Article pubs.acs.org/JAFC
anteiso-Fatty Acids in Brussels Sprouts (Brassica oleracea var. gemmifera L.): Quantities, Enantioselectivities, and Stable Carbon Isotope Ratios Dorothee Eibler, Carolin Seyfried, Stefanie Kaffarnik, and Walter Vetter* Institute of Food Chemistry (170b), University of Hohenheim, Garbenstrasse 28, 70599 Stuttgart, Germany S Supporting Information *
ABSTRACT: anteiso-Fatty acids (aFAs) are a class of branched-chain fatty acids that are characterized by one methyl branch on the antepenultimate carbon of the straight acyl chain. aFAs are mainly produced by bacteria, and sources in vegetables are scarce. This study reports the concentrations of odd-numbered aFAs (a15:0−a21:0) in Brussels sprout buds. Selective enrichment followed by enantioselective gas chromatography with mass spectrometry in the selected ion monitoring mode revealed that both a15:0 and a17:0 were (S)-enantiopure in Brussels sprout samples. δ13C values (‰) of a17:0 in Brussels sprouts were comparable with those of palmitic acid, indicating no different source for both fatty acids. KEYWORDS: Brussels sprout, branched-chain fatty acid, anteiso-fatty acid, enantiomer separation, stable carbon isotope ratio
■
Ginkgo biloba seed lipids,14 as well as nine iFAs and aFAs (C12− C22) in the phospholipids of yellow-white leaves and petals of Antirrhinum majus and Nicotiana tabacum plants.15 Additionally, 17 different aFAs (a13:0−a29:0) and 7 iFAs (i13:0−i18:0, i20:0) were identified in the acid moiety of wax esters from the cuticular waxes of green tobacco leaf.16 However, none of these plants serves as food for human nutrition. In this context, Brussels sprouts (Brassica oleracea var. gemmifera L.) seem to be special, due to the presence of aFAs, documented in the wax layer of their surface.17,18 Despite their relatively low abundance in selected food sources only, it is worth noting that the dietary daily intake of aFAs and iFAs in U.S. food (∼500 mg/day) exceeded 5-fold the summed share of 4,7,10,13,16,19-docosahexaenoic acid (22:6n-3, DHA) and 5,8,11,14,17-eicosapentaenoic acid (20:5n-3, EPA) of 100 mg/day.3 aFAs and iFAs have also been shown to be valuable bioactive minor fatty acids because they positively influence the gut (a reduction of the incidence of necrotizing enterocolitis in a neonatal rat model has been shown).19 Moreover, they are potential anticancer agents because they have inhibited tumor growth and induced apoptosis of human breast cancer cells in cultured cells.20,21 In this study we aimed to determine the concentrations of aFAs (and other fatty acids) in the total lipids and surface wax of Brussels sprout buds. Because the carbon at the branching of aFAs is substituted with four different substituents (i.e., carboxyalkyl, ethyl, methyl, and hydrogen, Figure 1), aFAs are chiral.9 A high excess of (S)-aFAs or even (S)-enantiopurity has been reported in fish and milk fat.1,2 For this reason we also wished to determine the enantioselectivity of aFAs in Brussels sprout buds to compare the results with literature data.
INTRODUCTION anteiso-Fatty acids (aFAs) are minor fatty acids in food and bear a methyl branch on the antepenultimate (n-2)-carbon of the otherwise straight acyl chain (Figure 1).1−3 In many instances,
Figure 1. Structure of 12-methyltetradecanoic acid (a15:0) with (a) (S)-a15:0 and (b) (R)-a15:0.
aFAs belong to the major fatty acids of bacteria.4,5 Especially in Bacillus sp., branched-chain fatty acids (BCFAs, here aFAs and iso-fatty acids, iFAs) may represent up to 95% of the fatty acids (FAs).4,6 In contrast, BCFAsif present at allcontribute only 97% of the extracted substances crystallized and only ∼3% was separated in the form of an oily residue. Subsequent analysis of the fatty acid pattern in the hexane extract and the oily and crystalline fractions differed by 96%” was the limit for a17:0-ME (Figure 3b). Samples in which the peak shape of the aFA methyl esters showed no noticeable deviation from the one of neat (S)-aFAME standards were classified enantiopure (ee > 98% for a15:0-ME and ee > 96% for a17:0-ME). In the case of the detection of (R)-enantiomers prior to the first eluting (S)-enantiomer in samples, ee was assigned on the basis of the closest match of the abundance of the shoulder peak with the one in the reference standards. This procedure was carried out with both m/z 74 and 87, and all samples were processed and analyzed at least in duplicate (by individual weighing in and including all enrichment steps). For verification of aFA methyl esters, SIM ion traces were superimposed to prove congruent run for both enantiomers (Figure S2, Supporting Information) and thereby excluding coelutions with other compounds.2 Racemic a19:0-ME could not be sufficiently resolved on the β-TBDM column, and its enantioselectivity in Brussels sprout buds could not be explored. Stable Carbon Isotope Analysis (δ13C Values in ‰). Compound-specific isotope analysis was performed with a Delta plus XP mass spectrometer (Thermo Finnigan MAT, Bremen, Germany) coupled to a 6890 GC (Agilent, Waldbronn, Germany) via a GCC III combustion/reduction interface (GC-C-IRMS). The oxidation reactor (CuO, NiO, Pt) was operated at 940 °C and the reduction reactor (Cu wires) at 650 °C.22 Removal of water was facilitated by a Nafion tube. Three ion currents (i.e., m/z 44, 45, and 46 were monitored throughout the run, and standardization of runs was achieved by injecting five times CO2 gas with a known δ13C value (−28.4‰) directly into the ion source of the mass spectrometer. A 30 m × 0.25 mm internal diameter column coated with a 0.25 μm film of 50% phenyl, 50% dimethylpolysiloxane (CP-Sil 24CB, Varian, Middelburg, The Netherlands), was used in combination with the following GC oven program: 60 °C (hold for 1 min), raised at 10 °C/min to 160 °C, raised at 2 °C/min to 200 °C, and raised at 10 °C/min to 270 °C (hold for 5 min). The injector temperature was set at 260 °C, and the
Figure 3. GC-MS-SIM chromatograms (m/z 74) of the enantioselective determination of a standard solution of (S)-anteiso-fatty acid methyl esters and admixtures with various amounts of racemic standards on β-TBDM after smoothing according to Savitzky and Golay:31 (a) a15:0-ME and (b) a17:0-ME with partial enlargement inserted with lines in blue, ee 100%; green, ee 96%; orange, ee 90%; red, ee 80%; and gray, ee 50%. helium carrier gas flow rate was 1.6 mL/min. HPLC fractions (step 8) of polar and nonpolar ASE extracts of four samples from Germany, two samples from Belgium, and one sample from The Netherlands (2 μL) that contained both a17:0-ME and 16:0-ME were used for the measurements. The HPLC fractions were adjusted to a17:0-ME ∼ 30 ng/μL, and the solution was measured five times. The second fatty acid, 16:0-ME, was detected in two HPLC fractions. After GC-IRMS measurement of a17:0-ME, this and the subsequent HPLC fraction were combined and diluted to give 16:0-ME = 30 ng/μL. The resulting solution was also measured five times in the samples. Two microliters of the samples was injected in splitless mode (split opened after 2 min), and data were processed with Isodat 2.0 software (Thermo Finnigan MAT, Bremen, Germany). Due to a rapid and quantitative derivatization, no kinetic isotope effects were expected in the formation of FAMEs.33,34 For quality control, selected samples were also analyzed without previous enrichment of a17:0-ME and 16:0-ME (injection of 30 ng of the analytes). Only this retention time range was transferred from the GC column to the combustion unit. The resulting δ13C value (in ‰) matched the one in the enriched samples. By this measure it was verified that sample enrichment was not accompanied by isotope discrimination. Bulk δ13C values (‰) of nonpolar ASE extracts as well as freezedried whole edible buds of Brussels sprouts were analyzed with a Euro EA 3000 elemental analyzer (Hekatech, Wegberg, Germany) coupled to the IRMS system mentioned above. The furnace temperatures for oxidation and reduction were set to 650 and 1000 °C, respectively, and the helium carrier gas pressure was set at 70 kPa with a purge flow at 80 mL/min. Each EA-IRMS run started by measuring three times the reference CO2 gas. For sample measurement about 2 mg of freeze8924
DOI: 10.1021/acs.jafc.5b03877 J. Agric. Food Chem. 2015, 63, 8921−8929
8925
a
−28.4 −35.2 −34.5 −34.0
2
0.45 1.01 0.86 0.01 0.15 1.54 0.01 0.23 0.25 0.04 0.02 0.10
wax
−28.3 −34.1 −33.5 −33.1
Genius
Germany
2
0.99 86.1 11.2 8.66 28.4 82.0 0.07d 6.98 1.87 0.17 0.19 0.18
lipid
Genius
Germany
3
0.48 0.82 1.05 0.02 0.10 0.88 0.01 0.24 0.26 0.04 0.01 0.08
wax
−27.5 −34.3 −33.5 −32.8
Genius
Germany
3
1.18 61.0 12.8 8.63 25.2 57.5 0.09d 6.32 1.92 0.17 0.14 0.29
lipid
Genius
Germany
4
4
Sirius 0.44 40.3 8.13 4.30 18.3 33.5 0.02 3.20 1.03 0.08 0.08 0.2
lipid
−28.6 −35.6 −35.0 −33.9
Genius
Germany
0.62 1.67 1.98 0.11 0.24 1.10 0.03d 0.26 0.32 0.04 0.03 0.17
wax
Germany
5
lipid
−29.6 −37.4 −37.1 −35.8
Sirius
Belgium
5
0.43 0.67 1.25 77.3 1.12 12.6 0.08 8.15 0.20 31.7 1.22 80.1 0.03d 0.05 0.18 5.74 0.19 1.35 0.02 0.07 0.02 0.21 0.13 0.52 δ13C values (‰)
wax
unknown
Belgium
0.40 0.66 0.79 0.01 0.09 0.52 0.01 0.20 0.20 0.03 0.01 0.07
−27.0 −35.0 −34.4 −34.1
unknown
Belgium
0.98 95.8 11.3 11.0 40.1 94.1 0.07d 7.1 1.65 0.12 0.20 0.32
lipid
unknown wax
6
6 Belgium
7
−29.1 −37.0 −37.3 −36.0
unknown
Netherlands
7
0.81 1.70 1.56 0.05 0.25 1.49 0.01 0.31 0.35 0.04 0.03 0.18
wax 0.89 84.8 9.37 9.87 31.5 75.8 0.03 4.70 1.48 0.12 0.14 0.31
lipid
unknown
Netherlands
8
−37.5 −36.0
unknown
lipid 0.91 85.3 8.13 11.0 34.2 76.5 0.03 4.35 1.21 0.09 0.11 0.29
Netherlands
8
0.84 1.95 1.62 0.06 0.31 1.51 0.02 0.49 0.47 0.01 0.04 0.22
wax
unknown
Netherlands
Summed nonpolar and polar ASE extract. bSum of the contribution of 8:0−15:0. cSum of the contribution of >16:0. d(S)-Enantiopure (ee >98%). e(S)-Enantiopure (ee >96%).
plant lipids 16:0-ME a17:0-ME
Maximus
cultivar:
δ13C, δ13C, δ13C, δ13C,
1
Germany
0.66 90.8 7.29 8.63 35.7 94.6 0.04 2.78 0.70 0 0.24 0.49
origin:
0.29 0.69 0.60 0.09 0.17 1.05 0.004 0.15 0.15 0.01 0.01 0.07
sample:
short chain, sat. 16:0 long chain, sat.c monoenic dienoic trienoic a15:0 a17:0e a19:0 a21:0 i16:0 i18:0
b
cultivar:
lipid
Maximus
origin:
wax
1
Germany
sample:
concentration (mg/100 g Brussels sprouts)
9
0.83 87.5 8.54 9.25 29.5 73.5 0.02 3.24 1.12 0.07 0.14 0.20
lipid
−37.8 −36.2
unknown
Netherlands
0.99 1.65 1.42 0.09 0.25 1.40 0.01 0.30 0.29 0.02 0.02 0.16
wax
unknown
Netherlands
9
Table 1. Country of Origin and δ13C Values of the Whole Freeze-Dried Plants, Nonpolar Lipid Extracts, Palmitic Acid Methyl Ester (16:0-ME), and anteiso-Heptadecanoic Acid Methyl Ester (a17:0-ME) as well as Amounts of Individual Fatty Acids in the Wax and Lipidsa of the Brussels Sprout Samples
Journal of Agricultural and Food Chemistry Article
DOI: 10.1021/acs.jafc.5b03877 J. Agric. Food Chem. 2015, 63, 8921−8929
Article
Journal of Agricultural and Food Chemistry dried Brussels sprouts was transferred by means of a spatula/syringe into tin capsules for solids, and an aliquot of the nonpolar ASE extract was placed in preweighed tin capsules for liquids. The capsules were stored overnight at room temperature for total evaporation of the solvent from the lipid samples, and the lipid amount was determined by reweighing of the fat-containing capsules.35 Each sample was analyzed in duplicate. USGS40 was used as secondary standard to calibrate the system.25 δ13C values (‰) were calculated by means of the international internal Standard VPDB (in ‰), where R is the measured isotope ratio 13C/12C eq 2:34
■
δ13C (‰) = (R sample − R standard)/R standard × 103
The BCFA pattern in our wax samples differed from two literature reports in the fact that long-chained BCFAs were more relevant in the previous studies.17,18 Baker and Holloway additionally detected a23:0 (0.6%) and even-numbered, longchained iFAs (up to i24:0) in Brussels sprout wax.17 Even longer homologues (a25:0−a31:0, dominated by a29:0, 29.8%) were detected in the free fatty acid fraction of the wax of a subglaucous mutant of B. oleracea.18 These examples indicate different patterns from sample to sample depending on growing conditions and genotype.17 The presence of BCFAs has been associated with their role in membrane fluidity.12,13,15 However, separate investigation of the nonpolar and polar ASE extracts did not reveal higher aFA contents in the polar fraction. The contribution of a17:0 to the fatty acid pattern in the total lipids of Brussels sprouts (∼2.6% of the fatty acids; Figure 4a) was higher than in the major sources of aFAs in food, that is, fish oil (∼0.5%)7,9 and milk fat of ruminants (∼1%).3,8,10 Apart from BCFAs, the fatty acid patterns of the wax12 and the total lipids of Brussels sprouts30,31 were similar to those reported in the literature. Short-chain saturated fatty acids (8:0−15:0) as well as the long-chain saturated fatty acids (>16:0) represented less than 10 and 25% of the fatty acid pattern, and 16:0 was the most abundant saturated fatty acid (>20%, Figure 4). The majority of unsaturated fatty acids originated from 18:3n-3 (trienoic acids >20%; Table 1; Figure 4,5a). The sole presence of odd-numbered aFAs in Brussels sprouts agreed with reports in fish7 and milk from ruminants.8 By contrast, even-numbered aFAs were described in alpine plants (Oxygraphos glacialis, Rhodiola pamiroalaica, and Swertia
(2)
RESULTS AND DISCUSSION Fatty Acid Pattern and Share of anteiso-Fatty Acids in Brussels Sprout Buds. Fatty acids were determined as methyl esters but will be discussed as fatty acids to keep the text simpler. Both aFAs and iFAs were detected in all Brussels sprout samples (i.e., the ASE extracts representing the edible part and n-hexane extracts representing the wax) (Table 1). Due to the low contribution of the wax to Brussels sprouts, the amount of aFAs was on average 10-fold higher in the total lipids (i.e., ∼3−7 mg a17:0/100 g Brussels sprouts) than in the wax layer (∼0.2−0.5 mg a17:0/100 g Brussels sprouts; Table 1). This was surprising because the presence of aFAs had hitherto been reported only in (surface) wax17,18 but not in the edible part of Brussels sprouts.36 In the present samples, the wax layer and the total lipids (sum of nonpolar and polar ASE extracts) were dominated by a17:0 and a19:0, whereas a15:0 and a21:0 occurred only as minor components (Table 1; Figure 4). Typically, a17:0 and a19:0 were on the same level in wax, whereas a17:0 was more abundant in the total lipids (Table 1; Figure 4). In addition, the samples also contained i16:0 and i18:0 (i.e., a total of six BCFAs). The main difference between total lipids and wax layer originated from differences in a17:0.
Figure 4. Share of different fatty acids in the fatty acid pattern in lipids (sum of nonpolar and polar ASE extracts) (a) and wax (b) of Brussels sprout buds with short-chain FA defined as the sum of the contribution of 8:0−15:0 and long-chain FA as the sum of the contribution of >16:0.
Figure 5. GC-EI-MS-SIM chromatograms (TIC) of the methyl esters from the polar lipid ASE extract of Brussels sprout buds with (a) sample before hydrogenation and urea complexation, (b) sample after hydrogenation, and (c) filtrate of the urea complexation. 8926
DOI: 10.1021/acs.jafc.5b03877 J. Agric. Food Chem. 2015, 63, 8921−8929
Article
Journal of Agricultural and Food Chemistry marginata)12 and in the phospholipids from petals and yellowwhite leaves from Antirrhiunum majus (a22:0).15 Only a17:0 was described in Pinaceae and Ginkgo biloba seed lipids.14 Enrichment of a15:0 and a17:0 from FAME Solutions of the Samples. Partly due to the low contributions of aFAs (Table 1) and presumably low contributions of (R)-aFAs in the samples,1,2 it was necessary to perform a selective enrichment of the aFAs by (i) hydrogenation, (ii) urea complexation, and (iii) RP-HPLC (see Materials and Methods). After hydrogenation, 18:3n-3 and 18:2n-6 (∼30−50% of the fatty acids in Brussels sprouts; Figure 4) were no more detected in the solutions (Figure 5a,b; Table S2, Supporting Information). Only low amounts of 18:1n-9 (
