomes cannot be inferred by simply assuming that the phospholipid is fully deprotonated at neutral pH. Changes in the structure of the bilayer in the presence of the cyt c, involving the postulated CL extended conformation,29 could reveal the charge that is partially screened by the intramolecular hydrogen bond. Indeed, the interaction of cyt c with acidic phospholipids and particularly the level of penetration of the protein into the bilayer seem to depend on a subtle balance between electrostatic and hydrophobic interactions.30 Other factors that could influence the interaction are the orientation of the protein31,32 and its possible partial unfolding at the surface of anionic phospholipids.3,33 Whatever the exact mechanism involved, the results presented in this article indicate that the POPE/CL system, with the nominal mole fraction of CL found in the inner mitochondrial membrane, provides additional evidence8,9 of the selective adsorption of cyt c on CL domains. Acknowledgment. This work has been supported by grants CTQ2005-07989 from the Ministerio de Ciencia y Tecnologı´a (MCYT) and SGR00664 from DURSI (Generalitat de Catalunya) Spain. Supporting Information Available: Excess area per molecule as a function of composition for pure and mixed POPE/CL monolayers at 30 mN m-1. The POPE/CL monolayer transferred at 30 mN m-1 (shown in Figure 3E) has been scratched over a small area to evaluate its thickness; it is shown along with a height profile analysis. This material is available free of charge via the Internet at http://pubs.acs.org. LA0634241 (26) Ryto¨maa, M.; Kinnunen, P. K. J. J. Biol. Chem. 1994, 269, 1770-1774. (27) Gorbenko, G. P.; Molotovsky, J. G.; Kinnunen, P. K. J. Biophys J. 2006, 90, 4093-4103. (28) Nichols-Simth, S.; Kuhl, T. Colloids Surf., B 2005, 41, 121-127. (29) Kinnunen, P. K. J. Chem. Phys. Lipids 1996, 81, 151-166. (30) Bernad, S.; Oellerich, S.; Soulimane, T.; Noinville, S.; Baron, M. H.; Patermostre, M.; Lecomte, S. Biophys J. 2004, 86, 3863-3872. (31) Gorbenko, G.; Domanov, Y. Biophys. Chem. 2003, 103, 239-249. (32) Domanov, Y. A.; Molotkovsky, J. G.; Gorbenko, G. P. Biochim. Biophys. Acta 2005, 1716, 49-58. (33) Spooner, P. J. R.; Watts, A. Biochemistry 1991, 30, 3871-3879.
