11284
Langmuir 2006, 22, 11284-11289
Ceramide Promotes Restructuring of Model Raft Membranes Ira and Linda J. Johnston* Steacie Institute for Molecular Sciences, National Research Council Canada, 100 Sussex DriVe, Ottawa, ON Canada K1A 0R6 ReceiVed June 7, 2006. In Final Form: September 21, 2006 The generation of ceramide in cellular membranes is believed to cause coalescence of small lipid raft domains to give large signaling platforms, thus providing a site for the oligomerization of cell surface receptors. We have used atomic force microscopy to study the effects of ceramide generation by in situ enzymatic hydrolysis of sphingomyelin in phase-separated lipid bilayers that have sphingomyelin/cholesterol-rich domains surrounded by a fluid phase. In situ generation of ceramide produces heterogeneous domains with many raised subdomains that are also formed in bilayers containing premixed ceramide. However, in situ ceramide generation also results in the restructuring of the bilayer to give (1) areas of fluid phase that are devoid of domains, (2) areas that have a distribution of domains similar to the original bilayer, and (3) areas containing clusters of domains. The observation of the ceramide-promoted heterogeneity and clustering of raft domains in a physiologically relevant model provides strong support for the ceramide-induced formation of signaling platforms in cell membranes.
Introduction Lipid rafts are sphingolipid- and cholesterol (chol)-rich liquidordered domains in the fluid, liquid-disordered phase of the bulk cell membrane. They are characterized by tight packing of chol with saturated sphingolipids and are believed to be involved in regulating a number of important membrane-associated processes, including signaling, membrane trafficking, and viral infection. Specific proteins partition preferentially into lipid rafts, and this influences their functions, either through interaction with other raft proteins, or by limiting access to proteins in the bulk membrane.1-3 The functional significance of rafts has focused attention on the characterization of their protein distributions and the lipidlipid interactions that control membrane phase behavior. In particular, ceramides have recently been shown to modulate the behavior of lipid rafts.4 Ceramides are hydrophobic sphingolipids that are generated by the sphingomyelinase (SMase)-catalyzed hydrolysis of sphingomyelin (SM). Although the role of ceramide as a mediator of cell signaling is well-known,5,6 recent work suggests that ceramide can also act as a messenger indirectly by modulating membrane properties. In model membranes, ceramides increase lipid order, give rise to ceramide-rich domains, promote the formation of nonlamellar structures,7,8 and induce transbilayer flip-flop.8,9 In cellular studies, the activation of SMase leads to the reorganization of raft domains into ceramide-enriched patches, which coalesce into larger platforms.10,11 These large
platforms or macrodomains have been hypothesized to provide sites for the oligomerization of cell surface receptors and the internalization of bacteria and to aid in the inhibition of HIV invasion of cells.10 Several early studies that investigated the effects of ceramide addition to model membranes provided clear evidence for ceramide-promoted phase separation and suggested a role for ceramide in signal transduction.12-15 More recently, bilayer membranes have been used to examine the mechanism for the formation of ceramide-enriched platforms. For example, treating phosphatidylcholine (PC)/SM giant vesicles with SMase led to the formation of ceramide patches that coalesced into ceramiderich macrodomains that pinched off to form vesicles.16,17 It was suggested that the fusogenic property of ceramide results from hydrogen bonding and van der Waal forces between ceramide molecules. The formation of ceramide-enriched domains has also been observed recently for several binary mixtures in which the ceramide was added either by direct incorporation or by enzymatic hydrolysis.18-20 Although the addition of ceramide to vesicles with coexisting liquid-ordered and fluid phases has been shown to displace chol from rafts,21 the ceramide-induced coalescence of small raft domains has yet to be observed directly in a model membrane using techniques that allow for direct visualization of domains before and after ceramide addition. Given the postulated small size of rafts and the complexity of cellular membranes, visualizing the effect of ceramide on sphingolipid/chol-rich domains in model membranes is a key
* Corresponding author. E-mail:
[email protected]. (1) Simons, K.; Ikonen, E. Nature 1997, 387, 569-72. (2) Brown, D. A.; London, E. J. Membr. Biol. 1998, 164, 103-14. (3) Laude, A.; Prior, I. A. Mol. Membr. Biol. 2004, 21, 193-205. (4) Cremesti, A. E.; Goni, F. M.; Kolesnick, R. FEBS Lett. 2002, 531, 47-53. (5) Mathias, S.; Pena, L. A.; Kolesnick, R. N. Biochem. J. 1998, 335 (3), 465-80. (6) Hannun, Y. A.; Luberto, C.; Argraves, K. M. Biochemistry 2001, 40, 4893903. (7) Kolesnick, R. N.; Goni, F. M.; Alonso, A. J. Cell Physiol. 2000, 184, 285-300. (8) Contreras, F. X.; Villar, A. V.; Alonso, A.; Kolesnick, R. N.; Goni, F. M. J. Biol. Chem. 2003, 278, 37169-74. (9) Contreras, F. X.; Basanez, G.; Alonso, A.; Herrmann, A.; Goni, F. M. Biophys. J. 2005, 88, 348-59. (10) Bollinger, C. R.; Teichgraber, V.; Gulbins, E. Biochim. Biophys. Acta 2005, 1746, 284-294. (11) Grassme, H.; Jendrossek, V.; Riehle, A.; von Kurthy, G.; Berger, J.; Schwarz, H.; Weller, M.; Kolesnick, R.; Gulbins, E. Nat. Med. 2003, 9, 322-30.
(12) Holopainen, J. M.; Subramanian, M.; Kinnunen, P. K. Biochem. 1998, 37, 17562-70. (13) Holopainen, J. M.; Lehtonen, J. Y. A.; Kinnunen, P. K. J. Chem. Phys. Lipids 1997, 88, 1-13. (14) Huang, H.-W.; Goldberg, E. M.; Zidovetzki, R. Biochem. Biophys. Res. Commun. 1996, 220, 834-8. (15) Huang, H.-W.; Goldberg, E. M.; Zidovetzki, R. Biophys. J. 1999, 77, 1489-97. (16) Holopainen, J. M.; Angelova, M. I.; Kinnunen, P. K. J. Biophys. J. 2000, 78, 830-8. (17) Nurminen, T. A.; Holopainen, J. M.; Zhao, H.; Kinnunen, P. K. J. Am. Chem. Soc. 2002, 124, 12129-34. (18) Hartel, S.; Fanani, M. L.; Maggio, B. Biophys. J. 2005, 88, 287-304. (19) Silva, L.; de Almeida, R. F. M.; Federov, A.; Matos, A. P. A.; Prieto, M. Mol. Membr. Biol. 2006, 23, 137-48. (20) Sot, J.; Bagatolli, L. A.; Goni, F. M.; Alonso, A. Biophys. J. 2006, 90, 903-14. (21) Megha; London, E. J. Biol. Chem. 2004, 279, 9997-10004.
10.1021/la061636s CCC: $33.50 Published 2006 by the American Chemical Society Published on Web 11/15/2006
Ceramide Restructuring of Model Raft Membranes
step toward understanding the more complex interactions in cells. Such studies require the direct detection of small domains with nanometer-scale resolution. We and others have shown that atomic force microscopy (AFM) is a powerful tool for studying phase separation in the phospholipid monolayers and bilayers of ternary lipid mixtures as models for lipid rafts.22-27 AFM has the advantage of allowing direct detection of nanometer-scale domains under physiological conditions in aqueous solution.28-30 Herein we report a study of the effect of in situ ceramide generation on liquid-ordered domains in supported bilayers of dioleolylphosphatidylcholine (DOPC), SM, and chol. Our results demonstrate a striking bilayer restructuring that includes increases in the heterogeneity and clustering of the domains, effects that are similar to the hypothesized ceramide-induced coalescence of rafts in cellular membranes. For comparison, we show that the direct incorporation of ceramide in the same lipid mixture during membrane preparation leads to only some of the changes in morphology produced by in situ ceramide generation. Experimental Section Materials. DOPC, dipalmitoylphosphatidylcholine (DPPC), egg SM (ESM), C16:0 SM, and C16:0 ceramide were obtained from Avanti Polar Lipids and were used as received. SMase isolated from Bacillus cereus and Staphylococcus aureus were from Sigma-Aldrich. All aqueous solutions were prepared with 18.3 MΩ‚cm Milli-Q water. Preparation of Small Unilamellar Vesicles. Small unilamellar vesicles were prepared as previously described.22 Briefly, chloroform solutions of phospholipids were mixed in the appropriate ratios, and the lipid films obtained after drying the solvent were hydrated in water and vortexed to obtain multilamellar vesicles. The sample was then sonicated in a water bath sonicator to clarity to form small unilamellar vesicles with a final lipid concentration of 1 mM. Vesicles were used immediately for bilayer preparation or stored at 4 °C for up to a week prior to use. Bilayer Preparation. Vesicle solution (150 µL) and 300 µL of CaCl2 (15 mM) were added to freshly cleaved mica clamped in a liquid cell. After incubation for 30 min, bilayers were rinsed extensively with water to remove unattached vesicles before imaging. The presence of occasional defects allowed us to measure the bilayer thickness, confirming the presence of a single bilayer. Control experiments in which vesicle and bilayer preparation was carried out in a nitrogen atmosphere showed that bilayer morphologies for the ternary and quaternary lipid mixtures were similar for samples prepared under air and nitrogen. This indicates that lipid oxidation does not affect the results. AFM Imaging. AFM images were obtained at room temperature (22 ( 1 °C) on a PicoSPM atomic force microscope (Molecular Imaging) in MAC-mode using magnetic coated silicon tips with spring constants of ∼0.5 N/m and resonance frequencies between 8 and 35 kHz in aqueous solutions. Either a 30 × 30 µm2 or 5 × 5 µm2 scanner was used with a scan rate between 0.7 and 1.3 Hz. All images shown are flattened raw data. Two or three independently prepared samples were imaged for each bilayer composition, and several areas were scanned for each sample. Domain heights were (22) Yuan, C.; Johnston, L. J. Biophys. J. 2001, 81, 1059-69. (23) Yuan, C.; Furlong, J.; Burgos, P.; Johnston, L. J. Biophys. J. 2002, 82, 2526-35. (24) Burgos, P.; Yuan, C.; Viriot, M.-L.; Johnston, L. J. Langmuir 2003, 19, 8002-9. (25) Milhiet, P.-E.; Giocondi, M.-C.; Baghdadi, O.; Ronzon, F.; Roux, B.; Le Grimellec, C. EMBO Rep. 2002, 3, 485-90. (26) van Duyl, B. Y.; Ganchev, D.; Chupin, V.; de Kruijff, B.; Killian, J. A. FEBS Lett. 2003, 547, 101-6. (27) Weerachatyanukul, W.; Ira; Kongmanas, K.; Tanphaichitr, N.; Johnston, L. J. Biochim. Biophys. Acta 2006, in press. (28) Rinia, H. A.; Boots, J. W.; Rijkers, D. T.; Kik, R. A.; Snel, M. M.; Demel, R. A.; Killian, J. A.; van der Eerden, J. P.; de Kruijff, B. Biochem. 2002, 41, 2814-24. (29) Dufrene, Y. F.; Lee, G. U. Biochim. Biophys. Acta 2000, 1509, 14-41. (30) Shaw, J. E.; Epand, R. F.; Epand, R. M.; Li, Z.; Bittman, R.; Yip, C. M. Biophys. J. 2006, 90, 2170-8.
Langmuir, Vol. 22, No. 26, 2006 11285 measured as the difference between the condensed phase domains and the fluid phase. Results from several different experiments for each composition were averaged to give the reported domain height plus/minus the standard deviation. For SMase treatment, the bilayers were imaged in water prior to replacing the water by buffered enzyme solution (HEPES buffer: 125 mM NaCl, 10 mM CaCl2, 2 mM MgCl2, 10 mM HEPES, pH 7.4). The bilayers were imaged immediately or incubated with enzyme at room temperature for 2-30 min, rinsed extensively with water, and re-imaged. In flow experiments, the water in the flow cell was replaced by flowing through a volume of enzyme solution immediately prior to imaging. B. cereus SMase was used in all experiments, unless otherwise noted. Ceramide Assay. A bilayer formed by vesicle fusion was washed with HEPES buffer and incubated with the desired concentration of SMase (from B. cereus) for 15-30 min. The amount of ceramide generated was quantitated by measuring the amount of soluble phosphorylcholine produced using an Amplex Red SMase assay kit (Molecular Probes). A 100 µL portion of fluid from the sample cell (after SMase treatment) was mixed with 100 µL of the reaction mixture (0.1 mM Amplex Red Reagent, 2U/mL HRP, 0.2 U/mL choline oxidase, 8U/mL alkaline phosphatase, 100 mM Tris-HCl, 10 mM MgCl2, pH 7.4). After incubation at 37 °C for 90 min, the phosphorylcholine concentration was assayed by measuring the fluorescence of the sample at 590 nm (λex ) 560 nm). Since B. cereus SMase as supplied may contain a phospholipase contaminant, a control experiment in which a DOPC/DPPC/chol bilayer was treated with SMase for the same length of time and then analyzed for phosphorylcholine was used to correct for residual phospholipase activity. This indicated that
