830
Langmuir 2008, 24, 830-836
Enantiospecific Interactions between Cholesterol and Phospholipids Juha-Matti Alakoskela,*,† Karen Sabatini,† Xin Jiang,‡ Venla Laitala,† Douglas F. Covey,‡ and Paavo K. J. Kinnunen†,§ Helsinki Biophysics and Biomembrane Group, Institute of Biomedicine/Biochemistry, P.O. Box 63, UniVersity of Helsinki, 00014 Helsinki, Finland, Department of Molecular Biology and Pharmacology, Washington UniVersity School of Medicine, St. Louis, Missouri 63110, and MEMPHYSsCenter for Biomembrane Physics, UniVersity of Southern Denmark, Odense, Denmark ReceiVed September 19, 2007. In Final Form: October 26, 2007 The effects of cholesterol on various membrane proteins have received considerable attention. An important question regarding each of these effects is whether the cholesterol exerts its influence by binding directly to membrane proteins or by changing the properties of lipid bilayers. Recently it was suggested that a difference in the effects of natural cholesterol and its enantiomer, ent-cholesterol, would originate from direct binding of cholesterol to a target protein. This strategy rests on the fact that ent-cholesterol has appeared to have effects on lipid films similar to those of cholesterol, yet fluorescence microscopy studies of phospholipid monolayers have provided striking demonstrations of the enantiomer effects, showing opposite chirality of domain shapes for phospholipid enantiomer pairs. We observed the shapes of ordered domains in phospholipid monolayers containing either cholesterol or ent-cholesterol and found that the phospholipid chirality had a great effect on the domain chirality, whereas a minor (quantitative) effect of cholesterol chirality could be observed only in monolayers with racemic dipalmitoylphosphatidylcholine. The latter is likely to derive from cholesterol-cholesterol interactions. Accordingly, cholesterol chirality has only a modest effect that is highly likely to require the presence of solidlike domains and, accordingly, is unlikely to play a role in biological membranes.
Introduction The demonstrations of enantiomer specificity in interactions between phospholipids and other small molecules are rare, with only a few reports in the literature,1-3 while most studies of enantiomer interactions with phospholipids have failed to demonstrate any difference between enantiomers.4-6 In contrast, enantiomer specificity is very frequent in interactions of compounds with proteins, since proteins typically have shaped surfaces devoid of symmetry and have multiple different interaction sites. The modulation of the functions of membrane proteins by lipids, lipophilic drugs, or other compounds interacting strongly with membranes always raises the question of whether this modulation derives from the direct binding of the lipid or drug onto membrane proteins or from the modulation of the physical properties of the membranes by these compounds. There are numerous ways in which the physical properties of membranes affect or could affect protein function (see, e.g., refs 7-11), and on the other hand, on the basis of lipid mobility studies and the * To whom correspondence should be addressed. Phone: +358 9 19125426. Fax: +358 9 1912544. E-mail:
[email protected]. † University of Helsinki. ‡ Washington University School of Medicine. § University of Southern Denmark. (1) Abood, L. G.; Hoss, W. P. Psychopharmacol. Commun. 1975, 1, 29-35. (2) Pathirana, S.; Neely, W. C.; Myers, L.; Vodyanoy, V. J. Am. Chem. Soc. 1992, 114, 1404-1405. (3) Nandi, N. J. Phys. Chem. A 2003, 107, 4588-4591. (4) Dickinson, B.; Franks, N. P.; Lieb, W. R. Biophys. J. 1994, 66, 20192023. (5) Tomlin, S. L.; Jenkins, A.; Lieb, W. R.; Franks, N. P. Anesthesiology 1998, 88, 708-717. (6) Alakoskela, J.-M.; Covey, D. F.; Kinnunen, P. K. J. Biochim. Biophys. Acta 2007, 1768, 131-145. (7) Mouritsen, O. G.; Bloom, M. Biophys. J. 1984, 46, 141-153. (8) Cantor, R. S. J. Phys. Chem. B 1997, 101, 1723-1725. (9) De Planque, M. R. R.; Killian, J. A. Mol. Membr. Biol. 2003, 20, 271-284. (10) Lee, A. G. Biochim. Biophys. Acta 2004, 1666, 62-87. (11) Alakoskela, J.-M. Interactions in Lipid-Water Interface Assessed by Fluorescence Spectroscopy. Academic Dissertation. University of Helsinki, Finland, 2006; http://urn.fi/URN:ISBN:952-10-2884-X.
increasing number of high-resolution structures of integral membrane proteins, these proteins are known to have several specific binding sites for different lipid species.12,13 One good way to separate specific effects from the nonspecific ones could be to use enantiomers of lipid or drug molecules: if the effects of these enantiomers on lipids themselves are equal, then any demonstration of enantiospecificity in the effects on proteins provides proof of binding to proteins. This strategy has been outlined in the review by Westover and Covey,14 suggesting that the enantiomer of cholesterol is an excellent tool to determine whether the effects of cholesterol on protein function derive from changes in the physical properties or from the binding of cholesterol on proteins. This suggestion, of course, relies on the fact that ent-cholesterol in phospholipid films has appeared to always produce equal effects compared to those of cholesterol.14 Yet, while no difference between ent-cholesterol and cholesterol has been observed so far, the most definitive test to demonstrate the lack or presence of the enantiospecific interactions would appear to be to compare ent-cholesterol and cholesterol under conditions where phospholipids themselves have been shown to have enantiospecific effects or where cholesterol chirality has been suggested to have some significance. If some conditions exist where lipid films devoid of proteins display enantioselectivity, then in such conditions the interpretation of the enantioselectivity as direct protein binding should be taken with care. A striking demonstration of enantiomer effects comes from fluorescence microscopy of dipalmitoylphosphatidylcholine (DPPC) monolayers, where the large solidlike domains display opposite chirality for (R)-DPPC (also known as L-DPPC) and (S)-DPPC (also known as D-DPPC).15 In the presence of small amounts of cholesterol they form spiral-shaped solid domains (12) Marsh, D.; Pa´li, T. Biochim. Biophys. Acta 2004, 1666, 118-141. (13) Qin, L.; Hiser, C.; Mulichak, A.; Garavito, R. M.; Ferguson-Miller, S. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 16117-16122. (14) Westover, E. J.; Covey, D. F. J. Membr. Biol. 2004, 202, 61-72. (15) Weis, R. M.; McConnell, H. M. Nature 1984, 310, 47-49.
10.1021/la702909q CCC: $40.75 © 2008 American Chemical Society Published on Web 01/03/2008
Cholesterol Enantiomers and Domain Chirality
whose sense was opposite for L-DPPC and D-DPPC.16 The fact that cholesterol produced similar effects in both systems suggested that its most important effect was to reduce line tension. Nevertheless, Weis and McConnell observed some chirality even for a racemic mixture and suggested that this residual chirality could derive from the chirality of cholesterol itself.16 If the chirality of cholesterol has significance in its lipid interactions, then we would expect its enantiomer to produce opposite effects, and natural cholesterol and ent-cholesterol should have quantitatively different effects for pure DPPC enantiomers and qualitatively different effects for racemic DPPC. Unfortunately, solidlike domains are required to have complex, chiral domain shapes, since liquidlike domains in a liquidlike matrix produce circular domain shapes. Therefore, only low cholesterol mole fractions may be assessed by these means, as large cholesterol fractions abolish the transition. Similarly, while the pretransition of dimyristoylphosphatidylcholine enantiomers has the lowest temperature for a racemic mixture,17 the disappearance of pretransition already at low cholesterol mole fractions18 also prevents the use of the phospholipid pretransition temperature as an indicator. Another feature where the pure enantiomers of phosphatidylcholines and their racemic mixture are different is the Maxwell displacement current of DPPC monolayers.19 Basically, the Maxwell displacement current measures how charge polarization changes as a function of the area per molecule when the area per molecule is changed (in experiments at a constant rate) as a function of time. Thus, it should be related to the derivative of the surface dipole potential, which measures the charge polarization as a function of the area per molecule, and the effect should be observable in the surface dipole potential as well. Surface potential measurements of mixtures of DPPC and cholesterol enantiomers at high cholesterol concentrations would therefore be likely to provide indications of enantiospecific interactions if any exist. This testing using the enantiomers of both the phospholipid and cholesterol is particularly useful, as it immediately offers a means to verify any real quantitative difference in interactions and distinguish them from any error sources such as small differences in purities or stock concentrations, as any stronger interaction of one cholesterol enantiomer with one phospholipid enantiomer, e.g., natural cholesterol with natural L-DPPC, should be reflected as the corresponding stronger interaction of the opposite enantiomers, e.g., ent-cholesterol with D-DPPC, and weaker interactions for both mixed pairs. Experimental Methods 1-Oleoyl-2-[12-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]dodecanoyl]glycero-sn-3-phosphocholine (NBD-PC) and 1,2-dipalmitoylglycero-sn-3-phosphocholine (L-DPPC) were from Avanti Polar Lipids (Alabaster, AL). 2,3-Dipalmitoylglycero-sn-1-phosphocholine (D-DPPC) was from Fluka, and racemic DPPC and natural cholesterol were from Sigma-Aldrich. The enantiomer of cholesterol, entcholesterol, was synthesized as described previously.20 For fluorescence microscopy a monolayer trough with a glass window at the bottom (Microtrough X, Kibron Inc., Espoo, Finland) was mounted on an inverted Zeiss IM-35 microscope. The stock solution of NBD-PC was prepared in chloroform, and the concentration was determined photometrically by applying a small volume of stock to ethanol ( ) 22 000 M-1 cm-1 at 465 nm). The stock solutions of pure, nonfluorescent lipids were prepared in chloroform,
Langmuir, Vol. 24, No. 3, 2008 831 and the lipid concentration of these stocks was determined gravimetrically using a high-precision microbalance (Cahn 2000), with deviation of
