Modulation of Dipalmitoylphosphatidylcholine Monolayers by Dimethyl Sulfoxide Aleksandra P. Dabkowska,†,⊥,# Louise E. Collins,† David J. Barlow,† Robert Barker,‡ Sylvia E. McLain,§ M. Jayne Lawrence,*,†,# and Christian D. Lorenz*,∥,# †
Institute of Pharmaceutical Science, School of Biomedical Sciences, King’s College London, 150 Stamford Street, London, SE1 9NH, United Kingdom ‡ Institut Laue-Langevin, B.P. 156, 38042 Grenoble Cedex 9, France § Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, United Kingdom ∥ Department of Physics, School of Natural and Mathematical Sciences, King’s College London, Strand, WC2R 2LS, United Kingdom S Supporting Information *
ABSTRACT: The action of the penetration-enhancing agent, dimethyl sulfoxide (DMSO), on phospholipid monolayers was investigated at the air−water interface using a combination of experimental techniques and molecular dynamics simulations. Brewster angle microscopy revealed that DPPC monolayers remained laterally homogeneous at subphase concentrations up to a mole fraction of 0.1 DMSO. Neutron reﬂectometry of the monolayers in combination with isotopic substitution enabled the determination of solvent proﬁles as a function of distance perpendicular to the interface for the diﬀerent DMSO subphase concentrations. These experimental results were compared to those obtained from molecular dynamic (MD) simulations of the corresponding monolayer systems. There was excellent agreement found between the MD-derived reﬂectivity curves and the measured data for all of the H/D contrast variations investigated. The MD provide a detailed description of the distribution of water and DMSO molecules around the phosphatidylcholine headgroup, and how this distribution changes with increasing DMSO concentrations. Signiﬁcantly, the measurements and simulations that are reported here support the hypothesis that DMSO acts by dehydrating the phosphatidylcholine headgroup, and as such provide the ﬁrst direct evidence that it does so primarily by displacing water molecules bound to the choline group.
INTRODUCTION Dimethyl sulfoxide (DMSO) is an organic solvent that has long been used in a wide range of biological and pharmaceutical applications, including enhancement of drug penetration across the skin, cryo-protection of cells, cell fusion, and radiation protection.1 While many of the biological applications of DMSO are believed to stem from its ability to alter the properties of cellular membranes, the mechanism by which it does so remains unclear, despite considerable research. Insight at the molecular level into the interaction of the amphiphilic solvent with membrane lipids is therefore needed to understand and fully utilize the capabilities of DMSO as a modulator of membrane structure and barrier function. It is well-established that DMSO alters membrane structure in a concentration-dependent manner.2 Experimentally, there have been four diﬀerent DMSO concentration regimes in which the behavior of DPPC membranes has been diﬀerentiated: 0 < XDMSO ≤ 0.133, 0.133 < XDMSO ≤ 0.3, 0.3 < XDMSO ≤ 0.8, 0.8 < XDMSO ≤ 1.0, where X is the mole fraction of DMSO. At low DMSO concentrations (0 < XDMSO ≤ 0.133), a wide variety of experimental techniques (i.e., X-ray diﬀraction and small-angle X-ray scattering) have been applied in various investigations2−6 © 2014 American Chemical Society
to demonstrate two common characteristics of DPPC membranes caused by the presence of DMSO: (a) the preand main transition temperatures have been found to increase and ﬁnally merge at XDMSO ≈ 0.10 from a variety of experimental techniques and (b) a signiﬁcant reduction of multilamellar repeat spacing has been found due to a reduction of solvent space separating opposing bilayers. For mole fractions of DMSO within the range of 0.133 < XDMSO ≤ 0.3, a small decrease in the repeat spacing between bilayers has been observed, and a transition to a disordered phase has been found in the packing of the hydrocarbon tails of the lipid molecules.2 The repeat distance and main transition temperature have been found to remain constant at 57 Å and 52 °C, respectively,2,3,5 when DPPC membranes are in contact with solvents with DMSO concentrations between 0.3 and 0.8 mole fraction. Finally, at very high DMSO concentrations (0.8 < XDMSO ≤ 1.0), DPPC membranes have been observed to undergo a transition toward a phase characterized by a small repeat Received: April 8, 2014 Revised: June 30, 2014 Published: July 7, 2014 8803
dx.doi.org/10.1021/la501275h | Langmuir 2014, 30, 8803−8811
Here, a combination of specular neutron reﬂection (SNR), Brewster angle microscopy (BAM) experiments, and molecular dynamics (MD) simulations has been used to investigate the action of 0.05 and 0.1 mole fraction DMSO on a DPPC monolayer. SNR with contrast variation reveals the structure of the monolayer in the direction perpendicular to the surface, while BAM monitors the lateral structure. SNR is a sensitive tool to investigate the localization of DMSO within a phospholipid monolayer at the air−water interface, as well as allowing the solvation of the lipid head groups to be investigated. In addition to the experimental data, MD simulations of the DPPC monolayer were performed to gain speciﬁc atomic-level insight into the action of DMSO on the lipid structure.
distance (∼52 Å) and a high transition temperature (74.2 °C).2,3,5 Molecular dynamics (MD) simulations have also been used to investigate the nature of the interactions between the DMSO and DPPC bilayers. Both atomistic (XDMSO > 0.15)7 and coarse-grain simulations (XDMSO > 0.27)8 have shown pore formation, which has been proposed as a mechanism resulting in the enhanced permeability that is observed experimentally at concentrations of XDMSO > 0.26.9 Additionally, simulations of DMSO in contact with ceramide membranes have also shown pore formation.10,11 In an attempt to determine the molecular mechanism that leads to these pores in DPPC membranes, a few simulation studies have been conducted with DMSO concentrations of less than 0.15 mole fraction. A common feature in most of the simulation studies is that the phosphatidylcholine (PC) headgroup becomes dehydrated in the presence of DMSO,7,8,12,13 a ﬁnding that is consistent with interpretations of infrared spectroscopic measurements.14 The addition of DMSO was found to cause an increase in the thickness of phospholipid monolayer in the form of a foam, which was attributed to a reduction in the number of water molecules in the vicinity of the headgroup due to the competition between the lipid headgroup and the DMSO for water.15 Using neutron diﬀraction enhanced by isotopic substitution and computer modeling of a short carbon chain (C3) phospholipid dissolved in a mixture of DMSO and water, we have recently determined that some of the water molecules hydrating the phosphatidylcholine head are replaced by DMSO.16 It is postulated, however, that the location of the DMSO within a membrane depends on the length of the lipid chain.6 Therefore, it is diﬃcult to make predictions about fulllength phospholipid behavior in a membrane from this previous work due to the short, three carbon chains of the lipid used. The bulk of reported investigations suggest that DMSO causes a dehydration of lipid head groups. The exact mechanism by which DMSO causes dehydration is, however, a matter of current debate; some researchers believe that DMSO is excluded from the monolayer or does not preferentially interact with the lipid headgroup, but instead forms strong electrostatic interactions with water molecules, thereby withdrawing water molecules from the head groups.3,5,15,17,18 Another theory is that DMSO can interact with the lipid headgroup directly and displace water molecules, and is therefore situated either at the solvent−lipid interface2,7,8,10,13,19−21 or found penetrating into the headgroup region.1,22 In light of the questions raised by previous studies, we have focused on the atomic scale interactions that occur between DMSO and lipid membranes, with a view toward understanding the exact mechanism of dehydration, and the consequential eﬀects on lipids in membranes. In this study, we used insoluble lipid monolayers at the air−water interface as these have been found to be suitable models of lipid membranes23 and have been previously used to gain insight into the interaction of DMSO with phospholipids.14,22,24 Previous monolayer studies using grazing incidence X-ray diﬀraction22 and vibrational sum frequency generation14 have determined that DMSO causes a condensation of phosphatidylcholine monolayers and a dehydration of the phosphate group. The location of the DMSO molecules within the phospholipid monolayer was, however, not accessible using these techniques.
MATERIALS AND METHODS
Materials. 1,2-Dihexadecanoyl-sn-glycero-3-phosphatidylcholine (DPPC; C40H80NO8P; mw 734.05 g mol−1; 16:0 PC; >99% purity), its chain-deuterated analogue (d62DPPC; C40H18NO8PD62; mw 796.42 g mol−1; >99% purity), and its fully-deuterated analog (d75DPPC; C40H5NO8PD75; mw 809.50 g mol−1; >99% purity) were purchased from Avanti Polar Lipids (Alabaster, AL). Lipids were dissolved in chloroform at a known concentration of approximately 1 mg mL−1. The purity of the various isotopic forms of the lipids was conﬁrmed by comparing their absorption isotherms with those reported in the literature.25 Spectroscopic grade ethanol and chloroform were purchased from AnalaR (BDH Chemical Ltd., Poole, UK). Ultrapure water was either bidistilled in a well-seasoned still or puriﬁed using a Millipore Milli-Q system to a resistivity of 18 MΩ cm, resulting in spectroscopically pure water with a measured surface tension of 72.8 ± 0.5 mN m−1. Deuterated water (D2O; 99.9% deuteration) was obtained from Euriso-top (Paris, France). Anhydrous dimethyl sulfoxide (hDMSO; (CH3)2SO; mw 78.13 mol−1; ≥99.9% purity;