Bioconjugate
Chemistry MAY/JUNE 1992 Volume 3, Number 3 0 Copyright 1992 by the American Chemical Society
Modification of Dry 1,Z-DipalmitoylphosphatidylcholinePhase Behavior with Synthetic Membrane-Bound Stabilizing Carbohydrates' Mary A. Testoff and Alan S. Rudolph' The Center for Biomolecular Science and Engineering, Code 6090,Naval Research Laboratory, Washington, D.C. 20375-5000.Received October 23, 1991
Carbohydrates, particularly disaccharides, have been shown to accumulate in organisms as protective solutes during periods of stress such as freezing and desiccation. Cholesterol and lipid derivatives containing the protective carbohydrates galactose or maltose, 0-[1l-(l-~-~-galactosyloxy)-3,6,9-trioxaundecanyllol (TEC-GAL),O-[ll-(l-~-~-maltosyloxy)-3,6,9-trioxaundecanyllol (TEC-MAL),and 14(galactosyloxy)-N,N-dimethyl-O-(dipalmitoylphosphatidyl)-6,9,12-trioxa-3-azoniatetradecanol (DPGAL), have been synthesized to investigate the interaction of a protective carbohydrate moiety tethered to the 1,2-dipalmitoylphosphatidylcholine(DPPC) bilayer surface. Toward this goal,we have investigated the calorimetric and infrared spectroscopic behavior of mixtures of DPPC codried with these glycolipids. The synthetic glycolipids are shown to decrease significantly the main transition temperature (max C,) of dry DPPC with a concomitant reduction in the cooperativity of the transition, as evidenced by a decrease in the enthalpy with increasing glycolipid. The decrease in transition temperature is shown to be related to chain melting monitored by the CH2 symmetric stretch frequency through the transition using FTIR. We also present evidence that the glycolipids interact with the interfacial region of DPPC, as shown by the decrease in the phosphate symmetric stretch intensity with increasing concentration of glycolipid. These observed effects are similar to the action of bulk protective sugars with DPPC; however, the concentration of glycolipid and the associated carbohydrate concentration needed to effect the observed changes are reduced compared to the quantity of bulk carbohydrate previously shown to give similar results with DPPC.
INTRODUCTION
The presence of water is essential for proper assembly of the constituents of biomembranes (i.e. phospholipids, sterols, membrane proteins) and the proper maintenance of critical membrane metabolic and physiological cellular processes such as energy transduction and molecular recognition. The interplay between water and biological membranes still remains an active area of research, as the nature and physiological role of membrane associated water has not been fully elucidated (1-5). Hydrogen bonding of water to the polar regions of membrane components is thought to provide the necessary structural and functional maintenance critical to biomembranes ( I , 2). Water removal during freezing and desiccation results Not subject to U.S. Copyright.
in functional and structural defects of biomembranes by inducing a number of changes in membrane structure. l Abbreviations used: DPPC, 1,2-dipalmitoylphosphatidylcholine; DPPC-dsz, 1,2-dipalmitoyl-d~z-sn-glycerol-3-phosphocholine; DOPE, dioleoylphosphatidylethanolamine;DP-GAL, 14~alact.~syloxy)-N~-dimethyl-O-(dipalmitoylphosphatidyl)-6,9,12trioxa-3-azoniatetradecanobTEC, triethoxycholesterol or 0-(11hydroxy-3,6,9-trioxaundecyl)cholesterol;TEC-GAL, triethoxycholesterol galactose or o-[ ll-(l-~-~-g~actosyloxy)-3,6,9trioxaundecanyl]cholesterol; TEC-MAL, triethoxycholesterol maltose or O-[ll-(l-j3-~-maltosyloxy)-3,6,9-trioxaundecanyl]cholesterob EDTA, ethylenediaminetetraacetic acid; TLC, thin-layer chromatography; DSC, differential scanning calorimetry; FTIR, Fourier transform infrared spectroscopy; FAB, fast atom bombardment; CI, chemical ionization.
Published 1992 by American Chemical Society
204
Testoff and Rudolph
Bloconjugate Chem., Vol. 3, No. 3, 1992
Dehydration- and freezing-induced phase separation of membrane components has been shown to result in damage to embedded membrane proteins as well as provide structural defects which may lead to membrane fusion and leakage of intracellular contents (1-3, 5 ) . In the search for mechanisms by which biomembranes are protected from damage by water removal, recent focus has been placed on the interaction of water replacement solutes found in organisms that survive freezing or desiccation with biomembranes (1,3-5). One class of such solutes are the polyhydroxy compounds of which the carbohydrates are found to commonly accumulate in large quantities during the induction of hydration stress. A variety of disaccharides (e.g. trehalose, sucrose) have been found to accumulate at high concentrations in biological systems during dehydration or freezing ( 1 , 3 , 4 ) . Infrared spectroscopic studies on the interaction of disaccharides and dry DPPC suggest that these solutes hydrogen bond to hydration sites of phospholipids, of which the phosphate group may be a potential site of interaction (I,2). This action results in the reduction in the observed transition temperature of dry DPPC dihydrate and in anhydrous DPPC (1). The maximal effect resulted in the reduction of the transition temperature of the dry system to around 24 "C, significantly below the transition temperature of DPPC in excess water (6). It has been found that the effect of trehalose on dry DPPC depends on the phase from which the lipid is dried, with the most significant reduction observed when DPPC is dried from the liquidcrystalline phase (6,7).The phase which arises due to the association of trehalose and dry DPPC has been designated Lk and has been characterized by phospholipids with restricted rotational motion of the acyl chains with increased intermolecular volume of the acyl chains compared to dry DPPC in the absence of trehalose (8, 9). One approach to the further study of the action of protective carbohydrates with lipid systems is to construct glycolipids that have protective carbohydrates covalently linked to the phospholipid head group. The strategy employed in the synthesis of cryoprotective glycolipids is to use an amphiphilic membrane anchor that will incorporate into a phospholipid bilayer. The covalent attachment of the carbohydrate to the amphiphile is extended through the hydrophobic region of the bilayers by a linker group (oxyethylene). The oxyethylene linkage is designed to give the carbohydrate mobility to localize at the interfacial region of the membrane surface and hydrogen bond to hydration sites on the bilayer. This model also allows maximal interaction of the hydrophilic carbohydrates with the membrane and does not require efforts to ensure homogeneous mixing of carbohydrate and lipid assemblies in bulk solution before the induction of freezing or desiccation. Incorporation of lipid molecules with inherent protective carbohydrates into such membrane systems (or cellular biomembranes) could induce stability to hydration stress which may be required for other technologies relying on lipid amphiphiles. Two glycolipids that have been synthesized previously (8)focused on cholesterol derivatives covalently attached to a protective sugar via a similar oxyethylene linkage, namely triethoxycholesterol galactose (TEC-GAL) and triethoxycholesterol maltose (TEC-MAL) (Figure 1) (IO). Goodrich et al. have studied the changes in the lipid phase behavior of hydrated DPPC and DOPE influenced by the addition of the linker triethoxycholesterol (TEC) and TECGAL, which were characterized by fluorescence polarization, 31PNMR, DSC, and freeze-fracture electron microscopy (10). The addition of TEC and TEC-GAL
HO-(CH2CHPO),H
TsO
tetraethylene glycol
1.
(1) A 11
~
2
w
CH,OAc
CH,OAc
CH,OAc
hzO/l&enzene, room lenp. N2(g) AcO&Br 0
2. NaOMdMeOH. pH 11-12
CHZOH
(A)
1
0'
I
OAc
CH,OH
f&o&Br OAchAc
I
OAc
CH,OH
(E)
Figure 1. Synthetic scheme and structure of sterol-based glycolipids: (A) TEC-GAL,O-[ll-( l-P-~-galactosyloxy)-3,6,9-trioxaundecanyllcholesterol;(B)TEC-MAL, O-[ll-(l-P-D-maltosyloxy)-3,6,9-trioxaundecanyl]cholesterol.TsO = p-toluenesulfonate and OAc = acetate.
resulted in a reduction in the measured anisotropy of the gel state and an increase in the liquid-crystalline phase similar to the order-disorder effect observed with cholesterol (10). The difference between the glycolipid sterol derivatives and cholesterol is the ability of the derivatives to decrease the gel to liquid-crystalline phase-transition temperature of DPPC and to increase the lamellar to hexagonal phase-transition temperature of DOPE in a concentration-dependent manner (10). Infrared and Raman spectroscopicinvestigations of the interaction of TECGAL and TEC-MAL with DPPC also reveal that the phosphate symmetric stretch frequency of DPPC is shifted in a similar fashion as that observed with bulk protective sugars (11). The cryoprotectant action of these compounds has also been recently demonstrated by examining the ability of TEC-GAL and TEC-MAL to inhibit freezethaw-induced fusion and leakage (12). As a further extension of this work, we have studied the calorimetric and spectroscopic phase behavior of dry films of DPPC codried with different mole percentages of TEC-GAL and TEC-MAL to observe the concentration-dependent capability of these molecules to alter the phase properties of dry DPPC. In addition, we have synthesized a new modified phospholipid membrane component composed of a carbohydrate (galactose) attached to DPPC by an oxyethylene linkage. This glycolipid, 14-(galactosyloxy)-N,N-dimethyl-0-(dipalmitoylphosphatidyl)-6,9,12-trioxa-3-azoniatetradecanol (DP-GAL) (Figure 2), and its effects on the lipid phase behavior of dry DPPC films have been investigated by DSC and FT-IR. Dry DPPC films were codried with various mole percentages of DP-GAL to study the importance of the concentration effects of the sugar moiety of DP-GAL on the acyl chain interactions of DPPC.
Bioconjugate Chem., Vol. 3, No. 3, 1992 205
Modlflcatlon of DPPC Phase Behavlor
HO--(CHzCH20)4H
tetraethylene glycol
-
2. (CH,)zNH e
C
H
3
1. NaH
0 "C
CHzOAc
OAC
0
II 1. HOCH&HzOH, ethylene glycol
phaspholipase D
CHOCR
c
CH3 CH20POCH2CHCkCH3
IF;
I
I 0-
CH3 1. PBr3 or (CH3),SiBr
0
CH,OAc
II
CH3,
CH,OCR
1;;
2.
,N(CHpCH,O),
CH3
CHOCR
OAc
*
CHzOPOCHzCH20H
I
00
II
CHzOCR
IF;
CHOCR
I F ;
CH~OPOCH~CH~+N(CH~CH~O)~ I I I CH, 0OH = (CH2)14CH3
Figure 2. Synthetic scheme and structure of DP-GAL, 14-(galactosyloxy)-NJV-dimethyl-O-(dipalmitoylphosphatidyl)-6,9,12trioxa-3-azoniatetradecanol. O A c = acetate. EXPERIMENTAL PROCEDURES
All reagent-grade chemicalsand solvents were purchased from Aldrich Chemical Co. (Milwaukee, WI). Tetrahydrofuran, chloroform, methanol, and dimethylformamide were further purified by distillation and dried over molecular sieves. DPPC and deuterated L-a-DPPC-dGZ, obtained through Avanti Polar Lipids, Inc. (Alabaster, AL), were stored at -70 "C in dry powder form or in chloroform. Tetra-0-acetyl-a-D-glucopyranosyl bromide (acetobromo-a-D-galactose)was purchased from Fluka Chemical Co. (Ronkonkona,NY). Hepta-O-acetyl-a-maltosyl bromide (acetobromo-a-maltose) was supplied by Sigma Chemical Co. (St. Louis, MO). Silica gel 60 (EM 70-230 mesh) was used for all column chromatography purification, and thin-layer chromatography (TLC) was performed on EM silica gel 60 plates, both supplied by Merck (Germany). The plates were visualized by nondestructive iodine vapors. Some plates were further treated and characterized by orcinol ferric chloride (Bial's reagent), ninhydrin, Dragendorff s reagent, and/or molybdenum blue stain supplied by the Sigma Chemical Co. Amberlite IR-120 (16-45 mesh) was obtained from
Fluka Chemical Co. and 200-400mesh AG 1-X2hydroxideform cation-exchange resin was supplied by Bio-Rad (Richmond, CAI. Analysis. FT-IR spectra were obtained on a Model 1800Perkin-Elmer Fourier transform spectrophotometer and data analyzed by a Model 7700 Perkin-Elmer data station (Perkin Elmer Corp., Norwalk, CT). Differential scanning calorimetry was performed using a Perkin-Elmer DSC7 (Perkin-Elmer Corp., Norwalk, CT). The transition temperatures, onsets, and enthalpies were calculated using the Perkin-Elmer thermal analysis software. The mass spectrometry was performed on a Finnigan triple quadrupole mass spectrometer (MAT, Inc., San Jose, CA). Residual moisture of samples was determined by the Karl Fischer titration utilizing the Brinkmann 684 KF coulometer (Metrohm Inc., Switzerland). Melting points were determined by an Electrothermal melting point apparatus and are uncorrected. Synthesis of TEC-Gal and TEC-MAL. TEC-GAL and TEC-MAL were synthesized according to the experimental procedure described by Goodrich et al. and Patel et al. (10,13). Cholesterylp-toluenesulfonate was refluxed with tetraethylene glycol in the presence of anhydrous dioxane. Acetobromo-a-D-galactoseor acetobromo-a-maltose in benzene was added to the sterol derivative in the presence of silver oxide, iodine, and 4-A powdered molecular sieves. The acetate groups of each carbohydrate moiety were removed with treatment of sodium methoxide. The suspension was neutralized when passed over an Amberlite IR-120 exchange resin. TEC-GAL and TECMAL were characterized by FT-IR,TLC, and mass spectra and agreed with previous results obtained by Goodrich et al. and Patel et al. (10, 13). Synthesis of DP-GAL. Synthesis of 11-Bromo-3,6,9trioxaundecanol. Phosphorous tribromide (8.928 g, 33 mmol) in 10mL of anhydrous dichloromethane was slowly added dropwise to a tetraethylene glycol (19.4 g, 100 mmol)/anhydrous dichloromethane (50 mL) suspension at 0 "C to ensure a low probability of synthesizing the dibromo product. The resultant light brown oil was purified over a silica gel column to give a pale yellow oil (20% yield, 5.14 8): R f (chloroform/methanol, 90:10, v/v) 0.82; IR (neat, cm-l) 3460, 2880, 1150, 670. Synthesis of 11-(Dimethylamino)-3,6,9-trioxaundecanol. A 60% solution of dimethylamine in water was slowly added to potassium hydroxide pellets with constant stirring at room temperature. The resultant dimethylamine was obtained by condensing the gaseous vapors in a side flask that was placed in a -10 to 0 "C dry ice/ethanol bath. Dimethylamine (2.3 g, 51 mmol) was added to ll-bromo3,6,9-trioxaundecanol (4.0 g, 16 mmol) and 1 mL of anhydrous toluene. The reaction mixture was sealed under nitrogen and placed in a 4-8 OC refrigerator overnight, during which time crystallization of white spindles in a yellow suspension occurred (82 % yield, 2.9 g): Rf (chloroform/methanol, 90:10, v/v) 0.27; IR (neat, cm-l) 3400, 2980-2750,2450,1475,1350,1250-1050; mp 205-208 "C; mass spectrum (FAB, glycerol) m/e = 222. Attachment of Acetobromo-a-D-galactose to ll-(Dimethylamino)-3,6,9-trioxaundecanol.Sodium hydride (40% by weight in oil, 55.2 mg, 2.3 mmol) was rinsed three times with 25 mL of anhydrous tetrahydrofuran (THF) and weighed. ll-(Dimethylamino)-3,6,9-trioxaundecanol (0.50 g, 2.3 mmol) in 5 mL of anhydrous THF was slowly added to the sodium hydride suspension. Acetobromoa-D-galactose (0.945 g, 2.3 mmol) was dissolved in 50 mL of anhydrous THF and slowly added to the alcoholreaction mixture under nitrogen at room temperature. This
208
Bloconjugate Chem., Vol. 3, No. 3, 1992
Testoff and Rudolph
crystallized by acetone precipitation to give beige crystls mixture was stirred constantly for 7 days at room tem(84% yield, 4.08 g): R f (chloroform/methanol/water, 65: perature. The crude product was purified over a silica gel 254,v/v) 0.64;IR (film, cm-l) 3500-3200,3000-2850,1735, column to give a pale yellow oil (39% yield, 0.495 g): Rj 1650,1475,1375,1250,1150-1050. (chloroform/methanol, 90:10,v/v) 0.76;IR (neat, cm-l) 3480,3OO~2850,1750,1650,1450,1375,1250,1150-1050, The acetylated glycolipid (4.112 g, 3.35 mmol) was 760;mp 92-95 "C; mass spectrum (FAB, glycerol) m / e = dissolved in ethanol/tetrahydrofuran (l:l,v/v) (33 mL of 109,127,169,264,331,543,552,594; (CI, isobutane) m / e solvent mixture/g of glycolipid) with gentle heating. The = 169,222,331,552. warm suspension of glycolipid was subjected to an AglX 2 exchange resin in ethanol (16 mL/glycolipid), the Synthesis of the Hydroxyl Phospholipid Derivative. method described by Chabala and Shen (19). After the Preparation of the enzyme extract phospholipase D from solvent was removed under reduced pressure, the crude white cabbage was performed as previously described (14). deacetylated glycolipid was recrystallized by acetone The protein content of the cabbage extract (2.6-3.0 mg/ precipitation to give light beige crystals (10% yield, 0.345 mL) was determined by Lowry's assay (15). The reaction parameters used to obtain the hydroxyl derivative of the g): Rj (chloroform/methanol/water, 65254,v/v) 0.14;IR phospholipid DPPC was described by Yang et al. (16). (film, cm-l) 3400-3200,3000-2850,1740,1560,1475,1440, DPPC (1.0g, 1.4mmol) was dissolved in 50 mL of diethyl 1150-1050. ether. In successive order, 3.4 mL of ethylene glycol, 20 Differential Scanning Calorimetry. DPPC was mL of enzymatic buffer (0.2 M sodium acetate, 0.08 M dissolved in chloroform a t the concentration of 25 mg/ calcium chloride, pH 5.6),and 34 mL of phospholipase D mL. The glycolipid containing the sterol derivatives TECextract (3 mg/mL) were added to the lipid suspension. GAL and TEC-MAL was taken up in chloroform a t a After 7 days, the reaction was quenched by removing the concentration of 25 mg/mL. The glycolipid containing solvent under reduced pressure and EDTA (0.05 M, pH the lipid derivative DP-GAL was dissolved in methanol/ 8.5) was added to bind calcium, prohibiting further water (2:1,v/v) at a concentration of 25 mg/mL. Samples enzymatic activity (14). After solvent removal, the crude were prepared by codrying the individual solutes with product was recrystallized by acetone precipitation to give DPPC at specific molar percentages as thin films into a white solid (85% yield, 0.823 g): Rj (chloroform/ previously weighed stainless steel pans under nitrogen on methanol/water, 65254,v/v) 0.84,IR (filmand KBR, cm-') a hot plate. Before deposition, methanol and water were 330O-32OO,296~2850,1730,1650,1470,1250,1150-1050, added to ensure complete solubility of all material of the 960. desired samples. Approximately 3 mg of each suspension Synthesis of the Brominated Phospholipid Derivative. containing the varying mole percentages of solute were Two methods were employed to brominate the hydroxyl dispensed into the stainless steel pans. To ensure complete lipid derivative. One method of bromination used reacted evaporation of solvents, the pans were then placed in a phosphorus tribromide with the hydroxyl derivative and vacuum oven at room temperature under nitrogen overthe second method was to treat the hydroxyl derivative night or until use. Each pan was sealed and weighed prior with trimethylsilyl bromide (13,14).Phosphorus tribroto scanning. The calorimetric scans were run at 2 "C/ mide (51.0mg, 0.19mmol) in 2 mL of anhydrous dichlomin. Duplicate scans were run for the samples. Enthalromethane was slowly added to the hydroxyl lipid derivpies and transition temperatures were calculated, norative of DPPC (0.40g, 0.578 mmol)/anhydrous dichloromalized (to the mass of DPPC present), and calibrated methane (5 mL) suspension under the experimental using multilamellar vesicles of DPPC. Determination of conditions described by Singh and Marchywka (17). After the percentage of water was performed by the Karl Fiscolumn purification, the crude material was recrystallized cher titration on DPPC samples containing TEC-GAL by acetone precipitation to produce pale white crystals and TEC-MAL. Aliquots of stock solutions of DPPC, (10% yield, 44.0 mg). TEC-GAL, and TEC-MAL (as described above) were mixed at specific molar concentrations. These samples Trimethylsilyl bromide was used as an alternative were codried as thin films under nitrogen and desiccated method to brominate the hydroxyl lipid derivative. Triuntil use. The DPPC/TEC-GAL and DPPC/TEC-MAL methylsilyl bromide (1.32g, 8.62mmol) was added directly mixtures were taken up in 100 pL of chloroform and to the hydroxyl lipid derivative of DPPC (3.0g, 4.34mmol) chloroform/methanol (1:2, v/v), respectively, at the total in 20 mL of anhydrous chloroform. The reaction was concentration of 75 mg/mL. The DPPC samples were allowed to proceed at 55 "C under nitrogen, as described taken up in 100pL of both solvent mixtures at the identical by Jung et al. (18). After column purification, the crude concentration for background comparison. Titrations were crystals were recrystallized by acetone precipitation to performed in triplicate using 10pL for each run. A solvent give pale white crystals (20% yield, 0.655g): Rj (chlorobackground was subtracted from the samples containing form/methanol/water, 65:25:4,v/v) 0.60;IR (film, cm-l) the appropriate solvent. 3000-2850,1740,1470,1250,1150-1050,720; mass spectrum (CI, isobutane) m/e = 335,409,471,647,753. Fourier Transform Infrared Spectroscopy. DPPC, DPPC-d62, TEC-GAL, and TEC-MAL were each susAttachment of the Galactose Linker to the Brominated pended in chloroform in a concentration of 25 mg/mL, as Phospholipid Derivative. The brominated lipid derivative described in the calorimetry section. The DP-GAL was of DPPC (3.0 g, 3.97 mmol) was dissolved in 30 mL of also suspended in methanol/water (2:1,v/v) at a concenanhydrous dimethylformamide with gentle heating. The tration of 25 mg/mL. The glycolipids were individually sugar-ll- (dimethylamino)-3,6,9-trioxaundecanol complex aliquoted at specific mole percentages with DPPC or (2.19g, 3.97 mmol) was dissolved in anhydrous dimethDPPC-de2. Methanol and water were added to desired ylformamide and added directly to the lipid mixture. The samples to ensure complete solubility of all material. The material was sealed and placed in a 60 "C oven for 7 days. Reactions were performed without anhydrous solvents but mixtures were codried and deposited as thin films on silver chloride, KRS-5, or calcium fluoride windows under were unsuccessful in producing the quaternary compound. The mixture was purified over a silica gel column and the nitrogen. Approximately 1-2 mg of each sample was solvent was removed under reduced pressure, which caused deposited. For complete evaporation of solvents, the crystallization of a light brown solid. The solid was resamples were placed in a vacuum oven at room temper-
Bioconjugate Chem., Voi. 3, No. 3, 1992 207
Modification of DPPC Phase Behavior
10
0
20
30
40
50
MOLE PERCENTAGES
Figure 3. Calorimetric effects of increasing mole percents of glycolipid on the phase transition of dry DPPC; scan rate, 2 "C/min: 0,TEC-GAL;0,TEC-MAL;A,DP-GAL; +,cholesterol. Table I. Enthalpic Changes in the Calorimetric Transition of Dry Mixtures of Cryoprotectant Glycolipids with DPPC
AH,J l g mol % glycolipid
TEC-GAL
TEC-MAL
DP-GAL
0
43.71 24.31 26.86 26.28 19.73
43.71 34.58 19.72 13.43 16.10 12.09 7.49
43.71 27.62 25.07 23.77 21.82 15.33
5 10 15 20 30 50
ature overnight or until use. The windows were placed in a thermostated, water-jacketed cell mount controlled by aRTE-110programmable water bath (Neslab Instruments, Inc., Newington, NH). The temperature was monitored by a thermocouple attached to the windows. The spectra (4000-450 cm-l) were collected through the temperature range of 25-80 "C at 2 "C intervals (scan rate of 1"C/min) over the transition temperature region. The data were analyzed after baseline corrections of specific regions, with frequencies and bandwidths (at half-absorbance) determined using a center of gravity algorithm (20, 21). Mass Spectroscopy. Mass spectra were determined by two techniques, fast atom bombardment (FAB) and desorption chemical ionization (CI) mass spectrometry. FAB mass spectrometry involved mixing the sample with a viscous matrix such as sodium iodide in the presence of glycerol and applying the sample directly to the tip of a copper probe. In the desorption chemical ionization process, the sample was coated on a copper wire and subjected to a chemical ionization reagent gas such as isobutane or ammonia. An electrical current was applied to the tip of the copper wire containing the sample, rapidly increasing the temperature, favoring desorption over decomposition. The desorbed molecules were protonated by the selected reagent gases. RESULTS The phase behavior of most biological lipids is dramatically influenced by the degree of hydration, due to their amphiphilic nature (22-26). The DSC scans of dry DPPC films in the present study resulted in a transition temperature of 71.2 "C, which suggests that this lipid is in a dihydrate form (24, 27, 28). Melting scans of dry DPPC films codried with TEC-GAL, TEC-MAL, and DPGAL revealed a decrease in the main chain melting transition temperature (max C,) with increasing mole percent of glycolipid (Figure 3). The largest decrease is observed with TEC-MAL at 30 mol % or greater where
2848 4 25
I 35
45 55 TEMPERATURE
65
75
(%)
Figure 4. The effect of the temperature-dependentfrequency of the CH2 symmetric stretching vibration of DPPC containing increasing mole percents of TEC-GAL 0 , 5 % ; .,lo%; A,15%; A, 20%;0,30%; B, 50%;v,dry DPPC. the transition temperature is reduced almost 20 "C from the transition temperature of the pure DPPC dihydrate. TEC-GAL showed nearly the same decrease in transition temperature (an approximate 18"C reduction in transition temperature at 20 mol %), while DP-GAL a t 30 mol 7% showed a 10 "C reduction in transition temperature. The phase transition of the glycolipid/DPPC systems also showed a decrease in enthalpy (AH) with increasing mole percentage of glycolipid (Table I). The observed reduction in the enthalpy of these films correlates with an increase in the transition width and thus reflects reduced cooperativity of the transition. The enthalpy of the dry DPPC film transition, 43.71 J/g, agrees with previous measurements for the melting of the dihydrate form (24,25,27, 28). TEC-GAL showed the greatest reduction in enthalpy at 5 mol %,with almost a 20 J/g reduction in the enthalpy of the transition. At higher mole percentages of glycolipid, the enthalpy was further reduced (19.73 J/g at 20 mol % for TEC-GAL, 7.49 J/g at 50 mol % for TEC-MAL, and 15.33 J/g at 30 mol % DP-GAL). Calorimetric scans of the pure glycolipids TEC-GAL, TEC-MAL, and DP-GAL did not show a cooperative phase transition at the sensitivity of the calorimeter in either the dry or hydrated state when tested through the same calorimetric temperature range. We have contrasted the affects of TEC-GAL, TECMAL, and DP-GAL with the calorimetric effect of increasing cholesterol in films codried with DPPC. A cooperative phase transition in the films was not observed after 5 mol '% (Figure 3). This concentration of cholesterol is far below the concentration required to inhibit the gel to liquid-crystalline phase transition of DPPC in excess water. Infrared spectroscopy has been used to monitor vibrational modes which are sensitive to the changes in the phase behavior of the two-component dry films. The frequency in the CH2 symmetric stretch of dry cast films of DPPC dihydrate shows a 1.4-cm-' increase in frequency over the temperature range in which the calorimetric transition is observed (approximately 63 to 77 "C) (Figure 4). The change in frequency observed in cast films of dry DPPC with increasing TEC-GAL begins at lower temperatures and occurs over a broader temperature range. For example, increasing the mole percentage of TEC-GAL to 30 mol '% results in an observed change in the CH2 symmetric stretch frequency, which begins a t 45 "C, compared to that of DPPC alone, which occurs at 65 "C. At lower TEC-GAL concentrations (5 and 10 mol %), the DPPC CH2 symmetric stretch of the dry films occurs over a narrower temperature range (47 to 57 "C). Above the phase transition temperature, at all concentrations of TEC-
200
Testoff and Rudolph
Bloconjugate Chem., Vol. 3, No. 3, 1992 2854
2098 1
I
t
% 0
u"
2093 2092 2091
LL
2090 2089 2848
75
35
45
55
65
75
2088 25
35
45 55 TEMPERATURE (%)
65
75
35
45 55 65 TEMPERATURE (OC)
75
TEMPERATURE e C )
Figure 5. The effect on the temperature-dependentfrequency of the CH2symmetric stretching vibration of DPPC containing increasing mole percents of TEC-MAL: 0 , 5 % ; 0,1076;A,15%; A, 20%; 0,30%; W, 50%; V, dry DPPC.
GAL examined the CH2 symmetric stretch is observed to be greater than the frequency observed in the pure DPPC films. In addition, above the phase transition, there is a slight increase (2 cm-l) in frequency with increasing temperature in the DPPC and TEC-GAL/DPPC mixtures. The infrared spectra of dry DPPC containing increasing mole percent of TEC-MAL also show a decrease in the temperature over which the change in CH2 symmetric is observed (Figure 5). The observed reduction is not as marked as is observed with TEC-GAL, which correlates with the calorimetric observation. Similar to TEC-GAL/ DPPC films, the TEC-MAL/DPPC films also show an increase in frequency above the phase transition, although in the TEC-MAL films this increase appears to be linearly related to the TEC-MAL concentration, with the maximal effect observed at 50 mol ?4 TEC-MAL. To study spectroscopically the effect of DP-GAL on the phase behavior of DPPC, we have used DPPC-d62 in the two-component system of DP-GAL/DPPC to monitor the acyl chain regions of the two individual components of the dry film. The CD2 symmetric stretch of the DPPCd62 cast films increased 4.5 cm-l (from 2089 to 2093.5 cm-l) over the temperature range of 54-63.5 "C (Figure 6A). This temperature range is much reduced compared to that of the perhydro-DPPC film. The CD2 symmetric stretch frequency above the transition temperature is increased compared to that of the pure DPPC-d62 films, similar to the effect observed with the TEC-GAL/DPPC and TECMAL/DPPC films. However, unlike TEC-GAL and TECMAL, the temperature over which the change in CD2 symmetric stretch frequency is observed in the DP-GAL/ DPPC-ds2 films is not significantly altered (Figure 6B). Changes in the CH2 symmetric stretch of the acyl chains of DP-GAL in the DP-GAL/DPPC films show that the temperature range over which the frequency increases is reduced by 13 OC compared to the range for DPPC alone (Figure 6B). The CH2 symmetric stretch for the DP-GAL acyl chains is also 1.8 cm-l higher (2851.8 cm-l) than the CH2 symmetric stretch for dry DPPC alone. This increase is observed throughout the entire temperature range examined. The change in frequency over the transition temperature range for the CH2 stretch of DP-GAL through the transition is 1.8 cm-l compared to 1.4 cm-' for the dry DPPC film. Notably, the acyl chains of DP-GAL in the deuterated DP-GAL/DPPC show a cooperative change in the CH2 symmetric stretch frequency despite the lack of an observable cooperative transition of DP-GAL in the DSC (Table I). A more direct measure of the interaction of TEC-MAL, TEC-GAL, and DP-GAL with the interfacial region of
2854,
4
2849 25
Figure 6. (A) The effect on the temperature-dependentfrequency of the CD2 symmetric stretching vibration of perdeuterated DPPC (DPPC-de2, 1,2-dipamitoyl-d62-sn-glycerol-3phosphocholine) containing increasing mole percents of DPGAL: 0, DPPC-ds2;.,lo%; A, 20%;A,30%;0 , 5 0 % . (B)The effect on the temperature-dependent frequency of the CH2 symmetric stretching vibration of dry DPPC and increasing mole percents of DP-GAL in a perdeuterated DPPC system: 0,10%; 0 , 20%;A, 30%; A, 50%; 0,dry DPPC.
DPPC is seen by examining potential sites of hydrogen bonding on the DPPC surface. One such site is the phosphate moiety, which has been previously suggested as the site of action of the bulk carbohydrates in dry lipid/ carbohydrate mixtures (2). Figure 7A-C shows a decrease in intensity of the asymmetric PO2 stretch (1250 cm-l) of dry DPPC with increasing mole percents of TEC-GAL, TEC-MAL, and DP-GAL. The reduction in intensity is accomplished by band broadening. In the case of DPGAL/DPPC (Figure 7C) there is a shift in frequency of the band to 1225 cm-1 with smaller shifts observed for TEC-GAL and TEC-MAL. The C-0-C stretch (11501000 cm-l) also shows a decrease in intensity as the mole percent of glycolipids is increased. Examination of this vibration in TEC-MAL shows an 8-cm-I shift in frequency a t 50 and 75 mol 76, indicating significant perturbation of the glycerol backbone by the sterol group. DISCUSSION The mixtures of dry DPPC with TEC-GAL, TEC-MAL, and DP-GAL have been examined to elucidate the effect of these glycolipids on the phase behavior of dry DPPC. The results from the DSC data indicate miscibility of the glycolipids with DPPC in the two-component system. It has been suggested that a decrease in enthalpy associated with a decreased transition temperature is indicative of reduced cooperative interactions of the acyl chains of the lipid system (29,30). The cooperativity of the transition in the DPPC mixtures of TEC-GAL, TEC-MAL, and DPGAL is markedly reduced with increasing concentration of the glycolipids. Increasing mole percents of DP-GAL
Bioconjugate Chem., Vol. 3, No. 3, 1992 209
Modlflcatlon of DPPC Phase Behavior
I
I
Y
V
z
4
m U
0 v)
m U
, 1300
ldw
1200
FREQUENCY (cm")
FREQUENCY (cm-')
1300
1Mo FREQUENCY (Cm-')
Figure 7. (A) FT-IR spectra of the asymmetric phosphate, P=O, stretch region (1250 cm-l) of dry DPPC containing increasing mole percents of TEC-GAL: A, dry DPPC; B, 10 mol % ; C, 20 mol % ;D, 50 mol % ; and E, 75 mol % . (B)FT-IR spectra of the asymmetric phosphate, P=O, stretch region (1250 cm-1) of dry DPPC containing increasing mole percents of TEC-MAL: A, dry DPPC; B, 10 mol % ; C, 20 mol %; D, 50 mol % ; and E, 75 mol %. (C)FT-IR spectra of the asymmetric phosphate, P=O, stretch region (1250 cm-1) of dry DPPC containing increasing mole percents of DP-GAL: A, dry DPPC; B, 10 mol % ; C, 20 mol %; D, 50 mol % ; and E, 75 mol %.
decreased the enthalpy and transition temperature (max C,) of DPPC to a lesser degree than TEC-GAL and TECMAL. The observed calorimetric differences in the reduction of enthalpy and transition temperature of DPGAL/DPPC mixtures as compared to the TEC-GAL/ DPPC and TEC-MAL/DPPC mixtures suggest that acyl chains of DP-GAL may cooperatively mix with the dry DPPC film. In the case of TEC-GAL/DPPC and TECMAL/DPPC, the sterol moiety intercalates into the bilayers of DPPC, reducing cooperativity of the melting of the acyl chains. The further reduction of the transition temperature of TEC-GAL/DPPC and TEC-MAL/DPPC mixtures compared to that of cholesterol/DPPC may indicate that the carbohydrate moiety of TEC-GAL and TEC-MAL may interact directly with the interfacial region of DPPC, enhancing the effect observed with the sterol alone. At low water concentration, some lipids exhibit highly ordered phases (23-25). For DPPC, the acyl chains become rigid, with a specific angle of tilt, and appear to be packed with some degree of rotational disorder (24). The addition of bulk water results in an increase in the surface area between phosphate head groups, decreasing the packing density of the hydrocarbon chains and decreasing the van der Waals attractions. Consequently, the transition temperature from the gel to liquid-crystalline phase decreases. Bulk carbohydrates have also been shown with calorimetry and FTIR to reduce the phase transition of dry DPPC ( I , 2 ) . Studies with X-ray diffraction and solid-state NMR (8,9) reveal that the acyl chains of DPPC/carbohydrate mixtures are disordered in the dry state, although there may be some restriction of acyl chain rotational motion. This has led to the designation of a new liquid-crystalline
phase, Lk (8,9). For example, previous studies have shown that at 0.45 g of trehalose/g of DPPC, a 1:l molar ratio, the phase temperature of dry DPPC resembles the transition temperature of fully hydrated DPPC (1,31). Further reduction in T, was observed when DPPC was lyophilized from a liquid-crystalline phase at 2 5 1 mole ratio of treha1oase:DPPC (6).The depression of the phase transition has been attributed to the hydrogen bonding of the disaccharide hydroxyl groups to the interfacial region of DPPC membranes, increasing the molecular area between phosphate head groups ( I , 4 ) . At the highest concentrations of TEC-MAL and TEC-GAL, a significant reduction in the transition temperature (approaching the hydrated transition temperature) was also observed. Changes in the CH2 stretch of the acyl chains of DPPC are indicative of changes in the trans/gauche conformers of the acyl chains. The increased frequency of the CH2 symmetric stretch as the temperature increases through the main chain melting temperature is indicative of increased gauche conformers. The partitioning of cholesterol in hydrated DPPC has been shown to increase the average number of gauche conformers (with an increase in the CH2 stretch frequency) in hydrocarbon chains in the gel phase and to decrease them in the liquid-crystalline phase (29,32-35). For mixtures of TEC-GAL/DPPC and TEC-MAL/DPPC, the temperature range over which additional gauche conformers are observed is increased, in agreement with the calorimetric observation of reduced cooperativity and transition temperature. The CH2 symmetric stretch frequency in the low-temperature phase of the dry mixtures of both TEC-GAL and TEC-MAL was not significantly different from that of DPPC alone. Above the phase transition, however, there was a concentration-
210
Testoff and Rudolph
Bloconlugete Chem., Vol. 3, No. 3, 1992
dependent increase in the frequency, indicating a larger number of gauche conformers and thus disorder in the high-temperature phase. We currently do not know how this effect is mediated, although future experiments will be aimed at determining the effect of different linker lengths and their mobility in the high temperature phase. These results agree with previous measurements of 30 % TEC-MAL incorporated into dry DPPC bilayers which showed a similar increase in the CH2 asymmetric stretch at a reduced T, compared to DPPC alone (11). The FT-IR spectra of the asymmetric phosphate stretch shows band broadening and a decrease in intensity as the mole percentage of the glycolipidsin dry DPPC is increased (Figure 7A-C). In addition, the frequency of this band shifts slightly to lower frequency. This effect is suggestive that the dielectric environment around the phosphate moiety is changed and could result from the hydrogen bonding of carbohydrate moieties of the glycolipids to the interfacial region of the bilayer. Alternatively, the tetraethoxy linkage could significantly change the dielectric environment of the phosphate. The changes observed in the phosphate infrared spectra of mixtures of TEC-GAL, TEC-MAL, and DP-GAL are consistent with previous measurements of 30 5% TEC-MAL/DPPC, which showed a shift to lower frequency when compared to measurements of dry DPPC alone (11). Raman spectra of the C-N asymmetric stretch were also consistent with an expanded interfacial region (11). One advantage to constructing lipids with cryoprotectant moieties may be the localization of the cryoprotectant moiety at the surface of the lipid surface. Thus, one might suppose that the concentration of carbohydrate required to mediate these effects may be less than that required for bulk carbohydrate action. For example, the optimal concentration of TEC-MAL found to reduce the phasetransition temperature of dry DPPC was approximately 50 mol 5%. The mole percentage of maltose per mole of lipid is 44.4 mol 5% . The mole percent of trehalose which lowered the phase-transition temperature of dry DPPC dihydrate to that of the hydrated DPPC was 87.3 mol % (I). Further reduction in the transition temperature was reached with anhydrous DPPC at somewhat lower mole ratios of trehalose (6). The additional bulk carbohydrates required to preserve membranes may be related to the need to form a glass. This may also be why the glycolipids do not show a reduction of the transition temperature down to the hydrated transition as do the bulk carbohydrates. Future experiments will be aimed at determining whether these glycolipids can form two-dimensionalglasses at the membrane surface. In addition, as the carbohydrates are known to be hygroscopic, we have measured residual moisture of the dry mixtures to examine whether the observed effects were based on differential water retention. The results indicate that the mixtures do not significantly differ in percent water t o percent lipid (approximately 3 % ) as compared to DPPC alone (data not shown). This is supported by previous measurements of the residual water content of dry liposomes preserved with bulk trehalose which demonstrate that the productive effect of the bulk sugar does not require retention of residual water (36). In summary, we have demonstrated the synthesis of three glycolipids that have cryoprotective carbohydrate moieties attached by a tetraethoxy linkage to either cholesterol or DPPC. Dry mixtures of DPPC and the glycolipids show some of the same effects as bulk carbohydrates in their calorimetric and spectroscopic effects. The utility of synthesizing new cryoprotectant groups may be
the controlled localization of the cryoprotectant at a fixed distance from the interfacial region of the lipid surface. Amphiphilic cryoprotectants such as these may also be easier to incorporate into membrane systems, imparting protection from freezing or desiccation. ACKNOWLEDGMENT
We gratefully acknowledge the support of the Office of Naval Research through its CORE funds in conjunction with the Naval Research Laboratory. We thank Dr. Barry Spargo, Dr. Alok Singh, and Dr. Beth Goins for their technical advice and helpful discussions and John Callahan for acquisition and analysis of the mass spectra. LITERATURE CITED (1) Crowe, L. M., Crowe, J. H., and Chapman, D. (1985) Interaction of Carbohydrates with Dry Dipalmitoylphosphatidylcholine. Arch. Biochem. Biophys. 236, 289-296. (2) Crowe, J.H., Crowe, L. M., and Chapman, D. (1984)Infrared spectroscopic Studies on Interactions of Water and Carbohydrates with a Biological Membrane. Arch. Biochem. Biophys. 232, 400-407. (3) Rudolph, A. S.,and Crowe, J. H. (1985)Membrane Stabilization during Freezing: The Role of Two Natural Cryoprotectants, Trehalose and Proline. Cryobiology 22, 367-377. (4) Crowe, J. H., Crowe, L. M., Carpenter, J. F., Rudolph, A. S., Wistrom, C. A., Spargo, B. J., and Anchordoguy, T. J. (1988) Interactions of Sugars with Membranes. Biochem. Biophys. Acta 947, 367-384. (5) Crowe, L. M., Crowe, J. H., Rudolph, A. S., Womersley, C., and Appel, L. (1985)Preservation of Freeze-Dried Liposomes by Trehalose. Arch. Biochem. Biophys. 242, 240-247. (6) Crowe, L. M., and Crowe, J. H. (1988)Trehalose and dry dipalmitoylphosphatidylcholinerevisited. Biochem. Biophys. Acta 946, 193-201. (7) Tsvetkova, N., Tenchov, B., Tsonev, L., and Tsvetkov, T. (1988)The Effect of Trehalose on Membrane Phospholipids Depends on the Initial Phase from Which Drying Occurs. Cryobiology 25, 256-263. ( 8 ) Lee, C.W. B., Waugh, J. S., and Griffin, R. G. (1986)SolidState NMR Study of Trehalose/l,2-Dipalmitoyl-sn-phosphatidylcholine Interactions. Biochemistry 25, 3737-3742. (9) Lee, C. W. B. Das Gupta, S. K., Mittai, J., Shipley, G. G., Abdel-Mageed, 0.H., Makriyannis, A., and Griffin, R. G. (1989) Characterization of the L Phase in Trehalose-Stabilized Dry Membranes by Solid-state NMR and X-ray Diffraction. Biochemistry 28, 5000-5009. (10) Goodrich, R. P., Handel, T. M., and Baldeschwieler, J. D. (1988)Modification of Lipid Phase Behavior with MembraneBound Cryoprotectants. Biochim. Biophys. Acta 938, 143-
154. (11) Goodrich, R. P., Crowe, J. H., Crowe, L. M., and Baldeschweiler, J. D. (1991) Alterations in Membrane Surfaces Induced by Attachment of Carbohydrates. Biochemistry 30, 5313-5318. (12) Goodrich, R. P., and Baldeschweiler, J. D. (1991)The Cryoprotective Action of Synthetic Glycolipids. Cryobiology 28, 327-334. (13) Patel, K. R., Li, M. P., Schuh, J. R., and Baldeschwieler, J. D. (1984)The Pharmacological Efficacy of a Rigid Non-Phospholipid Liposome Drug Delivery System. Biochim. Biophys. Acta 797, 20-26. (14) Eibl, H., and Kovatchen, S. (1981)Preparation of Phospholipids and Their Analogs by Phospholipase D. Methods Enzymol. 72, 632-639. (15) Lowery, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951)Protein Measurement with the Folin Phenol Reagent. J . Bio. Chem. 193, 265-275. (16) Yang, S.F., Freer, S., and Benson, A. A. (1967)Transphosphatidylation by Phospholipase D. J. Bio. Chem. 242, 477484.
Modification of DPPC Phase Behavior
(17) Singh, A., and Marchywka, S. (1989)Synthesis and Characterization of Headgroup Modified 1,3 Diacetylenic Phospholipids. Polym. Mater. Sci. Eng. 61,675-678. (18) Jung, M. E., and Hatfield, G. L. (1978) Preparation of Bromides from Alcohols via Treatment with Trimethylsilyl Bromide. Tetrahedron Lett. 46,4483-4486. (19) Chabala, J., and Shen, T. Y. (1978)The Preparatioin of 3-Cholesteryl6-(Glycosylthio)hexy1Ethers and Their Incorporation into Liposomes. Carbohydr. Res. 67,55-63. (20) Kauppinen, J. K., Moffatt, D. J., Cameron, D. G., and Mantach, H. H. (1981)Noise in Fourier Self-Deconvolution. Appl. Opt. 20, NO. 10,1866-1879. (21) Kauppinen, J. K., Moffatt, D. J., Mantach, H. H., and Cameron, D. G. (1981)Fourier Self-Deconvolution: A Method for Resolving Intrinsically Overlapped Bands. Appl. Spectrosc. 65,NO. 3, 271-276. (22) Small, D. M. (1986)Handbook of Lipid Research: The Physical Chemistry of Lipids (D. J. Hanahan, Ed.) Chapter 12,pp 475-485, Plenum Press Corp., New York. (23) Luzzati, V., and Tardieu, A. (1974)Lipid Phases: Structure and Structural Transitions. Annu. Rev. Phys. Chem. 25,7994. (24) Tardieu, A.,Luzzati, V., and Reman, F. C. (1973)Structure and Polymorphism of the Hydrocarbon Chains of Lipids: A Study of Lecithin-Water Phases. J. Mol. Biol. 75,711-733. (25) Chapman, D. (1983)Physicochemical Properties of Phospholipids and Lipid-Water Systems. Liposome Technology (G. Gregoriadis, Ed.) Chapter 1, Vol. 1,pp 1-18, CRC, Boca Raton, FL. (26) Hauser, H., Pascher, I., Pearson, R. H., and Sundell, S. (1981) Preferred Conformatioinand Molecular Packing of Phosphatidylethanolamineand Phosphatidylcholine. Biochim.Biophys. Acta 650,21-51. (27) Fringeli, U. P. (1981)A New Crystalline Phase of L-a-DipalmitoylphosphatidylcholineMonohydrate. Biophys. J.34, 173-187.
Bloconjugate Chem., Voi. 3, No. 3, 1992 211
(28) Phillips, M. C., Williams, R. M., and Chapman, D. (1969) On the Nature of Hydrocarbon Chain Motions in Lipid Liquid Crystals. Chem. Phys. Lipids 3, 234-244. (29) Brauner, J. N., and Mendelsohn, R. (1986)A Comparison of Differential Scanning Calorimetric and Fourier Transform Infrared Spectroscopic Determination of Mixing Behavior in Binary Phospholipid Systems. Biochim. Biophys. Acta 861, 16-24. (30) Mabrey, S.,Mateo, P. L., and Sturtevant, J.M. (1978)HighSensitivity Scanning Calorimetric Study of Mixtures of Cholesterol with Dimyristoyl- and Dipalmitoylphosphatidylcholines. Biochemistry 17,2464-2468. (31) Tsvetkov, T. D., Tsonev, L. I., Tsvetkova, N. M., Koynova, R. D., and Tenchov, B. G. (1989)Effect of Trehalose on the Phase Properties of Hydrated and Lyophilized Dipalmitoylphosphatidylcholine Multilayers. Cryobiology 26, 162169. (32) Umemura, J., Cameron, D.G., and Mantach, H. M. (1980) A Fourier Transform Infrared Spectroscopic Study of the Molecular Interaction of Cholesterol with 1,2-Dipalmitoyl-snglycerol-3-phosphocholine. Biochim. Biophys. Acta 602,3244. (33) Gennis, R. B. (1989)Biomembranes: Molecular Structure and Function (C. R. Cantor, Ed.) Chapters 2 and 5,pp 36-84 and 166-198, Springer-Verlag New York, Inc., New York. (34) Casal, H. L., and Mantach, H. H. (1984)Polymorphic Phase Behaviour of Phospholipid Membranes Studied by Infrared Spectroscopy. Biochim. Biophys. Acta 779, 381-401. (35) Davies, M. A.,Schuster, H. F., Brauner, J. W., and Mendelsohn, R. (1990)Effects of Cholesterol on Conformational Disorder in Dipalmitoylphcaphatidylcholine Bilayers. A Quantitative IR Study of the Depth Dependence. Biochemistry 29,4368-4373. (36) Crowe, J. H.,Spargo, B. J., and Crowe, L. M. (1987) Preservation of Dry Liposomes Does Not Require Retention of Residual Water. Proc. Natl. Acad. Sci. U.S.A. 84, 15371540.
