Journal of Chromatography A, 1394 (2015) 89–94
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Isolation of two 5 polymethylene interrupted fatty acids from Podocarpus falcatus by countercurrent chromatography夽 Simon Hammann, Markus Schröder, Carolin Schmidt, 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 30 October 2014 Received in revised form 17 March 2015 Accepted 18 March 2015 Available online 24 March 2015 Keywords: Countercurrent chromatography Podocarpus falcatus Polyunsaturated fatty acid Gas chromatography Mass spectrometry
a b s t r a c t The lipids of gymnosperms frequently feature unusual polyunsaturated fatty acids (PUFAs) such as sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) showing a first double bond on C-5 which is separated from the next double bond by five methylene units. Compared to “classic” fatty acids, these fatty acids are not easily commercially available and their prices are quite high. For this reason, we wished to isolate those fatty acids from the seed oil of Podocarpus falcatus by countercurrent chromatography (CCC) after conversion of the fatty acids to methyl esters (FAMEs). The contribution of sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) in the unfractionated sample was 10% and 6% respectively, while oleic acid (18:19) and linoleic acid (18:29,12) were the major fatty acids. After a first CCC run with FAMEs from Podocarpus falcatus, fractions enriched in the target compounds were chosen for subsequent isolation by means of two subsequent CCC runs. Initially, 13 mg of juniperonic acid was recovered with a purity of 92% according to analysis by gas chromatography with mass spectrometry (GC/MS). Further purification of this fraction yielded 2.7 mg with a purity of 99% according to GC/MS. The isolation of sciadonic acid was hampered by high amounts of linoleic acid with the same equivalent chain length in suitable fractions of the first CCC separation. After an enrichment step by CCC, the critical pair sciadonic acid and linoleic acid was finally separated as free fatty acids. After this step, 4.4 mg of sciadonic acid was recovered with 99% purity. The methodology could also be applied to isolate larger amounts of those fatty acids or for the isolation of other minor fatty acids. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Polyunsaturated fatty acids (PUFAs), especially of the n-3 family are food components with high nutritional value due to beneficial health effects which have been linked with their consumption [1,2]. Classic PUFAs are characterized by two or more double bonds, separated by one methylene unit, respectively. Relevant representatives of the n-3 family are ␣-linolenic acid (18:39,12,15), which is abundant in linseed oil as well as eicosapentaenoic acid (20:55,8,11,14,17; EPA) and docosahexaenoic acid (22:64,7,10,13,16,19; DHA), which are commonly encountered in fish lipids [3]. By contrast, PUFAs of gymnosperms frequently bear the first double bond on C-5 (5 unsaturation) which is separated by five methylene units from the next double bond [4,5]. Major
夽 Presented at the 8th International Conference on Countercurrent Chromatography - CCC 2014, 23–25 July 2014, Uxbridge, United Kingdom. ∗ Corresponding author. Tel.: +49 711 459 24016; fax: +49 711 459 24377. E-mail address: [email protected] (W. Vetter). http://dx.doi.org/10.1016/j.chroma.2015.03.042 0021-9673/© 2015 Elsevier B.V. All rights reserved.
representatives for this type of fatty acids are sciadonic acid (20:35,11,14, Fig. 1a) and juniperonic acid (20:45,11,14,17, Fig. 1b) [5]. Contrary to the classic PUFAs of the n-3-family, only little is known about the biological activity of these fatty acids. Initial studies suggested anti-inflammatory properties of sciadonic acid while juniperonic acid showed antiproliferative properties on swiss 3T3 cells [6,7]. In addition, sciadonic acid is one of the fatty acids used to calculate the diagnostic index for the determination of the origin of different pine nuts [8]. In contrast to fatty acids such as palmitic acid (16:0) or oleic acid (18:19), sciadonic and juniperonic acid are not readily available in larger amounts and prices for standards are high. Their isolation has been achieved by repeated (argentation) thin layer chromatography or argentation liquid chromatography in combination with solid phase extraction on florisil [6,9]. In addition, juniperonic acid was synthesized in eight steps from eicosapentaenoic acid [10]. Either way, the procedures were laborious and/or difficult to scale up. Goal of our work was to isolate sciadonic and juniperonic acid from the seed oil of Podocarpus falcatus by means of countercurrent chromatography (CCC). This technique, which is based on
90
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
a)
b)
Irel
c)
18:2∆9,12
18:1∆9
20:3∆ 5,11,14 ~10% 20:2∆11,14 18:3∆9,12,15 16:0
20:2 ∆5,11 20:1∆11
18:0
20:4 (∆ 5,11,14,17 ~6%
20:0
19.0
20.0
21.0
22.0
23.0
24.0
25.0
26.0
27.0
28.0
29.0
30.0
t [min]
Fig. 1. Structures of the target compounds (a) 20:35,11,14 (sciadonic acid) and (b) 20:45,11,14,17 (juniperonic acid) and (c) GC/MS full scan chromatogram of the transesterified sample before countercurrent chromatography separation.
the distribution of analytes in two immiscible liquid phases, has been successfully used for the isolation and enrichment of fatty acids and fatty acid esters [11–14]. Prior to the CCC fractionation, the fatty acids in the seed oil were converted into fatty acid methyl esters (FAMEs) because fractions containing FAMEs can be directly analyzed by gas chromatography with mass spectrometry (GC/MS).
2. Experimental 2.1. Chemicals and samples n-Hexane, methanol (both HPLC grade) and acetonitrile (ACN, >99.9%) were from Th. Geyer (Renningen, Germany). Technical grade ethanol (distilled prior to use) and sulfuric acid were from BASF (Ludwigshafen, Germany). Acetic acid (≥99.7%) and diethyl ether (purity >99.8%) were from Sigma Aldrich (Steinheim, Germany). Pyrrolidine (>99.8%) was from Acros Organics (Geel, Belgium) and HCl (32%) was from Merck (Darmstadt, Germany). Sodium chloride (>99.8%) and potassium hydroxide (KOH, >85%) were from Carl Roth (Karlsruhe/Germany). Deuterated chloroform (CDCl3 , 99.8%) was from Deutero (Kastellaun, Germany). Trimethylsulfoniumhydroxide TMSH, 0.2 M in methanol) was from Macherey-Nagel (Düren, Germany), Methyl-tert-butyl ether (MTBE; purest, distilled before use) was from Fisher Scientific (Leicestershire, Great Britain) and Helium 5.0 was from Westfalengas (Münster, Germany). Seed oil of Podocarpus falcatus was produced by local communities around the village of Robe (Arsi zone) in Oromia/Ethiopia in 2010 for personal consumption and local markets. The fruits were collected, the seeds were roasted, grinded and subsequently extracted with boiling water. The oil layer was then removed. The oil was placed in an amber bottle and stored at 4 ◦ C in a dark place prior to analysis.
2.2. Preparation of fatty acid methyl esters (FAMEs) from Podocarpus falcatus seed oil Oil of Podocarpus falcatus (2.19 g) was weighed into a 250 mL round bottom flask and 30 mL 1% sulfuric acid in methanol was added. The mixture was stirred at 50 ◦ C for 12 h. Afterwards, 20 mL of saturated sodium chloride solution was added. The FAME solution was extracted 3 times with 20 mL n-hexane. The organic phases were united and the solvent was removed with a rotary evaporator. An aliquot of this sample (0.4428 g) was diluted in 4 mL of each upper and lower phase of the solvent system and used for CCC separation. An aliquot corresponding with about 0.1 mg of FAMEs was diluted in 1 mL of n-hexane and analyzed by GC/MS to determine the fatty acid composition of the unfractionated sample. 2.3. Saponification of fatty acid methyl esters (FAMEs) For saponification of FAMEs, fractions 8-10 of CCC-4 (see below) were transferred into a 10 mL test tube and the solvent removed by a gentle stream of nitrogen. Afterwards, 1.8 mL ethanol and 0.2 mL 50% KOH in water were added. The tube was sealed and heated to 80 ◦ C for one hour. After cooling, 1 mL water was added. The solution was acidified with 32% HCl and the free fatty acids were extracted with 2 mL n-hexane. 2.4. Countercurrent chromatography (CCC) CCC separations were performed with a PTR-1000 CCC instrument (PharmaTech Research, Baltimore, MD/USA) equipped with a 325 mL coil and a 4 mL or 10 mL sample loop as described before [15]. Solvents were pumped with a G55A-12 pump (Merck, Darmstadt, Germany) and fractions were collected with an Isco Retriever 500 fraction collector (Teledyne Isco, Lincoln, NE, USA). Suitable solvent systems were selected from those previously investigated by Schröder and Vetter [13]. The following fractionations were carried out.
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
2.4.1. CCC-1: Enrichment of sciadonic acid and juniperonic acid FAMEs from P. falcatus (0.4428 g) were fractionated with n-hexane/methanol/water 700:350:4 (v/v/v) in the tail-to-head mode, i.e. the upper phase was used as mobile phase, with a flow rate of 1.0 mL min−1 . The rotation speed was set to 1000 rpm and a 10 mL sample loop was used for injection. Before injecting the sample, the system was equilibrated and 51 mL of stationary phase were displaced. Initially, 3 fractions of 14 mL were collected followed by 21 fractions of 11 mL. Accordingly, a total of 284 mL was collected which corresponds to 87% of the total coil volume. The solvent was removed in a heating block with a gentle stream of nitrogen. The residue was re-diluted in n-hexane and analyzed by GC/MS. Fractions high in the target compounds were used for further CCC separations.
2.4.2. CCC-2: Enrichment of juniperonic acid Fractions 14 and 15 of CCC-1 were combined (133 mg) and fractionated with n-hexane/ACN 1:1 (v/v) in the tail-to-head mode with a flow rate of 1.0 mL. The rotation speed was set to 1000 rpm and injections were made with a 10 mL sample loop. The displacement of stationary phase was 44 mL. In total, 4 fractions of 15 mL and 10 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen. The residue was re-diluted in n-hexane and analyzed by GC/MS.
2.4.3. CCC-3: Purification of juniperonic acid For separation of juniperonic acid from small amounts of ␣-linolenic acid, fractions 15 and 16 (9.6 mg) of CCC-2 were fractionated with the solvent system n-hexane/methanol/water 700:350:4 (v/v/v) in the tail-to-head mode. The flow rate was set to 1.0 mL min−1 and the rotation speed to 1000 rpm. The sample (9.6 mg) was injected with a 10 mL sample loop. The displacement of stationary phase was 60 mL. Ten fractions of 15 mL and 15 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-diluted in n-hexane and analyzed by GC/MS.
2.4.4. CCC-4: Enrichment of sciadonic acid An aliquot of fraction 14 (59 mg) of CCC-1 was fractionated with n-hexane/ACN 1:1 (v/v) in the tail-to-head mode with a flow rate of 1 mL min−1 . The rotation speed was set to 950 rpm and the sample was injected with a 4 mL sample loop. The displacement of stationary phase was 32 mL. After 50 min run time, 11 fractions of 8 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-dissolved in n-hexane and analyzed by GC/MS.
2.4.5. CCC-5: Purification of sciadonic acid FAMEs of fractions 8–10 of CCC-4 (15.3 mg) were saponified (Section 2.3), the free fatty acids were extracted with n-hexane and subjected to a further CCC separation. n-Heptane/methanol/water 400:364:36 (v/v/v) according to Bousquet and Le Goffic [11] was used as solvent system in head-to-tail mode with a flow rate of 2 mL min−1 . The displacement of stationary phase was 43 mL. The rotation speed was set to 1000 rpm and the sample was injected with a 4 mL sample loop. After 60 min, 15 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-diluted in n-hexane. For analysis, aliquots corresponding with about 50 g of FA were placed into a vial, the solvent was evaporated and the residue re-diluted in 1 mL of MTBE. After addition of 50 L TMSH solution in methanol, the sample was used for GC/MS analysis.
91
2.5. Preparation of pyrrolidide derivatives from fatty acid methyl esters To determine the double bond positions of unsaturated fatty acids, the FAMEs of the unfractionated sample as well as aliquots of fraction 12 of CCC-1, fraction 15 of CCC-3 and fraction 11 of CCC5 were converted into pyrrolidides as described before [14,16]. In short, FAMEs were diluted in pyrrolidine, acetic acid was added and the solution was heated. Afterwards, pyrrolidides were extracted with n-hexane/diethyl ether, the organic phase washed, dried and used for GC/MS analysis. 2.6. Gas chromatography with mass spectrometry (GC/MS) GC/MS analysis of FAMEs was performed on a 5890 series II GC coupled to a 5971A mass spectrometer and a 7673A autosampler (Hewlett-Packard/Agilent, Waldbronn, Germany). Two serially connected 30 m RTX-2330 columns (10% cyanopropylphenyl, 90% bis(cyanopropyl) polysiloxane, internal diameter 0.25 mm, film thickness 0.1 m, Restek, Bellefonte, PA, USA) were installed in the GC oven and He (5.0 quality) was used as the carrier gas at a flow rate of 1.0 mL min−1 . The injector temperature was set to 250 ◦ C and the transfer line temperature to 280 ◦ C. The GC oven was programmed as follows: After 1 min at 60 ◦ C, the temperature was increased at 6 ◦ C min−1 to 150 ◦ C, then at 4 ◦ C min−1 to 190 ◦ C and finally at 7 ◦ C min−1 to 250 ◦ C. This final temperature was held for 7 min. The total run duration was 41.6 min. GC/MS analysis was conducted in full scan mode (m/z 50–500) after a solvent delay of 7 min. Pyrrolidides were analyzed on a 6890 series GC coupled to a 5973 mass spectrometer and a 7683 autosampler (HewlettPackard/Agilent, Waldbronn, Germany). The injector temperature was set to 250 ◦ C and the transfer line temperature to 280 ◦ C. A 30 m HP-5MS column (internal diameter 0.25 mm, film thickness 0.25 m) was used for the separations. Helium 5.0 was used as carrier gas with a flow rate of 1.0 mL min−1 . The oven was programmed as follows: After 1 min at 60 ◦ C, the temperature was increased at 15 ◦ C min−1 to 200 ◦ C, and then at 2 ◦ C min−1 to 270 ◦ C. This final temperature was held for 5 min. The total run duration was 50.33 min. The analysis was conducted in full scan mode and data was collected from m/z 50-750 after a solvent delay of 15 min. 2.7. Nuclear magnetic resonance spectroscopy (NMR) For NMR analysis, a total of 4.5 mg of fractions 12 and 13 of CCC-5 (20:35,11,14) was converted into FAMEs as described before (section 2.2.). Also, 2.5 mg from fraction 15 of CCC-3 (20:45,11,14,17) were used. One-dimensional 1 H- and 13 C-NMR spectra were measured with a Varian Inova 300 MHz instrument fitted with a 5 mm ATB sample probe. All spectra were taken in CDCl3 as solvent and chemical shifts were reported relative to the solvent signal of CDCl3 on the TMS scale. 3. Results and discussion 3.1. Fatty acid composition of Podocarpus falcatus seed oil In the unfractionated sample of the FAME fraction of P. falcatus seed oil, 11 fatty acids could be identified, i.e. palmitic acid (16:0), stearic acid (18:0), oleic acid (18:19), linoleic acid (18:29,12), ␣-linolenic acid (18:39,12,15), arachidic acid (20:0), gondoic acid (20:111, Fig. S1, supporting material), keteleeronic acid (20:25,11, Fig. S2, supporting material), bishomolinoleic acid (20:211,14, Fig. S3, supporting material), sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) (Table 1). The double bond positions of monoenoic and dienoic fatty acids
92
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
Table 1 Fatty acid composition of the transesterified seed oil of Podocarpus falcatus determined by GC/MS and distribution of these fatty acids over fractions of CCC-1. “x” means a minor contribution to the fraction (0–10%), “xx” a significant contribution (10–40%) and “xxx” a major compound (>40% contribution) in the fraction. Bold fractions were used for further purifications. Please note, that only fatty acids, which were identified in the sample before CCC fractionation are listed. Fatty acid
contribution
16:0 18:0 18:19 18:29,12 18:39,12,15 20:0 20:111 20:25,11 20:211,14 20:35,11,14 20:45,11,14,17
4.3% 3.2% 37.1% 28.6% 2.2% 0.4% 2.4% 1.1% 4.9% 9.7% 6.0%
1
2
3
4
5
6
7
8
9 x
x
xxx x
10
11
12
13
14
xx x xxx xx
xx
x
xxx
x xxx xxx
xxx xx
x xxx
x xxx
were determined by GC/MS spectra using the pyrrolidide derivatives (section 2.5.). The two major fatty acids oleic acid and linoleic acid contributed >50% to the total fatty acids (Fig. 1c, Table 1). The target fatty acids sciadonic acid and juniperonic acid added another ∼10% and 6% to the fatty acids in the sample (Fig. 1c, Table 1). Saturated fatty acids (palmitic acid, stearic acid and arachidic acid) contributed only about 8% to the total fatty acids (Table 1). Sciadonic acid and juniperonic acid had been identified before in Podocarpus species such as Podocarpus nagi and Podocarpus anidus [9,17,18], and also Podocarpus falcatus showed appreciable amounts of both fatty acids. The seed oil of Podocarpus falcatus did not contain any isomers of sciadonic acid and juniperonic acid such as arachidonic acid (20:45,8,11,14) and thus was regarded a good sample for the isolation of the both fatty acids by CCC. To describe the elution of the fatty acids, we used the partition coefficient K (Eq. (1)). K=
VR − VM VS
xx xx xx
x xx xx
x xx
xx x
15
16
17
18
19
xxx x
xx xx
x xxx
xxx
x
x xx
xxx
xxx
x
21
22
23
24
25
which was lower abundant than juniperonic acid in the unfractionated sample (Table 1). Even though these two fatty acids eluted into nearly the same fractions, fractions 14 and 15 showed a high ratio of juniperonic acid to ␣-linolenic acid and could be used for further isolation (Table 1, Fig. 2, Fig. 3a). It has to be noted, that these fractions also contained 16 mg of sciadonic acid (36% of initial amount, Table 1). The general elution characteristics of the FAMEs corresponded well to previous work with this solvent system [12–14]. 3.3. Isolation of juniperonic acid 3.3.1. CCC-2 Fractions 14 and 15 (K = 0.37-0.45) of CCC-1 were combined for the isolation of juniperonic acid (Fig. 2). This FA contributed about 10% to the total fatty acids in these combined fractions and ␣-linolenic acid was the only FAME with the same ECL in these
(1)
0.4428 g of FAMEs
With VR = retention volume, VM = volume of the mobile phase in the coil and VS = volume of the stationary phase in the coil. 3.2. Enrichment of the target fatty acids with CCC (CCC-1) and general elution characteristics Separation of fatty acids by CCC is hampered by the equivalent chain length (ECL) rule (Eq. (2)): ECL = N − 2n
20
(2)
where N is the number of carbons in the acyl chain and n the number of double bonds [11]. Fatty acids with the same ECL are difficult to separate by CCC [11,12]. Since FAMEs with the same ECL as the target fatty acids were present in the sample, a direct isolation in one CCC run was not considered feasible. Hence, the initial CCC separation was applied to produce fractions rich in sciadonic acid and juniperonic acid, which could be used for purification. Performed in the tail-to-head mode, this method eluted long-chain saturated FAMEs first (Table 1). The first eluting fatty acids, 23:0 to 30:0, were detected in fractions 6 and 7 (K = 0.05–0.13). These long chain saturated fatty acids had not been detected in the unfractionated sample (section 3.1.). While fraction 8 (K = 0.13–0.17) contained saturated fatty acids from 20:0 to 25:0, the first unsaturated FAME, i.e. gondoic acid (20:111) was detected in fraction 9 (K = 0.17–0.21). Sciadonic acid eluted into fractions 12–15 (K = 0.29–0.45) and was generally accompanied by the abundant linoleic acid (Table 1) due to the same ECL of 14. Juniperonic acid eluted into fractions 14–18 (K = 0.37–0.57) and thus partly overlapped with sciadonic acid (Table 1). The only fatty acid present in the sample sharing the ECL of 12 of juniperonic acid was ␣-linolenic acid (K = 0.41–0.61),
CCC-1 25 fracons of 15 mL fracon 14+15 (133 mg, P=10%, R=50%) CCC-2 4 fracons of 15 mL, 10 fracons of 10 mL fracon 14+15 (9.6 mg, P=92%, R =36%) CCC-3 10 fracons of 15 mL, 15 fracons of 10 mL
fracon 13 (59 mg, P=22%, R=30%) CCC-4 11 fracons of 8 mL fracon 8-10 (15.3 mg, P=55%, R=19%) Saponificaon CCC-5 15 fracons of 10 mL
Juniperonic acid Sciadonic acid (2.7 mg, P=99%, R=10%) (4.4 mg, P=99%, R=10%) Fig. 2. Scheme of the five countercurrent chromatography (CCC) fractionations which were conducted for the isolation of 20:45,11,14,17 (juniperonic acid) (CCC1,2,3) and 20:35,11,14 (sciadonic acid) (CCC-1,4,5). Purities (P) and recoveries (R, relative to initial amount) of the target fatty acids as well as the fractions and total mass used for subsequent CCC fractionation are given next to the arrows.
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
Irel
Irel
18:2∆9,12
CCC-1, Fraction 15
19.0 21.0 CCC-2 Fraction 15
a)
Irel
20:4∆ 5,11,14,17 ~22% 18:3∆9,12,15 20:3 ∆ 5,11,14
23.0
25.0
27.0
29.0
18:1∆9
CCC-1 Fraction 13
20:3∆ 5,11,14 ~22%
18:2∆9,12
d)
16:0 20:2∆11,14
31.0 t [min]
20:4∆ 5,11,14,17 ~92%
93
b)
19.0
21.0
23.0
Irel CCC-4 Fraction 8
25.0
18:2∆9,12
27.0
29.0 t [min]
20:3∆ 5,11,14 ~55%
e)
18:3∆9,12,15
19.0 Irel
21.0
23.0
25.0
21.0
29.0
31.0 t [min]
20:4∆ 5,11,14,17 ~99%
CCC-3 Fraction 15
19.0
27.0
23.0
25.0
27.0
29.0
c)
31.0 t [min]
19.0 Irel
21.0
23.0
25.0
29.0 t [min]
20:3∆ 5,11,14 ~99%
CCC-5 Fraction 10
19.0
27.0
21.0
23.0
25.0
27.0
f)
29.0 t [min]
Fig. 3. GC/MS chromatogram of chosen CCC fractions showing the progress of purification of (a–c) 20:45,11,14,17 (juniperonic acid) and (d–f) of 20:35,11,14 (sciadonic acid). Please note, that the GC column was replaced during this work, which lead to slight changes in GC retention times.
fractions, contributing ∼0.5%. However, the isolation was complicated by the presence of linoleic acid which represented ∼75% of the fraction. Due to the different ECL (12 vs. 14) of juniperonic acid and linoleic acid, the separation was deemed manageable but it required additional attention for the next step. We chose nhexane/ACN as solvent system because previous shake flask tests had indicated a slightly different selectivity of this solvent system compared to n-hexane/methanol/water based systems [13]. Fractions 14–18 (K = 0.38–0.56) of CCC-2 contained a total of 13.2 mg of juniperonic acid with a purity of >90% (GC/MS) along with small amounts (10%) of ␣-linolenic acid (Fig. 3b, Table S1, supporting material). Compared to the amount initially injected into CCC-1 (26.6 mg), the yield of juniperonic acid was still 50% while the purity increased from 6% to 90%. 3.3.2. CCC-3 Fractions 15 and 16 of CCC-2 (9.6 mg) were used for a further purification to remove last residual amounts of ␣-linolenic acid. For removal of the latter with the same ECL, a small portion of water was added to the solvent system. While water in the solvent system decreases the solubility of FAMEs, it accelerates the elution of longer FAMEs with the same ECL. Accordingly, juniperonic acid was expected to elute slightly prior to ␣-linolenic acid. In agreement with that, fraction 14 and 15 of CCC-3 provided a total 2.7 mg of juniperonic acid with a purity >99% (GC/MS) (10% of initial amount) and the subsequent fraction 16 yielded another 2.7 mg with a purity of 95% (Fig. 3c, Table S2, supporting material). The identity of the fatty acid was confirmed by 1 H- and 13 C-NMR as well as GC/MS analysis of the pyrrolidide derivative (Table 2, Figs. S4 and S5, Table S5, supporting material). 3.4. Isolation of sciadonic acid 3.4.1. CCC-4 Fraction 13 of CCC-1 was found to be best suited for further enrichment of sciadonic acid since it showed an appreciable contribution of the target compound (22 mg; ∼50% of the initial amount). Isolation of sciadonic acid was considered to be more challenging than the isolation of juniperonic acid due to the presence of the abundant linoleic acid with the same ECL of 14 (Table 1, Fig. 3d). In
fraction 13, sciadonic acid contributed about 22% and linoleic acid about 19% to the total fatty acids in the fraction, while oleic acid, palmitic acid, and bishomolinoleic acid (20:211,14) contributed 49%, 7% and 3%, respectively. We thus followed the approach of Bousquet and Le Goffic, who achieved a separation of PUFA with the same ECL as free fatty acids with n-heptane/methanol/water as the solvent system [11]. As a result, fractionation of fraction 13 of CCC-1 provided a satisfactory separation of linoleic acid and sciadonic acid. However, palmitic acid and oleic acid were still eluted together with sciadonic acid (data not shown). Accordingly, these fatty acids had to be removed before the separation of the critical ECL pair. For this reason, 59 mg of fraction 13 of CCC-1 (as FAMEs) were subjected to CCC separation using n-hexane/ACN 1:1 (v/v) as solvent system (CCC-4, Fig. 2). With this solvent system, linoleic acid and sciadonic acid largely co-eluted (K = 0.17–0.32),
Table 2 13 C-NMR chemical shifts in ppm of the carbon atoms of 20:35,11,14 and 20:45,11,14,17) (both as methyl esters) in the present study and in literature [1,2]. Carbon number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 OCH3
20:35,11,14 methyl ester
20:45,11,14,17 methyl ester
Present study
Literature
Present study
Literature
174.30 33.63 25.03 26.69 128.63 131.10 27.36 29.49 29.46 27.27 130.40 128.29 25.79 128.04 130.08 27.27 29.44 31.68 22.72 14.21 51.61
174.15 33.54 25.00 26.68 128.49 130.92 27.23 29.43 29.43 27.23 130.21 128.16 25.76 128.16 130.21 27.23 29.43 31.64 22.68 14.16 51.43
174.15 33.46 24.88 26.54 128.48 130.91 27.19 29.26 29.30 27.10 130.11 127.80 25.61 128.22 128.30 25.54 127.10 131.96 20.54 14.26 51.46
174.15 33.55 25.00 26.64 128.48 130.92 27.23 29.39 29.39 27.23 130.13 127.79 25.70 128.26 128.26 25.70 127.19 131.95 20.66 14.39 51.42
94
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
while a good selectivity for the separation from palmitic acid and oleic acid (K = 0.09–0.17) was achieved (Table S3, supporting material). Accordingly, four fractions containing only linoleic acid and sciadonic acid could be collected after the fatty acids palmitic acid and oleic acid were eluted from the system (Table S3, supporting material). In these fractions, sciadonic acid made up around 55% (GC/MS) with a total of 8.4 mg (19% of the initial amount) (Fig. 3e).
a scale up of the method combined the usage of an oil source with a higher initial contribution of those fatty acids. Also, the reported methodology offers the possibility to enrich and isolate minor fatty acids, even when compounds with the same ECL are present. For the production of unusual and bioactive fatty acids, CCC definitely represents a valid alternative to synthesis or solid phase extraction methods.
3.4.2. CCC-5 After saponification of fractions 8–10 of CCC-4 (15.3 mg), sciadonic acid and linoleic acid could be separated as free fatty acids with n-heptane/methanol/water 400:364:36 (v/v/v) as solvent system (Fig. 2), and a total of 4.4 mg (∼10% of the initial amount) of sciadonic acid was obtained in fractions 9–13 of CCC-5 with a purity of >99% (GC/MS) (Fig. 3f, Table S4, supporting material). Identity and purity of the fatty acid was confirmed by 1 H- and 13 C-NMR as well as GC/MS analysis of the pyrrolidide derivative (Table 2, Figs. S6 and S7, Table S6, supporting material).
Acknowledgments
3.5. Comparison of the procedures for the isolation of sciadonic acid and juniperonic acid from seed oil of Podocarpus falcatus
We gratefully acknowledge financial support and a stipend grant to Simon Hammann by the Fonds der Chemischen Industrie. Furthermore, we would like to thank Jörg Volkmann and Dr. Veronika Scherbaum for making the sample available to us. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2015.03.042. References
Due to the structural similarity of several fatty acids in the sample, baseline separation could not be achieved. Several co-elutions were observed especially for fatty acids having the same ECL. On these occasions, we distributed the fatty acids into different fractions and selected appropriate fractions where the target fatty acids were enriched. This peak cutting technique helped to improve the purity of the target compounds at the expense of the recovery. This is a common problem in the chromatographic separation of structurally similar compounds. Accordingly, the major drop in recovery for both juniperonic acid and sciadonic occurred when CCC-1 fractions were chosen for further purification. In this initial step, ∼50% of juniperonic acid and ∼70% of sciadonic were not taken forward. This share of both compounds could be re-fractionated in order to improve the yield, but this procedure would significantly increase the required labor. At this point, the chromatographer has to decide whether the yield is the most relevant factor (e.g. because a certain amount is required for follow-up steps) or if partial sacrifice is acceptable because enough material is available. Since CCC benefits from higher sample loads compared to HPLC (and a higher separation power compared to column chromatography), it is frequently more reasonable to accept a partial loss of a compound in favor of a higher purity and a lower workload. Isolation of sciadonic acid was hampered by the presence of high amounts of the major fatty acid linoleic acid in all relevant fractions (Table 1). Since both fatty acids share the ECL of 14, their separation in form of FAMEs turned out to be difficult. Furthermore, linoleic acid was at least equally concentrated in the fractions of CCC-1, which necessitated a more complicated purification protocol. In the present study, we aimed for a high purity of the fatty acids, which almost necessarily leads to a lower yield. If a purity of 95% or 90% was deemed acceptable, the yield could be surely increased, as was seen for juniperonic acid. Here, the yield dropped from 50% to 10% of the initial amount when the purity was increased from 90% to 99%. 4. Conclusions The fatty acids sciadonic acid and juniperonic acid could be isolated from the seed oil of Podocarpus falcatus by multiple CCC separations with purities >99%. Identity of the fatty acids could be confirmed and those fatty acids could now be used as standard substances for analytical purposes. If higher amounts of the fatty acids are required for bioactivity studies, this may be achieved by
[1] C.H.S. Ruxton, S.C. Reed, M.J.A. Simpson, K.J. Millington, The health benefits of omega-3 polyunsaturated fatty acids: A review of the evidence, J. Hum. Nutr. Diet. 17 (2004) 449–459. [2] L.A. Horrocks, Y.K. Yeo, Health benefits of docosahexaenoic acid (DHA), Pharmacol. Res. 40 (1999) 211–225. [3] I. Gill, R. Valivety, Polyunsaturated fatty acids, part 1: Occurrence, biological activities and applications, Trends Biotechnol. 15 (1997) 401–409. [4] R.L. Wolff, O. Lavialle, F. Pédrono, E. Pasquier, L.G. Deluc, A.M. Marpeau, K. Aitzetmüller, Fatty acid composition of Pinaceae as taxonomic markers, Lipids 36 (2001) 439–451. [5] R.L. Wolff, W.W. Christie, Structures, practical sources (gymnosperm seeds), gas-liquid chromatographic data (equivalent chain lengths), and mass spectrometric characteristics of all-cis 5-olefinic acids, Eur. J. Lipid Sci. Tech. 104 (2002) 234–244. [6] A. Berger, I. Monnard, M. Baur, C. Charbonnet, I. Safonova, A. Jomard, Epidermal anti-inflammatory properties of 5,11,14 20:3: Effects on mouse ear edema, PGE2 levels in cultured keratinocytes, and PPAR activation, Lipids Health Dis. 1 (2002) 1–12. [7] J. Morishige, N. Amano, K. Hirano, H. Nishio, T. Tanaka, K. Satouchi, Inhibitory effect of juniperonic acid (-5c,11c,14c,17c-20:4, -3) on bombesininduced proliferation of Swiss 3T3 cells, Biol. Pharm. Bull. 31 (2008) 1786–1789. [8] F. Destaillats, C. Cruz-Hernandez, F. Giuffrida, F. Dionisi, Identification of the botanical origin of pine nuts found in food products by gas-liquid chromatography analysis of fatty acid profile, J. Agric. Food Chem. 58 (2010) 2082–2087. [9] T. Takagi, Y. Itabashi, cis-5-olefinic unusual fatty acids in seed lipids of gymnospermae and their distribution in triacylglycerols, Lipids 17 (1982) 716–723. [10] A. Vik, T.V. Hansen, A.K. Holmeide, L. Skattebøl, Synthesis of juniperonic acid, Tetrahedron Lett. 51 (2010) 2852–2854. [11] O. Bousquet, F. Le Goffic, Counter-current chromatographic separation of polyunsaturated fatty acids, J. Chromatogr. A 704 (1995) 211–216. [12] D. Li, M. Schröder, W. Vetter, Isolation of 6,9,12,15-hexadecatetraenoic fatty acid (16:4n-1) methyl ester from transesterified fish oil by HSCCC, Chromatographia 75 (2012) 1–6. [13] M. Schröder, W. Vetter, Detection of 430 fatty acid methyl esters from a transesterified butter sample, J. Am. Oil Chem. Soc. 90 (2013) 771–790. [14] S. Hammann, U. Tillmann, M. Schröder, W. Vetter, Profiling the fatty acids from a strain of the microalgae Alexandrium tamarense by means of high-speed counter-current chromatography and gas chromatography coupled with mass spectrometry, J. Chromatogr. A 1312 (2013) 93–103. [15] M. Schröder, W. Vetter, High-speed counter-current chromatographic separation of phytosterols, Anal. Bioanal. Chem. 400 (2011) 3615–3623. [16] J.F. Rontani, S. Christodoulou, M. Koblizek, GC-MS structural characterization of fatty acids from marine aerobic anoxygenic phototrophic bacteria, Lipids 40 (2005) 97–108. [17] T. Takagi, 5,11,14-Eicosatrienoic acid in podocarpus nagi seed oil, J. Am. Oil Chem. Soc. 41 (1964) 516–519. [18] R.L. Wolff, F. Pédrono, A.M. Marpeau, W.W. Christie, F.D. Gunstone, The seed fatty acid composition and the distribution of (5-olefinic acids in the triacylglycerols of some Taxaceae (Taxus and Torreya), J. Am. Oil Chem. Soc. 75 (1998) 1637–1641.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Isolation of two 5 polymethylene interrupted fatty acids from Podocarpus falcatus by countercurrent chromatography夽 Simon Hammann, Markus Schröder, Carolin Schmidt, 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 30 October 2014 Received in revised form 17 March 2015 Accepted 18 March 2015 Available online 24 March 2015 Keywords: Countercurrent chromatography Podocarpus falcatus Polyunsaturated fatty acid Gas chromatography Mass spectrometry
a b s t r a c t The lipids of gymnosperms frequently feature unusual polyunsaturated fatty acids (PUFAs) such as sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) showing a first double bond on C-5 which is separated from the next double bond by five methylene units. Compared to “classic” fatty acids, these fatty acids are not easily commercially available and their prices are quite high. For this reason, we wished to isolate those fatty acids from the seed oil of Podocarpus falcatus by countercurrent chromatography (CCC) after conversion of the fatty acids to methyl esters (FAMEs). The contribution of sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) in the unfractionated sample was 10% and 6% respectively, while oleic acid (18:19) and linoleic acid (18:29,12) were the major fatty acids. After a first CCC run with FAMEs from Podocarpus falcatus, fractions enriched in the target compounds were chosen for subsequent isolation by means of two subsequent CCC runs. Initially, 13 mg of juniperonic acid was recovered with a purity of 92% according to analysis by gas chromatography with mass spectrometry (GC/MS). Further purification of this fraction yielded 2.7 mg with a purity of 99% according to GC/MS. The isolation of sciadonic acid was hampered by high amounts of linoleic acid with the same equivalent chain length in suitable fractions of the first CCC separation. After an enrichment step by CCC, the critical pair sciadonic acid and linoleic acid was finally separated as free fatty acids. After this step, 4.4 mg of sciadonic acid was recovered with 99% purity. The methodology could also be applied to isolate larger amounts of those fatty acids or for the isolation of other minor fatty acids. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Polyunsaturated fatty acids (PUFAs), especially of the n-3 family are food components with high nutritional value due to beneficial health effects which have been linked with their consumption [1,2]. Classic PUFAs are characterized by two or more double bonds, separated by one methylene unit, respectively. Relevant representatives of the n-3 family are ␣-linolenic acid (18:39,12,15), which is abundant in linseed oil as well as eicosapentaenoic acid (20:55,8,11,14,17; EPA) and docosahexaenoic acid (22:64,7,10,13,16,19; DHA), which are commonly encountered in fish lipids [3]. By contrast, PUFAs of gymnosperms frequently bear the first double bond on C-5 (5 unsaturation) which is separated by five methylene units from the next double bond [4,5]. Major
夽 Presented at the 8th International Conference on Countercurrent Chromatography - CCC 2014, 23–25 July 2014, Uxbridge, United Kingdom. ∗ Corresponding author. Tel.: +49 711 459 24016; fax: +49 711 459 24377. E-mail address: [email protected] (W. Vetter). http://dx.doi.org/10.1016/j.chroma.2015.03.042 0021-9673/© 2015 Elsevier B.V. All rights reserved.
representatives for this type of fatty acids are sciadonic acid (20:35,11,14, Fig. 1a) and juniperonic acid (20:45,11,14,17, Fig. 1b) [5]. Contrary to the classic PUFAs of the n-3-family, only little is known about the biological activity of these fatty acids. Initial studies suggested anti-inflammatory properties of sciadonic acid while juniperonic acid showed antiproliferative properties on swiss 3T3 cells [6,7]. In addition, sciadonic acid is one of the fatty acids used to calculate the diagnostic index for the determination of the origin of different pine nuts [8]. In contrast to fatty acids such as palmitic acid (16:0) or oleic acid (18:19), sciadonic and juniperonic acid are not readily available in larger amounts and prices for standards are high. Their isolation has been achieved by repeated (argentation) thin layer chromatography or argentation liquid chromatography in combination with solid phase extraction on florisil [6,9]. In addition, juniperonic acid was synthesized in eight steps from eicosapentaenoic acid [10]. Either way, the procedures were laborious and/or difficult to scale up. Goal of our work was to isolate sciadonic and juniperonic acid from the seed oil of Podocarpus falcatus by means of countercurrent chromatography (CCC). This technique, which is based on
90
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
a)
b)
Irel
c)
18:2∆9,12
18:1∆9
20:3∆ 5,11,14 ~10% 20:2∆11,14 18:3∆9,12,15 16:0
20:2 ∆5,11 20:1∆11
18:0
20:4 (∆ 5,11,14,17 ~6%
20:0
19.0
20.0
21.0
22.0
23.0
24.0
25.0
26.0
27.0
28.0
29.0
30.0
t [min]
Fig. 1. Structures of the target compounds (a) 20:35,11,14 (sciadonic acid) and (b) 20:45,11,14,17 (juniperonic acid) and (c) GC/MS full scan chromatogram of the transesterified sample before countercurrent chromatography separation.
the distribution of analytes in two immiscible liquid phases, has been successfully used for the isolation and enrichment of fatty acids and fatty acid esters [11–14]. Prior to the CCC fractionation, the fatty acids in the seed oil were converted into fatty acid methyl esters (FAMEs) because fractions containing FAMEs can be directly analyzed by gas chromatography with mass spectrometry (GC/MS).
2. Experimental 2.1. Chemicals and samples n-Hexane, methanol (both HPLC grade) and acetonitrile (ACN, >99.9%) were from Th. Geyer (Renningen, Germany). Technical grade ethanol (distilled prior to use) and sulfuric acid were from BASF (Ludwigshafen, Germany). Acetic acid (≥99.7%) and diethyl ether (purity >99.8%) were from Sigma Aldrich (Steinheim, Germany). Pyrrolidine (>99.8%) was from Acros Organics (Geel, Belgium) and HCl (32%) was from Merck (Darmstadt, Germany). Sodium chloride (>99.8%) and potassium hydroxide (KOH, >85%) were from Carl Roth (Karlsruhe/Germany). Deuterated chloroform (CDCl3 , 99.8%) was from Deutero (Kastellaun, Germany). Trimethylsulfoniumhydroxide TMSH, 0.2 M in methanol) was from Macherey-Nagel (Düren, Germany), Methyl-tert-butyl ether (MTBE; purest, distilled before use) was from Fisher Scientific (Leicestershire, Great Britain) and Helium 5.0 was from Westfalengas (Münster, Germany). Seed oil of Podocarpus falcatus was produced by local communities around the village of Robe (Arsi zone) in Oromia/Ethiopia in 2010 for personal consumption and local markets. The fruits were collected, the seeds were roasted, grinded and subsequently extracted with boiling water. The oil layer was then removed. The oil was placed in an amber bottle and stored at 4 ◦ C in a dark place prior to analysis.
2.2. Preparation of fatty acid methyl esters (FAMEs) from Podocarpus falcatus seed oil Oil of Podocarpus falcatus (2.19 g) was weighed into a 250 mL round bottom flask and 30 mL 1% sulfuric acid in methanol was added. The mixture was stirred at 50 ◦ C for 12 h. Afterwards, 20 mL of saturated sodium chloride solution was added. The FAME solution was extracted 3 times with 20 mL n-hexane. The organic phases were united and the solvent was removed with a rotary evaporator. An aliquot of this sample (0.4428 g) was diluted in 4 mL of each upper and lower phase of the solvent system and used for CCC separation. An aliquot corresponding with about 0.1 mg of FAMEs was diluted in 1 mL of n-hexane and analyzed by GC/MS to determine the fatty acid composition of the unfractionated sample. 2.3. Saponification of fatty acid methyl esters (FAMEs) For saponification of FAMEs, fractions 8-10 of CCC-4 (see below) were transferred into a 10 mL test tube and the solvent removed by a gentle stream of nitrogen. Afterwards, 1.8 mL ethanol and 0.2 mL 50% KOH in water were added. The tube was sealed and heated to 80 ◦ C for one hour. After cooling, 1 mL water was added. The solution was acidified with 32% HCl and the free fatty acids were extracted with 2 mL n-hexane. 2.4. Countercurrent chromatography (CCC) CCC separations were performed with a PTR-1000 CCC instrument (PharmaTech Research, Baltimore, MD/USA) equipped with a 325 mL coil and a 4 mL or 10 mL sample loop as described before [15]. Solvents were pumped with a G55A-12 pump (Merck, Darmstadt, Germany) and fractions were collected with an Isco Retriever 500 fraction collector (Teledyne Isco, Lincoln, NE, USA). Suitable solvent systems were selected from those previously investigated by Schröder and Vetter [13]. The following fractionations were carried out.
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
2.4.1. CCC-1: Enrichment of sciadonic acid and juniperonic acid FAMEs from P. falcatus (0.4428 g) were fractionated with n-hexane/methanol/water 700:350:4 (v/v/v) in the tail-to-head mode, i.e. the upper phase was used as mobile phase, with a flow rate of 1.0 mL min−1 . The rotation speed was set to 1000 rpm and a 10 mL sample loop was used for injection. Before injecting the sample, the system was equilibrated and 51 mL of stationary phase were displaced. Initially, 3 fractions of 14 mL were collected followed by 21 fractions of 11 mL. Accordingly, a total of 284 mL was collected which corresponds to 87% of the total coil volume. The solvent was removed in a heating block with a gentle stream of nitrogen. The residue was re-diluted in n-hexane and analyzed by GC/MS. Fractions high in the target compounds were used for further CCC separations.
2.4.2. CCC-2: Enrichment of juniperonic acid Fractions 14 and 15 of CCC-1 were combined (133 mg) and fractionated with n-hexane/ACN 1:1 (v/v) in the tail-to-head mode with a flow rate of 1.0 mL. The rotation speed was set to 1000 rpm and injections were made with a 10 mL sample loop. The displacement of stationary phase was 44 mL. In total, 4 fractions of 15 mL and 10 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen. The residue was re-diluted in n-hexane and analyzed by GC/MS.
2.4.3. CCC-3: Purification of juniperonic acid For separation of juniperonic acid from small amounts of ␣-linolenic acid, fractions 15 and 16 (9.6 mg) of CCC-2 were fractionated with the solvent system n-hexane/methanol/water 700:350:4 (v/v/v) in the tail-to-head mode. The flow rate was set to 1.0 mL min−1 and the rotation speed to 1000 rpm. The sample (9.6 mg) was injected with a 10 mL sample loop. The displacement of stationary phase was 60 mL. Ten fractions of 15 mL and 15 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-diluted in n-hexane and analyzed by GC/MS.
2.4.4. CCC-4: Enrichment of sciadonic acid An aliquot of fraction 14 (59 mg) of CCC-1 was fractionated with n-hexane/ACN 1:1 (v/v) in the tail-to-head mode with a flow rate of 1 mL min−1 . The rotation speed was set to 950 rpm and the sample was injected with a 4 mL sample loop. The displacement of stationary phase was 32 mL. After 50 min run time, 11 fractions of 8 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-dissolved in n-hexane and analyzed by GC/MS.
2.4.5. CCC-5: Purification of sciadonic acid FAMEs of fractions 8–10 of CCC-4 (15.3 mg) were saponified (Section 2.3), the free fatty acids were extracted with n-hexane and subjected to a further CCC separation. n-Heptane/methanol/water 400:364:36 (v/v/v) according to Bousquet and Le Goffic [11] was used as solvent system in head-to-tail mode with a flow rate of 2 mL min−1 . The displacement of stationary phase was 43 mL. The rotation speed was set to 1000 rpm and the sample was injected with a 4 mL sample loop. After 60 min, 15 fractions of 10 mL were collected. The solvent was removed in a heating block with a gentle stream of nitrogen, the residue re-diluted in n-hexane. For analysis, aliquots corresponding with about 50 g of FA were placed into a vial, the solvent was evaporated and the residue re-diluted in 1 mL of MTBE. After addition of 50 L TMSH solution in methanol, the sample was used for GC/MS analysis.
91
2.5. Preparation of pyrrolidide derivatives from fatty acid methyl esters To determine the double bond positions of unsaturated fatty acids, the FAMEs of the unfractionated sample as well as aliquots of fraction 12 of CCC-1, fraction 15 of CCC-3 and fraction 11 of CCC5 were converted into pyrrolidides as described before [14,16]. In short, FAMEs were diluted in pyrrolidine, acetic acid was added and the solution was heated. Afterwards, pyrrolidides were extracted with n-hexane/diethyl ether, the organic phase washed, dried and used for GC/MS analysis. 2.6. Gas chromatography with mass spectrometry (GC/MS) GC/MS analysis of FAMEs was performed on a 5890 series II GC coupled to a 5971A mass spectrometer and a 7673A autosampler (Hewlett-Packard/Agilent, Waldbronn, Germany). Two serially connected 30 m RTX-2330 columns (10% cyanopropylphenyl, 90% bis(cyanopropyl) polysiloxane, internal diameter 0.25 mm, film thickness 0.1 m, Restek, Bellefonte, PA, USA) were installed in the GC oven and He (5.0 quality) was used as the carrier gas at a flow rate of 1.0 mL min−1 . The injector temperature was set to 250 ◦ C and the transfer line temperature to 280 ◦ C. The GC oven was programmed as follows: After 1 min at 60 ◦ C, the temperature was increased at 6 ◦ C min−1 to 150 ◦ C, then at 4 ◦ C min−1 to 190 ◦ C and finally at 7 ◦ C min−1 to 250 ◦ C. This final temperature was held for 7 min. The total run duration was 41.6 min. GC/MS analysis was conducted in full scan mode (m/z 50–500) after a solvent delay of 7 min. Pyrrolidides were analyzed on a 6890 series GC coupled to a 5973 mass spectrometer and a 7683 autosampler (HewlettPackard/Agilent, Waldbronn, Germany). The injector temperature was set to 250 ◦ C and the transfer line temperature to 280 ◦ C. A 30 m HP-5MS column (internal diameter 0.25 mm, film thickness 0.25 m) was used for the separations. Helium 5.0 was used as carrier gas with a flow rate of 1.0 mL min−1 . The oven was programmed as follows: After 1 min at 60 ◦ C, the temperature was increased at 15 ◦ C min−1 to 200 ◦ C, and then at 2 ◦ C min−1 to 270 ◦ C. This final temperature was held for 5 min. The total run duration was 50.33 min. The analysis was conducted in full scan mode and data was collected from m/z 50-750 after a solvent delay of 15 min. 2.7. Nuclear magnetic resonance spectroscopy (NMR) For NMR analysis, a total of 4.5 mg of fractions 12 and 13 of CCC-5 (20:35,11,14) was converted into FAMEs as described before (section 2.2.). Also, 2.5 mg from fraction 15 of CCC-3 (20:45,11,14,17) were used. One-dimensional 1 H- and 13 C-NMR spectra were measured with a Varian Inova 300 MHz instrument fitted with a 5 mm ATB sample probe. All spectra were taken in CDCl3 as solvent and chemical shifts were reported relative to the solvent signal of CDCl3 on the TMS scale. 3. Results and discussion 3.1. Fatty acid composition of Podocarpus falcatus seed oil In the unfractionated sample of the FAME fraction of P. falcatus seed oil, 11 fatty acids could be identified, i.e. palmitic acid (16:0), stearic acid (18:0), oleic acid (18:19), linoleic acid (18:29,12), ␣-linolenic acid (18:39,12,15), arachidic acid (20:0), gondoic acid (20:111, Fig. S1, supporting material), keteleeronic acid (20:25,11, Fig. S2, supporting material), bishomolinoleic acid (20:211,14, Fig. S3, supporting material), sciadonic acid (20:35,11,14) and juniperonic acid (20:45,11,14,17) (Table 1). The double bond positions of monoenoic and dienoic fatty acids
92
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
Table 1 Fatty acid composition of the transesterified seed oil of Podocarpus falcatus determined by GC/MS and distribution of these fatty acids over fractions of CCC-1. “x” means a minor contribution to the fraction (0–10%), “xx” a significant contribution (10–40%) and “xxx” a major compound (>40% contribution) in the fraction. Bold fractions were used for further purifications. Please note, that only fatty acids, which were identified in the sample before CCC fractionation are listed. Fatty acid
contribution
16:0 18:0 18:19 18:29,12 18:39,12,15 20:0 20:111 20:25,11 20:211,14 20:35,11,14 20:45,11,14,17
4.3% 3.2% 37.1% 28.6% 2.2% 0.4% 2.4% 1.1% 4.9% 9.7% 6.0%
1
2
3
4
5
6
7
8
9 x
x
xxx x
10
11
12
13
14
xx x xxx xx
xx
x
xxx
x xxx xxx
xxx xx
x xxx
x xxx
were determined by GC/MS spectra using the pyrrolidide derivatives (section 2.5.). The two major fatty acids oleic acid and linoleic acid contributed >50% to the total fatty acids (Fig. 1c, Table 1). The target fatty acids sciadonic acid and juniperonic acid added another ∼10% and 6% to the fatty acids in the sample (Fig. 1c, Table 1). Saturated fatty acids (palmitic acid, stearic acid and arachidic acid) contributed only about 8% to the total fatty acids (Table 1). Sciadonic acid and juniperonic acid had been identified before in Podocarpus species such as Podocarpus nagi and Podocarpus anidus [9,17,18], and also Podocarpus falcatus showed appreciable amounts of both fatty acids. The seed oil of Podocarpus falcatus did not contain any isomers of sciadonic acid and juniperonic acid such as arachidonic acid (20:45,8,11,14) and thus was regarded a good sample for the isolation of the both fatty acids by CCC. To describe the elution of the fatty acids, we used the partition coefficient K (Eq. (1)). K=
VR − VM VS
xx xx xx
x xx xx
x xx
xx x
15
16
17
18
19
xxx x
xx xx
x xxx
xxx
x
x xx
xxx
xxx
x
21
22
23
24
25
which was lower abundant than juniperonic acid in the unfractionated sample (Table 1). Even though these two fatty acids eluted into nearly the same fractions, fractions 14 and 15 showed a high ratio of juniperonic acid to ␣-linolenic acid and could be used for further isolation (Table 1, Fig. 2, Fig. 3a). It has to be noted, that these fractions also contained 16 mg of sciadonic acid (36% of initial amount, Table 1). The general elution characteristics of the FAMEs corresponded well to previous work with this solvent system [12–14]. 3.3. Isolation of juniperonic acid 3.3.1. CCC-2 Fractions 14 and 15 (K = 0.37-0.45) of CCC-1 were combined for the isolation of juniperonic acid (Fig. 2). This FA contributed about 10% to the total fatty acids in these combined fractions and ␣-linolenic acid was the only FAME with the same ECL in these
(1)
0.4428 g of FAMEs
With VR = retention volume, VM = volume of the mobile phase in the coil and VS = volume of the stationary phase in the coil. 3.2. Enrichment of the target fatty acids with CCC (CCC-1) and general elution characteristics Separation of fatty acids by CCC is hampered by the equivalent chain length (ECL) rule (Eq. (2)): ECL = N − 2n
20
(2)
where N is the number of carbons in the acyl chain and n the number of double bonds [11]. Fatty acids with the same ECL are difficult to separate by CCC [11,12]. Since FAMEs with the same ECL as the target fatty acids were present in the sample, a direct isolation in one CCC run was not considered feasible. Hence, the initial CCC separation was applied to produce fractions rich in sciadonic acid and juniperonic acid, which could be used for purification. Performed in the tail-to-head mode, this method eluted long-chain saturated FAMEs first (Table 1). The first eluting fatty acids, 23:0 to 30:0, were detected in fractions 6 and 7 (K = 0.05–0.13). These long chain saturated fatty acids had not been detected in the unfractionated sample (section 3.1.). While fraction 8 (K = 0.13–0.17) contained saturated fatty acids from 20:0 to 25:0, the first unsaturated FAME, i.e. gondoic acid (20:111) was detected in fraction 9 (K = 0.17–0.21). Sciadonic acid eluted into fractions 12–15 (K = 0.29–0.45) and was generally accompanied by the abundant linoleic acid (Table 1) due to the same ECL of 14. Juniperonic acid eluted into fractions 14–18 (K = 0.37–0.57) and thus partly overlapped with sciadonic acid (Table 1). The only fatty acid present in the sample sharing the ECL of 12 of juniperonic acid was ␣-linolenic acid (K = 0.41–0.61),
CCC-1 25 fracons of 15 mL fracon 14+15 (133 mg, P=10%, R=50%) CCC-2 4 fracons of 15 mL, 10 fracons of 10 mL fracon 14+15 (9.6 mg, P=92%, R =36%) CCC-3 10 fracons of 15 mL, 15 fracons of 10 mL
fracon 13 (59 mg, P=22%, R=30%) CCC-4 11 fracons of 8 mL fracon 8-10 (15.3 mg, P=55%, R=19%) Saponificaon CCC-5 15 fracons of 10 mL
Juniperonic acid Sciadonic acid (2.7 mg, P=99%, R=10%) (4.4 mg, P=99%, R=10%) Fig. 2. Scheme of the five countercurrent chromatography (CCC) fractionations which were conducted for the isolation of 20:45,11,14,17 (juniperonic acid) (CCC1,2,3) and 20:35,11,14 (sciadonic acid) (CCC-1,4,5). Purities (P) and recoveries (R, relative to initial amount) of the target fatty acids as well as the fractions and total mass used for subsequent CCC fractionation are given next to the arrows.
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
Irel
Irel
18:2∆9,12
CCC-1, Fraction 15
19.0 21.0 CCC-2 Fraction 15
a)
Irel
20:4∆ 5,11,14,17 ~22% 18:3∆9,12,15 20:3 ∆ 5,11,14
23.0
25.0
27.0
29.0
18:1∆9
CCC-1 Fraction 13
20:3∆ 5,11,14 ~22%
18:2∆9,12
d)
16:0 20:2∆11,14
31.0 t [min]
20:4∆ 5,11,14,17 ~92%
93
b)
19.0
21.0
23.0
Irel CCC-4 Fraction 8
25.0
18:2∆9,12
27.0
29.0 t [min]
20:3∆ 5,11,14 ~55%
e)
18:3∆9,12,15
19.0 Irel
21.0
23.0
25.0
21.0
29.0
31.0 t [min]
20:4∆ 5,11,14,17 ~99%
CCC-3 Fraction 15
19.0
27.0
23.0
25.0
27.0
29.0
c)
31.0 t [min]
19.0 Irel
21.0
23.0
25.0
29.0 t [min]
20:3∆ 5,11,14 ~99%
CCC-5 Fraction 10
19.0
27.0
21.0
23.0
25.0
27.0
f)
29.0 t [min]
Fig. 3. GC/MS chromatogram of chosen CCC fractions showing the progress of purification of (a–c) 20:45,11,14,17 (juniperonic acid) and (d–f) of 20:35,11,14 (sciadonic acid). Please note, that the GC column was replaced during this work, which lead to slight changes in GC retention times.
fractions, contributing ∼0.5%. However, the isolation was complicated by the presence of linoleic acid which represented ∼75% of the fraction. Due to the different ECL (12 vs. 14) of juniperonic acid and linoleic acid, the separation was deemed manageable but it required additional attention for the next step. We chose nhexane/ACN as solvent system because previous shake flask tests had indicated a slightly different selectivity of this solvent system compared to n-hexane/methanol/water based systems [13]. Fractions 14–18 (K = 0.38–0.56) of CCC-2 contained a total of 13.2 mg of juniperonic acid with a purity of >90% (GC/MS) along with small amounts (10%) of ␣-linolenic acid (Fig. 3b, Table S1, supporting material). Compared to the amount initially injected into CCC-1 (26.6 mg), the yield of juniperonic acid was still 50% while the purity increased from 6% to 90%. 3.3.2. CCC-3 Fractions 15 and 16 of CCC-2 (9.6 mg) were used for a further purification to remove last residual amounts of ␣-linolenic acid. For removal of the latter with the same ECL, a small portion of water was added to the solvent system. While water in the solvent system decreases the solubility of FAMEs, it accelerates the elution of longer FAMEs with the same ECL. Accordingly, juniperonic acid was expected to elute slightly prior to ␣-linolenic acid. In agreement with that, fraction 14 and 15 of CCC-3 provided a total 2.7 mg of juniperonic acid with a purity >99% (GC/MS) (10% of initial amount) and the subsequent fraction 16 yielded another 2.7 mg with a purity of 95% (Fig. 3c, Table S2, supporting material). The identity of the fatty acid was confirmed by 1 H- and 13 C-NMR as well as GC/MS analysis of the pyrrolidide derivative (Table 2, Figs. S4 and S5, Table S5, supporting material). 3.4. Isolation of sciadonic acid 3.4.1. CCC-4 Fraction 13 of CCC-1 was found to be best suited for further enrichment of sciadonic acid since it showed an appreciable contribution of the target compound (22 mg; ∼50% of the initial amount). Isolation of sciadonic acid was considered to be more challenging than the isolation of juniperonic acid due to the presence of the abundant linoleic acid with the same ECL of 14 (Table 1, Fig. 3d). In
fraction 13, sciadonic acid contributed about 22% and linoleic acid about 19% to the total fatty acids in the fraction, while oleic acid, palmitic acid, and bishomolinoleic acid (20:211,14) contributed 49%, 7% and 3%, respectively. We thus followed the approach of Bousquet and Le Goffic, who achieved a separation of PUFA with the same ECL as free fatty acids with n-heptane/methanol/water as the solvent system [11]. As a result, fractionation of fraction 13 of CCC-1 provided a satisfactory separation of linoleic acid and sciadonic acid. However, palmitic acid and oleic acid were still eluted together with sciadonic acid (data not shown). Accordingly, these fatty acids had to be removed before the separation of the critical ECL pair. For this reason, 59 mg of fraction 13 of CCC-1 (as FAMEs) were subjected to CCC separation using n-hexane/ACN 1:1 (v/v) as solvent system (CCC-4, Fig. 2). With this solvent system, linoleic acid and sciadonic acid largely co-eluted (K = 0.17–0.32),
Table 2 13 C-NMR chemical shifts in ppm of the carbon atoms of 20:35,11,14 and 20:45,11,14,17) (both as methyl esters) in the present study and in literature [1,2]. Carbon number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 OCH3
20:35,11,14 methyl ester
20:45,11,14,17 methyl ester
Present study
Literature
Present study
Literature
174.30 33.63 25.03 26.69 128.63 131.10 27.36 29.49 29.46 27.27 130.40 128.29 25.79 128.04 130.08 27.27 29.44 31.68 22.72 14.21 51.61
174.15 33.54 25.00 26.68 128.49 130.92 27.23 29.43 29.43 27.23 130.21 128.16 25.76 128.16 130.21 27.23 29.43 31.64 22.68 14.16 51.43
174.15 33.46 24.88 26.54 128.48 130.91 27.19 29.26 29.30 27.10 130.11 127.80 25.61 128.22 128.30 25.54 127.10 131.96 20.54 14.26 51.46
174.15 33.55 25.00 26.64 128.48 130.92 27.23 29.39 29.39 27.23 130.13 127.79 25.70 128.26 128.26 25.70 127.19 131.95 20.66 14.39 51.42
94
S. Hammann et al. / J. Chromatogr. A 1394 (2015) 89–94
while a good selectivity for the separation from palmitic acid and oleic acid (K = 0.09–0.17) was achieved (Table S3, supporting material). Accordingly, four fractions containing only linoleic acid and sciadonic acid could be collected after the fatty acids palmitic acid and oleic acid were eluted from the system (Table S3, supporting material). In these fractions, sciadonic acid made up around 55% (GC/MS) with a total of 8.4 mg (19% of the initial amount) (Fig. 3e).
a scale up of the method combined the usage of an oil source with a higher initial contribution of those fatty acids. Also, the reported methodology offers the possibility to enrich and isolate minor fatty acids, even when compounds with the same ECL are present. For the production of unusual and bioactive fatty acids, CCC definitely represents a valid alternative to synthesis or solid phase extraction methods.
3.4.2. CCC-5 After saponification of fractions 8–10 of CCC-4 (15.3 mg), sciadonic acid and linoleic acid could be separated as free fatty acids with n-heptane/methanol/water 400:364:36 (v/v/v) as solvent system (Fig. 2), and a total of 4.4 mg (∼10% of the initial amount) of sciadonic acid was obtained in fractions 9–13 of CCC-5 with a purity of >99% (GC/MS) (Fig. 3f, Table S4, supporting material). Identity and purity of the fatty acid was confirmed by 1 H- and 13 C-NMR as well as GC/MS analysis of the pyrrolidide derivative (Table 2, Figs. S6 and S7, Table S6, supporting material).
Acknowledgments
3.5. Comparison of the procedures for the isolation of sciadonic acid and juniperonic acid from seed oil of Podocarpus falcatus
We gratefully acknowledge financial support and a stipend grant to Simon Hammann by the Fonds der Chemischen Industrie. Furthermore, we would like to thank Jörg Volkmann and Dr. Veronika Scherbaum for making the sample available to us. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2015.03.042. References
Due to the structural similarity of several fatty acids in the sample, baseline separation could not be achieved. Several co-elutions were observed especially for fatty acids having the same ECL. On these occasions, we distributed the fatty acids into different fractions and selected appropriate fractions where the target fatty acids were enriched. This peak cutting technique helped to improve the purity of the target compounds at the expense of the recovery. This is a common problem in the chromatographic separation of structurally similar compounds. Accordingly, the major drop in recovery for both juniperonic acid and sciadonic occurred when CCC-1 fractions were chosen for further purification. In this initial step, ∼50% of juniperonic acid and ∼70% of sciadonic were not taken forward. This share of both compounds could be re-fractionated in order to improve the yield, but this procedure would significantly increase the required labor. At this point, the chromatographer has to decide whether the yield is the most relevant factor (e.g. because a certain amount is required for follow-up steps) or if partial sacrifice is acceptable because enough material is available. Since CCC benefits from higher sample loads compared to HPLC (and a higher separation power compared to column chromatography), it is frequently more reasonable to accept a partial loss of a compound in favor of a higher purity and a lower workload. Isolation of sciadonic acid was hampered by the presence of high amounts of the major fatty acid linoleic acid in all relevant fractions (Table 1). Since both fatty acids share the ECL of 14, their separation in form of FAMEs turned out to be difficult. Furthermore, linoleic acid was at least equally concentrated in the fractions of CCC-1, which necessitated a more complicated purification protocol. In the present study, we aimed for a high purity of the fatty acids, which almost necessarily leads to a lower yield. If a purity of 95% or 90% was deemed acceptable, the yield could be surely increased, as was seen for juniperonic acid. Here, the yield dropped from 50% to 10% of the initial amount when the purity was increased from 90% to 99%. 4. Conclusions The fatty acids sciadonic acid and juniperonic acid could be isolated from the seed oil of Podocarpus falcatus by multiple CCC separations with purities >99%. Identity of the fatty acids could be confirmed and those fatty acids could now be used as standard substances for analytical purposes. If higher amounts of the fatty acids are required for bioactivity studies, this may be achieved by
[1] C.H.S. Ruxton, S.C. Reed, M.J.A. Simpson, K.J. Millington, The health benefits of omega-3 polyunsaturated fatty acids: A review of the evidence, J. Hum. Nutr. Diet. 17 (2004) 449–459. [2] L.A. Horrocks, Y.K. Yeo, Health benefits of docosahexaenoic acid (DHA), Pharmacol. Res. 40 (1999) 211–225. [3] I. Gill, R. Valivety, Polyunsaturated fatty acids, part 1: Occurrence, biological activities and applications, Trends Biotechnol. 15 (1997) 401–409. [4] R.L. Wolff, O. Lavialle, F. Pédrono, E. Pasquier, L.G. Deluc, A.M. Marpeau, K. Aitzetmüller, Fatty acid composition of Pinaceae as taxonomic markers, Lipids 36 (2001) 439–451. [5] R.L. Wolff, W.W. Christie, Structures, practical sources (gymnosperm seeds), gas-liquid chromatographic data (equivalent chain lengths), and mass spectrometric characteristics of all-cis 5-olefinic acids, Eur. J. Lipid Sci. Tech. 104 (2002) 234–244. [6] A. Berger, I. Monnard, M. Baur, C. Charbonnet, I. Safonova, A. Jomard, Epidermal anti-inflammatory properties of 5,11,14 20:3: Effects on mouse ear edema, PGE2 levels in cultured keratinocytes, and PPAR activation, Lipids Health Dis. 1 (2002) 1–12. [7] J. Morishige, N. Amano, K. Hirano, H. Nishio, T. Tanaka, K. Satouchi, Inhibitory effect of juniperonic acid (-5c,11c,14c,17c-20:4, -3) on bombesininduced proliferation of Swiss 3T3 cells, Biol. Pharm. Bull. 31 (2008) 1786–1789. [8] F. Destaillats, C. Cruz-Hernandez, F. Giuffrida, F. Dionisi, Identification of the botanical origin of pine nuts found in food products by gas-liquid chromatography analysis of fatty acid profile, J. Agric. Food Chem. 58 (2010) 2082–2087. [9] T. Takagi, Y. Itabashi, cis-5-olefinic unusual fatty acids in seed lipids of gymnospermae and their distribution in triacylglycerols, Lipids 17 (1982) 716–723. [10] A. Vik, T.V. Hansen, A.K. Holmeide, L. Skattebøl, Synthesis of juniperonic acid, Tetrahedron Lett. 51 (2010) 2852–2854. [11] O. Bousquet, F. Le Goffic, Counter-current chromatographic separation of polyunsaturated fatty acids, J. Chromatogr. A 704 (1995) 211–216. [12] D. Li, M. Schröder, W. Vetter, Isolation of 6,9,12,15-hexadecatetraenoic fatty acid (16:4n-1) methyl ester from transesterified fish oil by HSCCC, Chromatographia 75 (2012) 1–6. [13] M. Schröder, W. Vetter, Detection of 430 fatty acid methyl esters from a transesterified butter sample, J. Am. Oil Chem. Soc. 90 (2013) 771–790. [14] S. Hammann, U. Tillmann, M. Schröder, W. Vetter, Profiling the fatty acids from a strain of the microalgae Alexandrium tamarense by means of high-speed counter-current chromatography and gas chromatography coupled with mass spectrometry, J. Chromatogr. A 1312 (2013) 93–103. [15] M. Schröder, W. Vetter, High-speed counter-current chromatographic separation of phytosterols, Anal. Bioanal. Chem. 400 (2011) 3615–3623. [16] J.F. Rontani, S. Christodoulou, M. Koblizek, GC-MS structural characterization of fatty acids from marine aerobic anoxygenic phototrophic bacteria, Lipids 40 (2005) 97–108. [17] T. Takagi, 5,11,14-Eicosatrienoic acid in podocarpus nagi seed oil, J. Am. Oil Chem. Soc. 41 (1964) 516–519. [18] R.L. Wolff, F. Pédrono, A.M. Marpeau, W.W. Christie, F.D. Gunstone, The seed fatty acid composition and the distribution of (5-olefinic acids in the triacylglycerols of some Taxaceae (Taxus and Torreya), J. Am. Oil Chem. Soc. 75 (1998) 1637–1641.
