Ceramides, which comprise more than 40% of lipid components in the stratum corneum, play an important barrier function in human skin.1,, 2 Lipid-depletion and replenishment experiments have revealed that sphingo-lipids, especially ceramides, play an essential role in the establishment and maintenance of the water-retaining properties of the skin. Since it is known that the content of ceramides in the skin reduces with age, it has been suggested that increased transepidermal water loss is the result of their reduced presence in the skin. The importance of ceramides in the water-retention capacity of the skin is well documented, and many investigators have clearly demonstrated that ceramides play a role in maintaining and repairing the skin's barrier property.2
From a pharmaceutical and cosmetic point of view, there is now good reason to believe that ceramides are among the most important compounds for skin protection.1,, 2 Therefore, ceramides have become important as cosmetic ingredients. Interpretation of the results of the numerous studies which have examined the effect of ceramides on the restoration of skin barrier function is complicated by the diversity of the nature of the ceramides which have been employed.3 These have included extracted stratum corneum lipids (from various species), ceramides derived from bovine brain sphingomyelin, cerobrosides, totally synthetic analogues of natural skin ceramides (which do not usually possess the correct three-dimensional molecular configuration), and ceramides obtained by biotechnological procedures which do have the same absolute molecular configuration as human skin ceramides (skin-identical ceramides). In addition, certain pseudoceramides are available; however, although these products claim to be ceramides, they lack the composition and structure typical of skin ceramides.
Animal tissues and organs are major sources of sphingolipids. About 300 different sphingolipids have been found in various mammalian types. In mammals, higher concentrations of sphingolipids are present in the brain and spinal cord. Using a multistage extraction procedure, a lipid mixture rich in sphingolipids can be obtained. However, the mad-cow disease incident in the UK has led to prohibition of use of sphingolipids from animal sources. There are some plants that also contain sphingolipids; however, the quantities available are too small to isolate sphingolipids or ceramides in an efficient and economic manner. Furthermore, plant sphingolipids generally have a glycosylated form, which is not found in the stratum corneum. As an alternative source for sphingolipids, the yeast Pichia ciferrii has been successfully exploited to produce certain types of ceramides in large quantity. The yeast Pichia ciferrii, which was first identified to produce extracellular sphingolipids in 1960,4 is a unique microbial source that can produce fairly large amounts of sphingolipids by fermentation. Furthermore, the stereochemical configuration of the yeast sphingolipids is identical to that found in human skin.
For use in cosmetic products at reasonable concentrations, the synthetic analogues and the skin-identical ceramides are practical sources. When choosing a ceramide for use in a product, the formulator should be aware of the origin of the ceramides and also of their chemical structure and chiral similarity to the natural ceramides. Therefore, a sensitive and selective method for accurate identification of molecular species of ceramide is necessary.
The ceramides of the stratum corneum are a highly heterogeneous family that can be fractionated into seven classes by thin-layer chromatography (TLC).5 A more informative nomenclature was suggested by Motta et al.6 This fractionation is based on the number and position of free hydroxyl groups among the different ceramide fractions, and thus does not take into account the variation of fatty acid and amine chain lengths. The complete study of each class requires further analysis, such as the hydrolysis of the amide bond for characterization of the base and the fatty acid moieties.7 TLC and HPLC are the main methods used for ceramide separation.8–10 Each separated compound must then be determined by its retention characteristics when studied by TLC or HPLC with known standards. Although HPLC has been widely applied to analysis, the qualitative information is so far very limited.11,, 12
A gas chromatography/mass spectrometry (GC/MS) method, used after hydrolysis of the ceramides, is currently used.13 Fatty acids as well as sphingoid bases can be separately analyzed by GC/MS. However, these data do not reveal information about the combination of the two moieties. The GC/MS methods for analyzing intact ceramides require prior trimethylsilylation, and first used packed columns in the early 1970s and, more recently, capillary columns.13,, 14 The requirement for derivatization is considered to be a disadvantage of GC/MS, especially in comparison with the recently developed LC/MS methods. The presence of the trimethylsilylated ceramides in a surplus of derivatization reagents may cause contamination of the GC/MS system and thereby affect quantification stability.14
Fast atom bombardment mass spectrometry (FAB-MS) and FAB-MS/MS have been demonstrated to be useful for structural elucidation of ceramides, which are relatively large, nonvolatile, and thermally labile.15–18 The FAB technique has a distinct disadvantage with respect to sensitivity in the analysis of complex multicomponent mixtures. Although compounds frequently used in cosmetics have been studied by HPLC employing continuous-flow FAB-MS, ceramides could not be detected in quantities less than ca. 500 ng.19
Recent papers have reported the application of LC/MS/MS20,, 21 to this problem. Couch et al. used LC/MS with an APCI (atmospheric pressure chemical ionization) interface for the determination of ceramides extracted from HL-60 cells.22 The thermal decomposition of the ceramides at the temperatures characteristic of APCI sources is possible via cleavage of the N-acyl moiety.
Electrospray ionization (ESI) offers several advantages over the FAB-MS/MS technique, including lower background signals because of the absence of matrix ions, the longer lasting and stable primary ion currents, the ease of sampling, and compatibility with liquid chromatography. More recently, ceramides derived from small quantities of biological or synthesized material have been examined using ESI-MS and ESI-MS/MS.22–25
To our knowledge the reversed-phase HPLC/ESI-MS method has not yet been used in the analysis of ceramides in cosmetics, probably because of the absence of general conditions applicable to the analysis of ceramides. The ceramides used in this study contain different amine bases (phytosphingosine, sphingosine) and fatty acids of different chain lengths and with different functionality (un-saturation). The present investigation describes the determination of ceramide molecular species in the cosmetic raw material DS-Y30 (ceramide 3B), originally derived from the fermentation of the yeast Pichia ciferrii4 in 1998 by Doosan.
A schematic of the industrial production process of ceramides is shown in Fig. 1. In the present work, the ceramides were separated and simultaneously determined by isocratic reversed-phase HPLC/ESI-MS/MS. Both positive and negative ESI modes were used to identify and quantify the molecular ions and structures of ceramides and of an unknown impurity in the commercial cosmetic raw materials.
Reagents and standards
The standards sphingosine (S), D-sphingosine (MW 299.5), N-acetyl-d-sphingosine (C2:0S; MW 341.5), N-hexanoyl-d-sphingosine (C6:0S; MW 397.6), N-palmitoyl-d-sphingosine (C16:0S; MW 537.9), N-oleyl-d-sphingosine (C18:1S; MW 563.9) were all purchased from Sigma (St. Quentin, Fallavier, France). Phytosphingosine (P; MW 317.5), N-oleoyl-d-phytosphingosine (ceramide IIIB; C18:1P; MW 582.0), N-linoleyl-d-phytosphingosine (C18:2P; MW 580.0), and N-octanoylphytosphingosine (C8:0P; MW 443.7) were generous gifts from Cosmoferm (Delft, Netherlands). Undecylenic phytosphingosine (C11:1P; MW 483.8), N-hexanoylphytosphingosine (C6:0P; MW 415.7), N-acetyl-d-phytosphingosine (C2:0P; MW 359.6) were generous gifts from Doosane (Korea). DS-Y30 (ceramide 3B; mainly N-oleoylphytosphingosine), the cosmetic raw material, was provided by Doosan. All standards and reagents were dissolved in chloroform/methanol (1:2) at concentrations of 0.5–0.005 mg/mL.
HPLC apparatus, column and apparatus
The liquid chromatograph consisted of a Shimadzu LC-10ADVP solvent delivery system (Tokyo, Japan), and a Rheodyne 7520 injector with a 0.5-µL sample loop (Rheodyne, CA, USA). The LC separations were performed on a micro-column, 250 × 1.0 mm i.d. Capcell Pak C18 (Shiseido, Japan). The flow rate used was 0.05 mL/min. Eluent A consisted of 0.1% formic acid in water. Eluent B consisted of 50 mM ammonium acetate (AcONH4) in methanol, 0.1% formic acid (FA) in acetonitrile (ACN), and tetrahydrofuran (THF), in the ratio 90:5:5 (v/v/v). The separation of most compounds was obtained using an isocratic mobile phase consisting of 7% of eluent A and 93% of eluent B.
Mass spectrometry was performed using a VG Quattro triple quadrupole mass spectrometer (Fisons Instruments, VG Organic, Altrincham, UK) equipped with a pneumatically assisted ESI source. The LC effluent entered the mass spectrometer through an electrospray capillary set at +3.00 kV (ESI+) or −3.00 kV (ESI−), with a source temperature of 80 °C. Nitrogen was used both as drying gas and nebulizing gas at flow rates of 500 and 15 L/h, respectively. In all modes, full-scan spectra of standard ceramides between m/z 100 and 800 were obtained at a scan speed of 167 Th per second with a unit mass resolution (FWHM definition).
In-source collision-induced dissociation (CID) involves the application of an acceleration voltage in the ESI interface region, and was optimized for best sensitivity in producing the more informative fragment ions. The sampling cone-skimmer voltage was varied between 20 and 80 V to produce different degrees of in-source CID. Data acquisition and processing were performed with a VG MassLynx data system.
RESULTS AND DISCUSSION
ESI-MS/MS fragmentation of ceramides with sphingosine and phytosphingosine
In order to characterize the mass spectral pattern of ceramides, the standards described in the Experimental section were used. The mass spectrometry software used could switch between positive and negative ion modes within the same HPLC run. First, a cone voltage (CV) of 20 V in the positive mode, and of 40 V in the negative ion mode, were used to give molecule related ions [M + H]+, [M − H2O + H]+, [M + Na]+, [M − H]−, [M + HCOO−]−, [M + AcO−]− (where AcO− = CH3COO−). For all ceramides under these conditions, the ESI mass spectra were quite simple, with few fragment ions observed.
A higher cone voltage was used to produce the more informative fragment ions in positive and negative ion modes. The nomenclature for the cleavages labeled in the spectra shown here is described in Fig. 2. These designations contain some modifications of the nomenclature proposed by Ann and Adams.15–17 We have added W, X, and a designation for the fragment ions proposed previously by Raith and Neubert.24 The fragment ions of N-acylsphingosine and N-acylphytosphingosine, observed by ESI-MS/MS, were found to be the same as those obtained by FAB-MS/MS.15–17
In positive ion mode at a low cone voltage (20 V), the mass spectrum consisted of the protonated molecule ([M + H]+), an ion corresponding to the loss of water ([M − H2O + H]+), and a small peak corresponding to the molecular sodium adduct ([M + Na]+). As the potential was increased to 40 V, in-source CID produced more fragmentation, and ions corresponding to loss of a second water molecule and loss of the fatty acid side chain were observed. At the highest cone voltage of 60 V, loss of water is one fragmentation route for [M + H]+ ions of ceramides that contain both sphingosine and phytosphingosine.21 There are also O′ (m/z 264) and O″ (m/z 282) ions that presumably arise via loss of one or two molecules of water from O (m/z 300) ions for ceramides with a sphingosine moiety, which are formed by cleavage of the amide bond with charge retention on nitrogen. For ceramides that contain phytosphingosine instead of sphingosine, the O (m/z 318) ion was detected, and O′ (m/z 300), O″ (m/z 282) ions and also the O″′ (m/z 264) ion, i.e. [O − 3H2O]+, was observed. The O, O″, and O″′ series of product ions provides information about the molecular weights of the sphingoid and fatty acyl residues.
In the negative ion mode a cone voltage of 20 V was first applied. In addition to small amounts of [M − H]−, the [M + HCOO−]− and [M + AcO−]− adduct ions were the predominant species. As the potential was increased to 40 V, the mass spectrum consisted of the [M − H]− ion and small peaks corresponding to the molecular adducts ([M + HCOO−]− and [M + AcO−]−). The optimized conditions for the negative in-source CID (cone voltage of 80 V) yielded fingerprint spectra providing an optimum of structural information about the fatty acid as well as the sphingoid base moiety. In this condition, ceramide ions that contain either sphingosine (Table 1) or phytosphingosine (Table 2) as sphingoid base produced the fragment ions labeled as P, S, T, U ions (Table 1, Fig. 2). The negative ion mass spectra of ceramides provided complementary information. All of the P, S, T, and U ions give information about the molecular weights of the fatty acyl and sphingoid chains. The mechanisms for formation of the S and T ions were previously proposed to follow Schemes 1 and 2.15,, 17
Table 1. Characteristic negative product ions from in-source CID of molecular or molecular adduct ion species of ceramides with sphingosine (CV = 80 V)
Characteristic in-source CID negative fragment ions (m/z) of ceramides containing sphingosine
[M − H]−
[M − H + AcOH]−
Table 2. Characteristic negative product ions from in-source CID of molecular or molecular adduct ion species of ceramides with phytosphingosine (CV = 80 V)
Characteristic in-source CID negative fragment ions (m/z) of ceramides with phytosphingosine
[M − H]−
[M − H + AcOH]−
For ceramides with sphingosine, the characteristic fragment ions corresponding to R (m/z 263) and V were produced. The R ion was not found in the product ion mass spectra of the C16:OS molecule obtained by using FAB-MS/MS.15,, 17 We suggest that the R ion (m/z 263) is probably the structure shown in Table 1. The characteristic ions R (m/z 267), Q, W, and X were observed in the product ion mass spectra of the ceramides with phytosphingosine instead of sphingosine. The W and X ions were previously proposed to be those shown in Fig. 2,21,, 24 and were found previously in the negative ion spectra (ESI-MS/MS) of C18:1P.24 The structure of the R (m/z 267) ion was proposed to be that shown in Fig. 2.
The characteristic product ions formed by ‘in-source’ CID may be briefly summarized as follows: ceramides with a sphingosine moiety yield fragment cations at m/z 264, 282 and 300, and those with phytosphingosine at m/z 264, 282, 300 and 318.21 Tables 1 and 2 show characteristic product anions for the ceramides with a phytosphingosine moiety at m/z 267 (R), m/z 255 (P) and m/z 225 (Q) and with a sphingosine moiety at m/z 263 (R) and m/z 237 (P), regardless of the length of the fatty acyl chains. These results suggest that the combination of the positive and negative mode fragments can be used for unambiguous characterization of ceramides.
Separation of ceramide standards with phytosphingosine
Because the DS-Y30 (ceramide 3B) product can be semi-synthesized from tetraacetylphytosphingosine (TAPS in Fig. 1), produced as yeast extracts,4 we first examined the separation of the ceramides with phytosphingosine. The standards used for ceramide separation were N-oleoyl-d-phytosphingosine (ceramide IIIB; C18:1P; MW 582.0), N-linoleyl-d-phytosphingosine (C18:2P; MW 580.0), undecylenicphytosphingosine (C11:1P; MW 483.8), N-octanoyl-phytoshingosine (C8:0P; MW 443.7), N-hexanoylphyto-sphingosine (C6:0 P; MW 415.7), and N-acetyl-d-phytosphingosine (C2:0P; MW 359.6). In this study, the separation of most compounds was obtained using an isocratic mobile phase consisting of 7% of HPLC eluent A (0.1% FA/H2O) and 93% of HPLC eluent B (50 mM AcONH4-MeOH/0.1% FA-ACN/THF 90:5:5). This mobile phase combined best sensitivity for ESI, short retention times and effective separation. In order to obtain high sensitivity and good separation of ceramides it was necessary to add the ammonium acetate to methanol and the formic acid (FA) to acetonitrile. It was also necessary to add the strong eluent tetrahydrofuran to methanol to obtain faster separations because ceramides have a strong affinity for the RP18 phase. The addition of 0.1% FA aqueous solution increases the retention times and improves the separation. This isocratic reversed-phased HPLC system provided a stable baseline and permitted good chromatographic separation and sensitivity for the mass spectrometric detection.
Selected-ion monitoring was used for detection of selected ceramides with greatly increased sensitivity. A rapid cone voltage switching method was also used to further maximize signals for the selected ions. For example, at an optimum CV of 20 V, the molecular ions [M + H]+ were detected as the major peaks in the positive ion mode. In the negative ion mode, the molecular ions [M − H]− were predominantly observed at a CV of 40 V. The ceramide standards have the following elution order: C2:0P, C6:0P, C8:0 P, C11:1P, C18:2P, and C18:1P. The elution order was dependent on the chain length and degree of unsaturation.
Analysis of the DS-Y30 (ceramide 3B) preparation in cosmetic raw material by LC/MS/MS
Selected-ion monitoring (SIM) was used for detection of the selected characteristic fragment ions with greatly increased sensitivity. In positive ion mode at a CV of 60 V, the characteristic fragment ions at m/z 318, 300 and 282 were the common significant peaks detected for the ceramides containing phytosphingosine. In negative ion mode at a CV of 80 V, the characteristic fragment ions at m/z 225, 255 and 267 were the common significant peaks. Thus, under the same chromatographic conditions, the positive (at m/z 318, 300 and 282) and negative (at m/z 225, 255 and 267) SIM chromatograms (Fig. 3) for DS-Y30 show the separation of six molecules of ceramides with phytosphingosine. However, these SIM experiments do not identify the ceramides.
The full-scan LC/ESI-MS spectra produced from a ceramide sample at various cone-skimmer potentials showed peaks for the protonated molecule and fragment ions that permitted structural characterization. In positive ion mode at a CV of 20 V, the mass spectra of the chromatographic peaks consisted of the protonated molecule and a small peak corresponding to the molecular sodium adduct. Under the same chromatographic conditions, full-scan negative ion ESI-MS analysis at a CV of 40 V revealed the major molecular ions [M − H]−, and the molecular adducts [M + AcO−]− and [M + HCOO−]−. SIM was also used for detection of selected molecular ions with greatly increased sensitivity. Therefore, using the same chromatographic conditions, the positive and negative SIM chromatograms for molecular ions (Fig. 4) show the separation of six molecular species in the DS-Y30 (ceramide 3B) shown to be phytosphingosine ceramides by the high cone-voltage SIM experiments (Fig. 3).
Analysis of the full-scan spectra of each chromatographic peak in both positive and negative ion modes, at CVs of 60 and 80 V, respectively, provided the structures of the individual ceramide species. The identification of individual molecular species was performed by simultaneously moni-toring their fragment ions.
Five chromatographic peaks (eluted between 14 to 25 min) are observed in the molecular ion SIM chromatograms (Fig. 4). The evidence for the identification of these species as ceramides containing phytosphingosine was provided by the full-scan mass spectra of each peak in positive and negative ion modes.
In positive ion mode at a CV of 60 V, the product ions obtained by in-source CID of the ceramide ions are shown in Figs 5(a)–(e). The peaks at m/z 318, 300, 282, and 264 are the characteristic O, O′, O″, and O″′ series of product ions of ceramides containing phytosphingosine (Fig. 2). The peaks at m/z 583, 581, 557, 555, and 529 represent the protonated molecules [M + H]+. The monosodiated ions [M + Na]+ were detected at m/z 605, 603, 579, 577, and 551. Molecular ion species ([M + H]+ and [M − H]−), obtained under positive and negative ion modes, were used to determine and confirm the molecular weights of these substances.
It was possible to ascertain the molecular weights of the sphingoid and fatty acyl residues in the negative ion mass spectra obtained in negative ion mode at a CV of 80 V. The in-source CID mass spectra of the ceramide ions are shown in Figs 6(a)–(e). The characteristic ions R (m/z 267), Q, W and X, observed in the mass spectra of the standard ceramides containing phytosphingosine, were also detected in the LC/ESI-MS/MS spectra of the sample (Fig. 6). The P, S, T and U ions were also found in these spectra. The P, Q, and R (m/z 267) ions give information about the molecular weights of the sphingoid chain. The peaks labeled S, T, U, W, and X all give information about the molecular weights of the fatty acyl groups; the molecular weight for the fatty acyl moiety of the ceramide ions is derived from the differences of molecular weight between the molecular ion and the S, T, U, W and X ions.
Comparison with the results obtained above for standards suggests that the peaks eluting at 23.7 and 17.9 min (Fig. 4) are C18:1P and C18:2P. Although there were no standard ceramides available for the others, based on comparison of the data in Figs 5 and 6, the peaks eluting at 14.6, 15.9, and 21.7 min (Fig. 4) were identified as C14:0P, C16:1P, and C 16:0P.
Identification of the unknown impurity in the DS-Y30 (ceramide 3B) preparation
The peak eluting at 4.2 (±0.4) min (Fig. 4) was not a ceramide containing phytosphingosine or sphingosine. The information for the identification of the impurity compound was provided by the mass spectra of the peak in the full-scan positive and negative ion modes.
In positive ion mode at a CV of 60 V, the product ions obtained by in-source CID of the impurity are shown in Fig. 7(a). The peaks at m/z 318, 300, 282, and 264 represent the O, O′, O″, and O″′ series of product ions characteristic of compounds containing the sphingoid base phytosphingosine. The protonated molecule [M + H]+ and the monosodiated ions [M + Na]+ were detected at m/z 347 and 369.
Molecular ion species ([M + H]+, [M − H]−) of the impurity compound obtained under positive and negative ion modes were used to characterize the molecular weight of this substance. It was possible to ascertain the molecular weight of the sphingoid base in the negative ion mass spectrum. In negative ion mode at a CV of 80 V, the mass spectrum of the impurity shown in Fig. 7(b) was obtained. The characteristic ions R (m/z 267), P, and Q of ceramides with phytosphingosine were detected in the in-source CID spectra of the substance (Fig. 7). The molecular ion, [M − H]−, and the molecular adducts, [M + HCOO−]− and [M + AcO−]−, were detected at m/z 345, 391 and 405. Therefore, the molecular weight of the unknown substance is 346 Da. Based on interpretation of the data in Figs 7(a) and 7(b), we suggest that the structure of the impurity is that shown in Fig. 7.
Figure 1 shows the process of production of phytosphingosine and related ceramides from the tetra-acetyl-phytosphingosine (TAPS) produced extracellularly by yeast fermentation.4 The DS-Y30 (ceramide 3B) was semi-synthesized from phytosphingosine (Fig. 1) by an N-acylation process. Therefore, the TAPS and phytosphingosine may possibly contain impurities from the production process. The proposed impurity (N-ethylphytosphingosine; MW 346) was not detected, however, in the TAPS and phytosphingosine, and must have been formed during the process of N-acylation of phytosphingosine, because only the DS-Y30 final product contained the impurity.
Quantification of the ceramide molecules and impurity
Quantitative determinations of ceramides in cosmetic preparations have been performed to a limited extent up to now. The limit of detection in full-scan mode is estimated to be in the nmol range for the ceramides, but the SIM mode enables a factor of up to 1000 improvement in sensitivity. The SIM method was used to determine the composition of ceramide molecular species in the DS-Y30 (ceramide 3B). Both the [M + H]+ and [M − H]− ions were used for the determination, and yielded similar results. For quantitative assessment, a constant amount of a non-naturally occurring internal standard (C 11:1P) was added to the lipid extracts. The calibration lines thus obtained showed high linearity for both [M + H]+ and [M − H]− of ceramides (Fig. 8). In order to avoid matrix effects and to compensate for machine instabilities which affect ionization efficiency, all values were expressed as the ratio of peak area for the ceramide molecule to that of the internal standard (C11:1P). For commercially unavailable ceramide species, the calibration curve for the closest related ceramide species was used for calculation, e.g., C18:1P for C16:0P, C16:1P and C14:0P and C2:0 for the impurity (N-ethylphytosphingosine; MW 346). The contents of C18:1P, C18:2P, C16:0P, C16:1P, C14:0P and N-ethylphytosphingosine were thus determined to be 75.7 ± 0.3%, 7.7 ± 0.5%, 5.9 ± 0.3%, 4.9 ± 0.4%, 2.7 ± 0.3% and 3.1 ± 0.3%, calculated from SIM chromatograms obtained in the positive mode. The contents of C18:1P and C18:2P, C16:0P, C16:1P, C14:0P and N-ethylphytosphingosine were 78.1 ± 0.3%, 6.1 ± 0.2%, 5.0 ± 0.2%, 4.1 ± 0.1%, 2.7 ± 0.2% and 4.0 ± 0.2%, calculated from the SIM chromatograms obtained in the negative mode. These results for the ceramides and impurity of DS-Y30 are briefly summarized in Table 3. Detection limits are estimated to be about 0.5 pmol for the SIM method (signal-to-noise (S/N) ratio = 3:1).
Table 3. Composition of ceramide molecules in the DS-Y30 (ceramide 3B) product
Molecular Species, % Composition, SD CV
75.7 ± 0.3 0.40
78.1 ± 0.3 0.38
7.7 ± 0.5 6.49
6.1 ± 0.2 3.28
5.9 ± 0.3 5.08
5.0 ± 0.2 4.00
4.9 ± 0.4 8.16
4.1 ± 0.1 2.43
2.7 ± 0.3 11.11
2.7 ± 0.2 7.41
(MW = 346 )
3.1 ± 0.3 9.68
4.0 ± 0.2 5.00
Reversed-phase LC/ESI-MS with in-source CID has been shown to be valuable for qualitative and semiquantitative analysis of different ceramide molecular species in the sample of cosmetic raw material. Molecular ions and adduct species ([M + H]+, [M + Na]+, [M − H]−, [M + HCOO−]−, and [M + AcO−]−) of ceramides, obtained using both positive and negative ion modes, were used to characterize the molecular weights of these substances.
The separation can be performed within 25 min and the detection limit was about 0.5 pmol for the SIM method. These ceramides were more sensitively detected by positive than negative ion mode.
The CID fragmentation patterns of molecular anion mode makes it possible to obtain information on the nature of the sphingoid base and fatty acyl chains. The product ion spectra in negative ion mode of ceramide species provide more structural information than those obtained in positive ion mode. C18:1P, C18:2P, C16:0P, C16:1P and C14:0P were found in the DS-Y30 (ceramide 3B) sample.
The structural determination of the unknown impurity was also performed by LC/ESI-MS with in-source CID in both positive and negative ion modes (Fig. 7). This method may be generally useful for the identification of unknown species and minor ceramide species in a mixture sample.
Also this method will be useful for quality control of the products contained in commercial cosmetic ceramides. From the results for identification and quantification of the ceramides by LC/ESI-MS/MS, it is possible to apply this method to the routine analysis of ceramides in different samples, such as natural products and cosmetic products.
This work was supported by the 21st Century Frontier Research Program funded by the Ministry of Science and Technology, Republic of Korea.