Langmuir 2004, 20, 8437-8441
8437
Effects of Small Neutral Molecules on Phospholipid Bicelle Ordering Xiaoxia Li and Boyd M. Goodson* 113 Neckers Hall, Department of Chemistry and Biochemistry, Southern Illinois University, Carbondale, Illinois 62901 Received May 5, 2004. In Final Form: July 31, 2004 The effects of small neutral molecules on the liquid-crystalline ordering of dimyristoyl-phosphatidylcholine (DMPC)/dihexanoyl-phosphatidylcholine (DHPC) bicelles (q ) 3.0 and 3.5) were studied via 2H, 31P, and 13C variable-temperature NMR. The addition of chloroform (up to ∼90 mM, with a lipid concentration of ∼120 mM) was observed to reduce the temperature onset of bicelle ordering by up to ∼10 °C, likely resulting from the depression of the DMPC phase transition temperature. The temperature for the collapse of the bicelle phase was also significantly reduced; the observed effects amount to a downward shift in temperature (and reduction in range) of the liquid-crystalline portion of the bicelle phase diagram with increasing dopant concentration. Other model dopants (e.g., tetrahydrofuran and benzene) yielded smaller effects. Additionally, the variable bicelle alignment permitted the characterization of the ordering of chloroform molecules within the lipid phase.
Introduction Aqueous mixtures of long-chain (bilayer-forming) and short-chain (micelle-forming) phospholipid molecules may be used to generate nematic liquid-crystalline phases commonly known as bicelles.1-3 Dilute bicellar solutions have gained considerable attention for achieving weak but tunable alignment of dissolved biological macromolecules, thereby increasing the information content of NMR structural determination methods (e.g., via the restoration of residual dipolar couplings and chemical shift anisotropies, CSAs).4-13 Moreover, bicelles have seen considerable use as model membranes, particularly for spectroscopic studies of membrane-associated polypeptides and proteins.2,14-20 * To whom correspondence should be addressed. Phone: 618453-6427. Fax: 618-453-6408. E-mail:
[email protected]. (1) Sanders, C. R.; Hare, B. J.; Howard, K. P.; Prestegard, J. H. Prog. NMR Spectrosc. 1994, 26, 421-444. (2) Sanders, C. R.; Landis, G. C. Biochemistry 1995, 34, 4030-4040. (3) NMR of Ordered Liquids; Burnell, E., de Lange, C., Eds.; Kluwer Academic: Dordrecht, 2003. (4) Tjandra, N.; Bax, A. Science 1997, 278, 1111-1120. (5) Ottiger, M.; Bax, A. J. Biomol. NMR 1998, 12, 361-372. (6) Clore, G. M.; Gronenborn, A. M.; Bax, A. J. Magn. Reson. 1998, 133, 216-221. (7) Clore, G. M.; Gronenborn, A. M. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 5891-5898. (8) Tjandra, N.; Tate, S.-i.; Ono, A.; Kainosho, M.; Bax, A. J. Am. Chem. Soc. 2000, 122, 6190-6200. (9) Peti, W.; Griesinger, C. J. Am. Chem. Soc. 2000, 122, 3975-3976. (10) Al-Hashimi, H. M.; Valafar, H.; Terrell, M.; Zartler, E. R.; Eidsness, M. K.; Prestegard, J. H. J. Magn. Reson. 2000, 143, 402-406. (11) Brunner, E.; Ogle, J.; Wenzler, M.; Kalbitzer, H. R. Biochem. Biophys. Res. Commun. 2000, 272, 694-698. (12) Tolman, J. R.; Al-Hashimi, H. M.; Kay, L. E.; Prestegard, J. H. J. Am. Chem. Soc. 2001, 123, 1416-1424. (13) Weise, C. F.; Weisshaar, J. C. J. Phys. Chem. B 2003, 107, 3265-3277. (14) Vold, R. R.; Prosser, R. S. J. Magn. Reson., Ser. B 1996, 113, 267-271. (15) Howard, K. P.; Opella, S. J. J. Magn. Reson., Ser. B 1996, 112, 91-94. (16) Czerski, L.; Sanders, C. R. Anal. Biochem. 2000, 284, 327-333. (17) Glover, K. J.; Whiles, J. A.; Wood, M. J.; Melacini, G.; Komives, E. A.; Vold, R. R. Biochemistry 2001, 40, 13137-13142. (18) Whiles, J. A.; Deems, R.; Vold, R. R.; Dennis, E. A. Bioorg. Chem. 2002, 30, 431-442. (19) Koenig, B. W. ChemBioChem 2002, 3, 975-980. (20) Zandomeneghi, G.; Williamson, P. T. F.; Hunkeler, A.; Meier, B. H. J. Biomol. NMR 2003, 25, 125-132.
One of the most common bicellar formulations utilizes mixtures of dimyristoyl-phosphatidylcholine (DMPC) and dihexanoyl-phosphatidylcholine (DHPC).5 For a given bicelle formulation, the temperature onset for liquidcrystalline behavior and the range of temperatures over which such behavior is observed are affected by a number of experimental variables, including the lipid molar ratio (q), the total lipid concentration, the salt concentration, and the presence of unsaturated lipids.3,5,21,22 In addition, doping bicelles with charged amphiphilic molecules can extend the liquid-crystalline range,23 and deuterating the lipid chains has recently been shown to effect a reduced temperature onset of bicelle ordering.24,25 Here we report that the addition of small, neutral molecules can also significantly affect the temperature onset of bicelle alignment as well as the observed range of liquid-crystalline behavior. The effects of such molecules on the ordering of DMPC/DHPC bicelles (q ) 3.0 and 3.5) were studied at 9.4 T via 2H, 31P, and 13C variabletemperature NMR. The addition of chloroform (up to ∼90 mM, with a lipid concentration of ∼120 mM) was reproducibly observed to reduce the temperature onset of bicelle ordering by up to nearly 10 °C. Collapse of the bicelle phase was also observed at significantly reduced temperatures; thus, the effects amounted to a downward shift in temperature (and reduction in overall range) of the liquid-crystalline portion of the bicelle phase diagram. The magnitudes of both effects grew monotonically with increasing chloroform concentration. Other model dopants (e.g., tetrahydrofuran (THF) and benzene) yielded smaller effects. The observed reduction in the temperature onset of ordering likely resulted from the depression of the gelto-liquid-crystalline phase transition (melting point, Tm) of the DMPC fatty-acid chains caused by the presence of the small neutral molecules (see below). Finally, measurement of 13CHCl3 splittings with varying liquid(21) Wang, H.; Eberstadt, M.; Olejniczak, E. T.; Meadows, R. P.; Fesik, S. W. J. Biomol. NMR 1998, 12, 443-446. (22) Raffard, G.; Steinbruckner, S.; Arnold, A.; Davis, J. H.; Dufourc, E. J. Langmuir 2000, 16, 7655-7662. (23) Losonczi, J. A.; Prestegard, J. H. J. Biomol. NMR 1998, 12, 447-451. (24) Sternin, E.; Nizza, D.; Gawrisch, K. Langmuir 2001, 17, 26102616. (25) Aussenac, F.; Laguerre, M.; Schmitter, J. M.; Dufourc, E. J. Langmuir 2003, 19, 10468-10479.
10.1021/la048886y CCC: $27.50 © 2004 American Chemical Society Published on Web 08/27/2004
8438
Langmuir, Vol. 20, No. 20, 2004
Letters
Figure 1. Selected 2H and 31P NMR spectra showing the influence of temperature on DMPC/DHPC bicelle solutions (q ) 3.5, 7.5% w/v), prepared without (A) and with (B) chloroform added (2.0 µL, ≈62 mM). 2H and 31P spectra were processed with 0.5 and 5.0 Hz of line broadening, respectively; all spectra are shown normalized by peak height for clarity (31P spectra are referenced to phosphate buffer).
crystalline alignment permitted the ordering of the chloroform molecules within the bicelle phase to be characterized. Experimental Section All phospholipids were purchased as dry powders from Avanti Polar Lipids, Inc., and used without further purification. Bicelle solutions were prepared generally following typical procedures reported in the literature (e.g., ref 5), described briefly as follows. Because of its hygroscopic nature, DHPC was carefully weighed in a dry (N2) atmosphere and dissolved in buffer (10 mM phosphate buffer in 93% H2O/7% D2O, with 0.15 mM sodium azide, at pH 6.6). A portion of the DHPC stock solution was added to a vial containing a premeasured quantity of DMPC sufficient to yield the desired molar ratio q ) [DMPC]/[DHPC] (≈3.0 and ≈3.5 for the present study, values chosen to fit within the range typically used in protein NMR5). Additional buffer was added to give a final total lipid [DMPC + DHPC] concentration of 150 mg/mL (15% w/v); 2.4 mM TTAB (myristyltrimethylammonium bromide, 99%, Acros) was then added to the solution as a stabilizing agent. Finally, the stock bicelle solution was treated with several brief (low-power) sonication cycles alternating between ice water and hot water (38 °C) baths. The DMPC/DHPC stock solution was stored at -78 °C when not in use. For experiments involving the addition of small neutral molecules (e.g., 13CHCl3, Cambridge Isotope Labs), first 200 µL of buffer was put into a vial chilled in ice water. Following the addition of the desired amount of chloroform, 200 µL of the DMPC/ DHPC stock solution was placed in the vial, yielding the desired final lipid concentration (here, 7.5% w/v). Then the vial was sealed and equilibrated for 10 min in ice water. At this stage, the samples were transparent, viscous liquid samples of uniform consistency (both those with and without the chloroform). Each sample (≈400 µL total volume) was then loaded in capped 5 mm NMR tubes immediately prior to the NMR experiment. NMR spectra were recorded on a two-channel 400 MHz Varian Unity Inova spectrometer using a Varian 10 mm (inverse) HX liquids probe retrofitted for z-gradient capability (JS Research, Quincy, MA). 2H spectra were acquired with the deuterium fieldfrequency lock turned off; for experiments involving alternating acquisition of spectra from 2H and 31P (or 13C) nuclei, the 2H spectra were obtained using the probe’s lock channel (and recabling) in order to avoid repeated manual retuning of the probe’s broadband channel. Shimming was performed on the solvent water (1H and 2H) line shape. Variable-temperature control was performed using the spectrometer’s VT controller (with temperature stability of (0.1 °C). The sample temperature was separately calibrated using a methanol standard; all temperatures reported here reflect a correction of (-)1 °C from the VT set temperature as a result of this calibration. At each
temperature setting, the samples were allowed to equilibrate for at least 10 min before collecting data. To minimize effects from thermal hysteresis,5,24 spectra were always acquired from lower to higher temperatures. All spectra were obtained via a simple pulse-acquire sequence (without 1H decoupling); the resulting 1-D NMR spectra were processed and analyzed on a PC using the MestRe-C (version 3.40) freeware NMR data processing program (www.mestrec.com).
Results and Discussion The microscopic environments within the DMPC/DHPC bicelle solutions were studied via 2H, 31P, and 13C variabletemperature NMR, providing spectra with complementary information regarding the sample ordering from the perspectives of the aqueous solvent, the phospholipids, and the added chloroform (respectively). Typical 2H and 31 P spectra taken from a q ) 3.5 sample are shown in Figure 1, with (B) and without (A) the addition of ≈62 mM of chloroform. The spectra in panel A show the expected temperature-dependent NMR signatures of bicelle ordering.5 For example, the 2H spectra exhibit a single resonance from the D2O solvent up to about 28 °C, at which point the line begins to split into a doublet. This characteristic doublet results from incomplete averaging of the deuterium quadrupolar interaction as the water molecules undergo rapid exchange between the (anisotropic) environment of the hydration shell associated with the oriented bicelles and the (isotropic) environment of the bulk solvent.5,23 Correspondingly, the 31P spectra obtained below the temperature onset of bicelle ordering exhibit a narrow resonance atop a broad CSA line shape (neglecting the buffer reference signal), primarily resulting from micellar DHPC and lamellar phases of DMPC, respectively.24 At higher temperatures, formation of the bicellar liquid-crystalline phase is clearly manifested by the two upfield CSA-shifted peaks, corresponding to DHPC and DMPC (rightmost peak), respectively, with integrated intensities that yield the molar ratio of the two phospholipids in the bicelles.5,26 Such spectra are consistent with the original picture of aligned bicelle structures (disks of DMPC bilayers edge-stabilized with DHPC, with normals perpendicular to the external field1) as well as the more recently accepted model: extended sheetlike structures (26) Bolze, J.; Fujisawa, T.; Nagao, T.; Norisada, K.; Saito, H.; Naito, A. Chem. Phys. Lett. 2000, 329, 215-220.
Letters
Figure 2. Plots of the magnitude of 2H splitting from deuterated water in DMPC/DHPC bicelle solutions (7.5% lipid w/v) as a function of temperature and amount of added chloroform (0 to ≈92 mM), showing the influence on alignment for q ) 3.5 (A) and for q ) 3.0 (B). The lines drawn over the stable bicelle region are meant to guide the eye. Multiple points at a given temperature indicate the presence of more than one phase.
of DMPC with “holes” lined with DHPC.27-29 The magnitudes of the 2H splitting and the upfield 31P shifts reflect the degree of alignment relative to the field.5 Following the usual bicellar behavior, the sample ordering reported by the 2H and 31P signals grows with increasing temperature, until the bicelle phase collapses (and the 2H and 31 P DMPC lines broaden considerably). Similar behavior is observed following the addition of chloroform to the bicelle solutions, with two primary differences manifested in the corresponding 2H and 31P spectra (Figure 1B): (1) a reduced temperature onset of the liquid-crystalline order; and (2) a reduced temperature range for the stable liquid-crystalline phase. Ordering begins ca. 5-6° “earlier” in panel B than in panel A, as indicated by the observed 2H splitting and corresponding (upfield) peaks in the 31P spectra; moreover, the stable bicelle phase probed in panel B is lost at a much lower temperature than in panel A, manifested primarily by the absence of the spectral signatures of bicelle formation in the 31P spectra and severe broadening (and biphasic behavior) in the 2H spectra.5,23 Results obtained from bicelle solutions prepared with varying amounts of added chloroform are summarized in Figure 2 (with data from bicelles with q ) 3.5 and q ) 3.0 shown in panels A and B, respectively). The curves show (27) Sanders, C. R.; Prosser, R. S. Structure 1998, 6, 1227-1234. (28) Gaemers, S.; Bax, A. J. Am. Chem. Soc. 2001, 123, 12343-12352. (29) Nieh, M.-P.; Glinka, C. J.; Krueger, S.; Prosser, R. S.; Katsaras, J. Langmuir 2001, 17, 2629-2638.
Langmuir, Vol. 20, No. 20, 2004 8439
the dependence of the 2H quadrupolar splitting on the temperature and amount of added chloroform and were generally reproducible to within a degree. Both effects observed in the spectra of Figure 1 (the reductions in temperature onset of ordering and range of liquidcrystalline behavior) grow monotonically stronger with increasing chloroform concentration, up to ≈92 mM; addition of significantly larger quantities generally yielded opaque white solutions without stable liquid-crystalline phases. Ordering was observed to begin in the q ) 3.5 samples about 3-4 °C earlier than in the q ) 3.0 samples (as expected5). Nevertheless, similar behavior was observed in the q ) 3.5 and q ) 3.0 runs: the average downward shift of the onset of ordering was roughly ∼0.8 °C per 10 mM of added chloroform (under the conditions of our experiments), with the effects of higher q and chloroform doping being essentially additive.30 Figure 3A shows selected 2H and 13C NMR spectra from a bicelle solution (q ) 3.5) following addition of ≈92 mM of 13CHCl3. As in Figure 1, the 2H spectra show the effects of bicelle alignment on the solvent water molecules; however, the increased amount of added chloroform causes the onset of quadrupolar splitting to shift down to ca. 19 °C, and the loss of stable liquid-crystalline order to be clearly manifested by 29 °C. The corresponding 13C spectra provide direct information regarding the ordering of the chloroform molecules themselves, which are partitioned primarily within the lipid phase. At low temperatures (below 19 °C), the 13C spectra exhibit a doublet signal with splitting slightly (
