- Email: [email protected]
oleate vesicles studied by spin labeling
Chemistry and Physics of Lipids 164 (2011) 83–88
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
On the surface properties of oleate micelles and oleic acid/oleate vesicles studied by spin labeling b ˇ Branka Dejanovic´ a , Vesna Noethig-Laslo a,∗ , Marjeta Sentjurc , Peter Walde c a b c
Department of Physical Chemistry, “Rud¯er Boˇskovi´c” Institute, 10002 Zagreb, Croatia “Jozef Stefan” Institute, Ljubljana, Slovenia ETH Zürich, Department of Materials, Zürich, Switzerland
a r t i c l e
i n f o
Article history: Received 29 April 2010 Received in revised form 5 October 2010 Accepted 3 November 2010 Available online 12 November 2010 Keywords: Spin labels Micelles Vesicles Oleic acid Sodium oleate Hydrogen bond network
a b s t r a c t Dilute aqueous systems composed of sodium oleate micelles and sodium oleate/oleic acid vesicles were investigated as a function of pH by electron spin resonance spectroscopy with TEMPO-stearate TEMPOstearamide as well as with a positively charged water soluble spin label, TEMPO-choline. The dynamics of the three TEMPO-spin labels were found to be sensitive to changes in the interfacial region of the aggregates as a function of pH. The results obtained are consistent with the formation of a hydrogen bond network (RCOO− ↔ HOOCR) at the surface of the sodium oleate/oleic acid system in the course of the transformation of micelles into the closed bilayers (vesicles). Vesicles formation below pH 10 was determined independently with a spin labeled glucose derivative. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Formation of vesicles from oleic acid is known since the pioneering work of Gebicki and Hicks (1973). At the appropriate concentration, vesicle formation is observed in dilute aqueous solution, if about half of the oleic acid molecules are deprotonated, which is the case at a measured pH value of about 8.5 (Gebicki and Hicks, 1973; Hargreaves and Deamer, 1978; Morigaki and Walde, 2007). In an earlier study, electron spin resonance (ESR) spectroscopy was used to investigate the pH-induced transformation of oleate micelles into mixed oleic acid/oleate vesicles, using the fatty acid spin label 16-doxylstearic acid located in the hydrophobic region of the micelles and vesicles (Fukuda et al., 2001). It was found that the experimental ESR spectra could only be simulated if the presence of two different environments of the doxyl group was considered, one in vesicle bilayers and one in non-vesicular (micellar) aggregates (Fukuda et al., 2001). In order to clarify whether micelles and vesicles coexist in the fatty acid system composed of shorter alkyl chains, ESR spectroscopy with different lipophilic spin labels was applied to the decanoic acid/decanoate system. Two different environments of the spin labels were determined and the presence of micelles and vesicles in
∗ Corresponding author. Tel.: +385 1 4561136; fax: +385 1 4680245. E-mail address: [email protected] (V. Noethig-Laslo). 0009-3084/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2010.11.002
equilibrium was confirmed in the interval 6.5 < pH < 7.8 (Dejanovic´ et al., 2008). The physicochemical properties of the surface of oleic acid/oleate vesicles or oleate micelles were not yet studied with ESR spectroscopy. We now report on such investigation which fully supports the proposed formation of a RCOO− ↔ HOOCR hydrogen bond network in the presence of both the ionized and the neutral form of the fatty acid (Apel et al., 2002). In order to obtain more information on the role of hydrogen bond networks for the transformation of oleic acid/sodium oleate micelles to vesicles, the TEMPO-stearate and TEMPOstearamide spin labels are used (Fig. 1). The nitroxide moiety of these lipophilic spin labels is located at the surface of oleate micelles and oleic acid/oleate vesicles covered with bound water molecules, while the long saturated hydrocarbon chains are located in the hydrophobic part of the aggregates. Therefore, these spin labels are expected to reflect properties of the Stern layer, and report on the hydrogen bond networks at the surface of vesicles and micelles. In addition, the water soluble positively charged spin label TEMPO-choline (ASL) (Fig. 1), is expected to bind to the surface, depending on the negative surface charge, which in turn was expected to correlate with the micelle to vesicle transformation. The dynamics of the spin labels incorporated into micelles and vesicles was studied as a function of pH, temperature and sodium oleate concentrations. The formation of vesicles was confirmed by determining the presence of a captured aqueous
84
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
amounts of 1 M HCl were then added to get final concentrations of oleic acid + sodium oleate of 0.025 M or 0.050 M, with the pH values of different samples varying between 12.5 and 7.5.
2.3. Preparation of samples containing spin labels
Fig. 1. Structural formulae of the spin labels used. (A) (2,2,6,6-Tetramethylpiperidine-1-oxyl-4-yl)octadecanoate, TEMPO-stearate; (B) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanamide, TEMPO-stearamide; (C) 1-oxyl-2,2,6,6tetramethyl-4-(2 ,3 ,4 ,5 ,6 -pentahydroxyhexanoyl-1 -amino)-piperidine, GluSL); (D) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-N,N-dimethylaminoethanol, ASL).
Acetone solutions of TEMPO-spin labels were added to the sodium oleate solution at pH 12.5 (molar ratio of the spin label to oleic acid/sodium oleate was 1:300). The solution was optically clear. The solution was vigorously hand shaken, and was left over night at room temperature. The micellar solution was titrated with 1 M HCl to a chosen pH value. The test tubes (1.5 mL of Na-oleate solution) were vortexed for 15 min and equilibrated at 25 ◦ C for 12 h before measurements. In another procedure acetone solutions of the TEMPO-spin labels were added to oleate solutions already titrated to different pH values. Both procedures gave similar result, i.e. same ESR spectra were obtained. Labeling with water soluble spin labels was performed by dissolving sodium oleate in 1 mM water solution of ASL or of GluSL. Further procedure was the same as for preparation of aqueous oleic acid systems.
2.4. CW-ESR measurements volume using a hydrophilic, spin-labeled glucose derivative, GluSL (Fig. 1). TEMPO-stearate was used recently in the study of interfacial ´ properties of phospholipid vesicles (liposomes) (Alves and Peric, 2006; Mirosavljevic´ and Noethig-Laslo, 2008; Alves et al., 2008). It was found that a negative charge in the liposome composition influences the amount of water bound at the surface, as well as the dynamics of the nitroxide moiety located at the liposome surface. Although the structural properties of the confined water in the phospholipid interface are different from that in the case of the fatty acid system (Disalvo et al., 2008), it was of interest to compare the interfacial properties of the surface of oleic acid/oleate vesicles and oleate micelles, with the interfacial properties of phospholipid vesicles, based on studies with TEMPO-spin labels. 2. Experimental
CW-ESR spectra were recorded with a X-band (9.5 GHz) Varian E 12 spectrometer equipped with a Bruker variable temperature control unit, and with the digital acquisition (EW-ESR WARE) (Moorse, 1987). Glass capillaries containing the sample were inserted into the standard 4 mm diameter quartz tubes and centered in a TE102 ESR cavity. Mn2+ in MgO and TEMPO in ethanol were used as external standards. ESR spectra of the spin-labeled oleic acid/sodium oleate system with lipophilic spin labels, characterized by anisotropic motions were simulated with the program “Chili” (Stoll and Schweiger, 2006). The spectra of ASL (the water soluble spin label) characterized by fast isotropic motion were simulated with “Garlic” of the EasySpin 3.1.1. software package (Stoll and Schweiger, 2006; McCarney et al., 2008). Best fit of the spectra was obtained by the least-square fitting method. The error 5% of the calculated parameters obtained from the simulated spectra was determined from the chi square values.
2.1. Materials Sodium oleate (99%) was purchased from Sigma–Aldrich (St. Louis, USA); sodium hydroxide (p.a.) and hydrogen chloride (p.a.) were from Kemika (Zagreb, Croatia). Aqueous samples were prepared in water of Milipore purity. The following spin labels were synthesized by Slavko Peˇcar, (Faculty of Pharmacy, University of Ljubljana, Ljubljana, Slovenia) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanoate, TEMPO-stearate; (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanamide, TEMPO-stearamide; (2,2,6,6-tetramethylpiperidine-1-oxyl-4-N,N-dimethylaminoethanol iodide) TEMPO-choline (ASL); and 1-oxyl-2,2,6,6tetramethyl-4-(2 ,3 ,4 ,5 ,6 -pentahydroxyhexanoyl-1 -amino)-piperidine, spin-labeled glucose (GluSL) (Babiˇc et al., 2003; Peˇcar et al., 1984). 2.2. Preparation of aqueous oleic acid samples Aqueous oleic acid/oleate samples in the absence of buffer salts were prepared by first dissolving sodium oleate in water (of Milipore purity). A small portion of 1 M NaOH was then added to this stock solution to obtain a pH of about 12.5. Appropriate
2.4.1. CW-ESR measurements of captured volume with GluSL The captured volume (determination of the presence of vesicles) (Peˇcar et al., 1984) was measured with the X-band ESR spectrometer (Bruker ELEXSYS E500) at room temperature. For this purpose, 30 L of 30 mM sodium ascorbate (Asc) was added to 30 L of 50 mM oleic acid samples with entrapped GluSL (1 mM). GluSL is a large molecule and cannot penetrate the vesicle membrane. Asc rapidly reduces the paramagnetic nitroxide group of the accessible GluSL molecules to the corresponding hydroxylamine, which does not give an ESR signal. It was supposed that due to its negative charge at neutral or alkaline pH, Asc – like the spin label GluSL – cannot penetrate intact membranes of the fatty acid vesicles. Therefore, only the non-entrapped GluSL molecules were reduced within the first few minutes after addition of Asc to the sodium oleate vesicles. The captured volume, Vi , was then obtained by measuring the intensity of the ESR spectra before and after addition of Asc. The captured volume (Vi ) is related to the total sample volume (V0 ) according to Vi /V0 = I/I0 , with I and I0 being the intensity of the ESR spectra immediately after Asc addition and before Asc addition, respectively.
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Fig. 2. ESR spectra of TEMPO-stearate taken at 300 K in oleate micelles at pH 11.5 (A) and in oleic acid/oleate vesicles at pH 8.66 (B). Narrow lines are experimental spectra and thick lines the spectral fit using the following parameters. (A) The spectra of both spin labels could be simulated supposing only one type of motion of the nitroxide moiety, that depended only on pH and temperature. (A) R⊥ = 3.6 × 108 s−1 , R =17.0 × 108 s−1 , ◦ = 0◦ ; (B) R⊥ = 2.3 × 108 s−1 , R = 17.0 × 108 s−1 , ◦ = 10◦ .
3. Results and discussion In order to gain insight into the surface properties of the sodium oleate micelles and oleic acid/sodium oleate vesicles, motional properties of TEMPO-stearate and TEMPO-stearamide incorporated into the micelles and vesicles were studied as a function of pH, temperature and total oleic acid + oleate concentration. At [oleic acid + oleate] = 25 mM and above pH 10, both lipophilic TEMPO-spin labels are fully incorporated into the micellar structures, because the uptake of free oleate by the micelles is significantly faster (with a correlation time <10−9 s) as compared to the release of oleate from the micelles into the bulk solution (with a correlation time >10−4 s) (Fukuda et al., 2001). Fast surfactant diffusion from water into micelles implies that the water exchange into and out of the micelles is even faster, with lipid to water association lifetimes of the order of <100 ps (McCarney et al., 2008). Formation of hydrogen bond networks in the course of the micelle to vesicle transformation on going from a high pH system to an intermediate pH will probably influence the dynamics of the TEMPO-spin labels located at the surface of the micelles or vesicles. With respect to this, it is expected that the TEMPO-stearate and TEMPO-stearamide spin labels (structural formulae are presented in Fig. 1), behave differently. The ester bond of TEMPO-stearate can form hydrogen bonds via an acceptor mechanism (>C O) while the amide group of TEMPO-stearamide may form both donor (>N–H) and acceptor (>C O) hydrogen bonds. Thus, a different dynamic behavior of the two TEMPO-spin labels is expected in the transformation of oleate micelles to oleic acid/oleate vesicles. ESR spectra of TEMPO-stearate and TEMPO-stearamide in oleate micelles (spectra A) and in oleic acid/oleate vesicles (spectra B) are shown in Figs. 2 and 3, respectively. The narrow lines are experimental data, and the thick lines the best fit of the spectra obtained with the program “Chili”, a part of the EasySpin 3.1.1.software package (Stoll and Schweiger, 2006). Spectra of both spin labels could be simulated supposing only one type of motion of the nitroxide moiety, that depended only on pH and temperature. From the simulation of the experimental spectra, the dynamic parameters of the two TEMPO-spin labels were obtained, i.e. the principal values (R , R⊥ ) of the rotational diffusion tensor, R, of the axial symmetry.
85
Fig. 3. ESR spectra of TEMPO-stearamide taken at 300 K in oleate micelles at pH 11.5 (A) and in oleic acid/oleate vesicles at pH 8.66 (B). Narrow lines are experimental spectra and thick lines the spectral fit using the following parameters. (A) R⊥ = 4.2 × 108 s−1 , R = 17.0 × 108 s−1 , ◦ = 10; (B) R⊥ = 1.57 × 108 s−1 , R = 16.0 × 108 s−1 , ◦ = 38◦ .
They represent the rotational rates of the nitroxide radical moiety around the axes parallel and perpendicular to the main symmetry axis of the radical (Ge and Freed, 1998). TEMPO-spin labels are “xordering” nitroxides (Van et al., 1974). The relatively free internal rotation of the nitroxide group about the C–O bond is determined by the rotational diffusion coefficient for an axial molecular motion, R , while the motion about the axis perpendicular to the C–O bond is determined by the off-axial rotational diffusion coefficients, R⊥ , (a rocking motion) (Ge and Freed, 1998, 2003). In our simulations, the Euler angles ˛ = 0◦ and ˇ = 90◦ were kept constant while the -angle describing deviations of the nitroxide moiety from the axial molecular motion was changed in order to get a satisfying fit of the spectra. Since the measurements of the principal components of the g and A tensors, neither for TEMPO-stearate nor for TEMPO-stearamide are not known, for simulations of the spectra, the starting values for the g and A tensors were taken from the works of Alves and Peric´ (2006). Table 1 lists the parameters required for the best fit of the experimental spectra (i.e. gxx , component of the g-tensor, Azz , the component of the hyperfine splitting tensor A, as well as the components of the rotational diffusion tensor R (R , R⊥ )) extracted from the optimal simulation for TEMPO-stearate in the oleic acid/oleate system as a function of pH. Table 2 lists the parameters for TEMPOstearamide. For both spin labels, a clear decrease of the off-axial rotational diffusion coefficients, R⊥ , was found on decreasing pH Table 1 pH dependence of the ESR parameters of TEMPO-stearate in 25 mM oleic acid/sodium oleate in 0.1 M NaCl, temperature 300 K. pH
10−8 R⊥ (s−1 )
10−8 R|| (s−1 )
gxx
a0 (Azz ) (mT)
(◦ )
11.52 10.70 10.30 10.00 9.85 9.80 8.75 8.66 8.30 7.34
3.60 3.60 3.60 3.45 2.90 2.80 2.30 2.30 2.00 1.90
17.0 17.0 17.0 17.0 18.0 18.0 17.0 17.0 17.0 17.0
2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.01005
1.65 (3.77) 1.65 (3.77) 1.65 (3.77) 1.65 (3.77) 1.64 (3.75) 1.64 (3.75) 1.64 (3.70) 1.64 (3.70) 1.65 (3.70) 1.65 (3.70)
0 0 0 0 10 10 10 10 10 10
86
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Table 2 pH dependence of the ESR parameters of TEMPO-stearamide in 25 mM oleic acid/sodium oleate in 0.1 M NaCl, temperature 300 K. pH
10−8 R⊥ (s−1 )
10−8 R|| (s−1 )
gxx
a0 (Azz ) (mT)
(◦ )
11.38 10.90 10.15 9.93 9.60 9.24 8.74 8.36 8.08 7.53
4.20 4.20 4.15 3.20 1.75 1.65 1.55 1.57 0.75 0.65
17.0 17.0 17.0 17.0 17.0 17.0 16.0 16.0 13.0 13.0
2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.01005
1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.67 (3.77) 1.67 (3.77) 1.67 (3.77) 1.66 (3.70) 1.66 (3.45)
10 10 10 10 25 28 35 40 38 38
value in the pH range 9.8 > pH > 8.0, while the rate of axial molecular motion, R , showed only a small decrease in the same pH region. In order to get the best fit of the experimental spectra recorded below pH 9.8 an increase of the -angle was from 0◦ to 10◦ was needed for TEMPO-stearate (Table 1), while for TEMPO-stearamide from 10◦ up to 40◦ (Table 2). Larger -angle in the case of the TEMPOstearamide is due to the fact that this spin label is both hydrogen bond donor (>N–H) and hydrogen bond acceptor (>C O). Deviation from the axial molecular motion of the nitroxide moiety, concomitant with a decrease of the off-axial rotational diffusion coefficients, R⊥ , determined with both spin labels suggest the formation of hydrogen bond network (RCOO− ↔ HOOCR) at the surface of the oleate system in the course of the micelle to vesicle transformation. In earlier studies, 16-doxylstearic acid was used as the spin probe. In the pH range 9.8 > pH > 8.0 motional properties of 16doxylstearic acid with the doxyl group located in the hydrophobic core of the micelles or vesicles could be understood by assuming two different spin label environments (Fukuda et al., 2001). This suggested the presence of both micelles and vesicles in equilibrium. The spin label 16-doxylstearic acid could distinguish the two types of aggregates. This is clearly different from the case of the two TEMPO-spin labels, TEMPO-stearate and TEMPO-stearamide. The ESR spectra of both TEMPO-spin labels could be fitted well by supposing only one type of the spin label environment. Since the TEMPO-spin labels report the dynamic properties at the surface of the oleic acid/oleate system, spectral fitting with only one type of spin label motion suggests similar surface properties of micelles and vesicles, known to coexist in the pH range 9.8 > pH > 8.0 (Hargreaves and Deamer, 1978; Morigaki and Walde, 2007; Fukuda et al., 2001). When TEMPO-spin labels (TEMPOoctanoate and TEMPO-decanoate) were applied to aqueous decanoic acid/sodium decanoate system, in the range 7.9 > pH > 6.8 for the best fit of the experimental spectra two different spin label environments had to be supposed, suggesting different surface properties of vesicles and micelles in that system (Dejanovic´ et al., 2008). Similar surface properties of micelles and vesicles in the oleic acid/oleate system in the pH range 9.8 > pH > 8.0 are probably due to longer and unsaturated aliphatic chains, as compared to the decanoic acid/decanote system. Temperature dependence of R and R⊥ extracted from the optimal simulation of the experimental spectra of TEMPO-stearamide at pH 9.9 (oleate micelles, squares) and at pH 8.3 (oleic acid/oleate vesicles, circles) is shown in Fig. 4. Linear temperature dependence of the values for R⊥ and R for both micelles and vesicles indicates that there are no temperature dependent structural changes in the surface of both micelles and vesicles. Smaller values for R⊥ at pH 8.3 (i.e. slower rate of motion, and larger -angle), where the number of vesicles prevail from that for R⊥ at pH 9.9 at all the temperatures examined, suggest stabile hydrogen bond networks (RCOO− ↔ HOOCR) in the pH region of coexistence of vesicles and micelles.
Fig. 4. Temperature dependence of the components of the rotational diffusion tensor, R⊥ and R , for TEMPO-stearamide in 25 mM oleic acid + oleate at pH 9.9 (R⊥ open squares; R|| closed squares) and at pH 8.4 (R⊥ open circles, R|| closed circles).
One of the important factors that might play a role in the transformation of the micelles into closed bilayers (vesicles) is the surface charge and the corresponding influence by the counterions. ASL is a positively charged water soluble spin label. It is expected to bind to the surface depending on the surface charge, and thus its binding property to correlate with the micelle to vesicle transformation. In Fig. 5, ESR spectra of ASL, in 25 mM sodium oleate at pH 11.16 (spectrum A) and at pH 8.47 (spectrum B) are compared. The spectra were fitted with the “Garlic” program (Stoll and Schweiger, 2006), so the motional properties of ASL are described by the rotational correlation time (/s). (The rotational diffusion tensor, R, and the rotational correlation time are related as R = 1/6). Starting values of the parameters required for the simulation of both spectral components were gxx = 2.0082, gyy = 2.0053, gzz = 2.0015;
Fig. 5. ESR spectra of TEMPO-choline (ASL) taken at 294 K in oleate system at: pH 11.2 (A); pH 8.47 (B). Narrow lines are experimental spectra and thick lines the spectral fit with the “Garlic” program. Spectrum A was fitted supposing 85% of spin labels with /s = 5 × 10−10 s; Azz = 3.51 mT, a01 = 1.62 mT, (component 1) and 15% with /s = 0.5 × 10−10 s, Azz = 3.70 mT, a02 = 1.68 mT (component 2); spectrum B, 30% of spin labels with 5 × 10−10 s (component 1) and 70% with 0.5 × 10−10 s (component 2).
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Fig. 6. pH dependence of the % of component 1 in the spectra, with rotational correlation time, /s = 5 × 10−10 s of TEMPO-choline (ASL) in 0.025 M sodium oleate + oleic acid. The spectra were taken at 294 K.
Axx = Ayy = 0.56 mT, Azz = 3.46 mT (Ge and Freed, 1998). Two different rotational correlation times (/s) of the spin label ASL were required for the best fit of the spectra A and B with a variable number of spin labels in the two spectral components. The number of ASL spin labels characterized with the rotational correlation time = 5.0 × 10−10 s (component 1) decreased from 85% at pH 11.16, to 45% was determined at pH 9.63 while to about 5% at pH 8.2 (Fig. 6). This observation can be explained as follows. Formation of hydrogen bonds at the micelles surface and reduction of the number of negative charges at the surface with decreasing the pH leads to a release of ASL molecules from the oleate surface to a more polar environment of the bulk water. The result is an increase in the number of ASL molecules with faster motion, i.e. with = 0.5 × 10−10 s (component 2). It is important to mention that below pH 9.9 (i.e. in the region in which vesicles formed) the two lipophilic TEMPO-spin labels examined in this work experienced a decrease in the nitroxide motion (R⊥ , a rocking motion) concomitant with an increase of the -angle which suggested formation of the hydrogen bond networks (RCOO− ↔ HOOCR) at the surface of the oleic acid/oleate aggregates. TEMPO-stearate and TEMPO-stearamide were applied recently in a study of interfacial properties of multilamellar liposomes
87
composed of egg phosphatidylcholine (Mirosavljevic´ and NoethigLaslo, 2008). It was found that both components of the rotational diffusion tensor, R(R and R⊥ ) extracted from the optimal simulation of the experimental spectra of both spin labels experienced slower rate of motion at all temperatures examined as compared with the dynamic parameters (R and R⊥ ) of these spin labels in the oleic acid/sodium oleate system. Thus, although oleic acid vesicles are multilamellar and structurally analogous to phospholipid vesicles, they have very different dynamic properties. This might be due to the fact that individual monomers of oleic acid are in rapid exchange between vesicles and micelles. In order to confirm that a decrease in the rate of motion of the two lipophilic TEMPO-spin labels is related to the formation of a closed volume (vesicles), pH dependent measurements of the captured volume were carried out with the spin label GluSL at the total oleic acid + sodium oleate concentration of 0.050 M. A captured volume was detected in the pH range 10 > pH > 7.5, with the highest entrapment yield in the pH range 9.9 > pH > 8.5 (Fig. 7). This is in qualitative agreement with earlier literature data (Hargreaves and Deamer, 1978; Morigaki and Walde, 2007; Fukuda et al., 2001; Walde et al., 1994). Fig. 7 indicates that vesicles are already present at pH 10, a pH region where micelles, however, dominate. A decrease of the captured volume to about 3% (Fig. 7) concomitant with increased amounts of a released ASL molecules from the surface of the oleic acid/oleate system (Fig. 6), followed by a slower motion of the TEMPO-stearamide below pH 8.1 (Table 2) suggests appearance of a second aggregate transition, i.e. namely from the oleic acid/oleate vesicles to oil droplets (Walde and Morigaki, 2007). The oil droplets are mainly composed of oleic acid molecules. 4. Conclusions Changes in the dynamic properties of both lipophilic TEMPOspin labels in oleic acid/sodium oleate system were determined in the region 9.9 > pH > 8.0, i.e. in the course of the micelle to vesicle transformations. Dynamic properties of both lipophilic TEMPO-spin labels suggest formation of hydrogen bond networks (RCOO− ↔ HOOCR) at the surface of the oleate system. Formation of hydrogen bonds at the aggregate surface below pH 10 was followed by measuring the release of positively charged ASL molecules from the aggregate surface to a more polar environment of the bulk water. Acknowledgements This work was supported by Croatian Ministry of Science, Education and Sport (Project No. 098-0982915-2939) and by the European Research Program on the Cooperation in Science and Technology within Europe: COST Chemistry, Action D 27 (Prebiotic Chemistry and Early Evolution). References
Fig. 7. Captured volume as a function of pH determined with GluSL in 0.050 M oleic acid + sodium oleate. The experiments were done at room temperature.
´ M., 2006. An EPR study of the interfacial properties of phosphatidylAlves, M., Peric, choline vesicles with different lipid chain lengths. Biophys. Chem. 122, 66–73. ´ M., 2008. Effect of lysophosphatidylcholine on the surface Alves, M., Bales, B.L., Peric, hydration of phospholipid vesicles. Biochim. Biophys. Acta 1778, 414–422. Apel, C.L., Deamer, D.W., Mautner, M.N., 2002. The formation of glycerol monodecanoate by a dehydration/condensation reaction: increasing the chemical complexity of amphiphiles on early Earth. Biochim. Biophys. Acta 1559, 1–9. Babiˇc, A., Ilaˇs, J., Peˇcar, S., 2003. Synthesis of hydrophilic and amphiphilic spin probes. Pharmazie 58, 599–600. ˇ ´ B., Mirosavljevic, ´ K., Noethig-Laslo, V., Sentjurc, Dejanovic, M., Peˇcar, S., Walde, P., 2008. An ESR characterization of micelles and vesicles formed in aqueous decanoic acid/sodium decanoate systems using different spin labels. Chem. Phys. Lipids 156, 17–25. Disalvo, E.A., Lairion, F., Martini, F., Tymcyszyn, E., Frias, M., Almaleck, H., Gordillo, G.J., 2008. Structural and chemical properties of hydration and
88
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
confined water in membrane interfaces. Biochim. Biophys. Acta 1778, 2655–2670. Fukuda, H., Goto, A., Yoshioka, H., Goto, R., Morigaki, K., Walde, P., 2001. Electron spin resonance study of the pH-induced transformation of micelles to vesicles in an aqueous oleic acid/oleate systems. Langmuir 17, 4223–4231. Ge, M., Freed, J.H., 1998. Polarity profile in oriented and dispersed phosphatidylcholine bilayers are different: an electron resonance spin study. Biophys. J. 74, 910–917. Ge, M., Freed, J.H., 2003. Hydration, structure and molecular interactions in the headgoup region in dioleoylphosphatidylcholine bilayers: an Electron Spin Resonance study. Biophys. J. 85, 4023–4040. Gebicki, J.M., Hicks, M., 1973. Ufasomes are stable particles surrounded by unsaturated fatty acid membranes. Nature 243, 232–234. Hargreaves, W.R., Deamer, D.W., 1978. Liposomes from ionic, single-chain amphiphiles. Biochemistry 17, 3759–3768. McCarney, E.R., Armstrong, B.D., Kausik, R., Han, S., 2008. Dynamic nuclear polarization enhanced nuclear magnetic resonance and electron spin resonance studies
of hydration and local water dynamics in micelle and vesicle assemblies. Langmuir 24, 10062–10072. ´ K., Noethig-Laslo, V., 2008. Spin labeling study of interfacial properties Mirosavljevic, of egg-phosphatidylcholine liposome as a function of cholesterol concentration. Chem. Phys. Lipids 155, 74–79. Morigaki, K., Walde, P., 2007. Fatty acid vesicles. Curr. Opin. Colloid Interface Sci. 12, 75–80. Moorse, P.D., 1987. EW-ESR ware. Biophys. J. 51, 440a. Peˇcar, S., Sorg, B., Schara, M., Hecker, E., 1984. Spin-labeled phorbol esters and their interaction with cellular membranes. I. Synthesis of spin-labeled phorbol. 12,13Diesters and related compounds. Chem. Phys. Lipids 35, 151–159. Stoll, S., Schweiger, A., 2006. EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 178, 42–55. Van, S.P., Birell, G.B., Griffith, G.B., 1974. Rapid anisotropic motion of spin labels. Models for motion averaging of ESR parameters. J. Magn. Reson. 15, 444–459. Walde, P., Wick, R., Fresta, M., Mangone, A., Luisi, P.L., 1994. Autopoietic selfreproduction of fatty acid vesicles. J. Am. Chem. Soc. 116, 11649–11654.
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
On the surface properties of oleate micelles and oleic acid/oleate vesicles studied by spin labeling b ˇ Branka Dejanovic´ a , Vesna Noethig-Laslo a,∗ , Marjeta Sentjurc , Peter Walde c a b c
Department of Physical Chemistry, “Rud¯er Boˇskovi´c” Institute, 10002 Zagreb, Croatia “Jozef Stefan” Institute, Ljubljana, Slovenia ETH Zürich, Department of Materials, Zürich, Switzerland
a r t i c l e
i n f o
Article history: Received 29 April 2010 Received in revised form 5 October 2010 Accepted 3 November 2010 Available online 12 November 2010 Keywords: Spin labels Micelles Vesicles Oleic acid Sodium oleate Hydrogen bond network
a b s t r a c t Dilute aqueous systems composed of sodium oleate micelles and sodium oleate/oleic acid vesicles were investigated as a function of pH by electron spin resonance spectroscopy with TEMPO-stearate TEMPOstearamide as well as with a positively charged water soluble spin label, TEMPO-choline. The dynamics of the three TEMPO-spin labels were found to be sensitive to changes in the interfacial region of the aggregates as a function of pH. The results obtained are consistent with the formation of a hydrogen bond network (RCOO− ↔ HOOCR) at the surface of the sodium oleate/oleic acid system in the course of the transformation of micelles into the closed bilayers (vesicles). Vesicles formation below pH 10 was determined independently with a spin labeled glucose derivative. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Formation of vesicles from oleic acid is known since the pioneering work of Gebicki and Hicks (1973). At the appropriate concentration, vesicle formation is observed in dilute aqueous solution, if about half of the oleic acid molecules are deprotonated, which is the case at a measured pH value of about 8.5 (Gebicki and Hicks, 1973; Hargreaves and Deamer, 1978; Morigaki and Walde, 2007). In an earlier study, electron spin resonance (ESR) spectroscopy was used to investigate the pH-induced transformation of oleate micelles into mixed oleic acid/oleate vesicles, using the fatty acid spin label 16-doxylstearic acid located in the hydrophobic region of the micelles and vesicles (Fukuda et al., 2001). It was found that the experimental ESR spectra could only be simulated if the presence of two different environments of the doxyl group was considered, one in vesicle bilayers and one in non-vesicular (micellar) aggregates (Fukuda et al., 2001). In order to clarify whether micelles and vesicles coexist in the fatty acid system composed of shorter alkyl chains, ESR spectroscopy with different lipophilic spin labels was applied to the decanoic acid/decanoate system. Two different environments of the spin labels were determined and the presence of micelles and vesicles in
∗ Corresponding author. Tel.: +385 1 4561136; fax: +385 1 4680245. E-mail address: [email protected] (V. Noethig-Laslo). 0009-3084/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2010.11.002
equilibrium was confirmed in the interval 6.5 < pH < 7.8 (Dejanovic´ et al., 2008). The physicochemical properties of the surface of oleic acid/oleate vesicles or oleate micelles were not yet studied with ESR spectroscopy. We now report on such investigation which fully supports the proposed formation of a RCOO− ↔ HOOCR hydrogen bond network in the presence of both the ionized and the neutral form of the fatty acid (Apel et al., 2002). In order to obtain more information on the role of hydrogen bond networks for the transformation of oleic acid/sodium oleate micelles to vesicles, the TEMPO-stearate and TEMPOstearamide spin labels are used (Fig. 1). The nitroxide moiety of these lipophilic spin labels is located at the surface of oleate micelles and oleic acid/oleate vesicles covered with bound water molecules, while the long saturated hydrocarbon chains are located in the hydrophobic part of the aggregates. Therefore, these spin labels are expected to reflect properties of the Stern layer, and report on the hydrogen bond networks at the surface of vesicles and micelles. In addition, the water soluble positively charged spin label TEMPO-choline (ASL) (Fig. 1), is expected to bind to the surface, depending on the negative surface charge, which in turn was expected to correlate with the micelle to vesicle transformation. The dynamics of the spin labels incorporated into micelles and vesicles was studied as a function of pH, temperature and sodium oleate concentrations. The formation of vesicles was confirmed by determining the presence of a captured aqueous
84
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
amounts of 1 M HCl were then added to get final concentrations of oleic acid + sodium oleate of 0.025 M or 0.050 M, with the pH values of different samples varying between 12.5 and 7.5.
2.3. Preparation of samples containing spin labels
Fig. 1. Structural formulae of the spin labels used. (A) (2,2,6,6-Tetramethylpiperidine-1-oxyl-4-yl)octadecanoate, TEMPO-stearate; (B) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanamide, TEMPO-stearamide; (C) 1-oxyl-2,2,6,6tetramethyl-4-(2 ,3 ,4 ,5 ,6 -pentahydroxyhexanoyl-1 -amino)-piperidine, GluSL); (D) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-N,N-dimethylaminoethanol, ASL).
Acetone solutions of TEMPO-spin labels were added to the sodium oleate solution at pH 12.5 (molar ratio of the spin label to oleic acid/sodium oleate was 1:300). The solution was optically clear. The solution was vigorously hand shaken, and was left over night at room temperature. The micellar solution was titrated with 1 M HCl to a chosen pH value. The test tubes (1.5 mL of Na-oleate solution) were vortexed for 15 min and equilibrated at 25 ◦ C for 12 h before measurements. In another procedure acetone solutions of the TEMPO-spin labels were added to oleate solutions already titrated to different pH values. Both procedures gave similar result, i.e. same ESR spectra were obtained. Labeling with water soluble spin labels was performed by dissolving sodium oleate in 1 mM water solution of ASL or of GluSL. Further procedure was the same as for preparation of aqueous oleic acid systems.
2.4. CW-ESR measurements volume using a hydrophilic, spin-labeled glucose derivative, GluSL (Fig. 1). TEMPO-stearate was used recently in the study of interfacial ´ properties of phospholipid vesicles (liposomes) (Alves and Peric, 2006; Mirosavljevic´ and Noethig-Laslo, 2008; Alves et al., 2008). It was found that a negative charge in the liposome composition influences the amount of water bound at the surface, as well as the dynamics of the nitroxide moiety located at the liposome surface. Although the structural properties of the confined water in the phospholipid interface are different from that in the case of the fatty acid system (Disalvo et al., 2008), it was of interest to compare the interfacial properties of the surface of oleic acid/oleate vesicles and oleate micelles, with the interfacial properties of phospholipid vesicles, based on studies with TEMPO-spin labels. 2. Experimental
CW-ESR spectra were recorded with a X-band (9.5 GHz) Varian E 12 spectrometer equipped with a Bruker variable temperature control unit, and with the digital acquisition (EW-ESR WARE) (Moorse, 1987). Glass capillaries containing the sample were inserted into the standard 4 mm diameter quartz tubes and centered in a TE102 ESR cavity. Mn2+ in MgO and TEMPO in ethanol were used as external standards. ESR spectra of the spin-labeled oleic acid/sodium oleate system with lipophilic spin labels, characterized by anisotropic motions were simulated with the program “Chili” (Stoll and Schweiger, 2006). The spectra of ASL (the water soluble spin label) characterized by fast isotropic motion were simulated with “Garlic” of the EasySpin 3.1.1. software package (Stoll and Schweiger, 2006; McCarney et al., 2008). Best fit of the spectra was obtained by the least-square fitting method. The error 5% of the calculated parameters obtained from the simulated spectra was determined from the chi square values.
2.1. Materials Sodium oleate (99%) was purchased from Sigma–Aldrich (St. Louis, USA); sodium hydroxide (p.a.) and hydrogen chloride (p.a.) were from Kemika (Zagreb, Croatia). Aqueous samples were prepared in water of Milipore purity. The following spin labels were synthesized by Slavko Peˇcar, (Faculty of Pharmacy, University of Ljubljana, Ljubljana, Slovenia) (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanoate, TEMPO-stearate; (2,2,6,6-tetramethylpiperidine-1-oxyl-4-yl)octadecanamide, TEMPO-stearamide; (2,2,6,6-tetramethylpiperidine-1-oxyl-4-N,N-dimethylaminoethanol iodide) TEMPO-choline (ASL); and 1-oxyl-2,2,6,6tetramethyl-4-(2 ,3 ,4 ,5 ,6 -pentahydroxyhexanoyl-1 -amino)-piperidine, spin-labeled glucose (GluSL) (Babiˇc et al., 2003; Peˇcar et al., 1984). 2.2. Preparation of aqueous oleic acid samples Aqueous oleic acid/oleate samples in the absence of buffer salts were prepared by first dissolving sodium oleate in water (of Milipore purity). A small portion of 1 M NaOH was then added to this stock solution to obtain a pH of about 12.5. Appropriate
2.4.1. CW-ESR measurements of captured volume with GluSL The captured volume (determination of the presence of vesicles) (Peˇcar et al., 1984) was measured with the X-band ESR spectrometer (Bruker ELEXSYS E500) at room temperature. For this purpose, 30 L of 30 mM sodium ascorbate (Asc) was added to 30 L of 50 mM oleic acid samples with entrapped GluSL (1 mM). GluSL is a large molecule and cannot penetrate the vesicle membrane. Asc rapidly reduces the paramagnetic nitroxide group of the accessible GluSL molecules to the corresponding hydroxylamine, which does not give an ESR signal. It was supposed that due to its negative charge at neutral or alkaline pH, Asc – like the spin label GluSL – cannot penetrate intact membranes of the fatty acid vesicles. Therefore, only the non-entrapped GluSL molecules were reduced within the first few minutes after addition of Asc to the sodium oleate vesicles. The captured volume, Vi , was then obtained by measuring the intensity of the ESR spectra before and after addition of Asc. The captured volume (Vi ) is related to the total sample volume (V0 ) according to Vi /V0 = I/I0 , with I and I0 being the intensity of the ESR spectra immediately after Asc addition and before Asc addition, respectively.
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Fig. 2. ESR spectra of TEMPO-stearate taken at 300 K in oleate micelles at pH 11.5 (A) and in oleic acid/oleate vesicles at pH 8.66 (B). Narrow lines are experimental spectra and thick lines the spectral fit using the following parameters. (A) The spectra of both spin labels could be simulated supposing only one type of motion of the nitroxide moiety, that depended only on pH and temperature. (A) R⊥ = 3.6 × 108 s−1 , R =17.0 × 108 s−1 , ◦ = 0◦ ; (B) R⊥ = 2.3 × 108 s−1 , R = 17.0 × 108 s−1 , ◦ = 10◦ .
3. Results and discussion In order to gain insight into the surface properties of the sodium oleate micelles and oleic acid/sodium oleate vesicles, motional properties of TEMPO-stearate and TEMPO-stearamide incorporated into the micelles and vesicles were studied as a function of pH, temperature and total oleic acid + oleate concentration. At [oleic acid + oleate] = 25 mM and above pH 10, both lipophilic TEMPO-spin labels are fully incorporated into the micellar structures, because the uptake of free oleate by the micelles is significantly faster (with a correlation time <10−9 s) as compared to the release of oleate from the micelles into the bulk solution (with a correlation time >10−4 s) (Fukuda et al., 2001). Fast surfactant diffusion from water into micelles implies that the water exchange into and out of the micelles is even faster, with lipid to water association lifetimes of the order of <100 ps (McCarney et al., 2008). Formation of hydrogen bond networks in the course of the micelle to vesicle transformation on going from a high pH system to an intermediate pH will probably influence the dynamics of the TEMPO-spin labels located at the surface of the micelles or vesicles. With respect to this, it is expected that the TEMPO-stearate and TEMPO-stearamide spin labels (structural formulae are presented in Fig. 1), behave differently. The ester bond of TEMPO-stearate can form hydrogen bonds via an acceptor mechanism (>C O) while the amide group of TEMPO-stearamide may form both donor (>N–H) and acceptor (>C O) hydrogen bonds. Thus, a different dynamic behavior of the two TEMPO-spin labels is expected in the transformation of oleate micelles to oleic acid/oleate vesicles. ESR spectra of TEMPO-stearate and TEMPO-stearamide in oleate micelles (spectra A) and in oleic acid/oleate vesicles (spectra B) are shown in Figs. 2 and 3, respectively. The narrow lines are experimental data, and the thick lines the best fit of the spectra obtained with the program “Chili”, a part of the EasySpin 3.1.1.software package (Stoll and Schweiger, 2006). Spectra of both spin labels could be simulated supposing only one type of motion of the nitroxide moiety, that depended only on pH and temperature. From the simulation of the experimental spectra, the dynamic parameters of the two TEMPO-spin labels were obtained, i.e. the principal values (R , R⊥ ) of the rotational diffusion tensor, R, of the axial symmetry.
85
Fig. 3. ESR spectra of TEMPO-stearamide taken at 300 K in oleate micelles at pH 11.5 (A) and in oleic acid/oleate vesicles at pH 8.66 (B). Narrow lines are experimental spectra and thick lines the spectral fit using the following parameters. (A) R⊥ = 4.2 × 108 s−1 , R = 17.0 × 108 s−1 , ◦ = 10; (B) R⊥ = 1.57 × 108 s−1 , R = 16.0 × 108 s−1 , ◦ = 38◦ .
They represent the rotational rates of the nitroxide radical moiety around the axes parallel and perpendicular to the main symmetry axis of the radical (Ge and Freed, 1998). TEMPO-spin labels are “xordering” nitroxides (Van et al., 1974). The relatively free internal rotation of the nitroxide group about the C–O bond is determined by the rotational diffusion coefficient for an axial molecular motion, R , while the motion about the axis perpendicular to the C–O bond is determined by the off-axial rotational diffusion coefficients, R⊥ , (a rocking motion) (Ge and Freed, 1998, 2003). In our simulations, the Euler angles ˛ = 0◦ and ˇ = 90◦ were kept constant while the -angle describing deviations of the nitroxide moiety from the axial molecular motion was changed in order to get a satisfying fit of the spectra. Since the measurements of the principal components of the g and A tensors, neither for TEMPO-stearate nor for TEMPO-stearamide are not known, for simulations of the spectra, the starting values for the g and A tensors were taken from the works of Alves and Peric´ (2006). Table 1 lists the parameters required for the best fit of the experimental spectra (i.e. gxx , component of the g-tensor, Azz , the component of the hyperfine splitting tensor A, as well as the components of the rotational diffusion tensor R (R , R⊥ )) extracted from the optimal simulation for TEMPO-stearate in the oleic acid/oleate system as a function of pH. Table 2 lists the parameters for TEMPOstearamide. For both spin labels, a clear decrease of the off-axial rotational diffusion coefficients, R⊥ , was found on decreasing pH Table 1 pH dependence of the ESR parameters of TEMPO-stearate in 25 mM oleic acid/sodium oleate in 0.1 M NaCl, temperature 300 K. pH
10−8 R⊥ (s−1 )
10−8 R|| (s−1 )
gxx
a0 (Azz ) (mT)
(◦ )
11.52 10.70 10.30 10.00 9.85 9.80 8.75 8.66 8.30 7.34
3.60 3.60 3.60 3.45 2.90 2.80 2.30 2.30 2.00 1.90
17.0 17.0 17.0 17.0 18.0 18.0 17.0 17.0 17.0 17.0
2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.01005
1.65 (3.77) 1.65 (3.77) 1.65 (3.77) 1.65 (3.77) 1.64 (3.75) 1.64 (3.75) 1.64 (3.70) 1.64 (3.70) 1.65 (3.70) 1.65 (3.70)
0 0 0 0 10 10 10 10 10 10
86
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Table 2 pH dependence of the ESR parameters of TEMPO-stearamide in 25 mM oleic acid/sodium oleate in 0.1 M NaCl, temperature 300 K. pH
10−8 R⊥ (s−1 )
10−8 R|| (s−1 )
gxx
a0 (Azz ) (mT)
(◦ )
11.38 10.90 10.15 9.93 9.60 9.24 8.74 8.36 8.08 7.53
4.20 4.20 4.15 3.20 1.75 1.65 1.55 1.57 0.75 0.65
17.0 17.0 17.0 17.0 17.0 17.0 16.0 16.0 13.0 13.0
2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.0100 2.01005
1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.66 (3.77) 1.67 (3.77) 1.67 (3.77) 1.67 (3.77) 1.66 (3.70) 1.66 (3.45)
10 10 10 10 25 28 35 40 38 38
value in the pH range 9.8 > pH > 8.0, while the rate of axial molecular motion, R , showed only a small decrease in the same pH region. In order to get the best fit of the experimental spectra recorded below pH 9.8 an increase of the -angle was from 0◦ to 10◦ was needed for TEMPO-stearate (Table 1), while for TEMPO-stearamide from 10◦ up to 40◦ (Table 2). Larger -angle in the case of the TEMPOstearamide is due to the fact that this spin label is both hydrogen bond donor (>N–H) and hydrogen bond acceptor (>C O). Deviation from the axial molecular motion of the nitroxide moiety, concomitant with a decrease of the off-axial rotational diffusion coefficients, R⊥ , determined with both spin labels suggest the formation of hydrogen bond network (RCOO− ↔ HOOCR) at the surface of the oleate system in the course of the micelle to vesicle transformation. In earlier studies, 16-doxylstearic acid was used as the spin probe. In the pH range 9.8 > pH > 8.0 motional properties of 16doxylstearic acid with the doxyl group located in the hydrophobic core of the micelles or vesicles could be understood by assuming two different spin label environments (Fukuda et al., 2001). This suggested the presence of both micelles and vesicles in equilibrium. The spin label 16-doxylstearic acid could distinguish the two types of aggregates. This is clearly different from the case of the two TEMPO-spin labels, TEMPO-stearate and TEMPO-stearamide. The ESR spectra of both TEMPO-spin labels could be fitted well by supposing only one type of the spin label environment. Since the TEMPO-spin labels report the dynamic properties at the surface of the oleic acid/oleate system, spectral fitting with only one type of spin label motion suggests similar surface properties of micelles and vesicles, known to coexist in the pH range 9.8 > pH > 8.0 (Hargreaves and Deamer, 1978; Morigaki and Walde, 2007; Fukuda et al., 2001). When TEMPO-spin labels (TEMPOoctanoate and TEMPO-decanoate) were applied to aqueous decanoic acid/sodium decanoate system, in the range 7.9 > pH > 6.8 for the best fit of the experimental spectra two different spin label environments had to be supposed, suggesting different surface properties of vesicles and micelles in that system (Dejanovic´ et al., 2008). Similar surface properties of micelles and vesicles in the oleic acid/oleate system in the pH range 9.8 > pH > 8.0 are probably due to longer and unsaturated aliphatic chains, as compared to the decanoic acid/decanote system. Temperature dependence of R and R⊥ extracted from the optimal simulation of the experimental spectra of TEMPO-stearamide at pH 9.9 (oleate micelles, squares) and at pH 8.3 (oleic acid/oleate vesicles, circles) is shown in Fig. 4. Linear temperature dependence of the values for R⊥ and R for both micelles and vesicles indicates that there are no temperature dependent structural changes in the surface of both micelles and vesicles. Smaller values for R⊥ at pH 8.3 (i.e. slower rate of motion, and larger -angle), where the number of vesicles prevail from that for R⊥ at pH 9.9 at all the temperatures examined, suggest stabile hydrogen bond networks (RCOO− ↔ HOOCR) in the pH region of coexistence of vesicles and micelles.
Fig. 4. Temperature dependence of the components of the rotational diffusion tensor, R⊥ and R , for TEMPO-stearamide in 25 mM oleic acid + oleate at pH 9.9 (R⊥ open squares; R|| closed squares) and at pH 8.4 (R⊥ open circles, R|| closed circles).
One of the important factors that might play a role in the transformation of the micelles into closed bilayers (vesicles) is the surface charge and the corresponding influence by the counterions. ASL is a positively charged water soluble spin label. It is expected to bind to the surface depending on the surface charge, and thus its binding property to correlate with the micelle to vesicle transformation. In Fig. 5, ESR spectra of ASL, in 25 mM sodium oleate at pH 11.16 (spectrum A) and at pH 8.47 (spectrum B) are compared. The spectra were fitted with the “Garlic” program (Stoll and Schweiger, 2006), so the motional properties of ASL are described by the rotational correlation time (/s). (The rotational diffusion tensor, R, and the rotational correlation time are related as R = 1/6). Starting values of the parameters required for the simulation of both spectral components were gxx = 2.0082, gyy = 2.0053, gzz = 2.0015;
Fig. 5. ESR spectra of TEMPO-choline (ASL) taken at 294 K in oleate system at: pH 11.2 (A); pH 8.47 (B). Narrow lines are experimental spectra and thick lines the spectral fit with the “Garlic” program. Spectrum A was fitted supposing 85% of spin labels with /s = 5 × 10−10 s; Azz = 3.51 mT, a01 = 1.62 mT, (component 1) and 15% with /s = 0.5 × 10−10 s, Azz = 3.70 mT, a02 = 1.68 mT (component 2); spectrum B, 30% of spin labels with 5 × 10−10 s (component 1) and 70% with 0.5 × 10−10 s (component 2).
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
Fig. 6. pH dependence of the % of component 1 in the spectra, with rotational correlation time, /s = 5 × 10−10 s of TEMPO-choline (ASL) in 0.025 M sodium oleate + oleic acid. The spectra were taken at 294 K.
Axx = Ayy = 0.56 mT, Azz = 3.46 mT (Ge and Freed, 1998). Two different rotational correlation times (/s) of the spin label ASL were required for the best fit of the spectra A and B with a variable number of spin labels in the two spectral components. The number of ASL spin labels characterized with the rotational correlation time = 5.0 × 10−10 s (component 1) decreased from 85% at pH 11.16, to 45% was determined at pH 9.63 while to about 5% at pH 8.2 (Fig. 6). This observation can be explained as follows. Formation of hydrogen bonds at the micelles surface and reduction of the number of negative charges at the surface with decreasing the pH leads to a release of ASL molecules from the oleate surface to a more polar environment of the bulk water. The result is an increase in the number of ASL molecules with faster motion, i.e. with = 0.5 × 10−10 s (component 2). It is important to mention that below pH 9.9 (i.e. in the region in which vesicles formed) the two lipophilic TEMPO-spin labels examined in this work experienced a decrease in the nitroxide motion (R⊥ , a rocking motion) concomitant with an increase of the -angle which suggested formation of the hydrogen bond networks (RCOO− ↔ HOOCR) at the surface of the oleic acid/oleate aggregates. TEMPO-stearate and TEMPO-stearamide were applied recently in a study of interfacial properties of multilamellar liposomes
87
composed of egg phosphatidylcholine (Mirosavljevic´ and NoethigLaslo, 2008). It was found that both components of the rotational diffusion tensor, R(R and R⊥ ) extracted from the optimal simulation of the experimental spectra of both spin labels experienced slower rate of motion at all temperatures examined as compared with the dynamic parameters (R and R⊥ ) of these spin labels in the oleic acid/sodium oleate system. Thus, although oleic acid vesicles are multilamellar and structurally analogous to phospholipid vesicles, they have very different dynamic properties. This might be due to the fact that individual monomers of oleic acid are in rapid exchange between vesicles and micelles. In order to confirm that a decrease in the rate of motion of the two lipophilic TEMPO-spin labels is related to the formation of a closed volume (vesicles), pH dependent measurements of the captured volume were carried out with the spin label GluSL at the total oleic acid + sodium oleate concentration of 0.050 M. A captured volume was detected in the pH range 10 > pH > 7.5, with the highest entrapment yield in the pH range 9.9 > pH > 8.5 (Fig. 7). This is in qualitative agreement with earlier literature data (Hargreaves and Deamer, 1978; Morigaki and Walde, 2007; Fukuda et al., 2001; Walde et al., 1994). Fig. 7 indicates that vesicles are already present at pH 10, a pH region where micelles, however, dominate. A decrease of the captured volume to about 3% (Fig. 7) concomitant with increased amounts of a released ASL molecules from the surface of the oleic acid/oleate system (Fig. 6), followed by a slower motion of the TEMPO-stearamide below pH 8.1 (Table 2) suggests appearance of a second aggregate transition, i.e. namely from the oleic acid/oleate vesicles to oil droplets (Walde and Morigaki, 2007). The oil droplets are mainly composed of oleic acid molecules. 4. Conclusions Changes in the dynamic properties of both lipophilic TEMPOspin labels in oleic acid/sodium oleate system were determined in the region 9.9 > pH > 8.0, i.e. in the course of the micelle to vesicle transformations. Dynamic properties of both lipophilic TEMPO-spin labels suggest formation of hydrogen bond networks (RCOO− ↔ HOOCR) at the surface of the oleate system. Formation of hydrogen bonds at the aggregate surface below pH 10 was followed by measuring the release of positively charged ASL molecules from the aggregate surface to a more polar environment of the bulk water. Acknowledgements This work was supported by Croatian Ministry of Science, Education and Sport (Project No. 098-0982915-2939) and by the European Research Program on the Cooperation in Science and Technology within Europe: COST Chemistry, Action D 27 (Prebiotic Chemistry and Early Evolution). References
Fig. 7. Captured volume as a function of pH determined with GluSL in 0.050 M oleic acid + sodium oleate. The experiments were done at room temperature.
´ M., 2006. An EPR study of the interfacial properties of phosphatidylAlves, M., Peric, choline vesicles with different lipid chain lengths. Biophys. Chem. 122, 66–73. ´ M., 2008. Effect of lysophosphatidylcholine on the surface Alves, M., Bales, B.L., Peric, hydration of phospholipid vesicles. Biochim. Biophys. Acta 1778, 414–422. Apel, C.L., Deamer, D.W., Mautner, M.N., 2002. The formation of glycerol monodecanoate by a dehydration/condensation reaction: increasing the chemical complexity of amphiphiles on early Earth. Biochim. Biophys. Acta 1559, 1–9. Babiˇc, A., Ilaˇs, J., Peˇcar, S., 2003. Synthesis of hydrophilic and amphiphilic spin probes. Pharmazie 58, 599–600. ˇ ´ B., Mirosavljevic, ´ K., Noethig-Laslo, V., Sentjurc, Dejanovic, M., Peˇcar, S., Walde, P., 2008. An ESR characterization of micelles and vesicles formed in aqueous decanoic acid/sodium decanoate systems using different spin labels. Chem. Phys. Lipids 156, 17–25. Disalvo, E.A., Lairion, F., Martini, F., Tymcyszyn, E., Frias, M., Almaleck, H., Gordillo, G.J., 2008. Structural and chemical properties of hydration and
88
B. Dejanovi´c et al. / Chemistry and Physics of Lipids 164 (2011) 83–88
confined water in membrane interfaces. Biochim. Biophys. Acta 1778, 2655–2670. Fukuda, H., Goto, A., Yoshioka, H., Goto, R., Morigaki, K., Walde, P., 2001. Electron spin resonance study of the pH-induced transformation of micelles to vesicles in an aqueous oleic acid/oleate systems. Langmuir 17, 4223–4231. Ge, M., Freed, J.H., 1998. Polarity profile in oriented and dispersed phosphatidylcholine bilayers are different: an electron resonance spin study. Biophys. J. 74, 910–917. Ge, M., Freed, J.H., 2003. Hydration, structure and molecular interactions in the headgoup region in dioleoylphosphatidylcholine bilayers: an Electron Spin Resonance study. Biophys. J. 85, 4023–4040. Gebicki, J.M., Hicks, M., 1973. Ufasomes are stable particles surrounded by unsaturated fatty acid membranes. Nature 243, 232–234. Hargreaves, W.R., Deamer, D.W., 1978. Liposomes from ionic, single-chain amphiphiles. Biochemistry 17, 3759–3768. McCarney, E.R., Armstrong, B.D., Kausik, R., Han, S., 2008. Dynamic nuclear polarization enhanced nuclear magnetic resonance and electron spin resonance studies
of hydration and local water dynamics in micelle and vesicle assemblies. Langmuir 24, 10062–10072. ´ K., Noethig-Laslo, V., 2008. Spin labeling study of interfacial properties Mirosavljevic, of egg-phosphatidylcholine liposome as a function of cholesterol concentration. Chem. Phys. Lipids 155, 74–79. Morigaki, K., Walde, P., 2007. Fatty acid vesicles. Curr. Opin. Colloid Interface Sci. 12, 75–80. Moorse, P.D., 1987. EW-ESR ware. Biophys. J. 51, 440a. Peˇcar, S., Sorg, B., Schara, M., Hecker, E., 1984. Spin-labeled phorbol esters and their interaction with cellular membranes. I. Synthesis of spin-labeled phorbol. 12,13Diesters and related compounds. Chem. Phys. Lipids 35, 151–159. Stoll, S., Schweiger, A., 2006. EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 178, 42–55. Van, S.P., Birell, G.B., Griffith, G.B., 1974. Rapid anisotropic motion of spin labels. Models for motion averaging of ESR parameters. J. Magn. Reson. 15, 444–459. Walde, P., Wick, R., Fresta, M., Mangone, A., Luisi, P.L., 1994. Autopoietic selfreproduction of fatty acid vesicles. J. Am. Chem. Soc. 116, 11649–11654.