Langmuir 2004, 20, 5801-5808
5801
Electron Paramagnetic Resonance Studies of Magnetically Aligned Phospholipid Bilayers Utilizing a Phospholipid Spin Label Paresh C. Dave, Johnson J. Inbaraj, and Gary A. Lorigan* Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056 Received December 16, 2003. In Final Form: April 6, 2004 X-band electron paramagnetic resonance (EPR) spectroscopy was used to study the structural and dynamic properties of magnetically aligned phospholipid bilayers utilizing a variety of phosphocholine spin labels (PCSL) as a function of temperature. 1-Palmitoyl-2-[n-(4,4-dimethyloxazolidine-N-oxyl)stearoyl]sn-glycero-3-phosphocholine (n-PCSL) in which a nitroxide group was attached to the different acyl chain positions of the phospholipid (n ) 5, 7, 12, and 14) were used as an EPR spin probe to investigate magnetically aligned phospholipid bilayers from the plateau (near to the headgroup) region to the end of the acyl chain (center of the bilayers). The addition of certain types of paramagnetic lanthanide ions changes the overall magnetic susceptibility anisotropy tensor of the bicelles, such that the bicelles flip with their bilayer normal either parallel or perpendicular to the magnetic field. The present study reveals for the first time that, in the case of the n-PCSL, the bilayer normal is aligned parallel and perpendicular to the magnetic field in the presence of lanthanide ions having positive ∆χ (e.g., Tm3+) and negative ∆χ (e.g., Dy3+), respectively. The magnetic alignment of the bilayers and the corresponding segmental molecular order parameter, Smol, were investigated as a function of the temperature. The Smol values decrease in the following order, 5-PCSL > 7-PCSL > 12-PCSL > 14-PCSL, for the magnetically aligned phospholipid bilayers. Also, the variable temperature study indicates that, by increasing the temperature, the order parameters Smol decreased for all the n-PCSLs. The results indicate that magnetically aligned phospholipid bilayers represent an excellent model membrane system for X-band EPR studies.
Introduction Nuclear magnetic resonance (NMR) spectroscopy, electron paramagnetic resonance (EPR) spectroscopy, and Fourier transform infrared are all biophysical techniques that can be used to study the characteristics of biological membranes.1-9 Among these techniques, EPR spectroscopy has proven to be very useful in providing insight into the dynamic and molecular structure of phospholipid membranes by utilizing appropriate spin labels.10-13 The motions of phospholipid membranes can be extensively investigated by means of careful analysis of the corresponding EPR line shape. Spin-label EPR spectroscopy has become a standard technique for studying not only lipid dynamics but also lipid-protein interactions in biological membranes.14-17 * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Borbat, P. P.; Costa-Filho, A. J.; Earle, K. A.; Moscicki, J. K.; Freed, J. H. Science 2001, 291, 266-269. (2) Degtyarev, Y. N.; Schlick, S. Langmuir 1999, 15, 5040-5047. (3) Goormaghtigh, E.; Raussems, V.; Ruysschaert, J. M. Biochim. Biophys. Acta 1999, 1422, 105-185. (4) Huster, D.; Dietrich, U.; Gutberlet, T.; Garwisch, K.; Arnold, K. Langmuir 2000, 16, 9225-9232. (5) Martin, I.; Goormaghtigh, E.; Ruysschaert, J. M. Biochim. Biophys. Acta 2003, 1614, 97-103. (6) Ouimet, J.; Croft, S.; Pare, C.; Katasaras, J.; Lafleur, M. Langmuir 2003, 19, 1089-1097. (7) Varkonyi, Z.; Masamoto, K.; Debreczeny, M.; Zsiros, O.; Ughy, B.; Gombos, Z.; Domonkos, I.; Farkas, T.; Wada, H.; Szalontai, B. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2410-2415. (8) Wolfangel, P.; Muller, K. J. Phys. Chem. B 2003, 107, 99189928. (9) Aussenac, F.; Laguerre, M.; Schmitter, J. M.; Dufourc, E. J. Langmuir 2003, 19, 10468-10479. (10) Freed, J. H. Annu. Rev. Phys. Chem. 2000, 51, 655-689. (11) Hubbell, W. L.; McConnell, H. M. J. Am. Chem. Soc. 1971, 93, 314-326. (12) Freed, J. H. In Spin labeling: Theory and applications; Berliner, L. J., Ed.; Academic Press: New York, 1976; pp 53-132. (13) Seelig, J. In Spin labeling: Theory and applications; Berliner, L. J., Ed.; Academic Press: New York, 1976; p 373-409.
The EPR spectrum of the nitroxide (NO) moiety is sensitive to spatial and temporal restrictions of analogue motion. Stearic acid spin labels (n-SASLs) are useful spin probes for testing the model and biological membranes.18,19 In model membranes, doxyl stearates (doxyl group attached to the stearic acid) tend to partition into lipid domains that are more fluid.20,21 However, as a result of the hydrophobic carboxyl group, doxyl stearates can also bind to proteins and tend to concentrate in the vicinity of the membrane proteins. In the last several years, a model has been developed that is suitable to describe the kinetics of the conformational transitions in flexible alkyl chains, and recently it has been adapted for a nitroxide spin label, linked by alkyl chains to a larger molecule (e.g., polypeptide) in isotropic solution.22-24 The structural and dynamic properties of the unoriented and oriented bilayer membranes have been studied incorporating various types of phospholipid spin labels. Nitroxide spin labels attached to phospholipids have also been used to study the lipid (14) Cornea, R. L.; Jones, L. R.; Autry, J. M.; Thomas, D. D. Biochemistry 1997, 36, 2960-2967. (15) Marsh, D.; Horvath, L. I.; Swamy, M. J.; Mantripragada, S.; Kleinschmidt, J. H. Mol. Membr. Biol. 2002, 19, 247-255. (16) Mihailescu, D.; Horvath, L. I. Eur. Biophys. J. 1999, 1999, 216221. (17) Persson, M.; Harbridge, J. R.; Hammarstrom, P.; Mitri, R.; Martensson, L. G.; Carlsson, U.; Eaton, G. R.; Eaton, S. S. Biophys. J. 2001, 80, 2886-2897. (18) Inoue, T.; Kawamura, H.; Matsuda, M.; Misono, Y.; Suzuki, M. Langmuir 2001, 17, 6915-6922. (19) Tajima, K.; Imai, Y.; Horiuchi, T.; Koshinuma, M.; Nakamura, A. Langmuir 1996, 12, 6651-6658. (20) Kocherginsky, N.; Swartz, H. M. Spin labels, Reactions in biology and chemistry; CRC Press: New York, 1995. (21) Vishnyakova, E. A.; Ruuge, A. E.; Golovina, E. A.; Hoekstra, F. A.; Tikhonov, A. N. Biochim. Biophys. Acta 2000, 1467, 380-394. (22) Ferrarini, A.; Moro, G. J.; Nordio, P. L. Mol. Phys. 1988, 63, 225-247. (23) Ferrarini, A.; Nordio, P. L.; Moro, G. J.; Crepeau, R. H.; Freed, J. H. J. Chem. Phys. 1989, 91, 5707-5721. (24) Cassol, R.; Ferrarini, A.; Nordio, P. L. J. Phys. Chem. 1993, 97, 2933-2940.
10.1021/la036377a CCC: $27.50 © 2004 American Chemical Society Published on Web 06/08/2004
5802
Langmuir, Vol. 20, No. 14, 2004
membrane expansion and micelle formation for polymergrafted lipid membrane systems.25,26 In the present work, this technique is extended and applied to magnetically oriented bilayer membranes. The results are compared to previous studies in the liquid crystalline phase of phospholipid bilayers utilizing 13C or 2H NMR at various carbon positions along the acyl chains of long-chain or shortchain phospholipids.23,27 Magnetically aligned phospholipid bilayers or bicelles have recently emerged as a model membrane system useful for a variety of NMR and EPR spectroscopic studies.28-34 Bicelles are formed upon mixing long-chain phospholipids, such as 1,2-dimyristoyl-sn-glyecrol phosphatidylcholine (DMPC), with short-chain phospholipids, such as 1,2-dihexanoyl-sn-glyecrol phosphatidylcholine (DHPC).35-37 A particularly useful characteristic of the magnetically aligned phospholipid bilayers is that the bilayer normal spontaneously aligns perpendicular to the static magnetic field (greater than 7.0 T) for NMR studies. However, they do not magnetically align in the comparably weak magnetic fields (approximately 0.64 T) used in EPR studies at the X-band without the addition of paramagnetic lanthanide ions.31,38-42 Recently, our group reported for the first time the magnetic alignment of phospholipid bilayers using an X-band EPR spectrometer.43,44 In this paper, we extend upon our previous work and describe for the first time the magnetic alignment of lipid bilayers at lower magnetic fields (0.64 T) utilizing a phospholipid spin label. The bilayers are aligned magnetically because of the overall magnetic susceptibility anisotropy tensor (∆χ) of the system. In the absence of lanthanide cations, the phospholipid bilayer has a negative magnetic susceptibility anisotropy tensor. Therefore, in the absence of lanthanide ions, the preferred alignment of the bilayer normal of the bicelle is perpendicular to the magnetic field. The addition of paramagnetic lanthanide cations with a negative magnetic susceptibility anisotropy tensor (∆χ), for example, in the present study Dy3+, increases the overall negative magnetic susceptibility anisotropy of the bicelle samples and helps to achieve (25) Belsito, S.; Bartucci, R.; Sportelli, L. Biophys. Chem. 2001, 93, 11-22. (26) Bartucci, R.; Belsito, S.; Sportelli, L. Chem. Phys. Lipids 2003, 124, 111-122. (27) Cassol, R.; Ge, M. T.; Ferrarini, A.; Freed, J. H. J. Phys. Chem. B 1997, 101, 8782-8789. (28) Whiles, J. A.; Glover, K. J.; Vold, R. R.; Komives, E. A. J. Magn. Reson. 2002, 158, 149-156. (29) Tiburu, E. K.; Moton, D. M.; Lorigan, G. A. Biochim. Biophys. Acta 2001, 1512, 206-214. (30) Mangels, M. L.; Cardon, T. B.; Harper, A. C.; Howard, K. P.; Lorigan, G. A. J. Am. Chem. Soc. 2000, 122, 7052-7058. (31) Mangels, M. L.; Harper, A. C.; Smirnov, A. I.; Howard, K. P.; Lorigan, G. A. J. Magn. Reson. 2001, 151, 253-259. (32) Marcotte, I.; Dufourc, E. J.; Ouellet, M.; Auger, M. Biophys. J. 2003, 85, 328-339. (33) Sternin, E.; Nizza, D.; Garwisch, K. Langmuir 2001, 17, 26102616. (34) Raffard, G.; Steinbruckner, S.; Arnold, A.; Davis, J. H.; Dufourc, E. J. Langmuir 2000, 16, 7655-7662. (35) Sanders, C. R.; Schwonek, J. P. Biochemistry 1992, 31, 88988905. (36) Sanders, C. R.; Hare, B. J.; Howard, K. P.; Prestegard, J. H. Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26, 421-444. (37) Rowe, B. A.; Neal, S. L. Langmuir 2003, 19, 2039-2048. (38) Tjandra, N.; Bax, A. Science 1997, 278, 1111-1114. (39) Sanders, C. R.; Landis, G. C. Biochemistry 1994, 34, 40304040. (40) Struppe, J.; Vold, R. R. J. Magn. Reson. 1998, 135, 541-546. (41) Sanders, C. R.; Prestegard, J. H. Biophys. J. 1990, 58, 447-460. (42) Prosser, R. S.; Shiyanovskaya, I. V. Concepts Magn. Reson. 2001, 13, 19-31. (43) Garber, S. M.; Lorigan, G. A.; Howard, K. P. J. Am. Chem. Soc. 1999, 121, 3240-3241. (44) Cardon, T. B.; Tiburu, E. K.; Padmanabhan, A.; Howard, K. P.; Lorigan, G. A. J. Am. Chem. Soc. 2001, 123, 2913-2914.
Dave et al.
perpendicular orientation (bilayer normal) at the low magnetic fields used in EPR studies. Conversely, the addition of paramagnetic lanthanide cations with a large positive ∆χ, e.g., in the present paper Tm3+, can cause the bicelles to flip 90° such that the average bilayer normal is collinear with the direction of the magnetic field.42,45,46 Also, the segmental order parameter (Smol) profiles were obtained by inserting the EPR spin-labeled probes phosphocholine spin labels (n-PCSL), with the doxyl group attached to the different positions of the acyl chain of the phospholipids incorporated into magnetically aligned phospholipid bilayers. The order parameter Smol measures the average orientation of the chain segment at the corresponding position of the doxyl group.11,19,47,48 Smol strongly depended on the depth of the labeled position from the polar headgroup of phospholipid bilayers. The order parameter reflects the motion of the long molecular axis of the probes and indicates the extent of microflexibility of the hydrocarbon chains of lipids, and the depth profiles of Smol have been referred to as the flexibility gradient. Materials and Method Materials. DMPC, DHPC, and 1-palmitoyl-2-[n-(4,4-dimethyloxazolidine-N-oxyl)stearoyl]-sn-glycero-3-phosphocholine (n-PCSL) were purchased from Avanti Polar Lipids (Alabaster, AL). All phospholipids were dissolved in chloroform and stored at 253 K prior to use. Thulium(III) chloride hexahydrate (TmCl3‚6H2O), dysprosium(III) chloride hexahydrate (DyCl3‚ 6H2O), and HEPES (N-[2-hydroxyethyl]piperazine-N′-2-ethanesulfonic acid) were obtained from Sigma-Aldrich (St. Louis, MO). Deuterium-depleted water was obtained from Isotec (Miamisburg, OH). Cholesterol powder was purchased from Alfa Aesar (Ward Hill, MA). Aqueous solutions of HEPES buffer and lanthanide ions were prepared on the day of sample preparation and were adjusted to pH 7. All aqueous solutions were prepared with deuterium-depleted water. The standard bicelle samples, consisting of 25% (w/w) phospholipid-solution with a q ratio of 3.5 (DMPC/DHPC mole ratio), were made in 15-mL or 25-mL pear-shaped flasks. In the flask, DMPC, 1,2-dimyristoyl-sn-glycerolphosphatidylethanolamineN-[methoxy(poly(ethylene glycol))-2000] (PEG2000-DMPE), DHPC, cholesterol, and the phospholipid spin label were mixed in mole ratios of 3.5:0.035:1:0.35:0.0196, respectively.30,49 The chloroform in the flask was rotovaped off using a Buchi R-3000 rotovap at room temperature (approximately 20 min), and then the flask was placed under a high vacuum overnight. The following day, an appropriate amount of 100 mM HEPES buffer at pH 7.0 was added to the flask. The samples were chilled in an ice bath and vortexed until all of the lipids were solubilized. The samples were chilled frequently in an ice bath to maintain fluidity. Next, the samples were sonicated with a Fisher Scientific FS30 bath sonicator (Florence, KY) for 30 min with the heater turned off and ice added to the bath. The samples were subjected to two or three freeze (77 K)/thaw (room temperature) cycles to homogenize the sample and remove any air bubbles. Finally, at 0 °C (ice bucket), an appropriate aliquot (20 mol % of lanthanide ions with respect to DMPC) of a concentrated aqueous solution of TmCl3 or DyCl3 was added and mixed into the sample. Typically, the total mass of the prepared samples was 200 mg. The bicelle samples were drawn into 1 mm i.d. capillary tubes (Kimax) via a syringe. Both ends of the capillary tube were sealed with Critoseal (Fisher Scientific) and placed inside standard quartz EPR tubes (Wilmad, 707-SQ-250M) filled with light mineral oil. (45) Caporini, M. A.; Padmanabhan, A.; Cardon, T. B.; Lorigan, G. A. Biochim. Biophys. Acta 2003, 1612, 52-58. (46) Prosser, R. S.; Hunt, A.; DiNatale, J. A.; Vold, R. R. J. Am. Chem. Soc. 1996, 118, 269-270. (47) Seelig, J.; Niederberger, W. Biochemistry 1974, 13, 1583-1588. (48) Seelig, J. Q. Rev. Biophys. 1977, 10, 353-418. (49) Cardon, T. B.; Tiburu, E. K.; Lorigan, G. A. J. Magn. Reson. 2003, 161, 77-90.
EPR Studies of Phospholipid Bilayers
Langmuir, Vol. 20, No. 14, 2004 5803
EPR Spectroscopy. All EPR experiments were carried out on a Bruker EMX X-band EPR spectrometer consisting of an ER 041XG microwave bridge and a TE102 cavity coupled with a BVT 3000 nitrogen gas temperature controller (temperature stability of (0.2 °C). All EPR spectra were gathered with a center field 3350 G (0.3350 T), sweep width of 140 G, microwave frequency of approximately 9.39 GHz, modulation frequency of 100 kHz, and modulation amplitude of 1 G and at a power of 2.0 mW. All oriented samples were aligned by warming the sample from 298 to 318 K at a maximum magnetic field strength of 6400 G (0.64 T) for approximately 20 min.30,44,49 Next, the temperature was reduced to 308 K over a 10-min span. EPR spectra were collected over a temperature range from 308 to 348 K in 5 K increments. The sample equilibrated at the appropriate temperature for 10 min before a spectrum was collected. The EPR spectra were processed on a G4 iMac computer utilizing the Igor Carbon software package (Wavemetrics, Lake Oswego, OR). All of the spectra presented in the figures were overlayed and normalized to the central EPR spectral line. Order Parameter Calculations. The EPR spectra of a nitroxide spin label consists of three lines as a result of a system with S ) 1/2 (unpaired electron) coupled to an I ) 1 (14N) nucleus. The magnetic principal axes have the x-axis along the nitroxide N-O bond, the z-axis is along the 2p π orbital of the nitrogen, and the y-axis is perpendicular to the other two. The order parameter S33 can be determined by measuring the resultant hyperfine splittings of the bicelle-aligned spectra using the following equation:11,48,50,51
S33 ) (A| - A⊥/Azz - Axx)(aN/aN′)
(1)
Figure 1. Chemical structures of n-PCSL (phospholipid spin labels) used in this work. The broken arrow indicates the position of the doxyl labels attached to the acyl chain of the phospholipids. quadratic terms of the spectral line width. The following equations were used:
where aN ) 1/3(Axx + Ayy + Azz) is the isotropic hyperfine splitting constants in a rigid molecular frame and is sensitive to the solvent polarity and while aN′ ) 1/3(A| + 2A⊥) is the solvent polarity correction of the hyperfine splitting. Axx ) Ayy ) 5.0 G and Azz ) 33.0 G were taken from a previously reported analysis in the rigid limit.27,52 It is assumed that the hyperfine principal values do not change upon binding of either Tm3+ or Dy3+ to the phospholipid headgroups of the bicelles. The hyperfine splittings A| and A⊥ were measured between the mI ) +1 and 0 spectral lines from the parallel and perpendicular oriented EPR spectra, respectively. The Smol molecular order parameter corresponding to the long molecular axis can be calculated from the equation:13
where h+1 is the peak-to-peak height of the low field line. If τB ) τC, then this is an indication of isotropic motion of the spin label. In the case of anisotropic motion, τB * τC. The difference between τB and τC can be used to evaluate the degree of anisotropy.53
Smol ) S33[(3 cos2 θ - 1)/2]-1
Results and Discussion
(2)
where θ denotes the angle between the long molecular axis and the corresponding NO axis (z-axis). In our case, the z-axis is always collinear to the long molecular axis (θ ) 0°); thus, Smol ≈ S33. Rotational Correlation Time. In the rapid motional regime, the rotational correlation time can be obtained from the corresponding spectral line widths using the following equation:53
τr ) 6.5 × 10-10∆H(0)[(h0/h-1)1/2 - 1] s
(3)
where h0 and h-1 are the peak-to-peak heights of the central and high-field lines, respectively. ∆H(0) is the peak-to-peak line width of the central line (MI ) 0). The above equation is used to calculate the τr value for 12- and 14-PCSL on the basis of the assumption that 12- and 14-PCSL are in an isotropic medium. To verify the anisotropy of the motion of the 12- and 14-PCSL inside the DMPC/DHPC bicelle, rotational correlation times were calculated using two different formulas based upon linear and (50) Seelig, J. J. Am. Chem. Soc. 1971, 93, 5017-5022. (51) Seelig, J.; Limacher, H.; Bader, P. J. Am. Chem. Soc. 1972, 94, 6364-6371. (52) Ge, M.; Field, K. A.; Aneja, R.; Holowka, D.; Baird, B.; Freed, J. H. Biophys. J. 1999, 77, 925-933. (53) Berliner, L. J. Methods Enzymol. 1978, 49, 466-470. (54) Dufourc, E. J.; Parish, E. J.; Chitrakorn, S.; Smith, I. C. P. Biochemistry 1984, 23, 6062-6071.
τB ) 6.5 × 10-10∆H(0)[(h0/h+1)1/2 - (h0/h-1)1/2] s τC ) 6.5 × 10-10∆H(0)[(h0/h+1)1/2 + (h0/h-1)1/2 - 2] s
(4)
(5)
Figure 1 shows the chemical structures of the phospholipid spin labels, n-PCSL, used in the present study as EPR probes to investigate the structural and dynamic properties of the label inside magnetically aligned bilayers. These molecules are labeled with the paramagnetic doxyl group at different positions along the hydrocarbon chain of n-PCSL (carbon nos. 5, 7, 12, and 14). Thus, spectral parameters of oriented n-PCSL, plotted as a function of position (n) of the nitroxide group, can be used to assay the profiles of ordering and molecular mobility of the acyl chains of the membrane lipids. To probe the bilayer alignment for X-band EPR studies, the standard bicelle samples with a q ratio (mole ratio of DMPC/DHPC) of 3.5 incorporated with n-PCSL, with the doxyl group attached to the different positions, were used. In addition to the inclusion of n-PCSL, there are two other components also included in the present study. The bicelle samples were enriched with 10 mol % cholesterol to increase the local order in the membrane.54-56 Cholesterol restricts the movement of the spin probe, and the effects of macroscopic bilayer alignment are highlighted in the observed orientational-dependent EPR spectra. Also, a (55) Ohvo-Rekila, H.; Ramstedt, B.; Leppimaki, P.; Slotte, J. P. Prog. in Lipid Res. 2002, 41, 66-97. (56) Dave, P. C.; Tiburu, E. K.; Nusair, N. A.; Lorigan, G. A. Solid State Nucl. Magn. Reson. 2003, 24, 137-149.
5804
Langmuir, Vol. 20, No. 14, 2004
Figure 2. Temperature dependence of the normalized EPR spectra of 5-PCSL incorporated into oriented and unoriented phospholipid bilayers. The sample consisted of 25% (w/w) lipid DMPC, PEG2000-PE, DHPC, cholesterol, and phospholipid spin label in molar ratios of 3.5:0.035:1:0.35:0.0196 in 100 mM HEPES buffer, pH 7.0. The dotted line spectra represent the perpendicular alignment of the bilayer normal in the presence of Dy3+, and the dot-dashed line spectra represent the parallel alignment of the bilayer normal in the presence of Tm3+. The solid line spectra represent unaligned DMPC/DHPC bilayers in the absence of lanthanide ions.
small amount of a phospholipid that had a soluble poly(ethylene glycol) polymer tail attached to its headgroup, PEG2000-DMPE, was added to improve the stability of the lanthanide-doped bicelle samples.43,57 The EPR spectra of 5-PCSL incorporated into DMPC/DHPC bicelles are displayed in Figure 2 as a function of the temperature. The dotted line and the dot-dashed line spectra represent the perpendicular and parallel aligned phospholipids bilayers in the presence of lanthanide ions Dy3+ and Tm3+, respectively. The solid line spectra represent the randomly dispersed DMPC/DHPC bilayers in the absence of lanthanide ions. The X-band EPR spectral line shape clearly indicates that the phospholipid bilayers are aligned either parallel or perpendicular to the static magnetic field in the presence of Tm3+ or Dy3+, respectively, at temperatures (57) King, V.; Parker, M.; Howard, K. P. J. Magn. Reson. 2000, 142, 177-182.
Dave et al.
318 K and above. Similar EPR spectral line shape analysis has also been observed previously upon the magnetic alignment of phospholipid bilayers when the 5-SASL and the cholestane spin label are incorporated into DMPC/ DHPC bicelle samples.30,44 Interestingly, by increasing the temperature, the intensity of all three spectral lines are increased. At higher temperatures, the phospholipid spin label incorporated into the phospholipid bilayers undergoes more rapid motions than at lower temperatures.30 EPR spectra of the Tm3+ doped bicelle samples indicate that the hyperfine splittings decreases as the temperature increases. Conversely, the hyperfine splittings increase for the Dy3+ doped bicelle samples as the temperature increases. The presence of paramagnetic lanthanide ions that are required for bicelle alignment could cause paramagnetic line broadening and complicate detailed analysis of spinlabeled EPR spectra. Thus, to explore this possibility, we performed a control experiment with an unoriented DMPC/DHPC bicelle sample of 5- and 7-PCSL both with and without Tm3+ or Dy3+ in the gel (298 K) and liquid crystalline phases (318 K). We did not observe any significant changes in the line width for both 5-PCSL and 7-PCSL with and without Tm3+ or Dy3+ (data not shown). Additionally, our line width data agrees well with similar spectra obtained from n-doxyl stearic acids interacting with lanthanide ions, which showed no significant changes in the spectral line width.57 Thus, the addition of Tm3+ or Dy3+ to DMPC/DHPC bicelles does not significantly alter the EPR spectral line widths. To better understand the molecular dynamics of magnetically aligned phospholipid bilayers in the different hydrophobic acyl chain regions, EPR experiments were carried out using nitroxide spin labels attached to different positions of the acyl chain. Figure 3 shows the EPR spectra of phospholipid spin labels incorporated into the DMPC/ DHPC bicelle samples as a function of the position (n) at 308 K. The dotted line and dot-dashed line EPR spectra represent the perpendicular and parallel oriented bilayers in the presence of Dy3+ and Tm3+ lanthanide ions, respectively. The solid line spectra represent the randomly dispersed bicelle samples in the absence of lanthanide ions. The bicelle sample compositions are the same as discussed previously. Analysis of the EPR spectral line shape of the doxyl group attached to positions 5 and 7 (5-PCSL and 7-PCSL) of the acyl chain of the phospholipid spin label incorporated into the bicelles suggests that it is almost identical and possesses more anisotropic motion. In this case, the 5-PCSL and 7-PCSL doxyl groups are situated closer to the lipid polar headgroup region where the molecular motions are more restricted. In contrast to this, three symmetrical EPR spectral lines corresponding to less anisotropic motions for the doxyl group attached to the 12 and 14 positions (12- and 14-PCSL) were observed. The doxyl group being attached near to the end of the acyl chains makes it free from other side chain restrictions. Also, there is more disorder and motion in the center and at the ends of the acyl chains when compared to the headgroup regions of the phospholipid bilayers. Similar spectral line shapes have also been observed recently when the doxyl group attached to the different positions of an acyl chain of phospholipid spin labels was incorporated into the DMPC bilayers,59 into the PEG2000-DPPE (1,2-dipalmitoyl-sn-glycerophosphatidylethanolamine-N-[methoxy(poly(ethylene glycol))(58) Cheng, Y.; Chen, B. W.; Lu, J.; Wang, K. J. Inorg. Biochem. 1998, 69, 1-7. (59) Livshits, V. A.; Dzikovski, B. G.; Marsh, D. J. Magn. Reson. 2003, 162, 429-442.
EPR Studies of Phospholipid Bilayers
Langmuir, Vol. 20, No. 14, 2004 5805
2000]) micelles,26 and into DPPC (1,2-dipalmitoyl-snglycerolphosphatidylcholine) bilayers.26 Inspection of Figure 3 indicates that, in the case of 5-PCSL and 7-PCSL, a small broad feature (hump) is observed on the top of the first spectral line (mI ) +1) in the perpendicular aligned bicelle sample (dotted line spectra), while one small shoulder is observed between the first and second spectral lines (between mI ) +1 and mI ) 0) in the parallel oriented EPR spectrum (dot-dashed line spectra). The small shoulder between the first and second EPR spectral line (between mI ) +1 and mI ) 0) has also been observed previously for the 5-SASL incorporated into mechanically oriented multibilayer films at room temperature (293296 K).60 Also, Figure 3 indicates that, in the case of the 12-PCSL and the 14-PCSL bicelle in the parallel orientation (dot-dashed line spectra), the third high-field EPR spectral line (for mI ) -1) splits into two lines. Recently, the overlap of the two spectral lines in the high-field position (for mI ) -1) has been reported for 16-PCSL incorporated into 1,2-dimyristoyl-sn-phosphatidylglycerol bilayers between the temperatures ranging from 293 to
296 K.61 The slight splitting has been characterized as an intermediate phase transition from the fluid to the liquid crystalline phase. The presence of similar EPR spectral features in earlier papers for 5-PCSL and 16-PCSL below the liquid crystalline temperature suggests that in the present case bicelle samples at 308 K are in an intermediate transition phase. The spectral position for the broad feature and small line in the case of the parallel oriented bicelle samples at 308 K is approximately the same as the position of Amin (A⊥). These results indicate the presence of small contributions from the perpendicular oriented bicelles with respect to the static magnetic field at 308 K. The bicelle samples used in the present study contain 10 mol % cholesterol, and it has been well-known that cholesterol increases the phase transition temperature of the phospholipids bilayers.55 Recently, we reported that the macroscopic alignment of the Tm3+ doped bicelle system gradually increased as the temperature was raised slowly from 307 to 318 K in the presence of an applied magnetic field.49 This occurs presumably because of the decrease in the viscosity with increasing temperature and an increase in the order of the bicelle system upon going from the nematic phase to the smectic phase.49 At 308 K, the lanthanide doped (either Tm3+ or Dy3+) bicelle system is in a nematic phase and may not be in a perfectly aligned smectic phase, which is reflected in the corresponding EPR spectra of the 5-PCSL and 7-PCSL (Figure 3). Figure 4 represents the EPR spectra of the same DMPC/ DHPC bicelle samples as represented in Figure 3, but these spectra are presented at 348 K to show the effect of temperature on the magnetic alignment of the bilayers. Three sharp EPR spectral lines are observed for the bicelle samples containing lanthanide ions for all the spin labels (5-, 7-, 12-, and 14-PCSL) in either parallel or perpendicular oriented phospholipid bilayers with respect to the magnetic field. The data clearly indicates that at 348 K the bicelles are well aligned in the liquid crystalline phase. The EPR spectra for all the spin labels incorporated into the bicelle samples show very well aligned line shape over the temperature range from 313 to 343 K (spectra not shown). The hyperfine splittings decrease for the parallel oriented bicelle samples and increase for the perpendicular oriented bicelle samples as the doxyl group position moves from the headgroup region to the center of the bilayers. Interestingly, the EPR spectral line shape and hyperfine splittings of 14-PCSL in the presence of lanthanide ions, either Tm3+ or Dy3+, and even in the absence of lanthanide ions are almost identical. This result indicates significant motion and suggests that to study the magnetic alignment of phospholipid bilayers using phospholipid spin labels, it is necessary to use the phospholipid spin label in which the doxyl group is attached to the acyl chain closer to the plateau region (near to the headgroup region); otherwise, the data may be misleading. The relative intensity between the EPR spectral lines reveals pertinent information on the motion of the nitroxide group along the molecular axis. Interestingly, the relative intensity differences between the three EPR lines are less, which indicates that the doxyl group moves freely. Conversely, if the relative intensity difference between the EPR spectral lines is large, this indicates that the doxyl group is experiencing more restricted motions.12 The EPR spectral line shapes in Figure 4 clearly indicate that the relative intensity differences decrease from 5- to 14-PCSL. This further confirms that the nitroxide group attached near to the end of acyl chain moves more freely when compared
(60) Schreier-Muccillo, S.; Marsh, D.; Dugas, H.; Schneider, H.; Smith, I. C. P. Chem. Phys. Lipids 1973, 10, 11-27.
(61) Riske, K. A.; Fernandez, R. M.; Nascimento, O. R.; Bales, B. L.; Teresa Lamy-Freund, M. Chem. Phys. Lipids 2003, 124, 69-80.
Figure 3. Stacked plot of normalized n-PCSL EPR spectra of unoriented and oriented bicelles showing the different positions of the doxyl group attached to the acyl chain at 308 K. The sample composition is the same as in Figure 2. The dotted line spectra represent the perpendicular alignment of the bilayer normal in the presence of Dy3+, and the dot-dashed line spectra represent the parallel alignment of the bilayer normal in the presence of Tm3+. The solid line spectra represent DMPC/DHPC unaligned bilayers in the absence of lanthanide ions.
5806
Langmuir, Vol. 20, No. 14, 2004
Dave et al. Table 1. Rotational Correlation Time for Spin-Labeled Phosphatidylcholines, n-PCSL, in the DMPC/DHPC Bicelle Sample at 308 and 348 K Obtained from the Line Width of the Corresponding EPR Spectra 308 K
Figure 4. Stacked plot of normalized n-PCSL EPR spectra of unoriented and oriented bicelles showing the different positions of the doxyl group attached to the acyl chain at 348 K. The sample composition is the same as in Figure 2. The dotted line spectra represent the perpendicular alignment of the bilayer normal in the presence of Dy3+, and the dot-dashed line spectra represent the parallel alignment of the bilayer normal in the presence of Tm3+. The solid line spectra represent unaligned DMPC/DHPC bilayers in the absence of lanthanide ions.
to the nitroxide group present near the top of the headgroup region. This can be discussed more clearly based upon the difference in the rotational correlation time, which is obtained from the EPR spectral line width of the spin label at different acyl chain positions. In the fast motional regime (τr ≈ 10-11-10-9 s) on the X-band EPR time scale, the spin label motion completely averages out the anisotropy of the g and hyperfine tensors and results in an isotropic spectrum. Axx and Azz (or gxx and gzz) of a PCSL spin-labeled bicelle EPR spectrum will be averaged out if the frequency of rotation about the long molecular axis is fast when compared with (Azz - Axx) ≈ 108 s-1 (at 0.34 T). In the absence of complete Azz - Axx motional averaging, the Azz and Axx 14N hyperfine splittings are resolved in the corresponding EPR spectrum (5-PCSL and 7-PCSL). A recent study has pointed out that the rotational motion of 8-PCSL/DMPC about the membrane normal is approximately 107 s-1.62 In the case of 5-PCSL and 7-PCSL, the doxyl groups are situated closer like 8-PCSL to the lipid polar headgroup region where more restricted molecular motion is expected. Thus, the spectral line shape reveals that the motion about the long molecular (62) Gaffney, J. B.; Marsh, D. Proc. Natl. Acad. Sci. 1998, 95, 1294012943. (63) Ge, M.; Freed, J. H. Biophys. J. 1998, 74, 910-917.
τr spin label (ns)
τB (ns)
12-PCSL 14-PCSL
2.63 1.78 1.71 1.34
2.2 1.5
τC (ns)
348 K τ B - τC 0.85 0.37
τr (ns)
τB (ns)
τC (ns)
0.69 0.71 0.67 0.29 0.30 0.28
τ B - τC 0.04 0.02
axis is sufficiently anisotropic to resolve the inner and outer hyperfine tensoral components. In contrast, the rotational rate varies relatively with the position of the label down the chain, such as for 12-PCSL and 14-PCSL. The calculated rotational correlation time (using eq 3) for 12- and 14-PCSL from EPR spectra of unoriented DMPC/ DHPC bicelles at 308 and 348 K are given in Table 1. The estimated effective correlation time for rotation about the membrane normal, τr, for 12- and 14-PCSL is in the region of 2.2 and 1.5 × 10-9 s, respectively. If the motion is not truly isotropic, then the τr values obtained are not true correlation times. Thus, to evaluate the degree of anisotropy, the effective rotational correlation times, τB and τC (calculated using eqs 4 and 5) and the difference between these two values are presented in Table 1. The small discrepancy in the τ values may arise from the anisotropic motion of the spin label. The τ values are larger for 12PCSL when compared to 14-PCSL at 308 K, indicating that 12-PCSL is undergoing anisotropic motion in addition to that of 14-PCSL. At 348 K where the rotational motion is expected to be even higher than 308 K, the difference between two τ values becomes almost zero (Table 1), indicating that the spectra looks more isotropic in nature. This result clearly suggests systematic changes in the spectral anisotropy of the spin label based upon membrane depth in going from 5-PCSL to 14-PCSL. In the following paragraphs, we discuss these changes in the spectral anisotropy reflected in the observed differences in the hyperfine splitting of parallel and perpendicularly aligned EPR spectra of lanthanide-doped DMPC/DHPC bicelle samples. Figure 5A shows a plot of the hyperfine splittings obtained from the bicelle samples oriented parallel to the magnetic field in the presence of Tm3+ as a function of temperature for different PCSLs. The principle magnetic tensor values were obtained from the analysis of aligned membranes in the rigid limit spectra values obtained from the literature at 138 K; they are gxx, gyy, gzz ) 2.0089, 2.0058, 2.0021 and Axx, Ayy, Azz ) 5, 5, 33 G.27,52 Thus, the isotropic hyperfine splitting (Aiso) is (Axx + Ayy + Azz)/3 ) 14.33 G. The magnetic principal axes have the x-axis along the nitroxide N-O bond, the z-axis is along the 2p π orbital of the nitrogen, and the y-axis is perpendicular to the other two. The dynamics of the n-PCSL is complicated because the rapid segmental motion occurs as a result of a large number of the gauche-trans interconversions of the acyl chain of the phospholipid spin label resulting in a partial averaging of the hyperfine tensors. Clearly, Figure 5A indicates that the hyperfine splittings for all the bicelle samples oriented parallel to the magnetic field are larger than the Aiso (14.33 G), irrespective of the positions of the spin labels to the acyl chains. The molecular axis of the n-PCSL is parallel to the magnetic field when the bilayers are aligned parallel to the magnetic field. Thus, at this orientation, the z-axis is parallel to the magnetic field while the x-axis and y-axis are perpendicular to the magnetic field. Higher values of the hyperfine splittings at lower temperatures indicate that the contribution from the z-tensoral component is higher than
EPR Studies of Phospholipid Bilayers
Figure 5. Temperature dependence of the hyperfine splitting of n-PCSL incorporated into the bicelle samples. (A) Hyperfine splittings measured from the parallel oriented bicelles with respect to the static magnetic field in the presence of Tm3+. (B) Hyperfine splittings measured from the perpendicular oriented bicelles with respect to the static magnetic field in the presence of Dy3+.
the x- and y-tensoral components. These results clearly indicate that the doxyl group rotation around the z-axis is greater when compared to those around the x-axis and y-axis. Also, the hyperfine splittings decrease when the doxyl group position on the acyl chain moves from the top (near to the headgroup region) to the end of the acyl chain of the phospholipid (near to the center of the bilayer). For example, a splitting of 22.6 G was found for 5-PCSL, whereas a splitting of 15.25 G for 14-PCSL was incorporated into the bicelles at 313 K. These results indicate that the acyl chain motion of the phospholipids in the center of the bilayer possess more free motion when compared to the headgroup region, where it experiences more restricted motion.26,56 In other words, the contribution from the z-tensoral component decreases or contributions from the x- and y-tensoral components increase for the 5-PCSL to the 14-PCSL samples.Interestingly, the z-tensoral component of the nitrogen nuclear hyperfine interaction of the nitroxide spin labels is very sensitive to the polarity of the surroundings, which can provide valuable information on the degree of hydration of the immediate environment around the spin labels.52,63 Obviously, our results also indicate that the degree of hydration in bicelles is higher in the polar headgroup region, when
Langmuir, Vol. 20, No. 14, 2004 5807
compared to the hydrophobic acyl chain bilayer central region. Figure 5A also reveals that as the temperature increases the hyperfine splittings decrease for all spin labels (n-PCSL) incorporated into the bicelle samples oriented parallel to the static magnetic field. The data clearly indicates that the mobility of the acyl chains increases as the temperature increases. Figure 5B shows a plot of the hyperfine splittings obtained from DMPC/DHPC bicelle samples aligned perpendicular to the static magnetic field in the presence of Dy3+ as a function of temperature. Figure 5B clearly indicates that the hyperfine splittings for all the bicelle samples oriented perpendicular to the magnetic field are smaller than the corresponding theoretical isotropic hyperfine splitting (Aiso) irrespective to the positions of the spin labels on the acyl chains. When the bilayer normal is aligned perpendicular to the magnetic field, the molecular axis and the z-axis of the phospholipid spin label are perpendicular to the magnetic field. Also, the xy plane of the doxyl group is parallel to the magnetic field. At lower temperatures, the hyperfine splitting for bicelle samples containing 5-PCSL is 11.5 G, which reveals that the contributions from the x- and y-tensoral components are averaged and the contribution from the z-tensoral component is much less. Interestingly, the hyperfine splittings increase as the position of the doxyl group attached to the acyl chains of the phospholipid move close to the center of the bilayer. The hyperfine splittings increase in the order of 5-, 7-, 12-, and 14-PCSL, respectively, while they decrease in the same order when the bilayer normal is aligned parallel to the magnetic field. These results suggest that the contributions from the z-tensoral components increases as the doxyl group attached to the acyl chains moves from the more rigid headgroup region to the more mobile center region of the bilayer. Hubbell and McConnell have also reported the increasing of A| in the order of 4-PCSL, 8-PCSL, and 12PCSL and that A⊥ decreases in the order of 4-PCSL, 8-PCSL, and 12-PCSL when phospholipid spin labels are incorporated into egglecithin-cholesterol samples.11 Our results agree with that work and indicate that magnetically aligned phospholipid bilayers represent an excellent model membrane system for EPR studies. Also, as the temperature increases the hyperfine splittings increase for all the bicelle samples aligned perpendicular to the magnetic field. The data further supports our conclusion that as the temperature increases the mobility of the acyl chains increases and the contribution from z-tensoral component increases. The contribution from all three tensoral (x, y, and z) values is averaged as the temperature increases from 308 to 348 K. At higher temperatures, the n-PCSLs incorporated into either the parallel or the perpendicular orientations undergo more rapid molecular motions than at lower temperatures.30 Thus, all the cases studied have the hyperfine splitting approaching an isotropic value of (Axx + Ayy + Azz)/3, which is approximately equal to 14.33 G. In the case of the parallel oriented phospholipid bilayers, the hyperfine splittings decrease, while the hyperfine splitting for the perpendicular oriented phospholipid bilayers increase as the temperature increases. The molecular order parameter, Smol, describes the local orientational or dynamic perturbations of the doxyl group from its standard state due to perturbations of n-PCSL conformations or dynamics as a result of the incorporation of the phospholipid spin labels into DMPC/DHPC bicelle samples. Figure 6 depicts the molecular orientational order parameter profiles of the n-PCSL embedded into DMPC/ DHPC bicelles derived from the EPR spectra of the bicelle
5808
Langmuir, Vol. 20, No. 14, 2004
Dave et al.
the Smol molecular order parameters decrease for the phospholipid spin labels incorporated into the bicelle samples in the following decreasing order of 5-PCSL > 7-PCSL > 12-PCSL > 14-PCSL. Similar trends have also been observed recently when n-PCSLs were incorporated into PEG2000-DPPE micelles and into DPPC bilayers.26 The data indicate that there is more disorder and motion in the center of the bilayer and at the ends of the acyl chains when compared to the headgroup regions. Also, the order parameter profiles indicate that by increasing the temperature, the value of Smol decreases and the mobility of the acyl chains increase. Finally, the Smol values are fairly close to those of the previous studies, indicating that magnetically aligned phospholipid bilayers are an excellent model membrane system for EPR studies.26,64 Conclusion
Figure 6. Temperature-dependent Smol order parameter profiles for the n-PCSL phospholipid spin labels incorporated into magnetically aligned phospholipid bilayers.
samples oriented parallel (A|) and perpendicular (A⊥) to the magnetic field as a function of temperature. From these hyperfine splitting values, Smol was calculated using eqs 1 and 2 as described in Materials and Method. The molecular order parameter Smol values obtained for 5-PCSL, 7-PCSL, 12-PCSL, and 14-PCSL from unoriented DMPC samples was 0.46, 0.38, 0.20, and 0.06, respectively, at 312 K,64 while the Smol values obtained for the magnetically aligned bilayer sample were 0.37, 0.28, 0.11, and 0.06 for 5-PCSL, 7-PCSL, 12-PCSL, and 14-PCSL, respectively, at 313 K. At 308 K, Smol could not be calculated accurately because the bicelles were not perfectly aligned with respect to the magnetic field. The level of hydration is higher in the present bicelle study when compared to the unoriented DMPC samples. The hydration level and the usage of DHPC for the magnetically aligned bilayer samples most likely explains the slightly lower Smol values of our work when compared to the previous study. Recently, X-band EPR and 2H NMR studies clearly indicated that magnetically aligned bilayers at 313 K was ascribed as the lanthanide-induced smectic LR phase.44 In the smectic phase, bicelles fuse together to form extended, more stable structures. Furthermore, lanthanide ions bind to the two phospholipid headgroups, causing tighter packing, and this effect may cause the order parameters obtained from magnetically aligned bilayers to deviate slightly from previously reported order parameters from unoriented DMPC bilayers. Figure 6 clearly indicates that (64) Dzikovski, B. G.; Livshits, V. A.; Marsh, D. Biophys. J. 2003, 85, 1005-1012.
For the first time, EPR spectral studies have been conducted on n-PCSL phospholipid spin labels inserted into DMPC/DHPC bicelle samples aligned either parallel or perpendicular in a weak magnetic field (0.64 T, X-band EPR) in the presence of lanthanide ions Tm3+ and Dy3+, respectively. The EPR spectral line shape, the hyperfine splittings, and the molecular order parameters obtained from the bicelle data indicate that the motion of the acyl chains increases from the polar headgroup region to the central bilayer region, and these results agree well with previous studies.26,54,56 The hyperfine splittings of the spin labels decreases in the order 5-PCSL > 7-PCSL > 12PCSL > 14-PCSL when the bilayer normal is aligned parallel (A|) to the static magnetic field. Conversely, the hyperfine splittings increase in the order 5-PCSL < 7-PCSL < 12-PCSL < 14-PCSL when the phospholipid bilayers are aligned perpendicular (A⊥) to the magnetic field. To the best of our knowledge, this is the first paper that calculates the corresponding molecular order parameters of phospholipid spin labels incorporated into magnetically aligned bilayer samples utilizing X-band EPR spectroscopy. The Smol order parameter profiles for the n-PCSL phospholipid spin labels suggest that the mobility of the acyl chains increases from the polar rigid headgroup region to the hydrophobic central bilayer region. Also, the variable temperature study indicates that the mobility of the acyl chains increases with increasing temperature. The present study clearly indicates that EPR spectral investigations of bicelle systems can be carried out utilizing phospholipid spin labels at lower magnetic fields. Acknowledgment. This work was supported by a NSF CAREER Award (CHE-0133433) and a National Institutes of Health Grant (GM60259-01). LA036377A
