Gas chromatographic separation of fatty acid esters of cholesterol and phytosterols on an ionic liquid capillary column

Journal of Chromatography B, 1007 (2015) 67–71

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Short communication

Gas chromatographic separation of fatty acid esters of cholesterol and phytosterols on an ionic liquid capillary column Simon Hammann, Walter Vetter ∗ University of Hohenheim, Institute of Food Chemistry (170b), Stuttgart, Germany

a r t i c l e

i n f o

Article history: Received 25 March 2015 Received in revised form 26 October 2015 Accepted 3 November 2015 Available online 7 November 2015 Keywords: Steryl ester Ionic liquid Gas chromatography Mass spectrometry

a b s t r a c t Steryl esters are high molecular weight compounds (600–700 g/mol) regularly present as a minor lipid class in animal and plant lipids. Different sterol backbones (e.g., cholesterol, ␤-sitosterol and brassicasterol) which can be esterified with various fatty acids can result in highly complex steryl ester patterns in food samples. The gas chromatographic (GC) analysis of intact steryl esters is challenging, since high elution temperatures are required for their elution. On nonpolar GC phases, steryl esters with fatty acids with differing degree of unsaturation (e.g., oleate and linoleate) cannot be separated and there are only few polar columns available with sufficient temperature stability. In this study, we used gas chromatography with mass spectrometry (GC/MS) and analyzed intact steryl esters on a commercial room temperature ionic liquid (RTIL) column which was shortened to a length of 12 m. The column separated the steryl esters both by total carbon number and by degree of unsaturation of the fatty acid. For instance, cholesteryl esters with stearic acid (18:0), oleic acid (18:1n-9), linoleic acid (18:2n-6) and ␣-linolenic acid (18:3n-3) could be resolved (R ≥ 1.3) from each other. By analysis of synthesized standard substances, the elution orders for different steryl backbones and different fatty acids on a given sterol backbone could be determined. Analysis of spreads and plant oils allowed to determine retention times for 37 steryl esters, although a few co-elutions were observed. The ionic liquid column proved to be well-suited for the analysis of intact steryl esters. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Steryl esters (SE) are a minor lipid class regularly found in animal and plant lipids. They are derived from cholesterol (in animal fats) and phytosterols (in plant oils) by esterification of the hydroxyl group on C-3 with fatty acids [1,2]. Also, SE are used in several foodstuffs due to their ability to lower serum cholesterol in humans [1,2]. Sterols are usually analyzed after saponification of the extracted lipids. In this step, SE are hydrolyzed and the formerly esterified sterols are determined together with the free sterol fraction to give the total sterol content. In plant oils, up to 70% of the total sterols may be bound as SE [2]. Intact SE can either be analyzed by gas chromatography (GC) or reversed-phase high performance liquid chromatography (RP-HPLC). Both techniques have advantages and drawbacks. RP-HPLC benefits from low analysis temperatures. Yet, with either C18 or hexyl-phenyl modified stationary phases, coelutions have been noted to occur (most pronounced between the steryl esters of palmitic acid and oleic acid

∗ Corresponding author. Fax: +49 711 459 24377. E-mail address: [email protected] (W. Vetter). http://dx.doi.org/10.1016/j.jchromb.2015.11.007 1570-0232/© 2015 Elsevier B.V. All rights reserved.

[3,4]. Recently, a C8 modified stationary phase enabled the separation of phytostanyl esters, but coelutions of esters with different sterol backbones could not be completely omitted [5]. In general, the resolution power of GC is better than in HPLC. However, the GC analysis of intact SE is a challenging task due to the high molecular weight (∼600–700 g/mol) and low volatility of these compounds. As a consequence, short GC columns along with high temperatures (300 ◦ C or higher) are necessary for their elution. Nonpolar GC phases, such as 100% dimethyl polysiloxane, were frequently employed in SE analysis due to their high thermal stability [6]. With this setup, SE with fatty acids of different chain lengths can be separated. Yet, SE containing fatty acids with the same carbon number but a different number of double bonds are difficult to separate on nonpolar GC columns [7]. Switching to polar GC phases significantly improves the separation power for SE [8,9]. However, with increasing polarity the temperature stability of classic GC columns is usually decreasing. On the one hand, medium-polar GC phases coated with trifluoropropyl polysiloxane operated at temperatures >300 ◦ C failed to separate SE with saturated and monounsaturated fatty acids (e.g., sitosteryl stearate and oleate) [10,11]. On the other hand, polar GC phases with cyanopropyl polysiloxanes or polyethy-

68

S. Hammann, W. Vetter / J. Chromatogr. B 1007 (2015) 67–71

lene glycol have a lower thermal stability which makes it difficult to elute these high-boiling compounds from the GC system. In this study, we used gas chromatography with mass spectrometry (GC/MS) in combination with a commercial GC column coated with a room-temperature ionic liquid (RTIL) for the separation of SE. RTILs are molten salts, which are liquid at room temperature and which can be used in GC at high temperatures (300 ◦ C) [12]. These polar phases partly showed non-conventional interactions with different types of analytes such as fatty acid methyl esters or essential oil constituents [12–14]. RTIL columns have been noted to differ in their separation characteristics from polysiloxane-based GC phases [12]. To test the separation characteristics of the RTIL phase, we synthesized different SE standards for analysis and isolated SE from food samples.

R=1.1

Irel

a)

R=2.7

2 4

5

R=2.5

11

R=2.4

10

18

2. Material and methods

27.0

28.0

29.0

24

30.0

31.0

32.0

33.0

t [min]

2.1. Chemicals, standard substances and samples Cholesterol (puriss.), technical ␤-sitosterol (consisting of 55.5% ␤-sitosterol, 34.2% campesterol and 2.8% stigmasterol [15]) and stearic acid (puriss.) were from Merck (Darmstadt/Germany). Palmitic acid, myristic acid (both > 98%), oleic acid (>99%), linoleic acid (puriss.), ␣-linolenic acid (∼70%), pentadecanoic acid (99%) and ethyl acetate (distilled prior to use) were from Sigma–Aldrich (Steinheim/Germany). n-Hexane (HPLC grade) was from Th. Geyer (Renningen, Germany) and potassium hydroxide (>85%) from Carl Roth (Karlsruhe/Germany). t-Butyl methyl ether (MTBE) was from Fisher Scientific (Leicestershire, England) and was distilled prior to use. Food samples (two spreads enriched with phytosteryl esters, corn germ oil and oil from cod liver) were purchased in local supermarkets in Stuttgart/Germany. 2.2. Synthesis of steryl ester standards Steryl esters were prepared according to Barnsteiner et al. [11]. In short, ∼0.25 mmol of the sterol (cholesterol or technical ␤sitosterol) and a two-fold molar excess of a fatty acid (∼0.5 mmol) were heated without solvent in a nitrogen flushed reaction vessel to 180 ◦ C for 25 h. Afterwards, 2.5 mL 1 M KOH solution was added and SE were extracted three times with 2.5 mL n-hexane/MTBE (3:2, v/v) [11]. Purities of the solutions were determined by means of GC/MS. 2.3. Isolation of steryl esters from food samples Steryl esters were isolated from other lipid classes by solid phase extraction as described elsewhere [16]. In brief, lipid extracts or plant oils were fractionated on 5 g deactivated silica gel. SE were eluted after hydrocarbons (30 mL n-hexane, Fraction 1) with 40 mL n-hexane/ethyl acetate 99:1 (v/v) (Fraction 2). 2.4. Gas chromatography with electron ionization mass spectrometry (GC/MS) SE were analyzed on a 6890/5973 GC/MS system (HewlettPackard/Agilent, Waldbronn, Germany) equipped with a cool-oncolumn inlet and a pre-column (2 m, 0.53 mm i.d., deactivated with 1,3-diphenyl-1,1,3,3-tetramethyldisilazane, BGB Analytics, Boeckten/Switzerland), which was connected by a press-fit to an IL-59 GC column (12 m, 0.25 mm i.d., 0.25 ␮m film thickness, kindly provided by Supelco, Bellefonte, PA/USA). Helium 5.0 was used as the carrier gas with a flow rate of 1.3 mL/min. The oven heater was programmed as follows: After 1 min at 50 ◦ C, the temperature was raised by 10 ◦ C/min to 300 ◦ C and this final temperature was held for 20 min. The temperature of the transfer line, ion source and

Irel

18:1n-9-CE (16)

b)

R=1.3 18:0-CE (12) 18:2n-6-CE (19) 18:3n-3-CE (26)

27.0

28.0

29.0

30.0

R=2.1 R=2.6 31.0 32.0

33.0

t [min]

Fig. 1. GC/MS full scan chromatograms of synthesized standard compounds of (a) steryl esters with saturated fatty acids and (b) cholesteryl ester of fatty acids with different degree of unsaturation. Oven program in (b): 50 ◦ C (1 min), 10 ◦ C/min to 280 ◦ C, 2 ◦ C/min to 300 ◦ C (15 min). The peak numbering refers to Table 1.

quadrupole were set to 300 ◦ C, 150 ◦ C and 230 ◦ C, respectively. Data was recorded in full scan mode from m/z 50-800. 3. Results and discussion 3.1. General separation characteristics of the RTIL phase and impact of the sterol moiety The general elution characteristics of SE on the RTIL phase (elution temperatures 300 ◦ C) were determined using synthesized standard compounds. In agreement with other stationary phases, the elution was primarily based on the molecular weight of the SE (Fig. 1a, Table 1). Accordingly, individual esters of a given sterol eluted the later the longer the fatty acid moiety was (e.g., tR 14:0-sitosteryl ester < 16:0-sitosteryl ester < 18:0-sitosteryl ester). Despite of the quite high molecular weight, SE eluted as narrow peaks (pw50 = 0.08 min for 14:0-sitosteryl ester and 0.13 min for 18:0 sitosteryl ester), but the peak width increased with increasing retention time. As expected, campesteryl esters (28 carbons in the sterol moiety) eluted slightly earlier than sitosteryl esters (29 carbons in the sterol moiety) of the same fatty acid and those peaks were baseline separated (resolution between R = 2.7 and 2.4, Fig. 1a). Noteworthy, 15:0-cholesteryl ester (15 + 27 = 42 carbons) eluted slightly earlier than the isobaric 14:0-campesteryl ester (14 + 28 = 42 carbons) (R = 1.1, Fig. 1a). The retention times (tR ) of SE with the same fatty acid increased in the order brassicasterol < cholesterol < stigmasterol ≈ campesterol < ␤sitosterol < avenasterol < cycloartenol (Table 1).

S. Hammann, W. Vetter / J. Chromatogr. B 1007 (2015) 67–71

69

Table 1 Relative retention times of steryl ester standard compounds and steryl esters identified in samples. Peak number

Compound

Carbons in the sterol backbone

Total carbon number (double bonds in the sterol moiety; in the fatty acid moiety)

Relative retention timea

Identified byb

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

4,8,12-Trimethyl-13:0-cholesteryl ester 15:0-Cholesteryl ester 14:0-Stigmasteryl ester 14:0-Campesteryl ester 14:0-Sitosteryl ester 16:0-Brassicasteryl ester 16:0-Cholesteryl ester 16:1-Cholesteryl ester 16:0-Stigmasteryl ester 16:0-Campesteryl ester 16:0-Sitosteryl ester 18:0-Cholesteryl ester 18:0-Brassicasteryl ester 18:1n-9-Cholesteryl ester 16:0-Avenasteryl ester 18:1n-9-Brassicasteryl ester 18:0-Stigmasteryl ester 18:0-Campesteryl ester 18:2n-6-Cholesteryl ester 18:2n-6-Brassicasteryl ester 16:0-Cycloartenyl ester 18:1n-9-Stigmasteryl ester 18:1n-9-Campesteryl ester 18:0-Sitosteryl ester 20:1-Cholesteryl ester 18:3n-3-Cholesteryl ester 18:2n-6-Stigmasteryl ester 18:2n-6-Campesteryl ester 18:1n-9-Sitosteryl ester 18:1n-9-Avenasteryl ester 18:2n-6-Sitosteryl ester 18:2n-6-Avenasteryl ester 22:1-Cholesteryl ester 11D3-Cholesteryl ester 18:1n-9-Cycloartenyl ester 18:2n-6-Cycloartenyl ester 11D5-Cholesteryl ester

27 27 29 28 29 28 27 27 29 28 29 27 28 27 29 28 29 28 27 28 30 29 28 29 27 27 29 28 29 29 29 29 27 27 30 30 27

43(1;0) 42(1;0) 43(2;0) 42(1;0) 43(1;0) 44(2;0) 43(1;0) 43(1;1) 45(2;0) 44(1;0) 45(1;0) 45 (1;0) 46(2;0) 45 (1;1) 45(2;0) 46(2;1) 47(2;0) 46(1;0) 45 (1;2) 46(2;2) 46(1;0) 47(2;1) 46(1;1) 47(1;0) 47 (1;1) 45(1;3) 47(2;2) 46(1;2) 47(1,1) 47(2;1) 47(1;2) 47(2;2) 49 (1;1) 47(1;2) 48 (1;1) 48(1;2) 49(1;2)

0.98 1.00 1.00 1.01 1.02 1.02 1.03 1.04 1.04 1.04 1.06 1.06 1.07 1.07 1.07 1.08 1.09 1.09 1.10 1.10 1.10 1.11 1.11 1.11 1.12 1.13 1.13 1.13 1.13 1.15 1.16 1.18 1.19 1.21 1.22 1.23 1.27

MS, FAME Standard, MS Standard, MS Standard, MS Standard, MS MS, tR MS, FAME. MS, FAME Standard, MS Standard, MS Standard, MS Standard, MS MS, tR Standard, MS MS, tR, ten. MS, tR Standard, MS Standard, MS, tR Standard MS, tR MS MS, tR MS, tR Standard, MS, tR MS, FAME Standard MS, tR MS, tR MS, tR MS, tR MS, tR MS, tR MS, FAME MS MS MS MS

a b

Relative to the internal standard 15:0-cholesteryl ester (26.96 min). Standard: authentic standard compound; tR : GC retention time, MS: GC/EI-MS spectrum; FAME: identification after transesterification as methyl ester.

3.2. Separation of steryl esters with fatty acids with different number of double bonds Cholesteryl esters (CE) with C18 -fatty acids resulted in the four well-resolved peaks of 18:0-CE, 18:1n-9-CE, 18:2n-6-CE and 18:3n3-CE (R ≥ 1.3, Fig. 1b). Especially the resolution of 18:0-CE and 18:1n-9-CE is remarkable because this pair cannot be separated on most commercially available polysiloxane phases of the trifluoropropyl [10] or cyanopropyl OV-1701 type (own observation) but only on polar polysiloxane columns with high cyanopropyl content (e.g., SP-2330 type) which are characterized by lower maximum operation temperatures [9]. The same result was also obtained for esters with other sterols (see Section 3.3). 3.3. Separation of intact steryl esters on the ionic liquid GC column GC/MS analysis of the steryl ester fraction isolated from a cod liver sample (Section 2.3) on the RTIL column resulted in wellresolved peaks. Ten different CE (Fig. 2a) were detected in the sample but no esters with phytosterols (which was expected from animal samples). Noteworthy, the abundant cholesteryl esters with furan fatty acids, i.e., with 11-(3,4-dimethyl-5-pentylfuran-2-yl)undecanoic acid (11D5) (37) and 11-(3,4-dimethyl-5-propylfuran2-yl)-undecanoic acid (11D3) (34), were stronger retained than on nonpolar GC phases. For instance, 11D3-CE (34; 20 carbons for the fatty acid) and 11D5-CE (37; 22 carbons for the acid) eluted after the

cholesteryl ester with 22:1 (33). On nonpolar columns of the 100% dimethylpolysiloxane type, CE with these furan fatty acids eluted together with the cholesteryl esters of 20 carbon and 22 carbon fatty acids, respectively [16]. Plant oils are characterized by the presence of different phytosterols which can be esterified with different fatty acids. The RTIL column also resolved phytosterol esters with the critical pairs 18:0 and 18:1n-9, and the relative difference in tR between the homologues could be predicted in a very narrow range. Despite this success on the RTIL column, a few co-elutions were noticed between different phytosterols esterified with different fatty acids (Fig. 2b-d). Since the mass of the sterol backbone could be determined with GC/MS data several co-eluting SE could be distinguished by means of diagnostic fragment ions [17]. For instance, 18:2n-6-campesteryl ester (28) co-eluted with the regularly more abundant 18:1n-9-sitosteryl ester (29) (Fig. 2b-d, Table 1). For campesteryl esters, the diagnostic ion is observed at m/z 382, and this fragment ion was only present with low abundance (∼4% of the diagnostic m/z 396) in the mass spectra of sitosteryl esters (Fig. 2e). If this ratio is found to be significantly larger (as seen in Fig. 2e) the presence of (28) can be verified in spite of the co-elution and the contribution roughly estimated. The two spread samples differed in the additional presence of esters with brassicasterol (peaks (6), (13), (16) and (20)) in sample 2 (Fig. 2b,c). Since brassicasterol is a common constituent of rapeseed oil [2], the presence of brassicasteryl esters hinted at the use of rapeseed oil in sample 2 but not in sample 1. In addition,

70

S. Hammann, W. Vetter / J. Chromatogr. B 1007 (2015) 67–71

Irel

2

a)

14

1 7 *8

* Irel

27.0

25

12 28.0

29.0

33

30.0

31.0

b)

37

34

32.0

33.0

34.0 t [min]

32.0

33.0

34.0

32.0

33.0

34.0 t [min]

31 28,29

2 11 10

27.0 Irel

c)

28.0

29.0

30.0

31.0

t [min]

31

6 27.0

d)

28,29

2

11,13

Irel

23,24

18

10

20,23,24 16 18

28.0

29.0

30.0

31.0

2

30,31 27,28,29

9,10 Irel

27.0

11 15

28.0

21,22,23

29.0

32

30.0

31.0

e)

30

32.0

35

36

33.0

34.0 t [min] m/z 382 m/z 394 m/z 396

29 28

27.0

910 28.0

11

22 29.0

23 27 30.0

31.0

32.0

33.0

34.0 t [min]

Fig. 2. GC/MS full scan chromatograms of the steryl ester fraction of the lipid extract of (a) cod liver, (b) spread sample 1, (c) spread sample 2 and (d) corn germ oil. The peak numbering refers to Table 1. The peak marked with an asterisk in (a) is no steryl ester.

esters of saturated homologues of campesterol and ␤-sitosterol (i.e., campestanol and sitostanol) could be detected by means of their characteristic fragment ions (m/z 215, 384, 398), but these stanyl esters could not be resolved from the corresponding SE. The corn germ oil featured esters of five phytosterols including avenasterol (15, 30, 32) and cycloartenol (21, 35, 36). However, no esters with 18:0 were detected in this sample (missing peaks 19,24). This sample showed the highest variety of SE (16 SE in total). Screening of the four food samples and SE standards allowed us to establish a total of (relative) retention times for 37 steryl esters on the RTIL column (Table 1). The maximum operating temperature of 300 ◦ C was sufficient to elute all SE in the samples from the

GC/MS system. Since retention times slightly varied from run-torun, retention times relative to the internal standard 15:0-CE (2) were more applicable. 4. Conclusions and outlook On the used RTIL GC column, intact steryl esters could be very well separated both by carbon number and number of double bonds in the fatty acid. Cutting the column to a length of 12 m enabled the elution of all relevant steryl esters using the maximum operation temperature of 300 ◦ C. This setup provided a good resolution of SE with different fatty acids. Yet, this better resolution (=more peaks to

S. Hammann, W. Vetter / J. Chromatogr. B 1007 (2015) 67–71

separate than on most other GC columns) required the application of GC/MS to distinguish co-eluting pairs of different SE with different fatty acids. GC with flame ionization detector (GC/FID) which allows using hydrogen as the carrier gas in routine analysis, may enable to use longer RTIL columns, but the structural information provided by the mass spectrometer is lost. Using the RTIL column, a better resolution of individual SE could be achieved than with non-polar or most other frequently used GC columns. Compared to HPLC separations, where the critical pair palmitate/oleate was hard to separate, these compounds were easily separated here. Overall, the resolution can be compared to the very good separation recently achieved by RP-HPLC with a C8 modified stationary phase. Our measurements verify that ionic liquid columns can serve as a valuable tool for the analysis of intact steryl esters in food and biological samples. Acknowledgments We gratefully acknowledge financial support in form of a stipend grant to Simon Hammann by the Fonds der Chemischen Industrie. Furthermore, we would like to thank Supelco/Sigma–Aldrich for providing the RTIL GC column for our research. References [1] R.A. Moreau, B.D. Whitaker, K.B. Hicks, Phytosterols, phytostanols, and their conjugates in foods: structural diversity, quantitative analysis, and health-promoting uses, Prog. Lipid Res. 41 (2002) 457–500. [2] K.M. Phillips, D.M. Ruggio, J.I. Toivo, M.A. Swank, A.H. Simpkins, Free and esterified sterol composition of edible oils and fats, J. Food Compos. Anal. 15 (2002) 123–142. [3] I. Mezine, H. Zhang, C. Macku, R. Lijana, Analysis of plant sterol and stanol esters in cholesterol-lowering spreads and beverages using high-performance liquid chromatography-atmospheric pressure chemical ionization-mass spectroscopy, J. Agric. Food Chem. 51 (2003) 5639–5646.

71

[4] M.F. Caboni, G. Iafelice, M. Pelillo, E. Marconi, Analysis of fatty acid steryl esters in tetraploid and hexaploid wheats: identification and comparison between chromatographic methods, J. Agric. Food Chem. 53 (2005) 7465–7472. [5] B. Scholz, A. Barnsteiner, K. Feist, W. Schmid, K. Engel, Analysis of phytostanyl fatty acid esters in enriched foods via UHPLC-APCI-MS, J. Agric. Food Chem. 62 (2014) 4268–4275. [6] R.P. Evershed, V.L. Male, L.J. Goad, Strategy for the analysis of steryl esters from plant and animal tissues, J. Chromatogr. A 400 (1987) 187–205. [7] W. Kamm, F. Dionisi, L. Fay, C. Hischenhuber, H. Schmarr, K. Engel, Analysis of steryl esters in cocoa butter by on-line liquid chromatography–gas chromatography, J. Chromatogr. A 918 (2001) 341–349. [8] A. Kuksis, J.J. Myher, L. Marai, J.A. Little, R.G. McArthur, D.A.K. Roncari, Fatty acid composition of individual plasma steryl esters in phytosterolemia and xanthomatosis, Lipids 21 (1986) 371–377. [9] N.B. Smith, Gas-liquid chromatography of cholesteryl esters on non-polar and polar capillary columns following on-column injection, J. Chromatogr. A 254 (1983) 195–202. [10] A. Barnsteiner, T. Lubinus, A. Di Gianvito, W. Schmid, K.- Engel, GC-based analysis of plant stanyl fatty acid esters in enriched foods, J. Agric. Food Chem. 59 (2011) 5204–5214. [11] A. Barnsteiner, R. Esche, A. di Gianvito, E. Chiavaro, W. Schmid, K.- Engel, Capillary gas chromatographic analysis of complex phytosteryl/-stanyl ester mixtures in enriched skimmed milk-drinking yoghurts, Food Control 27 (2012) 275–283. [12] J.L. Anderson, D.W. Armstrong, High-stability ionic liquids. A new class of stationary phases for gas chromatography, Anal. Chem. 75 (2003) 4851–4858. [13] C. Ragonese, D. Sciarrone, P.Q. Tranchida, P. Dugo, G. Dugo, L. Mondello, Evaluation of a medium-polarity ionic liquid stationary phase in the analysis of flavor and fragrance compounds, Anal. Chem. 83 (2011) 7947–7954. [14] P. Delmonte, A. Fardin Kia, J.K.G. Kramer, M.M. Mossoba, L. Sidisky, J.I. Rader, Separation characteristics of fatty acid methyl esters using SLB-IL111, a new ionic liquid coated capillary gas chromatographic column, J. Chromatogr. A 1218 (2011) 545–554. [15] M. Schröder, W. Vetter, High-speed counter-current chromatographic separation of phytosterols, Anal. Bioanal. Chem. 400 (2011) 3615–3623. [16] S. Hammann, C. Wendlinger, W. Vetter, Analysis of intact cholesteryl esters of furan fatty acids in cod liver, Lipids 50 (2015) 611–620. [17] W.R. Lusby, M.J. Thompson, J. Kochansky, Analysis of sterol esters by capillary gas chromatography-electron impact and chemical ionization-mass spectrometry, Lipids 19 (1984) 888–901.