Bicellar Mixture Phase Behavior Examined by Variable-Pressure

Jun 10, 2010 - Collin Knight , Ashkan Rahmani , and Michael R. Morrow. Langmuir ... Alanna Flynn , Michael Ducey , Anand Yethiraj , and Michael R. Mor...
0 downloads 0 Views 1MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Bicellar Mixture Phase Behavior Examined by Variable-Pressure Deuterium NMR and Ambient Pressure DSC Md. Nasir Uddin and Michael R. Morrow* Department of Physics and Physical Oceanography, Memorial University of Newfoundland, St. John’s, Newfoundland and Labrador, Canada A1B 3X7 Received April 11, 2010. Revised Manuscript Received May 23, 2010 Variable-pressure deuterium nuclear magnetic resonance (2H NMR) has been used to study the pressure-temperature phase diagram of bicellar mixtures containing 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 1,2dihexanoyl-sn-glycero-3-phosphocholine (DHPC). Spectra were obtained for DMPC-d54/DHPC (3:1), DMPC-d54/ DHPC (4.4:1), DMPC/DHPC-d22 (3:1), and DMPC/DHPC-d22 (4.4:1) in the range 10-68 °C at ambient pressure, 66 MPa, 102 MPa, and 135 MPa. Isotropic-to-nematic and nematic-to-lamellar transition temperatures were found to rise with pressure at ∼0.15 and ∼0.14 °C/MPa, respectively, for DMPC-d54/DHPC (3:1) and at at ∼0.19 and ∼0.18 °C/MPa, respectively, for DMPC-d54/DHPC (4.4:1). Pressure had little effect on the range of DMPC-d54 chain orientational order through the nematic phase temperature range, but the behavior of chain orientational order at the nematic-tolamellar transition was found to vary slightly with pressure. Comparison of differential scanning calorimetry (DSC) observations with ambient-pressure 2H NMR observations of DMPC-d54 in the bicellar mixtures suggests that absorption of heat persists for a few degrees above the onset of axially symmetric DMPC-d54 reorientation.

Introduction Depending on temperature and composition, mixtures of a short-chain phospholipid, such as 1,2-dihexanoyl-sn-glycero-3phosphocholine (DHPC), with a longer chain lipid, such as 1,2dimyristoyl-sn-glycero-3-phosphocholine (DMPC), can assemble into structures in which the longer chain lipids preferentially aggregate in planar bilayer regions while the shorter chain lipids aggregate in highly curved regions that close the edges of such planar bilayers.1-5 We will follow a recent suggestion that such mixtures be generally referred to as bicellar mixtures.6 Because of the tendency of some bicellar lipid assemblies to orient in magnetic fields under specific conditions, such systems are of interest as membrane-mimetic platforms in which membraneassociated proteins can be studied by nuclear magnetic resonance (NMR).7,8 Deuterium NMR (2H NMR) observations of chain-perdeuterated DMPC in DMPC-d54/DHPC mixtures typically display spectra characteristic of fast isotropic reorientation at lower temperature, axially symmetric reorientation about spherically distributed bilayer normal orientations at high temperature, and axially symmetric reorientation about an axis oriented perpendicular to the applied magnetic field over a small, intermediate range of temperature.1,4,5,9 Observations of chain-perdeuterated DHPC in corresponding DMPC/DHPC-d22 mixtures suggest that, at intermediate and high temperatures, one fraction of the *To whom correspondence should be addressed. E-mail: mmorrow@ mun.ca. (1) Sanders, C. R.; Schwonek, J. P. Biochemistry 1992, 31, 8898–8905. (2) Struppe, J.; Vold, R. R. J. Magn. Reson. 1998, 135, 541–546. (3) Sanders, C. R.; Prosser, R. S. Structure 1998, 6, 1227–1234. (4) Prosser, R. S.; Hwang, J. S.; Vold, R. R. Biophys. J. 1998, 74, 2405–2418. (5) Sternin, E.; Nizza, D.; Gawrisch, K. Langmuir 2001, 17, 2610–2616. (6) Soong, R.; Nieh, M.-P.; Nicholson, E.; Katsaras, J.; Macdonald, P M. Langmuir 2010, 26, 2630–2638. (7) Ram, P.; Prestegard, J. H. Biochim. Biophys. Acta 1988, 940, 289–294. (8) Marcotte, I.; Auger, M. Concepts Magn. Reson., Part A 2005, 24A, 17–37. (9) Aussenac, F.; Laguerre, M.; Schmitter, J.-M.; Dufourc, E. J. Langmuir 2003, 19, 10468–10479.

12104 DOI: 10.1021/la1014362

short-chain lipid component may undergo isotropic reorientation while other fractions exist in anisotropic environments with small but nonzero orientational order.4,5 It was originally suggested that the observation of preferred bilayer orientation in a magnetic field reflected a bicelle morphology consisting of DMPC-rich bilayer disks with highly curved, DHPC-rich edges.1,3,5 In this scheme, the isotropic phase was pictured as a micelle phase and nonoriented anisotropic phases were identified as bilayer vesicle phases with differing degrees of DMPC/DHPC mixing. It was also suggested that in the presence of lanthanide ions the magnetically alignable phase might be a smectic phase of perforated bilayers.4,10,11 There is now evidence from small-angle neutron scattering for a scheme in which the phases encountered on warming DMPC/ DHPC mixtures, in particular, from low to high temperature, are an isotropic phase, consisting of bicellar particles undergoing rapid thermal reorientation, a magnetically alignable chiral nematic phase consisting of entangled wormlike micelles, and a multilamellar bilayer vesicle phase.12-14 While there is growing consensus regarding the morphologies of the lipid assemblies associated with the isotropic, nematic, and lamellar phases of bicellar mixtures, some interesting questions can be raised regarding the thermal properties of these systems and the way in which the observed morphology changes are driven by interactions among mixture components.15 Differential scanning calorimetry and variable-pressure nuclear magnetic (10) Nieh, M.-P.; Glinka, C. J.; Krueger, S.; Prosser, R. S.; Katsaras, J. Langmuir 2001, 17, 2629–2638. (11) Nieh, M.-P.; Glinka, C. J.; Krueger, S.; Prosser, R. S.; Katsaras, J. Biophys. J. 2002, 82, 2487–2498. (12) Nieh, M.-P.; Raghunathan, V. A.; Glinka, C. J.; Harroun, T. A.; Pabst, G.; Katsaras, J. Langmuir 2004, 20, 7893–7897. (13) Katsaras, J.; Harroun, T. A.; Pencer, J.; Nieh, M.-P. Naturwissenschaften 2005, 92, 355–366. (14) Harroun, T. A.; Koslowsky, M.; Nieh, M.-P.; de Lannoy, C.-F.; Raghunathan, V. A.; Katsaras, J. Langmuir 2005, 21, 5356–5361. (15) Gutberlet, T.; Hoell, A.; Kammel, M.; Frank, J.; Katsaras, J. Appl. Phys. A: Mater. Sci. Process. 2002, 74, S1260–S1261.

Published on Web 06/10/2010

Langmuir 2010, 26(14), 12104–12111

Uddin and Morrow

resonance (NMR) spectroscopy can be used to gain insight into such questions. The effects of hydrostatic pressure on phospholipid bilayers have been investigated using a variety of techniques including neutron scattering, X-ray diffraction, optical transmission, and 2 H NMR.16-23 For multilamellar vesicle morphology, in which the local lipid bilayer organization is planar, increased hydrostatic pressure increases lipid-chain orientational order and raises the main gel-to-liquid crystalline phase transtition by ∼0.2 °C/ MPa.24,25 High pressure can also stabilize phases, such as an interdigitated gel phase, that are not normally observable at ambient pressure.19,20,24,26 In multilamellar bilayer vesicles containing a single lipid species, susceptibility to pressure-induced ordering of the phospholipid chains depends on proximity of the temperature to the ambientpressure main phase transition temperature for bilayers of that lipid.26 In multilamellar vesicles comprising binary lipid mixtures, both species display susceptibilities to pressure-induced ordering that fall between the values of either species individually, presumably reflecting the influence of an average ordering potential within the mixed lipid bilayer.27 In general, the ways in which lipid assemblies respond to pressure are sensitive to interactions in both the head and chain regions of the bilayer.25 Suggested morphologies of DMPC/ DHPC mixtures under some conditions are characterized by regions of high local curvature that are enriched in the shorter chain lipid component. In effect, such structures contain edges which are not present in multilamellar vesicle samples. In order to learn whether such differences from multilamellar vesicle morphology might be reflected in the response of bicellar mixtures to applied hydrostatic pressure, we have used 2H NMR to compare the ordering and mixing of long- and short-chain lipid components in two bicellar mixtures, DMPC/DHPC (3:1) and DMPC/ DHPC (4.4:1), at pressures up to 135 MPa. We have also used differential scanning calorimetry (DSC) to characterize changes in enthalpy associated with the observed transitions at ambient pressure. The effect of pressure on bicellar lipid mixtures was previously studied using residual dipolar couplings.28 A tentative pressure-temperature phase diagram for the DMPC/DHPC system, obtained using small-angle X-ray scattering, was recently reported.29 Some DSC observations have also been reported.15,30,31 Considered together, the DSC and variable-pressure NMR observations described here suggest a way of understanding how interactions between the long- and short-chain lipids in the bicellar mixture drive the observed changes in morphology. (16) Braganza, L. F.; Worcester, D. L. Biochemistry 1986, 25, 7484–7488. (17) Czeslik, C.; Reis, O.; Winter, R.; Rapp, G. Chem. Phys. Lipids 1998, 91, 135–144. (18) Wong, P. T. T.; Siminovitch, D. J.; Mantsch, H. H. Biochim. Biophys. Acta 1988, 947, 139–171. (19) Winter, R.; Pilgrim, W.-C. Ber. Bunsenges. Phys. Chem. 1989, 93, 708–717. (20) Driscoll, D. A.; Jonas, J.; Jonas, A. Chem. Phys. Lipids 1991, 58, 97–104. (21) Kaneshina, S.; Tamura, K.; Kawakami, H.; Matsuki, H. Chem. Lett. 1992, 1963–1966. (22) Maruyama, S.; Hata, T.; Matsuki, H.; Kaneshina, S. Biochim. Biophys. Acta 1997, 1325, 272–280. (23) B€ottner, M.; Ceh, D.; Jacobs, U.; Winter, R. Z. Phys. Chem. (Muenchen, Ger.) 1994, 184, 205–218. (24) Braganza, L. F.; Worcester, D. L. Biochemistry 1986, 25, 2591–2595. (25) Bonev, B. B.; Morrow, M. R. Phys. Rev. E 1997, 55, 5825–5833. (26) Singh, H.; Emberley, J.; Morrow, M. R. Eur. Biophys. J. 2008, 37, 783–792. (27) Brown, A.; Skanes, I.; Morrow, M. R. Phys. Rev. E 2004, 69, 011913. (28) Brunner, E.; Arnold, M. R.; Kremer, W.; Kalbitzer, H. R. J. Biomol. NMR 2001, 21, 173–176. (29) Winter, R.; Jeworrek, C. Soft Matter 2009, 5, 3157–3173. (30) Kozak, M.; Kempka, M.; Szpotkowski, K.; Jurga, S. J. Non-Cryst. Solids 2007, 353, 4246–4251. (31) Takajo, Y.; Matsuki, H.; Matsubara, H.; Tsuchiya, K.; Aratono, M.; Yamanaka, M. Colloids Surf., B 2010, 76, 571–576.

Langmuir 2010, 26(14), 12104–12111

Article

Experimental Section Materials. DMPC (1,2-dimyristoyl-sn-glycero-3-phosphocholine) and DHPC (1,2-dihexanoyl-sn-glycero-3-phosphocholine) along with the chain perdeuterated versions of these lipids, DMPC-d54 and DHPC-d22, respectively, were purchased from Avanti Polar Lipids, Alabaster, AL, in powder form and used without further purification. Sample Preparation. Dry lipids, with the acyl chains of either DMPC or DHPC fully deuterated, were combined to give DMPC/DHPC molar ratios of either 3:1 or 4.4:1 and total lipid masses of 17-30 mg. Dry lipid mixtures were then dissolved in 2:1 (v/v) chloroform/methanol. Solvent was removed by rotary evaporation at 45 °C and then by application of vacuum for ∼8 h. Samples were then hydrated in 300-500 μL of 100 mM HEPES buffer (pH = 7), vortexed, and further mixed using a sonicator bath for 15 min followed by 4-5 cycles of freezing and thawing. Samples were sealed into polyethylene capsules formed by heatsealing short segments of disposable pipet. 2 H NMR Spectroscopy. Wide-line deuterium NMR spectra at selected temperatures and pressures were collected using a spectrometer comprising a 3.5 T superconducting magnet (Nalorac Cryogenics, Martinez, CA), a locally assembled data acquisition system, and a variable-pressure NMR probe.32 Pressures were measured using a Bourdon tube gauge. The temperature of the sample pressure cell was measured using a thermocouple and controlled to (0.1 °C using an Omega (Laval, QC) CYC3200 temperature controller. Samples were allowed to equilibrate for at least 40 min at each temperature before the start of data collection. Spectra were obtained by Fourier transforming free induction decays collected using a quadrupole echo pulse sequence33 characterized by π/2 pulse durations of 3-4 μs and a 35 μs pulse separation. Depending on signal strength, a given spectrum was obtained by averaging 8000-20 000 transients collected with a repetition time of 0.7 s. Effective digitization dwell times after oversampling34 were 4 μs for DMPC-d54/DHPC samples and 10 μs for DMPC/DHPC-d22 samples unless otherwise noted. No line broadening was applied to DMPC-d54/DHPC spectra. For DMPC/DHPC-d22, 200 Hz line broadening was applied only to some high-pressure nematic phase spectra. Observations were made at selected temperatures for ambient pressure, 66 MPa, 102 MPa, and 135 MPa. At ambient pressure, the transition observed on cooling from the nematic to the isotropic phase was found to be as much as 6 °C lower than the isotropic-to-nematic transition observed on warming. For this reason, observations reported for each pressure were carried out sequentially from low to high temperature so that effects of thermal hysteresis were consistent between spectral series. At the end of each sequence, samples were cycled back to low temperature and monitored for equilibration in the isotropic phase prior to observation at the next pressure. At ambient pressure, data collection started at 10 °C and extended to 65 °C in either 2 or 5 °C steps depending on proximity to a phase change. At higher pressures, the starting temperature was adjusted to accommodate pressure-induced changes in phase boundaries as described below. For each mixture containing DMPC-d54, two independent samples were examined to confirm reproducibility. In phases with axially symmetric reorientation about the bilayer normal on a time scale short compared to the characteristic time of the deuterium NMR experiment, the spectrum from deuterons on a given chain segment is a doublet split by35 !  3 e2 qQ 3 cos2 β - 1 3 cos2 θ - 1 Δν ¼ ð1Þ 2 h 2 2 (32) Bonev, B. B.; Morrow, M. R. Rev. Sci. Instrum. 1997, 68, 1827–1830. (33) Davis, J. H.; Jeffrey, K. R.; Bloom, M.; Valic, M. I.; Higgs, T. P. Chem. Phys. Lett. 1976, 42, 390–394. (34) Prosser, R. S.; Davis, J. H.; MacKay, A. L. Biochemistry 1991, 30, 4687–4696. (35) Davis, J. H. Biochim. Biophys. Acta 1983, 737, 117–171.

DOI: 10.1021/la1014362

12105

Article

Uddin and Morrow

where β is the angle between the bilayer normal and the applied magnetic field and θ is the angle between the carbon-deuterium bond and the bilayer normal. The angular brackets denote an average over chain motions that modulate θ on the time scale of the quadrupole echo experiment. That average is referred to as the orientational order parameter, SCD, for that chain segment. For samples in which the orientation of the bilayer normal is spherically distributed, the superposition of contributions from all bilayer orientations broadens each doublet into a Pake powder pattern with prominent edges at the splitting corresponding to β = 90°. Differential Scanning Calorimetry. Lipid mixture samples, prepared as described above, were diluted to ∼2 mg/mL in 100 mM HEPES buffer and scanned using a Microcal (Northhampton, MA) MC-2 DSC instrument. Each sample was equilibrated at ∼10 °C and then scanned at a rate of 30°/h. Origin 6.1 (OriginLab Corp., Northhampton, MA) was used to fit and subtract scan baselines and to obtain transition enthalpies by integration of observed heat capacity peaks. Each sample was scanned three times. For comparison, small fractions of lipid material were also extracted during preparation of NMR samples, diluted, and scanned by DSC. The total lipid content of samples prepared specifically for DSC was considered to be more reliable, and quoted enthalpies are based on those samples.

Results 2

Figure 1 shows H NMR spectra at ambient pressure and selected temperatures for DMPC-d54/DHPC and DMPC/ DHPC-d22 at molar ratios of 3:1 and 4.4:1. The sequences of spectra obtained are qualitatively similar to those reported previously for the DMPC-d54 component1,4,5,9 and for the DHPC-d22 component4,5 of the mixtures. At the lowest temperatures observed for each DMPC/DHPC ratio, both lipid components display spectra characteristic of isotropic reorientation on the quadrupole echo experiment time scale. It should be noted that complete chain perdeuteration of a lipid bilayer sample typically lowers observed phase transitions by a few degrees. In particular, for DMPC-d54/DHPC, where the deuterated component accounts for either ∼75% or ∼81% of the lipid content, deuteration is expected to depress transition temperatures mixtures by 23 °C. For DMPC/DHPC-d22 samples, where the deuterated component carries fewer deuterons and accounts for less than 25% of the lipid content, deuteration is expected to depress transitions by a few tenths of a degree at most. In comparing spectra with one or the other lipid component deuterated, this difference must be taken into account. At 23 °C, the DMPC-d54/DHPC spectra for both ratios (Figure 1a,c) become characteristic of anisotropic chain reorientation that is not axially symmetric on the time scale of the quadrupole echo experiment. Between 25 and 28 °C, the DMPCd54/DHPC spectra become increasingly characteristic of axially symmetric chain reorientation, predominantly about an axis perpendicular to the applied magnetic field. This behavior persists to between 36 and 38 °C for DMPC-d54/DHPC (3:1) and to between 34 and 36 °C for DMPC-d54/DHPC (4.4:1). The narrow peak at the spectral midpoint shows that a small fraction of DMPC-d54 reorients isotropically at these temperatures. For DMPC-d54/DHPC at ambient pressure, the spectra observed between 28 °C and either 36 or 34 °C for the (3:1) or (4.4:1) samples, respectively, are characteristic of significant, but not complete, bilayer orientation. This may reflect the low magnetic field (3.55 T) available in this variable-pressure spectrometer, or it may reflect some orientational metastability in the highly viscous wormlike micelle phase that has previously been identified as the magnetically orientable morphology of this system.12 12106 DOI: 10.1021/la1014362

Figure 1. 2H NMR spectra at ambient pressure and selected temperatures for dispersions of (a) DMPC-d54/DHPC (3:1), (b) DMPC/DHPC-d22 (3:1), (c) DMPC-d54/DHPC (4.4:1), and (d) DMPC/DHPC-d22 (4.4:1) in 100 mM HEPES buffer (pH = 7.0). Vertical bars indicate the approximate range of temperatures for which each sample is presumed to be in the nematic phase based on inspection of the DMPC-d54/DHPC spectra.

In the range 28-34 °C, spectra from DMPC/DHPC-d22 (4.4:1) at ambient pressure (Figure 1d) indicate that almost all of the DHPC-d22 undergoes reorientation with a small, but nonzero, orientational order parameter. The spectra in this range are qualitatively similar to those reported by Sternin et al.5 in the corresponding temperature range. For DMPC/DHPC-d22 (3:1) at ambient pressure (Figure 1b), the spectra in the intermediate temperature range reflect a small DHPC-d22 fraction that reorients isotropically in addition to the population undergoing anisotropic reorientation. Above 36 °C for DMPC-d54/DHPC (3:1) and above 34 °C for DMPC-d54/DHPC (4.4:1), spectra from the DMPC-d54 component are characteristic of axially symmetric reorientation about a spherically distributed symmetry axis. The corresponding spectra for DMPC/DHPC-d22 are superpositions of doublets with distinct splittings, suggesting a distribution of orientational order parameter values along the DHPC-d22 chain. The 60 °C spectra for DMPC/DHPC-d22 are similar to that shown by Sternin et al. for similar conditions.5 The maximum DHPC-d22 splitting in the more ordered environment is ∼44% of the maximum DMPC-d54 splitting in the corresponding sample. Langmuir 2010, 26(14), 12104–12111

Uddin and Morrow

Article

Figure 2. Representative 2H NMR spectra at ambient pressure for DMPC-d54/DHPC (4.4:1) and DMPC/DHPC-d22 (4.4:1) in (a, e) the isotropic phase, (b, f) the nematic phase below the onset of axially symmetric DMPC reorientation, (c, g) the nematic phase above the onset of axially symmetric DMPC reorientation, and (d, h) the lamellar phase.

Figure 3. 2H NMR spectra at selected temperatures for DMPCd54/DHPC (3:1) at (a) 66 MPa, (b) 102 MPa, and (c) 135 MPa and for DMPC-d54/DHPC (4.4:1) at (d) 66 MPa, (e) 102 MPa, and (f) 135 MPa. Samples are dispersed in 100 mM HEPES buffer (pH = 7.0). Vertical bars indicate the approximate range of temperatures for which each sample is presumed to be in the nematic phase based on inspection of the DMPC-d54/DHPC spectra.

Based on the scheme inferred from recent small-angle neutron scattering studies,12-14 the spectral changes observed on warming Langmuir 2010, 26(14), 12104–12111

Figure 4. 2H NMR spectra at 102 MPa and selected temperatures for dispersions of (a) DMPC/DHPC-d22 (3:1) and (b) DMPC/ DHPC-d22 (4.4:1) in 100 mM HEPES buffer (pH = 7.0). Vertical bars indicate the approximate range of temperatures for which each sample is presumed to be in the nematic phase based on inspection of the DMPC-d54/DHPC spectra.

these DMPC/DHPC mixtures from low to high temperature are presumed to reflect transitions from an isotropic phase, consisting of bicellar particles undergoing rapid thermal reorientation to a magnetically alignable chiral nematic phase consisting of entangled wormlike micelles and then to a multilamellar bilayer vesicle phase. Figure 2 shows representative DMPC-d54/ DHPC and DMPC/DHPC-d22 spectra for the isotropic phase, the nematic phase below the onset of axially symmetric DMPC reorientation, the nematic phase above the onset of axially symmetric DMPC reorientation, and the lamellar phase. As discussed below, the persistence of isotropic DHPC-d22 reorientation into the nematic phase temperature range (Figure 2f) likely reflects an exclusion of DHPC from more ordered DMPC-rich environments at temperatures below the onset of axially symmetric DMPC reorientation. Vertical bars next to each spectral sequence in Figures 1, 3, and 4 indicate the expected range of nematic phase morphology based on inspection of corresponding DMPC-d54/ DHPC spectra and, for DMPC/DHPC-d22 spectra, an estimate of the differential effects of deuteration on transition temperatures. Figure 3 show 2H NMR spectra at 66, 102, and 135 MPa for DMPC-d54/DHPC at molar ratios of 3:1 and 4.4:1. Before discussing specific observations at each pressure, some general trends are noted. On warming from the isotropic phase into the wormlike micelle phase, the temperature at which anisotropic reorientation is first apparent in DMPC-d54/DHPC spectra rises at ∼0.15 °C/ MPa for the 3:1 mixture and ∼0.19 °C/MPa for the 4.4:1 mixture. These rates are comparable with the rate of ∼0.19 °C/MPa that can be inferred from the pressure-temperature phase diagram reported from small-angle X-ray scattering (SAXS) observations of DMPC/DHPC (3.2:1).29 It should be noted that the main gel to liquid crystal transition for DMPC bilayers, as observed by variable-pressure 2H NMR also rises at ∼0.19 °C/MPa.25 Spectra for DMPC-d54/DHPC (3:1) at 66, 102, and 135 MPa are shown in parts a, b, and c of Figure 3, respectively. Spectra for DMPC-d54/DHPC (4.4:1) at 66, 102, and 135 MPa are shown in parts d, e, and f of Figure 3, respectively. For the 3:1 mixture, in which DHPC accounts for ∼25% of the molar lipid content, application of pressure reduces the extent to which DMPC-d54 is oriented, but the upper limit of the partial orientation temperature range is still identifiable and is found to rise at ∼0.14 °C/MPa. DOI: 10.1021/la1014362

12107

Article

For the 4.4:1 sample, in which DHPC accounts for ∼19% of the molar lipid content, clear evidence for orientation of DMPC-d54 is seen at all pressures. The apparent rate at which the nematic-tolamellar transition temperature changes with pressure in DMPCd54/DHPC (4.4:1) is ∼0.18 °C/MPa. The difference between how pressure affects magnetic orientation in the DMPC-d54/DHPC (3:1) and DMPC-d54/DHPC (4.4:1) mixtures was confirmed by repeating selected observations on independent samples of these two lipid compositions. Comparison of spectra for DMPC/DHPC-d22 (3:1) and DMPC/ DHPC-d22 (4.4:1) suggests that the lipid molar ratio may also affect the distribution of short-chain lipid between isotropic and more ordered environments at a given pressure in these mixtures. Spectra for DMPC/DHPC-d22 (3:1) and for DMPC/DHPC-d22 (4.4:1) at 102 MPa are shown in parts a and b of Figure 4, respectively. Spectra for these mixtures at 66 and 135 MPa are provided in Appendix 1 of the Supporting Information. While differences in quadrupole echo decay time and optimization of the spectrometer for wide-line spectroscopy preclude quantitative analysis of the partitioning of lipid components between isotropic and anisotropic environments, the DMPC/DHPC-d22 (3:1) spectra at ambient pressure (Figure 1b) show a fraction of DHPC-d22 continuing to display isotropic reorientation through the nematic phase temperature range. At 102 MPa the DMPC/ DHPC-d22 (3:1) spectral sequence (Figure 4a) shows decreasing amounts of isotropically reorienting DHPC-d22 as temperature is raised through the nematic range. For the DMPC/DHPC-d22 (3:1) spectral sequences at elevated pressure, the isotropically reorienting fraction of DHPC-d22 is a minimum just below the nematic-to-lamellar transition and increases sharply as the sample is warmed through the transition into the lamellar phase. Based on these observations, the fraction of isotropically reorienting DHPC-d22 in the nematic phase of DMPC/DHPC-d22 (3:1) appears to be more significant at ambient pressure than at elevated pressure. In contrast, the fraction of isotropically reorienting DHPC-d22 in the nematic phase of DMPC/DHPC-d22 (4.4:1) appears to be lower at ambient pressure (Figure 1d) than at elevated pressure (Figure 4b). In Figure 5, chain order behavior in the anisotropic phases is illustrated by plots of quadrupole splitting versus temperature for the 2H NMR spectral doublets corresponding to the methyl groups and the most ordered methylenes on DMPC-d54 acyl chains in the 3:1 and 4.4:1 mixtures at each pressure studied. As shown by eq 1, quadrupole splitting is proportional to the orientational order parameter for a particular methyl or methylene group. Arrows indicate the approximate temperatures of the nematic-to-lamellar transition in the given DMPC-d54/DHPC mixture at each pressure. For gel-like spectral components corresponding to reorientation that is not axially symmetric on the experimental time scale, the quadrupole splitting was approximated by the separation of inflection points on the spectral component shoulders, and the estimated uncertainty in splitting was correspondingly higher. At all pressures, DMPC-d54 methyl splittings (Figure 5a,c) decreased with increasing temperature through the nematic phase temperature range. The temperature dependence of the methyl splitting is slightly weaker at ambient pressure than at elevated pressure, but the methyl splitting at the nematic-to-lamellar transition is effectively independent of pressure. This suggests that, regardless of pressure, the nematic phases can only accommodate a limited degree of chain disorder without becoming unstable relative to the lamellar phase. At ambient pressure for both lipid molar ratios, there is a slight plateau in the methyl splittings at the nematic-to-lamellar transition. This is also seen at 66 MPa for the 12108 DOI: 10.1021/la1014362

Uddin and Morrow

Figure 5. Temperature dependence of quadrupole splittings at ambient pressure (open circle), 66 MPa (solid triangle), 102 MPa (open square), and 135 MPa (solid diamond) for (a) the methyl group doublet in DMPC-d54/DHPC (3:1), (b) the maximally split doublet in DMPC-d54/DHPC (3:1), (c) the methyl group doublet in DMPC-d54/DHPC (4.4:1), and (d) the maximally split doublet in DMPC-d54/DHPC (4.4:1). Arrows indicate approximate temperatures of the nematic-to-lamellar transition for each mixture at the labeled pressure in MPa.

4.4:1 mixture but is not apparent in the other high-pressure methyl splitting plots. The largest DMPC-d54 methylene splittings (Figure 5b,d) correspond to the plateau region of the orientational order parameter profile. Changes in plateau splitting reflect changes in area per lipid in a bilayer phase.36 At ambient pressure, spectra from 28 to 36 °C for the 3:1 mixture and from 28 to 34 °C for the 4.4:1 mixture are characteristic of axially symmetric reorientation in the magnetically oriented nematic phase. Over these temperature ranges, quadrupole splitting for the most ordered DMPC-d54 methylene group decreases only slightly. For the ambient pressure series, splitting increases at the nematic-to-lamellar transition as the spectrum changes from being characteristic of reorientation about a preferred bilayer normal direction to reorientation about spherically distributed bilayer normals. This presumably arises from the same effect as the small methyl splitting plateau seen at the same transition and may reflect a small decrease in area per lipid as the wormlike micelles of the nematic phase reorganize into more extended lamellae. In this regard, it is interesting to note that the DMPC/DHPC-d22 spectra generally show an increase in the fraction of isotropically reorienting DHPC-d22 as samples are warmed through the nematic-to-lamellar transition. A vestige of this feature in the maximum splitting versus temperature plot remains at 66 MPa for both lipid molar ratios. For higher pressures, the maximum splitting decreases monotonically through the nematic-to-lamellar transition for both samples. For the 4.4:1 sample at 102 and 135 MPa, there is a more abrupt (36) Nagle, J. F. Biophys. J. 1993, 64, 1476–1481.

Langmuir 2010, 26(14), 12104–12111

Uddin and Morrow

Article

Figure 7. Differential scanning calorimetry traces for dispersions of (a) DMPC/DHPC (3:1), (b) DMPC/DHPC (4.4:1), (c) DMPC/ DHPC-d22 (4.4:1), and (d) DMPC-d54/DHPC (4.4:1). Scans have been normalized to show excess heat capacity per mole of DMPC in each sample. For each sample, scans from two consecutive DSC runs are superimposed.

Figure 6. Pressure-temperature

phase diagrams obtained from inspection of spectra for (a) DMPC-d54/DHPC (3:1) and (b) DMPC-d54/DHPC (4.4:1) obtained while warming from low to high temperature. Phases distinguished are isotropic (solid triangle), nematic (open circle or square), and lamellar (inverted solid triangle). Within the nematic phase, squares and circles denote spectra in which DMPC-d54 reorientation was observed to be axially symmetric or not, respectively, on the time scale of the 2H NMR experiment. Superimposed symbols denote spectra which could not be unambiguously identified with a single phase.

change in the slope of splitting versus temperature at the nematicto-lamellar transition than is seen for the 3:1 sample. It may be significant that the 4.4:1 sample also displays a higher degree of magnetic alignment at high pressure. Finally, it should be noted that, for both lipid molar ratios over the temperature range corresponding to the oriented axially symmetric phase, the largest DMPC-d54 splitting in each spectrum changes only slightly with temperature and that the value of this splitting at the nematicto-lamellar transition is effectively independent of pressure. Figure 6 shows pressure-temperature phase diagrams derived from comparison of DMPC-d54/DHPC (3:1) and DMPC-d54/ DHPC (4.4:1) spectra with the representative spectra shown in Figure 2. Open symbols denote spectra most characteristic of the nematic phase. Within the nematic phase, spectra characteristic of axially or nonaxially symmetric reorientation are distinguished as squares or circles, respectively. A small number of spectra which could not be assigned unambiguously to a single phase are denoted by overlapping symbols. For each sample, the upper and lower boundaries of the nematic phase differ only slightly in Langmuir 2010, 26(14), 12104–12111

their pressure dependence. As noted above, the boundary temperatures for DMPC-d54/DHPC (4.4:1) are slightly more sensitive to pressure than those of DMPC-d54/DHPC (3:1). Figure 7 shows differential scanning calorimetry traces for samples of DMPC/DHPC (3:1) and DMPC/DHPC (4.4:1) prepared specifically for DSC (Figure 7a,b) and for small samples extracted during preparation of the DMPC-d54/DHPC (4.4:1) and DMPC/DHPC-d22 (4.4:1) mixtures for NMR (Figure 7c,d). The heat capacity curves have been normalized to show excess heat capacity per mole of DMPC in each sample. The curves have also been shifted vertically to aid visibility. The excess heat capacity traces display a lower temperature cusp spanning the temperature range over which the DMPC-d54/DHPC spectra change from being more gel-like to being more characteristic of axially symmetric reorientation. Heat continues to be absorbed beyond that temperature, and the excess heat capacity feature ends with another cusp ∼3°-5° above the onset of axially symmetric reorientation. It should be noted that over the upper 3°-5° range in which heat continues to be absorbed DMPC-d54 chain order decreases only slightly. Widths of the excess heat capacity features from samples prepared specifically for DSC and from samples extracted from the NMR preparations were similar. Average enthalpies associated with the excess heat capacity features, measured relative to the DMPC content of each sample, were 24.7 kJ/mol (DMPC) and 23.0 kJ/mol (DMPC) for the DMPC/DHPC (3:1) sample (Figure 7a) and DMPC/DHPC (4.4:1) sample (Figure 7b), respectively. For comparison, the transition enthalpy for the main gel-to-liquid crystal transition of DMPC bilayers is 22.6 kJ/mol. Transition enthalpies normalized to the total lipid content of each sample were also calculated and found to be 18.8 kJ/mol for both the DMPC/DHPC (3:1) and DMPC/DHPC (4.4:1) samples (Figure 7, a and b, respectively).

Discussion The observations described above can be discussed on two levels. Specific details of the observed spectra, and their pressure dependence, reflect behavior and environments of the long-chain and short-chain lipid components at ambient and high pressure. The DSC observations and the pressure dependence of the DOI: 10.1021/la1014362

12109

Article

observed transitions provide more general insight into the nature of transitions in the DMPC/DHPC system. In the nematic phase, the edges of the putative wormlike micelles are presumed to be enriched with the shorter chain lipid, DHPC. In considering the DHPC environment in these samples, it should be noted that, for all of the DMPC/DHPC-d22 spectral series, quadrupole splittings generally increase with increasing temperature above the onset of the nematic phase while the splittings for DMPC-d54 over the same temperature range generally decrease slightly with increasing temperature. This suggests that partial ordering of DHPC-d22 for temperatures above the onset of the nematic phase reflects exchange of DHPC-d22 between highly curved “edge” regions, characterized by low orientational order, and flatter DMPC-rich regions characterized by greater orientational order. The observed temperature dependence suggests that the fraction of time spent by a DHPC molecule in the DMPC-rich environment generally increases with increasing temperature. As noted by Sternin et al.,5 the onset of anisotropic reorientation is observed at lower temperature for DMPC-d54 than for DHPC-d22. Isotropic reorientation of DHPC-d22 persists up to the temperature at which DMPC-d54 begins to display spectra characteristic of axially symmetric reorientation, suggesting that DHPC diffuses more easily into fluidlike as opposed to more ordered DMPC-rich regions. At ambient pressure, the spectra for DMPC/DHPC-d22 (4.4:1) over the rest of the nematic temperature range (Figure 1d) contain almost no component characteristic of isotropic reorientation, suggesting that, at this lipid ratio, effectively all of the DHPC is partially oriented. In contrast to the (4.4:1) mixture, the ambient pressure spectra for DMPC/DHPCd22 (3:1) contain a small component corresponding to isotropic DHPC reorientation at all temperatures. As pressure is increased, the fraction of isotropically reorienting DHPC-d22 in the nematic temperature range decreases for DMPC/DHPC-d22 (3:1) but increases for DMPC/DHPC-d22 (4.4:1). The Clausius-Clapeyron equation describes how pressure dependence of a transition depends on the transition enthalpy and the change in molar volume of the material through the transition. While the excess heat capacity feature observed for the DMPC/DHPC samples in this work extends, by a few degrees, beyond the temperature at which the DMPC-d54 spectra display axially symmetric reorientation characteristic of a fluid bilayer phase, the DMPC-normalized transition enthalpy, observed by DSC, is close to that reported for the main gel-to-liquid crystal transition of DMPC bilayers. The dependences of the isotropicto-nematic and nematic-to-lamellar transition temperatures on pressure also approximate that of the main transition temperature for bilayers of DMPC alone. The correspondences of the mixture transition enthalpies and transition temperature pressure dependence with those of a DMPC bilayer suggest a coupling between the DMPC main transition and the changes in morphology of DMPC/DHPC bicellar mixtures as was suggested by Harroun et al.14 Indeed, it is possible that onset of the isotropic-to-nematic transition reflects an instability of bicelle morphology as ordered DMPC chains begin to melt and the resulting increase in area becomes inconsistent with constraints imposed by the fixed circumference of the DHPC-rich bicelle edge. However, the observation that heat continues to be absorbed through temperatures over which DMPC-d54 orientational order does not change appreciably suggests that disordering of DMPC chains is not the only process absorbing heat above the isotropic-to-nematic transition. The DHPC chains are presumed to have little orientational order in the bicelle phase, and the DHPC-d22 spectra confirm that their orientational order remains low at the onset of the nematic 12110 DOI: 10.1021/la1014362

Uddin and Morrow

phase. It is possible, however, that hydration of the DHPC headgroups does change as the mixture transforms from bicelle morphology, through the nematic phase, to more lamellar morphology. In the bicelle morphology, DHPC molecules are in a very highly curved environment which probably increases their accessibility to interaction with water. In the less curved environments of the nematic and ultimately lamellar phases, the headgroups likely become less accessible to water. The loss of DHPC headgroup hydration through this process may account for some of the heat absorbed through the nematic range of temperatures. It is also possible that the shape and width of the excess heat capacity feature may be more directly related to changes in geometry of the lipid aggregates through the isotropic-to-nematic transition. Heimburg has demonstrated how changes in the curvature of lipid structures through a transition can broaden and split the excess heat capacity profile.37 It is possible that NMR studies of bicellar mixtures containing headgroup-deuterated lipids might provide some additional insight into the processes responsible for the shape of the excess heat capacity profile at the isotropicto-nematic transition. Quadrupole splittings of the DMPC-d54 methyl and most ordered methylene groups decrease slightly as temperature is raised about the onset of axially symmetric reorientation in the nematic phase. At ambient pressure, there is a slight increase in splitting associated with the nematic-to-lamellar transition. This implies a decrease in lateral area per lipid at that transition. This behavior may be related to the observation that the fraction of isotropically reorienting DHPC-d22 increases at the nematicto-lamellar transition for both molar ratios and all pressures. This suggests that some of the short-chain lipid interacting with DMPC in the nematic phase segregates into less ordered environments, possibly pore edges, in the lamellar phase, allowing for a slight increase in DMPC order just above the transition. At high pressure, DMPC-d54 chain order decreases more monotonically through the nematic-to-lamellar transition, suggesting that the distribution of DHPC between more or less ordered environments is sensitive to applied hydrostatic pressure. There are some differences between the ways in which the 3:1 and 4.4:1 lipid molar ratio samples respond to pressure. The fraction of isotropically reorienting DHPC-d22 in the nematic phase decreases with increasing pressure for the 3:1 mixture but increases with pressure for the 4.4:1 mixture. In order to determine whether this difference reflects a systematic dependence on lipid molar ratio or sensitivity to the details of how temperature and pressure are cycled, further systematic studies over a range of lipid molar ratios will likely be needed. Nevertheless, some general comments about the behavior of DHPC-d22 in these mixtures are possible. DHPC-d22 spectral components characteristic of anisotropic reorientation must reflect exchange of DHPC-d22 between “edge” environments and more ordered DMPC-rich environments since the DHPC-d22 splittings generally increases with increasing temperature through the nematic and lamellar phase ranges. Above the isotropic-to-nematic transition, spectral components characteristic of isotropically reorienting DHPC-d22 reflect DHPC-d22 that is confined to highly disordered environments such as layer or pore edges or, possibly, residual small isotropically reorienting particles. It should be noted that both of the DMPC/DHPC-d22 samples displayed varying fractions of isotropically reorienting DHPC-d22 in their nematic phases as temperature and pressure were cycled. Regardless of the distribution of DHPC-d22 between isotropic and anisotropic environments (37) Heimburg, T. In Planar Lipid Bilayers (BLMs) and Their Applications; Tien, H. T., Ottava-Leitmannova, A., Eds.; Elsevier: Amsterdam, 2003; pp 269-293.

Langmuir 2010, 26(14), 12104–12111

Uddin and Morrow

in the nematic phase, the lamellar phase DHPC-d22 spectra for both samples at all pressures reflected coexistence of isotropically and anisotropically reorienting DHPC-d22. Above the nematicto-lamellar transition, the more isotropic DHPC-d22 component likely reflects lipids confined to pore edges. The observation of isotropically reorienting DHPC-d22 under some conditions in the nematic phase might reflect the onset of pore formation in nematic structures under some conditions. One noticeable difference between the two lipid mixtures is the effect of pressure on the extent to which the nematic phases orient at high pressure. The formation of a nematic phase in the 3:1 sample at 135 MPa is confirmed by the DMPC/DHPC-d22 (3:1) spectra (Appendix 1), but there is almost no evidence of orientation in the DMPC-d54/DHPC (3:1) spectra at that pressure. This is in contrast to the clear evidence for orientation in the DMPC-d54/ DHPC (4.4:1) spectra at 135 MPa. While Figure 5b,d shows slight differences in the behavior of DMPC-d54 acyl chain quadrupole splitting at the nematic-to-lamellar transition, the similarity of the splittings for the 3:1 and 4.4:1 samples suggests that, in general, the DMPC-d54 environments, including bilayer thickness, are similar for the two lipid molar ratios and thus unlikely to account for the difference in susceptibility to magnetic orientation at high pressure. One way in which DMPC/DHPC ratio might affect these samples, however, is particle size. For bicelles or wormlike micelles with edges enriched in DHPC, increasing the fraction of DHPC in a sample might be expected to increase the perimeter to area ratio and thus decrease particle size. At higher pressures, the nematic phase temperature range for a given sample is higher. As shown from the quadrupole splittings of Figure 5, however, the combined effects of pressure and temperature leave DMPC-d54 chain order roughly constant in the nematic phase. Because of differences in particle size, however, the higher temperature required to form the nematic phase at high pressure may affect orientation of the nematic phase structures for different lipid molar ratios.

Langmuir 2010, 26(14), 12104–12111

Article

Summary DMPC/DHPC (3:1) and (4.4:1) mixtures, with one or the other lipid chain-perdeuterated, have been examined by DSC and variable-pressure 2H NMR. By DSC, heat absorption is found to persist above the temperature, in the nematic phase, at which the DMPC chains are effectively melted. This suggests that heat is also absorbed as the short-chain lipid increasingly samples the more ordered DMPC-rich environment with increasing temperature in the nematic phase. The isotropic-to-nematic and nematicto-lamellar transition temperatures are found to increase with pressure at rates similar to that seen for the main transition in bilayers of DMPC alone. This suggests that the phase changes seen in DMPC/DHPC bicellar mixtures are strongly coupled to the gel-to-liquid crystalline phase transition of DMPC bilayers. The effect of pressure on behavior of DMPC-d54 quadrupole splittings at the nematic-to-isotropic phase transition may indicate that redistribution of the short-chain lipid at this transition is affected slightly by pressure. Acknowledgment. This work was supported by the Natural Sciences and Engineering Research Council of Canada. The authours thank John Katsaras and Mu-Ping Nieh for helpful discussions. The authors also thank Valerie Booth and Donna Jackman for access to the differential scanning calorimeter and technical assistance with its operation and Alanna Flynn for assistance with spectroscopy. Supporting Information Available: Appendix 1 showing spectra at selected temperatures for DMPC-d54/DHPC (3:1), DMPC/DHPC-d22 (3:1), DMPC-d54/DHPC (4.4:1), and DMPC/DHPC-d22 (4.4:1) at ambient pressure, 66 MPa, 102 MPa, and 135 MPa. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la1014362

12111