High Enantiomeric Excess of the Flavor Relevant 4-Alkyl-Branched

Dec 27, 2014 - The 4-alkyl-branched fatty acids, 4-methyloctanoic acid (4-Me-8:0), ... well as on octakis(2,3-di-O-methyl-6-O-tert-butyldimethylsilyl)...
0 downloads 0 Views 904KB Size
Download PDF
Article pubs.acs.org/JAFC

High Enantiomeric Excess of the Flavor Relevant 4‑Alkyl-Branched Fatty Acids in Milk Fat and Subcutaneous Adipose Tissue of Sheep and Goat Stefanie Kaffarnik, Carolina Heid, Yasemin Kayademir, Dorothee Eibler, and Walter Vetter* University of Hohenheim, Institute of Food Chemistry (170b), Garbenstraße 28, 70593 Stuttgart, Germany ABSTRACT: Volatile 4-alkyl-branched fatty acids are characteristic flavor compounds of sheep and goat. Due to the methyl branch, the carbon C-4 represents a stereogenic center with the possible presence of one or both enantiomers in the respective samples. In this study, we used enantioselective gas chromatography to study the enantiomeric composition of 4-methyloctanoic acid (4-Me-8:0) and 4-ethyloctanoic acid (4-Et-8:0) in milk and dairy products from sheep and goat as well as in goat subcutaneous tissue. Different columns coated with modified cyclodextrins were tested to resolve racemic 4-alkyl-branched fatty acid methyl ester standards. The best enantiomer resolution was obtained on 25% octakis(2,3,6-tri-O-ethyl)-γ-cyclodextrin (γTECD) diluted in OV-1701. For analysis of the food samples, the lipids were extracted and fatty acids in the extracts were converted into fatty acid methyl esters. Non-aqueous reversed-phase high-performance liquid chromatography was used to fractionate the samples in order to gain one solution enriched in 4-Me-8:0 methyl ester and one solution enriched with 4-Et-8:0 methyl ester. Subsequent analysis by enantioselective gas chromatography with mass spectrometry allowed only the detection of one enantiomer of 4-Me-8:0 and 4-Et-8:0 in the samples. By means of a non-racemic standard of 4-Me-8:0, it was found that the predominant enantiomer was (R)-4-Me-8:0. KEYWORDS: alkyl-branched-chain fatty acids, enantioselective analysis, modified cyclodextrins, goat, sheep, milk

■

INTRODUCTION The 4-alkyl-branched fatty acids, 4-methyloctanoic acid (4-Me8:0), 4-ethyloctanoic acid (4-Et-8:0), and 4-methylnonanoic acid (4-Me-9:0), were found to be responsible for the characteristic sweet, goaty and sheepy aroma of sheep and goat.1−4 The carbon at the branching point (i.e., C-4) represents a stereogenic center, due to its substitution with four chemically different moieties. Accordingly, 4-alkylbranched fatty acids may occur in both S- and R-forms in nature. In such situations, enantioselective studies may contribute to the understanding of the impact because enantiomers of the same molecule may have different properties including flavor notes. In addition, enantioselective analyses can provide insights into the biosynthesis of chiral compounds. Different modified cyclodextrins have been used for the enantiomer separation of chiral compounds by gas chromatography (GC). Cyclodextrins (CD) are cyclic oligosaccharides composed of six (α-cyclodextrin), seven (β-cyclodextrin), or eight (γ-cyclodextrin) (1,4)-linked α-glucopyranose units which have the three-dimensional shape of a hollow torus.5−7 Each glucopyranose unit of the cyclodextrin has hydroxyl groups in the 2-, 3-, and 6-positions which need to be modified for their applicability in GC. Suitable substituents are alkyl groups, acetyl groups, or silyl groups (especially tert-butyldimethylsilyl substituents).8,9 Unfortunately, the potential of a chiral stationary phase (CSP) for a successful enantiomer separation cannot be predicted and has to be tested individually. Previous research has shown that the enantiomer separation of chiral fatty acids becomes more difficult if the stereogenic center is remote from the functional carboxylic group. For instance, the food relevant anteiso-fatty acids (fatty acids with a methyl © 2014 American Chemical Society

branch on the antepenultimate carbon which is the stereogenic center) 12-methyltetradecanoic acid (a15:0) and 14-methylhexadecanoic acid (a17:0) have 10 methylene groups or more between the headgroup and the methyl branch and are difficult to enantioseparate.10 By contrast, the enantioseparation of racemic standards of free 2-, 3-, and 4-alkyl-branched fatty acids has been achieved on heptakis(2,3-di-O-methyl-6-O-tert-butyldimethylsilyl)-β-cyclodextrin (β-TBDM) as well as on octakis(2,3-di-O-methyl-6-O-tert-butyldimethylsilyl)-γ-cyclodextrin (γ-TBDM).11−14 The aim of this study was to analyze the enantiomer composition of 4-Me-8:0 and 4-Et-8:0 in sheep and goat samples. For the enantioselective analysis of racemic standards of 4-Me-8:0, 4-Et-8:0, and 4-Me-9:0 methyl esters, we tested four CSPs and finally applied a CSP consisting of 25% octakis(2,3,6-tri-O-ethyl)-γ-cyclodextrin (γ-TECD) diluted in OV-1701. For the analysis of samples, volatile fatty acids had to be enriched because they are present only as minor components in milk fat of goat and sheep as well as in sheep subcutaneous adipose tissue.15−19 For this purpose, the lipid phase was isolated and the fatty acids in the fraction were transferred into fatty acid methyl esters which were fractionated by non-aqueous reversed-phase high-performance liquid chromatography (RP-HPLC) similarly to the method of Hauff et al.10 Enriched fractions of the fatty acids were then analyzed by GC with mass spectrometry operated in the selected ion monitoring mode (GC/MS-SIM). Received: Revised: Accepted: Published: 469

July 29, 2014 December 22, 2014 December 27, 2014 December 27, 2014 DOI: 10.1021/jf505452u J. Agric. Food Chem. 2015, 63, 469−475

Article

Journal of Agricultural and Food Chemistry

■

particle size from Supelco (Taufkirchen, Germany) and one 25 cm length × 4.6 mm inner diameter C30-RP-HPLC column with 5 μm particle size from Trentec (Gerlingen, Germany). Because methyl esters of saturated fatty acids do not provide a response in UV or fluorescence detectors, the samples were fractionated without a detector. For this reason, 14 fractions were collected in intervals of 4 min (corresponding to a 2 mL volume). Subsequently, 0.5 mL of water and 0.5 mL of saturated aqueous sodium chloride solution as well as 200 μL of n-hexane were added to each fraction for extraction. The organic phase containing the fatty acid methyl esters was separated and analyzed by GC/MS. Further concentration of fractions 5 and 6 by a factor of 2 was carried out as follows. In order to prevent loss of the volatile analytes, 15 μL of methanol was added to 40 μL of the sample and the resulting azeotropic mixture of n-hexane and methanol (72:28; v/v) was carefully evaporated with a slight stream of nitrogen until the final volume was 20 μL. Column Testing by Gas Chromatography with Flame Ionization Detection (GC/FID). Enantioseparations of racemic standards of 4-methyl-branched fatty acid methyl esters on different CSPs were performed with a Hewlett-Packard 5890 series II GC equipped with a flame ionization detector (GC/FID), a 7673 auto sampler, and a split/splitless injector. The injection volume was set to 1 μL. Four columns with modified cyclodextrins were tested. The first two contained the same amount of the same modified cyclodextrin dissolved in two different polysiloxanes, i.e., 20% heptakis(6-O-tertbutyldimethylsilyl-2,3-di-O-methyl)-β-cyclodextrin (β-TBDM) in (i) SE-52 or (ii) PS086 (30 m length, 0.25 mm inner diameter, 0.1 μm film thickness, BGB Analytik, Adliswil, Switzerland). In addition, we used (iii) 66% heptakis(6-O-tert-butyldimethylsilyl-2,3-di-O-methyl)-βcyclodextrin (β-TBDM) in OV-1701 (30 m length, 0.25 mm inner diameter, 0.1 μm film thickness, BGB Analytik, Adliswil, Switzerland) and (iv) 25% triethylated-γ-cyclodextrin (γ-TECD) in OV-1701 (12 m, 0.25 mm inner diameter, film thickness not specified). The γTECD column (Table 1) was prepared in the laboratories of Prof. Dr. Michael Oehme (University of Basel, Switzerland).21−23 Nitrogen was used as the carrier gas with a column head pressure of 10 psi (30 m columns) or 6 psi (12 m γ-TECD). For enantioselective analysis on the β-TBDM column, the oven program started at 60 °C for 10 min, ramped 0.5 °C/min to a final temperature of 80 °C (βTBDM: method A). The performance on the γ-TECD column differed from that on the β-TBDM column. On γ-TECD, the isothermal temperature oven program was at 40 °C for 280 min (γTECD: method B). Enantioselective GC/MS-SIM Analysis of 4-Alkyl-Branched Fatty Acids as Methyl Esters in Samples. Enantioselective analyses were performed with a CP-3800/1200 GC/MS system equipped with a CP-8410 auto sampler (Varian, Darmstadt, Germany). The split/ splitless injector was heated to 250 °C and used in splitless mode (split opened after 2 min). The injection volume was 1, 2, or 3 μL, depending on the fatty acid concentration. Helium was used as the carrier gas with a flow rate of 1.2 mL/min. Electron ionization was carried out with 70 eV. The transfer line temperature was set at 40 °C, while the temperature of the ion source was 200 °C. The GC column (iv) (γ-TECD) was installed in the GC oven using a retention gap (1 m, 0.25 inner diameter). The analyses were performed isothermally at 40 °C. After the elution of the analytes, a second run was started for the elution of less volatile compounds in the sample solution. This run started for 8 min at 40 °C before the temperature was raised at 10 °C/ min to 150 °C (hold time 1 min). The new injection also served to rinse the injector by the injection and volatilization of the solvent. This program was repeated three times in order to elute higher boiling compounds from the column. During the third run, we could not observe peaks anymore in the chromatogram. In the selected ion monitoring (SIM) mode, m/z 74, 87, 99, 115, 129, 141, and 172 (M+ of 4-Me-8:0 methyl ester) were measured from 10 to 60 min followed by m/z 74, 87, 113, 115, 129, 155, and 186 (M+ of 4-Me-9:0-ME and 4-Et-8:0-ME) from 60 to 90 min (method C). The limit of quantitation (LOQ) was determined to be 0.34 mg/g for 4-Me-8:0ME as well as 0.78 mg/g for 4-Et-8:0-ME and 4-Me-9:0-ME. The limit of detection (LOD) was 0.10 mg/g for 4-Me-8:0-ME and 0.23 mg/g

MATERIALS AND METHODS

Chemicals. Methanol (purity >95%), n-hexane (purity >95%), and acetonitrile (purity 99.9%) were from Th. Geyer (Renningen, Germany). Concentrated sulfuric acid (>98%) was from BASF (Ludwigshafen, Germany), whereas sodium chloride was from Carl Roth (Karlsruhe, Germany). Helium (purity 5.0) and nitrogen (purity 5.0) were from Westfalen (Münster, Germany). Racemic standards of 4-methyloctanoic acid (4-Me-8:0), 4-ethyloctanoic acid (4-Et-8:0), and 4-methylnonanoic acid (4-Me-9:0) were from Endeavor Specialty Chemicals Ltd. (Daventry, U.K.). Racemic 4-methylhexanoic acid (4Me-6:0) was from Sigma-Aldrich (Steinheim, Germany). Two different standard mixtures were prepared and used in this study. The racemic standard mixture 1 consisted of the methyl esters of 4Me-6:0, 4-Me-8:0, and 4-Me-9:0, whereas the racemic standard mixture 2 consisted of the methyl esters of 4-Me-8:0, 4-Me-9:0, and 4Et-8:0. The concentration of each compound in the mixtures was 10 ng/μL. Because pure enantiomer standards were not commercially available, a non-racemic standard of free 4-methyloctanoic acid was produced as follows. Racemic 4-Me-8:0 was derivatized with (−)-menthol to give the corresponding 4-Me-8:0 (−)-menthyl ester (i.e., a mixture of two diastereomers). The diastereomers were separated by reversed-phase HPLC using four serially linked columns (similarly to the method shown below for the isolation of individual fatty acids) and 100% acetonitrile as the mobile phase.20 From 80 to 100 min, 40 fractions were taken every 0.5 min and the (−)-menthyl esters were saponified with KOH, the solutions acidified with HCl, and the free 4-Me-8:0 extracted with n-hexane. Enantioselective analysis of the RP-HPLC fractions allowed selecting one solution showing a distinct non-racemic composition of 4-Me-8:0. Aliquots of this fraction were also transferred in the corresponding methyl ester. Samples. In a recent study, we developed a method for the direct quantitation of volatile 4-alkyl-branched fatty acid methyl ester in sheep and goat samples.18 The samples with quantitative data for 4Me-8:0, 4-Et-8:0, and 4-Me-9:0 were used in the present study for enantioselective determinations.18,19 In brief, we used two goat milk and two sheep milk samples and nine cheese samples made from goat or sheep milk. These samples were bought in retail stores close to Stuttgart (Germany) in 2013/2014. In addition, two samples of homogenized sheep subcutaneous adipose tissue were analyzed in this study. The milk and cheese samples were lyophilized and extracted by accelerated solvent extraction before transesterification, as previously described,19 while the sheep subcutaneous adipose tissue was directly subjected to transesterification.18 Transesterification Procedure. An aliquot of 3 mL of the milk fat extract (fat content ∼1 g, solvent evaporated thereafter) or ∼1 g of subcutaneous adipose tissue and 50 mL of 1% sulfuric acid in methanol was placed in a 100 mL round flask. The sample solution was stirred and refluxed for 2 h at 80 °C. After cooling the mixture on ice, it followed the addition of 20 mL of distilled water, 30 mL of saturated aqueous sodium chloride solution, and 500 μL of n-hexane. The two-phase system was vigorously shaken, and the upper organic phase was separated for RP-HPLC enrichment. Non-Aqueous Reversed-Phase High-Performance Liquid Chromatography (RP-HPLC). The fractionation was based on the method of Hauff et al.10 developed for anteiso-fatty acids with modifications. About 0.2 mL of acetonitrile was placed in a 1.5 mL GC vial, and 0.2 mL of the sample extract in n-hexane was added. After shaking, the upper n-hexane phase was evaporated by a gentle stream of nitrogen at 30 °C. The fatty acid methyl esters remained in the more polar acetonitrile phase. This solution could directly be used for RP-HPLC analysis, because acetonitrile is miscible with the mobile phase methanol. Enrichment of 4-methyl-branched fatty acid methyl esters was performed with a Waters HPLC system (Milford, USA) consisting of a 717 plus auto sampler, a 616 pump, a 996 photodiode array detector (DAD), and a 600S controller. Sample solutions (20 μL) were injected at room temperature (24 °C). The flow rate of the mobile phase methanol was set to 0.5 mL/min. Three serially linked columns were used for separation, i.e., two Supelcosil LC-18-DB columns of 16 and 25 cm length, 4.6 mm inner diameter, and 5 μm 470

DOI: 10.1021/jf505452u J. Agric. Food Chem. 2015, 63, 469−475

Article

4-Me-9:0

60 42 35 61 77 ± 2.9

4-Et-8:0

a 81 ± 2.0

(1)

As will be shown later, in all cases, only one enantiomer could be detected in the samples. Due to the low concentrations of the 4-alkylbranched fatty acids in the sample, we could not generally rule out the presence of trace amounts of the undetected enantiomer in the sample solutions (which would have been the requisite for ee = 100%). For this reason, the limit of detection (LOD) had to be determined and formed the basis to conclude on the minimum ee of the predominant enantiomer in the samples. Accordingly, ee was calculated by assigning the LOD to the nondetected enantiomer according to the literature.24 Then, it was noted that the ee was larger than the resulting ee value. Enantioselective GC/MS-SIM Analysis of Free 4-AlkylBranched Fatty Acids and the Corresponding Methyl Esters in a Non-Racemic Standard. Analyses were performed with a G1800B GCD-Plus system (Hewlett-Packard) equipped with a HP6890 autosampler and a split/splitless injector operated in splitless mode. Helium was used as the carrier gas with a constant flow of 1.0 mL/min. The temperatures of the injector and the transfer line were set to 250 °C and 280 °C, respectively. β-TBDM (ii) (see above) was used. The oven program for the enantioselective analysis of the free 4Me-8:0 started at 50 °C for 1 min; then, the temperature was increased at 1 °C/min to 200 °C (10 min hold time). In the SIM mode, m/z 57, m/z 73, m/z 83, m/z 85, m/z 99, m/z 101, and m/z 158 (M+) were recorded throughout the run. In addition, 4-Me-8:0-ME was analyzed with the following program: 50 °C for 1 min, then at 10 °C/min to 65 °C and after 150 min at 10 °C/min to 200 °C (10 min hold time). Seven ions (m/z 74, m/z 87, m/z 99, m/z 115, m/z 129, m/z 141, and m/z 172) were measured in SIM mode. Analysis of the enantioenriched 4-Me-8:0 standard (see above) on the β-TBDM column (ii) showed an enantiomer ratio of the first to second enantiomer of 3:1 (ee = 50%). In a previous study of Dietrich et al., who synthesized both enantiomers, it was found that the (R)enantiomer of 4-Me-8:0 eluted first from β-TBDM.25 Transferred on our measurements on β-TBDM, the non-racemic standard showed an enantiomeric excess of (R)-4-Me-8:0. An aliquot of the non-racemic standard was converted into the corresponding methyl ester and used for the determination of the elution order of the methyl ester on the CSPs. Determination of the Efficiency of the Enantiomer Separation and Statistical Analysis. The efficiency of the enantiomeric separation was evaluated by means of the separation factor α (eq 2) and the percentage valley between the two enantiomers (eq 3).

26 21 26 70 77 ± 2.8

Analyzed by GC/MS (all other measurements were performed by GC/FID); mean values (racemic standard) n = 10.

′ /t R1 ′ α = t R2

(2)

with α = separation factor; tR1 ′ = reduced retention time of the first eluting enantiomer, and t′R2 = reduced retention time of the second eluting enantiomer. V (%) =

1 (HE1 + HE2) − HV 2 *100 1 (HE1 + HE2) 2

(3)

with V (%) = valley between the two enantiomers, HE1 = peak height of enantiomer 1, HE2 = peak height of enantiomer 2, and HV = height of the valley between the two peaks. The statistical analysis was performed by one-way analysis of variance (ANOVA), using α = 0.05. Evaluation by means of the pvalue which describes the level of significance was based on p < 0.05 for statistical significance and p < 0.001 for high statistical significance.

a

4-Me-8:0

ee = (E1 − E2)/(E1 + E2) in %

61 52 58 0 a

4-Me-6:0 4-Me-9:0

for 4-Et-8:0-ME and 4-Me-9:0-ME. These concentrations were ∼100fold higher compared to the nonchiral measurements.18 This loss of sensitivity on the γ-TECD column was due to broader peaks at longer retention times as a consequence of the low isothermal oven temperature required for the resolution of the enantiomers. The quality of the enantiomer purity was expressed by the enantiomeric excess (ee) (eq 1):10

1.018 1.012 1.010 1.034 1.028 ± 0.001 1.000 1.000 1.000 a 1.030 ± 0.001

4-Et-8:0 4-Me-8:0

1.010 1.007 1.014 1.038 1.029 ± 0.001 1.014 1.015 1.023 1.000 a A A A B C 30 30 30 12 12

4-Me-6:0 GC method length (m) solvent

SE-52 PS086 OV1701 OV1701 OV1701 β-TBDM β-TBDM β-TBDM γ-TECD γ-TECD 20% 20% 66% 25% 25%

column no.

1 (i) 2 (ii) 3 (iii) 4 (iv) 4a (iv)

valley between the enantiomer peaks (%) α-value

Table 1. Enantioseparation Characteristics (Separation Factor α and Valley between the Peaks in %) of 4-Alkyl-Branched Fatty Acid Methyl Esters on Modified Cyclodextrin Columns

Journal of Agricultural and Food Chemistry

471

DOI: 10.1021/jf505452u J. Agric. Food Chem. 2015, 63, 469−475

Article

Journal of Agricultural and Food Chemistry

Table 2. Concentrations and Enantiomeric Excess of 4-Methyloctanoic Acid (4-Me-8:0), 4-Ethyloctanoic Acid (4-Et-8:0), and 4-Methylnonanoic Acid (4-Me-9:0) in Goat and Sheep Milk, Dairy Products, and Subcutaneous Adipose Tissue μg/g milk, cheese, and adipose fat18,19 ± average deviation (n = 2)

first eluting enantiomera ee (%)b

no.

sample

type of milk

4-Me-8:0

4-Et-8:0

4-Me-9:0

4-Me-8:0

4-Et-8:0

4-Me-9:0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

goat milk goat milk, organic goat, cream cheese goat, French soft cheese goat, organic butter cheese goat, hard cheese old sheep milk sheep milk, organic sheep, soft cheese sheep, Pecorino hard cheese sheep, semihard cheese sheep, blue-veined cheese sheep and goat, greek Feta cheese subcutaneous adipose fat 1 subcutaneous adipose fat 2

goat milk goat milk goat milk goat milk goat milk goat milk sheep milk sheep milk sheep milk sheep milk sheep milk sheep milk sheep and sheep fat sheep fat sheep fat

5.9 ± 0.08 6.9 ± 0.12 15 ± 0.54 32 ± 0.11 59 ± 0.59 185 ± 12 6.2 ± 0.18 2.5 ± 0.04 14 ± 0.00 21 ± 0.41 25 ± 0.61 29 ± 6.0 15 ± 0.76 48 ± 8.1 23 ± 1.3

3.9 ± 0.04 4 ± 0.06 11 ± 0.41 23 ± 0.65 24 ± 0.74 68 ± 4.5 0.1 ± 0.02 0.2 ± 0.00 0.9 ± 0.01 0.8 ± 0.24 1.5 ± 0.09 1.5 ± 0.21 1.3 ± 0.11 26 ± 3.5 13 ± 1.7

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 5.6 ± 0.7 2.9 ± 0.4c

>96.5 >96.5 >95.9 >97.7 >98.2 >98.0 >95.7 >96.7 >87.5 >93.3 >88.9 >91.8 >91.6 >93.8 >96.9

>87.8 >90.6 >64.7 >93.4 >80.6 >79.3