Langmuir 2007, 23, 1359-1364
1359
Thermotropic Behavior of Ceramides and Their Isolation from Wool Sandra Me´ndez, Meritxell Martı´, Clara Barba, Jose´ Luis Parra, and Luisa Coderch* IIQAB (CSIC), Jordi Girona 18-26, 08034 Barcelona, Spain ReceiVed July 21, 2006. In Final Form: October 27, 2006 The composition of internal wool lipids (IWL) resembles that of lipids present in other keratinic tissues such as human hair or the stratum corneum. Advances in the isolation of ceramides from wool and in the characterization of their thermotropic properties could facilitate their application in human skin care treatments. IWL are solvents extracted from wool fibers. Ceramide isolation is carried out by medium-pressure liquid chomatrography. The different fractions obtained were analyzed quantitatively by thin layer chromatography coupled to an automated flame ionization detector and by high-performance thin layer chromatography using a densitometric detector. Two important fractions were isolated: one was a mixture of different ceramides and the other was exclusively made of ceramide 2 (nonhydroxy acid sphingosine [NS]). The thermotropic behavior of IWL and their isolated fractions were studied by thermogravimetric analysis, differential scanning calorimetry (DSC), and by attenuated total reflection Fourier-transform infrared spectroscopy (ATR-FTIR) methodologies. The transition temperature (Tm) obtained was compared with the results of the IWL extract, stratum corneum lipids, and the values found in the bibliography for isolated ceramides. The Tm obtained for IWL (48 °C) was lower than that achieved for SCL (65 °C). This discrepancy could be due to the different ceramide pattern and to the larger amounts of free fatty acids present in the IWL extract. Although the isolated ceramides had higher Tm values, they resembled the values reported in the bibliography. The suitability of the fraction composed exclusively of ceramide 2 [NS] for skin applications was confirmed by ATR-FTIR and DSC.
Introduction Wool is a natural fiber mainly composed of protein with an external lipid content (lanolin) and a minor internal lipid content (1.5%). Internal wool lipids (IWL) are rich in cholesterol, free fatty acids, cholesteryl sulfate, cerebrosides, and ceramides. These lipids resemble in composition the ones present in other keratinic tissues such as human hair or the stratum corneum (SC).1,2 Their composition yields a highly ordered arrangement of lipids known as lamellar lipid bilayers. The intercellular lipids of the SC, especially ceramides, play an important role in the barrier function of the skin, preventing the penetration of external agents and controlling the transpidermal water loss to maintain the physiological skin content of water.3 Recent studies4,5 have demonstrated that the topical application of liposomes composed of IWL on disturbed skin offers a more satisfactory recovery of water barrier function than stratum corneum lipids. These studies also show that the application of these liposomes on intact skin reinforces barrier integrity, thereby improving its water-holding capacity. The lamellar lipid bilayers are responsible for the permeability of water through the hydrophobic barrier of the SC. The physical state of the nonpolar regions of the hydrocarbon chains plays a key role. This is due to the high melting point of the aliphatic chains of the ceramides and free fatty acids. Most lipid chains are in trans configuration, which is characteristic of the solid crystal state, at physiological temperature; there is a low lateral diffusion that results in a low permeability because of the interaction between adjacent hydrophobic chains. On the other * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Coderch, L.; Soriano, S.; de la Maza, A.; Erra, P.; Parra, J. L. J. Am. Oil Chem. Soc. 1995, 72, 715-720. (2) Schaefer, H.; Redelmaier, T. E. Skin Barrier: Principles in Percutaneous Penetration; Karger: Basel, 1996; pp 55-58. (3) Elias, P. M. Arch. Dermatol. Res. 1981, 270, 95-117. (4) De Pera, M.; Coderch, L.; Fonollosa, J.; De la Maza, A.; Parra, J. L. Skin Pharmacol. Appl. Physiol. 2000, 13, 188-195. (5) Coderch, L.; de Pera, M.; Fonollosa, J.; de la Maza, A.; Parra, J. L. Contact Dermatitis 2002, 47, 139-146.
Figure 1. Scheme of the experimental section of this work.
hand, some lipid chains change to a gauche configuration phase when the temperature rises, forming a gel state phase with a small increase in the diffusion. Most of the lipid chains in the crystalline liquid phase are in a gauche configuration. Given that there is no specific packing, the lateral diffusion is faster.6 Intercellular wool lipids arranged as concentrated liposomes were shown to be a good intercellular lipid model, allowing their study by X-ray diffraction techniques.7 The lipid conformation and molecular organization in biological membranes and lipid systems can also be followed up by using physical techniques, such as attenuated total reflection Fourier transform infrared (ATR-FTIR) spectrometry and differential scanning calorimetry (DSC). CH stretching frequencies have been the subject of ATR(6) Forslind, B.; Engstro¨m, S.; Engblom, J.; Norlen, L. J. Dermatol. Sci. 1997, 14, 115-125. (7) Fonollosa, J.; Campos, L.; Martı´, M.; de la Maza, A.; Parra, J. L.; Coderch, L. Chem. Phys. Lipids 2004, 130, 159-166.
10.1021/la0621315 CCC: $37.00 © 2007 American Chemical Society Published on Web 12/06/2006
1360 Langmuir, Vol. 23, No. 3, 2007
Me´ ndez et al.
Table 1. Gradients Used by MPLC To Separate the Ceramides Procedure A: MPLC Using 3 Successive Eluents for 58 min F
t (min)
AF1 AF2
10 2 10 4 32
AF3 AF4
% chloroform
% acetone
100 50
50 100 50
% methanol
50 100
Procedure B: MPLC Using 2 Successive Eluents for 61 min F
t (min)
% chloroform
BF1 BF2
15 5 5 2 2 2 2 2 2 24
100 99.5 99 98 95 90 80 50 25
BF3
% methanol 0.5 1 2 5 10 20 50 75 100
FTIR studies. These frequencies are sensitive to the relative population ratio of trans and gauche isomers along the alkyl chain, yielding the transition temperature (Tm) of the two phases. A shift in peak absorbance toward higher wavenumbers indicates a change in the conformer population, increasing the number of gauche conformers; this phenomenon can be related to the greater freedom and greater fluidity of the system.8 DSC, FTIR, and electron paramagnetic resonance (EPR) methodologies have been applied to IWL extracts structured in liposome vesicles as a model for a wool lipid membrane, demonstrating that the membrane is less permeable at the surface hydrophilic ends than in its hydrophobic core.9 In terms of thermotropic behavior, the IWL membrane presents two phase transitions: a main phase transition at 40 °C (Tm), produced by a loss of the conformational order, increasing the fluidity of the membrane with the disorder; and a second minor and broader irreversible transition phase from 60 to 65 °C 9. Although most ceramides used in cosmetic or dermatological formulations are synthesized by biotechnological methods, they do not have the composition or diversity found in keratinized tissues such as skin, hair, or wool. Recent physicochemical studies based on intermolecular and intramolecular lipid organization provide a basis for studies of the specific role that each ceramide species have in SC organization and function.10,11 This demands the preparation of an equilibrated mixture of ceramides with a composition and molar proportions similar to the ones present in skin.12 Isolation of the ceramides from the other lipidic compounds, especially ceramide 2 [NS], which is found in large quantities, is important not only for their use as an active principle in cosmetic and dermatologic preparations but also for the study of its role in the lamellar structures of the keratinized tissues. HPLC coupled to mass spectrometry is a sensitive and selective method for the analysis of ceramides.13,14 A sample rich in ceramide is necessary to perform this kind of study. For this (8) Coderch, L.; Bondı´a, I.; Fonollosa, J.; Me´ndez, S.; Parra, J. L. IFSCC 2003, 6, 117-123. (9) Fonollosa, J.; Martı´, M.; Sabe´s, M.; de la Maza, A.; Parra, J. L.; Coderch, L. Langmuir 2000, 16, 4808-4812. (10) Moore, D. J.; Rerek, M. E. Proc. Cos. Sci. Conf. (Barcelona) 2000, 1, 36-43. (11) Chen, H.; Mendelsohn, R.; Rerek, M. E.; Moore, D. J. Biochim. Biophys. Acta 2000, 1468, 293-303. (12) Coderch, L.; Fonollosa, J.; Martı´, M.; Garde, F.; De la Maza, A.; Parra, J. L. J. Am. Oil Chem. Soc. 2002, 79, 1215-1220. (13) Vietzke, J. P.; Strassner, M.; Hintze, U. Chromatographia 1999, 50, 15-20. (14) Han, X. Anal. Biochem. 2002, 302, 199-212.
Table 2. Compounds Found in the Quantitative Analysis of IWL Extract (Cholesterol Ester (Chol-Est), Triglycerides (TG), Free Fatty Acids (FFA), Fatty Alcohol (R-OH), Cholesterol (Chol), Ceramide (Ceram), Glycosil Ceramides (GC), and Cholesterol Sulfate (CholS)) component
IWL (%)
Chol-Est TG FFA R-OH Chol Ceram. 2 [NS] Ceram. 3 [NP], 5 [AS], 7 [AP] and 7-OH-Chol GC CholS
3.0 3.1 11.2 4.0 15.0 16.9 16.8 0.6 10.8
reason, we want to isolate ceramides by MPLC. Some studies use techniques like high-performance liquid chromatography (HPLC),15 gas chromatography,16 or a specific enzymatic method17 to quantify ceramides but these have a long and difficult preparation and derivatization of the samples. In this work, this quantification was carried out by TLC-FID and HPTLCdensitometry without sample preparation before the analysis. Therefore, the main aim of this work was the preparation of isolated ceramides from a natural source (IWL) using liquid chromatography. The resulting samples were evaluated by thermogravimetric analysis (TGA), DSC, and ATR-FTIR to study their thermotropic behavior. Experimental Section Raw Spanish Merino wool samples supplied by SAIPEL (Terrassa, Spain) were industrially cleaned before the solvent extraction. Samples of 7 kg of raw wool were extracted at a pilot plant level. The extraction procedure consisted of a pump-forced reflow system and took 3 h at 50 °C with a wool/methanol 1/25 solvent ratio. The solvent was finally distilled to obtain 1 L of IWL extract. Before separation, the extract was evaluated by quantitative (TLC-FID and HPTLC-densitometry) and qualitative (TGA, ATR-FTIR, and DSC) methods for comparison with the isolated fractions enriched in ceramides. A short scheme of this work is shown in Figure 1. Medium-Pressure Liquid Chromatography. The ceramide isolation was carried out by medium-pressure liquid chromatography (MPLC). This technique demands the use of a big column (15-100 mm of diameter). The size of the injected sample should be between 10 mg and 1 g (bigger than that in normal liquid chromatography) and the use of a precolumn is necessary to ensure that the sample application and the solvent flow are linear over time. The instrument of MPLC consists of a pump (B-688, Bu¨chi, Switzerland) and a gradient former (B-687), pre-column, and column (B-685) filled up with silica gel with pore size of 60 Å, UV-vis detector (Knauer number A0297) and registrator (Labograph E586 Methrohm), and finally, a fraction collector (B-684) to collect the different separated fractions. The experimental conditions used were as follows: a registrator sensitivity of 200 mV, pressure of 6-8 bar, and a flow between 10 and 20 mL/min. One gram of the dry IWL extract was dissolved in 15 mL of chloroform before being applied in the column. Two different gradients called procedures A and B, respectively, were used to separate the ceramides; these are shown in Table 1. At the end of the experiment, the columns were cleaned with pure methanol for 15 min. The different fractions obtained by MPLC were grouped together after carrying out a qualitative analysis by thin layer chromatography. This grouping was based on the isolation of ceramide compounds. (15) Camera, D.; Picardo, M.; Presutti, C.; Catarcini, P.; Fanali, S. J. Sep. Sci. 2004, 27 (12), 971-976. (16) Tserng, K.; Griffin, R. Anal. Biochem. 2003, 323, 84-93. (17) Bektas, M.; Jolly, P. S.; Milstien, S.; Spiegel, S. Anal. Biochem. 2003, 320, 259-265.
BehaVior of Ceramides and Their Isolation from Wool
Langmuir, Vol. 23, No. 3, 2007 1361
Figure 2. Plate of the collected fractions and standards (S) using procedure A (A) and procedure B (B) by MPLC. The standards used were cholesteryl ester (CholEst), cholesterol (Chol), free fatty acids (FFA), ceramides (Cer), glycosil ceramides (GCer), and cholesteryl sulfate (CholS). Table 3. Quantitative Analysis of Lipid Fractions (Procedures A and B) by TLC-FID and HPTLC-Densitometrya % l.i.f IWL extract AF1 AF2 AF3 AF4 A-total BF1 BF2 BF3 B-total
34.2 12.6 9.0 34.0 89.8 25.2 3.1 43.7 72.0
Ester Chol
TG
FFA
R-OH
Chol
3.0 1.8
3.1 1.4
11.2
4.0 3.6
15.0 8.0
1.8 0.7
1.4 1.1
0.7
1.1
6.6 6.6 8.5 8.5
3.6 1.6
8.0 3.1
1.6
3.1
Ceram. 2 [NS] 16.9 5.4 5.2 0.06 10.7 1.8 2.3 0.4 4.1
Ceram. 3 [NP],5 [AS], 7[AP], OH-Chol
GCS
CholS
% l.a.f
16.8 4.7 3.3 0.7 7.0 15.7 3.9
0.6
10.8
1.0 1.0
0.8 3.6 4.4
15.4 19.1
0.5 0.5
1.8 1.8
81.4 24.9 8.5 2.6 17.2 53.2 12.2 2.3 26.6 41.1
a See abbreviation descriptions in Table 2. “l.i.f” is the percentage with respect to the total lipid isolated and “l.a.f” the lipids analyzed with respect to the lipid extract.
For procedure A, the first fractions composed of cholesterol esters, triglycerides, and cholesterol were pooled into F1, the fractions with ceramides into F2, the fractions with free fatty acids, ceramides, glycosil ceramides, and cholesterol sulfate into F3, and free fatty acids, ceramides, and cholesterol sulfate into F4. For procedure B, the F1 contained cholesterol esters, triglycerides, and cholesterol. F2 consisted mainly of ceramide 2 [NS], and F3 contained ceramides, free fatty acids, glycosil ceramides, and cholesterol sulfate. All these fractions obtained in this way were analyzed quantitatively. The nonpolar lipids were measured by TLC-FID whereas the HPTLCdensitometry was used to evaluate the polar lipids. Qualitative and Quantitavive Analysis of Fractions Obtained. Qualitative analysis was carried out by TLC. The samples and the standards were applied with an automatic device (Linomat 5, Camag, Muttenz, Switzerland) in a plate (Merck, Darmstadt, Germany, Silicagel 60 10 × 10 cm). The experiment was performed using two different solvents: CHCl3/MeOH (9:1) and CHCl3/diethyl ether/ ethyl acetate (80:4:16). The chromatographic layer was soaked in 20 mg/mL of CuSO4‚5H2O solution in water for 2 min and then dried at 160 °C for 15 min for lipid detection. The quantitative analyses of nonpolar lipids from IWL extract as well as the different fractions were performed by thin layer chromatography coupled to an automated ionization detector (Iatroscan MK-5 analyzer, Iatron, Tokyo, Japan). Samples were applied on Silica gel S-III Chromarods using a SES (Nieder-Olm, Germany) 3202/15-01 sample spotter. Before performing a total scan, the rods were developed to a distance of 2.5 cm with chloroform/ methanol/water (57:12:0.6), after this to 8 cm with hexane/diethyl ether/formic acid (50:20:0.3), and finally to 10 cm with hexane/ benzene (35:35). High-performance thin layer chromatography (HPTLC) and densitometry were used to evaluate the polar lipids of the IWL extract and the separated fractions by a Phosphoimager Fluor-S Multimager of Bio-Rad; the plate was scanned and then analyzed by the Quantity One software. The plate was developed with CHCl3/ MeOH (9:1) and CHCl3/diethyl ether/ethyl acetate (80:4:16) and finally soaked in CuSO4‚5H2O for 2 min and dried at 160 °C for 15 min.
Thermotropic Analyses of Ceramide-Rich Fractions. The samples enriched in ceramides were analyzed by TGA, DSC, and FTIR-ATR to determine the thermotropic behavior of ceramides and compare their transition temperature (Tm) with those of other authors.18-20 Thermogravimetric analysis (TGA) of the IWL extract and lipidic fractions rich in ceramides were performed using a TG-50 Analyser Mettler Toledo (Columbus, OH). Around 6 mg of sample was prepared by evaporation to dryness and acclimatization in 65% RH for 24 h. A thermogram was performed from 30 to 700 °C at 10 °C/min with N2 at 200 mL/min. Fourier transform infrared spectroscopy (ATR-FTIR) of the IWL extract and lipidic fractions rich in ceramides were performed using a 360-FTIR spectrophotometer Nicolet Avatar (Nicolet Instruments, Inc., Madison, WI). Approximately 60 mg of lipid films was prepared by evaporation from the chloroform/methanol solution (2:1) in a 45° ZnSe thermal horizontal attenuated total reflection (ATR) crystal mounted in a trough plate placed in a temperature-controlled to carry out the FTIR experiments. These films were hydrated by covering them with 2 mL of TRIS buffer, pH 7.5. To induce lipids to achieve their natural organization, the temperature was changed from the refrigerator (7 °C, 1 h) to room temperature (20 °C, 2 h). Spectra were generated by co-addition of 256 interferograms in 80 s collected at 2 cm-1 resolution. IR spectra were acquired between 20 and 80 °C at 10 °C intervals before subtracting the buffer signal. High-sensitivity differential scanning calorimetry (DSC) was used to study the behavior of the IWL extract and fractions enriched in ceramides by means of the temperature using a DSC III Setaram (Calvire, France). The dried sample was heated from 0 to 100 °C, then cooled, and reheated. (18) Krill, S. L.; Knutson, K.; Higuchi, W. I. J. Controlled Release 1993, 25, 31-42. (19) Glombitza, B.; Mu¨ller-Goymann, C. C. Chem. Phys. Lipids 2002, 117, 29-44. (20) Shah, J.; Atienza, J. M.; Rawlings, A. V.; Shipley, G. G. J. Lipid Res. 1995, 36, 1945-1955.
1362 Langmuir, Vol. 23, No. 3, 2007
Me´ ndez et al.
Figure 4. Thermotropic responses of IWL extract (A), AF2 (B), and BF2 (C) by ATR-FTIR. Figure 3. Thermotropic responses of IWL extract (A), AF2 (B), and BF2 (C) by TGA.
Results and Discussion Lipid Analysis and Ceramide Isolation by MPLC. Based on the quantitative analyses, the main components of IWL were ceramides, cholesterol, and free fatty acids (see Table 2). The ceramides represented approximately 30% of the total lipids, ceramide 2 [NS] being the most abundant (16.9%). Consequently, IWL extract should be regarded as an important natural source of ceramides. Isolation of the ceramides from IWL extract was carried out with a chromatographic method, MPLC. First, a gradient with chloroform, acetone, and methanol (procedure A) was used and 46 different fractions were collected, although the UV/vis detection was only followed up in the first and last fractions. With procedure B, acetone was eliminated given that its use as a solvent hindered the follow-up of separation by UV/vis. Moreover, ceramides have been separated by using only
chloroform and methanol.21 Accordingly, a gradient of these eluents (procedure B) was used to improve the ceramide separation. Eighteen different fractions were collected without interference in the UV/vis spectrum in this case. Qualitative Analysis. Qualitative analyses of the most colored fractions obtained under procedure A were performed by thin layer chromatography TLC, the resulting plate being shown in Figure 2A. The nonpolar compounds (cholesterol, cholesterol esther, triglycerides, and some derivates of cholesterol) were the first compounds that emerged from the column following separation with 100% chloroform. Using a gradient of acetone/ methanol, ceramides, FFA, 7-OH-cholesterol, and the most polar lipids (GCS and CHolS) were fractioned. The new group of fractions consisted only of four fractions (AF1, AF2, AF3, and AF4) and is also shown in Figure 2A. The formation of these groups depended on the chemical nature of their compounds. AF1 contains the nonpolar compounds and AF3 and AF4 are (21) Oku, H.; Mimura, K.; Tokitsu, Y.; Onaga, K.; Iwasaki, H.; Chinen, I. Lipids 2000, 35, 373-381.
BehaVior of Ceramides and Their Isolation from Wool
Langmuir, Vol. 23, No. 3, 2007 1363
Figure 5. Thermotropic responses of IWL extract (A), AF2 (B), and BF2 (C) by DSC.
made up of the most polar compounds. AF2 is the most important fraction given in that it contains ceramides 2, 3, 5, 7, and 7-OHChol. Qualitative analyses of the most colored fractions obtained under procedure B were also performed by TLC. The plate is shown in Figure 2B, where the cholesterol esther, cholesterol, triglycerides, and derivatives of cholesterol appear in the first fraction; these compounds were separated by 100% chloroform. With the CHCl3/MeOH 99.5:0.5, ceramide 2 [NS] was isolated from the polar compounds and the other ceramides, which were extracted first with a gradient of CHCl3/MeOH and then with 100% MeOH. In light of these results, it was decided to regroup these fractions into only three (BF1, BF2, and BF3) depending on their chemical composition. This separation is shown in Figure 2B: BF1 is made up of nonpolar compounds, BF3 consists of polar compounds (7-OH-Cholesterol, FFA, ceramides, GCS, and CholS) and BF2 is the most important fraction because it is exclusively made up of ceramide 2 [NS]. Quantitative Analysis of Fractions. All fractions obtained by MPLC under the two procedures A and B were quantitatively evaluated by TLC-FID and HPTLC-densitometry. The percentages of the different lipidic compounds in each fraction are listed in Table 3 where the “l.i.f” accounts for the total amount of lipids isolated in each fraction and “l.a.f” accounts for the lipid analyzed in each fraction by TLC-FID and HPTLC-densitometry. The best isolation of ceramides was obtained in fraction AF2 with procedure A and in fraction BF2 for procedure B. Even though AF2 accounts for a larger amount of ceramides, almost 8.5%, the best isolation of only ceramide 2 [NS] was obtained in fraction BF2. Thermotropic Study of the Lipidic Samples by TGA, ATRFTIR, and DSC. In this work the thermotropic behavior of the IWL extract and the different fractions rich in ceramides AF2 and BF2 obtained by MPLC were studied by TGA, ATR-FTIR, and DSC methodologies. Thermogravimetrical Analysis. The weight loss of the sample with the increase in temperature can be followed by thermogravimetrical analysis. In Figure 3, a different pattern was obtained for the TGA when the IWL extract was analyzed with respect to the AF2 and BF2 fractions of procedures A and B, respectively. Three main decomposition temperatures were obtained for the initial sample of the IWL extract of about 120, 363, and 476 °C. However, fractionation of the lipidic compounds results in a unique main decomposition diagram at 362 °C for AF2 and 356 °C for BF2 in which only ceramides are present. Therefore, this decomposition temperature of about 360 °C may be due to the ceramides. Spectroscopic Analysis of Peak Shifts. The thermal behavior of hydrated lipid films of IWL extract was investigated by ATRFTIR, taking into account the asymmetric and symmetric CH
stretching frequencies (about 2920 and 2850 cm-1, respectively) of the CH2 groups. An increase in the value of these frequencies, from 2919 and 2848 to 2922 and 2853 cm-1, respectively, was observed when the temperature increased. Moreover, an inflection point was observed when these wavelengths were plotted versus temperature (Figure 4A), indicating the transition temperature (Tm) of the IWL. It seems that a decrease in chain packing in the hydrocarbon chains had occurred: at low temperatures the hydrocarbon lipid chains are in trans conformation and when T exceeds Tm, their conformation changes to gauche.8,18 Then, there is a transition from a gel state to a liquid crystal phase as temperature increases. The Tm found for the IWL extract was 48 °C, which is lower than the one obtained for SCL (stratum corneum lipids), 65 °C.8,22 This difference could be due to the different ceramide pattern and also to the higher amounts of FFA present in the IWL extract. The AF2 fraction obtained with procedure A, which is rich in ceramides but also contains a cholesterol derivative, was studied by ATR-FTIR. An increase in vibration frequencies of CH2 was observed (Figure 4B) with the temperature. Furthermore, two inflection points were obtained: the first (30-35 °C) could be due to a change from a metastable to a more stable phase.23 The second peak (59 °C) corresponded to the phase transition temperature of the hydrocarbon lipid chains. The Tm of this ceramide mixture is higher than the Tm of the IWL extract (48 °C). The absence of FFA and Chol in the separated fraction accounts for this difference, although the Tm of these ceramides is still lower than the value found in the literature, between 60 and 90 °C depending on the kind of ceramides present in the lipidic fraction.18 The same experiment was performed with procedure B for the BF2 fraction, which was enriched in ceramide 2 [NS]. Again, an increase was observed in the vibration frequencies of CH2 with the temperature. There were two inflection points similar to those detected for AF2 (Figure 4C). Again, the first inflection point could correspond to the nonstable transition whereas the second inflection point, 77 °C, could be related to the Tm value for the CH2 of lipid chains. This value is the highest one obtained in this work. A value of approximately 80 °C was found in the literature13 for N-ceramides (like ceramide 2 [NS]). Analysis by Differential Scanning Calorimetry. Finally, the thermotropic behavior of dried lipids was studied by differential scanning calorimetry (DSC). The changes in enthalpy detected between 0 and 100 °C are shown in Figure 5A for the IWL extract; on first heating, the sample showed two endothermic transitions at 21.3 and 46.2 °C. The second value (46.2 °C) agrees with the Tm value of IWL extract yielded by ATR-FTIR. (22) Lo´pez, O.; de la Maza, A.; Coderch, L.; Parra, J. L. J. Am. Oil Chem. Soc. 1996, 73, 443-448. (23) Mimeault, M.; Bonenfant, D. Talanta 2002, 56, 395-405.
1364 Langmuir, Vol. 23, No. 3, 2007
This technique was unable to detect the first value (21.3 °C) because the experiment was performed at a temperature higher than 20 °C. The subsequent cooling process shows two exothermic transitions at 33.1 and 39 °C. On reheating of the sample, only one endothermic transition was detected at 39.5 °C. This confirms the reversibility of this phase transition. Thus, the reversible phase transition of IWL took place at about 47 °C using both techniques. The DSC curves of the AF2 fraction are shown in Figure 5B. On the first heating thermogram, two endothermic transitions were detected: 46.4 and 68.9 °C. In principle, the second peak should correspond to the transition from a gel phase to a crystalline liquid phase, but this value is lower than the typical Tm of ceramides (∼70-130 °C19). The presence of cholesterol derivatives could account for the behavior detected by ATR-FTIR. Only a single exothermic transition peak at 51.9 °C appears on the cooling run and on the reheating thermogram of the sample. Two transitions were observed at 54.2 and 69.4 °C, which indicates the reversibility of the process. The differences observed between the values yielded by ATR-FTIR and DSC could be due to the different treatment of the lipidic sample, i.e., hydrated in the former technique and dry in the latter. The BF2 fraction was also studied by DSC and the curves obtained are shown in Figure 5C. Three peaks appeared on the first heating: 37.5, 69.6, and 81.2 °C. The 69.6 and 81.2 °C peaks are common in the N-ceramides:15 the 69.6 °C corresponds to a solid-solid transition, this peak being irreversible because it disappeared after cooling of the sample. The 81.2 °C indicates the Tm, a reversible transition (it reappeared in the second heating). The Tm agrees with the value found in the literature14 for ceramide 2 [NS].
Me´ ndez et al.
As a consequence, the Tm values obtained by ATR-FTIR and DSC confirmed that ceramide 2 [NS] was isolated from the IWL extract in BF2 fraction in the absence of Chol, FFA, or other compounds that could affect the thermotropic behavior of this ceramide.
Conclusions MPLC has proved to be a suitable technique to separate a large amount of IWL, which is necessary to isolate a large quantity of ceramides for thermogravimetric study. One fraction containing a mixture of ceramides was obtained with an optimized procedure using chloroform, acetone, and methanol as successive eluents and another fraction rich in ceramide 2 (nonhydroxy acid sphingosine [NS]) was obtained with another optimized procedure using chloroform with increasing amounts of methanol. Moreover, the thermotropic behavior of these fractions and the IWL extract was studied by TGA, ATR-FTIR, and DSC techniques: ATR-FTIR yielded some insight into the transition temperature of the ceramides, and DSC allowed us to differentiate between isolated ceramides (Tm ) 81.2 °C) and ceramides combined with other compounds (Tm < 80 °C). The isolation of ceramide 2 [NS] and the knowledge of the thermotropic behavior of the fractions obtained from IWL extract prove to be useful in reinforcing the lipidic barrier function of the skin. Acknowledgment. We acknowledge the expert assessment of J. Carilla and technical assistance of A. Lo´pez in the DSC experiments. The authors acknowledge also the expert support of G. von Knorring. LA0621315
