Article pubs.acs.org/JAFC

Studies on the Reaction of trans-2-Heptenal with Peanut Proteins Martin Globisch, Marco Schindler, Jana Kreßler, and Thomas Henle* Institute of Food Chemistry, Technische Universität Dresden, D-01062 Dresden, Germany ABSTRACT: Hexanal, 2-heptenal, and nonanal were identified as relevant reaction products formed in the course of the lipid peroxidation of heated peanut oil. For the identification of potential amino acid side chain adducts, kinetic studies between Nα-benzoylglycyl-L-lysine as a model for proteinbound lysine and trans-2-heptenal were performed, showing a strong decrease of the lysine-derivative whereupon the loss of trans-2-heptenal was moderate. Following acid hydrolysis of the incubation mixture of Nα-acetyl-L-lysine and trans-2heptenal, two UV-active major lipation products were observed, isolated and identified as isomeric pyridiniumderivatives, namely (Z)- and (E)-1-(5-amino-5-carboxypentyl)4-butyl-3-(pent-1-en-1-yl)pyridin-1-ium (cis- and trans-BPPlysine). After heating of a native peanut protein extract with trans-2-heptenal, both derivatives were quantitated by LC-ESI-MS/ MS after acid hydrolysis and the modification of lysine was measured by amino acid analysis. At low, “food-relevant”, concentrations of trans-2-heptenal, up to 80% of the lysine modification could be explained by the formation of cis- and transBPP-lysine, showing that these two lipation derivatives represent good markers for a protein modification by the lipid peroxidation product trans-2-heptenal. KEYWORDS: peanuts, lipid peroxidation, 2-heptenal, lipation, pyridinium derivatives, BPP-lysine, carbonyl-protein-reaction

■

INTRODUCTION Peanuts are a common food and serve as ingredients for many foods, being a rich source of protein (25%) and fat (48%).1 In the U.S. and Europe, peanuts are commonly consumed roasted, whereas in Asian countries peanuts are mostly consumed boiled or fried.2,3 Heat-induced protein modifications occurring during food processing can influence the allergenic potential of peanuts. Findings concerning a relationship between heating of peanuts and their allergenic potential are currently contradictory. On the one hand, no difference in the immunoreactivity between proteins isolated from raw and commercially roasted peanuts was observed4 and even a reduction of the allergenic potential of Ara h 1 was measured following boiling of the isolated protein.5 On the other hand, extracts from roasted peanuts exhibited at 90-fold higher binding to serum IgE of allergic patients when compared to the raw extracts.6 Additionally, Ara h 1 and Ara h 3 modified with advanced glycation endproducts (AGEs) were found to interact with the receptor for AGEs (RAGE), and thus may be involved in allergic sensitization.7 After heating of Ara h 1 at low moisture in the presence or absence of glucose, the degranulation capacity of basophils was increased, whereas for Ara h 2 the degranulation capacity was decreased.8 These results show that heating in the presence or absence of reactive carbonyls can have different effects, depending on the protein structure. Among the reactions that can occur during the roasting process, besides glycation, the lipid peroxidation is of particular importance. Recently, it could be shown that up to 30−40% of lysine was modified as a result of peanut roasting.9 © 2014 American Chemical Society

However, only about one tenth of the totally observed lysine modification could be explained by the formation of the Maillard reaction products Nε-fructosyllysine, pyrraline and Nεcarboxymethyllysine. Taking into account the relatively high content of fat in peanuts and the fact that about 85% of the fatty acids are unsaturated,10 it is highly likely that lipid peroxidation occurs in the course of the peanut roasting process. Actually, a large variety of secondary lipid peroxidation products have been detected by analyzing peanut oils obtained from raw and roasted peanuts,11 ranging for monocarbonyl compounds from 8 to 26 μmol per 100 g oil, respectively.12 Well known secondary products are α,β-unsaturated aldehydes like 4-hydroxyalkenals and 2-alkenals.13,14 Due to their high reactivity, these electrophiles readily form covalent adducts, for example with nucleophilic amino acid side chains of proteins.15 Having two electrophilic centers, 2-alkenals initially can form Michael- and Schiff base-adducts which can undergo further reactions. In model mixtures consisting of trans-2-hexenal and Nα-acetylglycyllysine methyl ester, several pyridinium derivatives resulting from the reaction of alkenals with one amino group and arising cross-links were detected, but no information about formation of corresponding compounds at proteins under conditions relevant for food processing were given.16 Under physiological conditions, trans-2-nonenal was found to Received: Revised: Accepted: Published: 8500

May 26, 2014 July 14, 2014 July 26, 2014 July 26, 2014 dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

gas washing bottle filled with 100 mL dichloromethane. After evaporation of the dichloromethane to a volume of 5 mL using a rotary evaporator at atmospheric pressure at 46 °C, an aliquot was used for GC-MS (EI) analysis. This was performed using an Agilent 7890A System, consisting of an 7683 Series injector with a sample tray, a 5975C MS detector working in EI mode, a HP-5MS capillary column (30.0 m × 0.25 mm ID, 0.25 μm film thickness), all from Agilent (Böblingen, Germany), and a Zebron Z-guard column (deactivated, 5.0 m × 0.25 mm ID, Phenomenex, Germany). Helium was used as carrier gas with a constant flow of 1 mL/min. The injector temperature was set to 250 °C and 1 μL sample was injected by using the pulsed splitless mode. The auxiliary temperature was set to 250 °C, ion source and quadrupole to 230 and 150 °C, respectively. The initial oven temperature was set to 40 °C and held for 9 min, then raised at 3 °C/ min to 61 °C and held for 3 min, then raised at 3 °C/min to 76 °C, at 6 °C/min to 90 °C, at 20 °C/min to 110 °C and held for 1 min and finally raised at 30 °C/min to 300 °C. The post run time was set to 3 min at 300 °C. The mass spectrometer was working in electron impact mode at 70 eV in the SCAN mode (mass range: m/z 30.0−300.0). Identification of the analytes was performed by comparing retention times and mass spectra to corresponding standard samples. Semiquantitation of the volatile compounds was performed by evaluating the peak areas of the analytes related to the peak areas of the internal standards. Depending on their boiling points, hexanal, 2-hexenal, and heptanal were evaluated related to p-xylene whereas 2-heptenal, octanal, and nonanal were evaluated related to 1,3-diethylbenzene. Identification of compounds was obtained by reference compounds. Kinetic Studies on the Incubation of trans-2-Heptenal with Nα-Benzoylglycyl-L-lysine. Nα-benzoylglycyl-L-lysine and trans-2heptenal, 4 mM each in methanol, were incubated for 24 h at 37 °C under nitrogen atmosphere. Every 2 h an aliquot of 400 μL was taken and mixed with 1200 μL of water to stop further reactions. Samples were filtered (0.45 μm) and trans-2-heptenal and Nα-benzoylglycyl-Llysine quantitated using a Knauer HPLC system (Berlin, Germany), which consisted of a degasser, a solvent organizer K-1500, a WellChrom pump K-1001, a dynamic mixing chamber K-5001, an automatic injector Basic Marathon, a column oven and a diode array detector WellChrom K-2700. Samples of 20 μL were separated using an Eurospher 100−5 C18 column (250 × 4.6 mm2) from Knauer (Berlin, Germany) at 30 °C at a flow rate of 0.9 mL/min and a detection wavelength of 227 nm. A gradient was used with solvent A (water) and B (acetonitrile), each containing 13 mM heptafluorobutyric acid. Gradient was as follows: Increased from 10% B to 70% B within 30 min, to 90% B within 5 min, held at 90% B for 10 min, decreased within 2 min to 10% B and held at 10% B for 10 min. For external calibration, trans-2-heptenal and Nα-benzoylglycyl-L-lysine were used. Nuclear Magnetic Resonance spectroscopy (NMR). 1H NMR was recorded on a Bruker Avance 600 instrument (Rheinstetten, Germany) at 600,16 MHz. 13C NMR was recorded on a Bruker DRX 500 instrument at 150,9 MHz. Tetradeuteromethanol (MeOH-d4) was used as solvent. All chemical shifts are given in parts per million (ppm) relative to the solvent signal serving as internal standard. The following 2D-NMR experiments were performed additionally: COSY (correlation spectroscopy), NOESY (nuclear overhauser enhancement spectroscopy), HSQC (heteronuclear single quantum coherence), and HMBC (heteronuclear multiple bond correlation). Elemental Analysis. Elemental analysis was performed using an EuroEA3000 (Eurovector, Milan, Italy) to quantitate the product content in the purified standards. Cis- and trans-BPP-lysine were isolated as salts of heptafluorobutyric acid, so the molecular ratio of cationic pyridinium derivative and heptafluorobutyric acid in the isolated products was not known. Therefore, the analyzed percentage of nitrogen was compared to the calculated nitrogen percentage. The pure amount of cationic pyridinium derivative in the product without counterion is expressed in percent by weight. Acid and Enzymatic Hydrolysis. For acid hydrolysis of proteins, a ratio of 3 mg protein per 1.0 mL 6 M hydrochloric acid was used. Samples were hydrolyzed under nitrogen for 23 h at 110 °C. For enzymatic hydrolysis,21 3.0 mg of protein were dissolved in 1.0 mL of

modify human and bovine serum albumin, resulting mainly in a modification of lysine residues and the formation of two cis− trans-isomeric pyridinium derivatives.17 After acid hydrolysis, both derivatives could be quantitated in model proteins as well as in oxidatively modified LDL (oxLDL) and mice kidneys after exposure of the animals to ferric nitrilotriacetate.17 Regarding reactions in food, very little information is available concerning modification of proteins by lipid peroxidation products. Using ELISA, protein adducts in beef and rice seeds were detected, which resulted from reactions with 4-hydroxynonenal-, malondialdehyde- and alkenals with 6 to 9 C atoms.18 In 22 fresh food products ε-N-pyrrolylnorleucine, representing a reaction product of 4,5-epoxy-2-alkenals and the ε-amino group of lysine,19 was quantified by capillary electrophoresis after alkaline hydrolysis.20 The concentration ranged from 0.24 to 6.36 μmol/g of food and the amount in peanuts was 2.89 μmol/g. The authors concluded that ε-Npyrrolylnorleucine is a normal component of fresh food products. The aim of this study, therefore, was to identify newly formed lipation products in food proteins and to develop an LC-ESI-MS/MS method for direct quantification. To identify potential precursors for amino acid side chain modifications arising during the peanut roasting process, peanut oil was heated under roasting conditions and volatile secondary products were semiquantitated by GC-MS. In kinetic studies, the reaction behavior of trans-2-heptenal and the lysine derivative Nα-benzoylglycyl-L-lysine was studied by HPLC. For identification of relevant reaction products with the εamino group of lysine, Nα-acetyl-L-lysine was incubated with trans-2-heptenal and reaction products were identified by LCESI-MS/MS. After isolation of two isomeric pyridinium derivatives arising from trans-2-heptenal and the ε-amino group of lysine, both derivatives were quantitated for the first time in peanut proteins by LC-ESI-MS/MS. Additionally, modification of lysine was measured by amino acid analysis.

■

MATERIALS AND METHODS

Materials. Heptanal, pepsin (EC 3.4.23.1), Pronase E (EC 3.4.24.4), and hydrochloric acid (37%) were obtained from Merck (Darmstadt, Germany). Leucine aminopeptidase (EC 3.4.11.1), prolidase (EC 3.4.13.9), dialysis tubing cellulose membrane (MWCO 14 kDa), trans-2-heptenal (≥95%), hexanal, 2-hexenal, octanal, p-xylene, 1,3-diethylbenzene, tetradeuteromethanol, and petroleum ether (boiling point 40−60 °C) were obtained from Sigma-Aldrich (Taufkirchen, Germany). Nα-acetyl-L-lysine, Nα-benzoylglycyl-L-lysine, and N-benzoylglycyl-L-phenylalanine were obtained from Bachem (Bubendorf, Switzerland). Nonanal, methanol, dichloromethane, heptafluorobutyric acid, and cyclohexanol were obtained from VWR (Darmstadt, Germany). Acetonitrile was obtained from Fischer Scientific (Schwerte, Germany). Sodium bicarbonate, monosodium phosphate, and disodium phosphate were obtained from Grüssing (Filsum, Germany). Tris(hydroxymethyl)aminomethan was obtained from Serva (Heidelberg, Germany). Thymol was obtained from Carl Roth (Karlsruhe, Germany). All chemicals were of highest purity unless otherwise indicated. For all experiments, ultrapure water was used prepared by an ELGA LabWater Purelab Plus water system (Celle, Germany). For LC-ESI-MS/MS-measurements, double-distilled water, prepared in the presence of potassium permanganate, was used. Refined peanut oil was obtained from a local market. GC-MS (EI) Analysis of Volatile Compounds Formed during Heating of Peanut Oil. After adding 75 μL of an internal standard solution containing p-xylene and 1,3-diethylbenzene in dichloromethane (6 mM each) to 50 g of peanut oil, the sample was heated in a three-neck flask while stirring for 20 min at 170 °C using a gentle airstream to transfer the volatile compounds in the head space into a 8501

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

hydrochloric acid were added and heated for 23 h at 110 °C. Afterward, the hydrochloric acid was removed using a water-jet vacuum pump at 40 °C. cis- and trans-BPP-lysine were purified using a semipreparative HPLC system. The system consisted of a Smartline solvent manager 5000, a Smartline pump 1000, and a Smartline UVdetector 2500 (all from Knauer, Berlin, Germany). Two mg of the unpurified reaction mixture were dissolved in 1.0 mL of methanol and separated after filtration (0.45 μm) using an Eurospher 100−10 C18 column (250 × 16 mm2; Knauer, Berlin, Germany) at 23 °C room temperature, a flow rate of 3 mL/min and a detection wavelength of 227 nm. A gradient was used with solvent A (water) and B (acetonitrile), each containing 13 mM heptafluorobutyric acid. The gradient program started with 25% solvent B for 5 min, increased to 50% B within 20 min, followed by isocratic elution with 50% B for 31 min, a flow reduction to 0.7 mL/min from 35 to 55 min, increased to 95% B within 2 min at a flow rate of 3 mL/min, isocratic elution with 95% B for 5 min, decreased to 25% B within 2 min, isocratic elution with 25% B for 15 min. cis- and trans-BPP-lysine eluted between 37 and 41 min and 41 and 51 min, respectively. Multiple separations were performed, relevant fractions containing the two compounds combined and finally the solvent was evaporated in vacuo at 40 °C. To increase purity, both compounds were purified for a second time prior to HPLC-ESI-MS/MS and one- and two-dimensional NMR analysis. For cis-BPP-lysine, results were as follows: ESI-MS (positive mode), [M]+ m/z 333.3; 1H NMR (600 MHz, MeOH-d4): δ 0.98 (t, 3H, J = 7.5 Hz, H-16), 1.04 (t, 3H, J = 7.5 Hz, H-20), 1.51 (m, 2H, H19), 1.57 (m, 2H, H-15), 1.72 (m, 2H, H-18), 1.60−1.69 (m, 2H, H8A,B), 2.03−2.10 (m, 2H, H-9A,B), 2.13 (m, 2H, H-7), 2.19 (q, 2H, J = 7.5 Hz, H-14), 2.96 (t, 2H, J = 7.9 Hz, H-17), 4.05 (t, 1H, J = 6.4 Hz, H-10), 4.67 (t, 2H, J = 7.5 Hz, H-6), 6.24 (dt, 1H, J = 7.5 Hz, J = 11.3 Hz, H-13), 6.60 (d, 1H, J = 11.3 Hz, H-12), 8.02 (d, 1H, J = 6.4 Hz, H-5), 8.75 (s, 1H, H-3), 8.81 (d, 1H, J = 6.4 Hz, H-4). Purity of cation based on elemental analysis: 42.9%. For trans-BPP-lysine, results were as follows: ESI-MS (positive mode), [M]+ m/z 333.3; 1H NMR (600 MHz, MeOH-d4): δ 1.06 (t, 3H, J = 7.5 Hz, H-20), 1.08 (t, 3H, J = 7.5 Hz, H-16), 1.53 (m, 2H, H-19), 1.60−1.71 (m, 2H, H8A,B) 1.66 (m, 2H, H-15), 1.72 (m, 2H, H-18), 2.04−2.09 (m, 2H, H9A,B), 2.15 (m, 2H, H-7), 2.41 (q, 2H, J = 7.2 Hz, H-14), 3.02 (t, 2H, J = 7.9 Hz, H-17), 4.07 (t, 1H, J = 6.4 Hz, H-10), 4.65 (t, 2H, J = 7.5 Hz, H-6), 6.62 (dt, 1H, J = 6.8 Hz, J = 15.8 Hz, H-13), 6.76 (d, 1H, J = 15.8 Hz, H-12), 7.93 (d, 1H, J = 6.4 Hz, H-5), 8.71 (d, 1H, J = 6.4 H, H-4), 9.04 (s, 1H, H-3). Purity of cation based on elemental analysis: 35.7%. HPLC-DAD-ESI-MS/MS Analysis. For identification of major reaction products between trans-2-heptenal and Nα-acetyl-L-lysine, for characterization of isolated BPP-standards and for quantitation of cisand trans-BPP-lysine in modified peanut proteins, HPLC-DAD-ESIMS/MS analysis were performed. The HPLC system consisted of a degasser, a pump, an autosampler and a diode array detector, all from Agilent Technologies 1200 Series (Böblingen, Germany). For MS/ MS-measurements, a Triple Quad LC/MS 6410 from Agilent Technologies was used. For identification of major reaction products between trans-2-heptenal and Nα-acetyl-L-lysine, 5 μL of samples were separated using an Eurospher 100−5 C18 column (250 × 3.0 mm2; Knauer, Berlin, Germany) at 30 °C, a flow rate of 0.38 mL/min and a detection wavelength of 227 nm. A gradient was used with solvent A (water) and B (acetonitrile), each containing 2 mM heptafluorobutyric acid and 11 mM acetic acid. The gradient increased from 5% B to 70% B within 50 min, to 90% B within 15 min, held at 90% B for 20 min, followed by a decrease within 3 min to 5% B and isocratic elution at 5% B for 10 min. At first full scan analyses and then product ion scans were performed. Full scan analyses were performed from 5.1 to 78.0 min in positive mode, scan ranges were from m/z 70 to 1000, scan time 500 ms, fragmentor voltage was set to 135 V, gas temperature to 300 °C, gas flow to 11 mL/min and nebulizer pressure to 15 psi. Product ion scans were performed for the hydrolyzed sample from 25 to 45 min with precursor ion m/z 333.3, using the same conditions as described above. Product ion scan ranges were from m/z 60 to 500 with a scan time of 200 ms, a fragmentor voltage of 135 V and a collision voltage of 25 V for cis- and 20 V for trans-BPP-lysine,

0.02 M hydrochloric acid (HCl) containing a crystal of thymol and incubated with 1 FIP-U of pepsin for 24 h at 37 °C. After adding 0.25 mL of 2 M TRIS-HCl buffer (pH 8.2) and 400 PU of Pronase E, samples were incubated for 24 h at 37 °C. Following addition of 0.3 U of leucine aminopeptidase and 2 U of prolidase, samples were incubated for another 24 h at 37 °C and finally lyophilized. Amino Acid Analysis. Amino acid analyses were performed using a SYKAM S4300 amino acid analyzer (Fürstenfeldbruck, Germany) after acid hydrolysis for the acid-stable amino acids valine, histidine, lysine, and arginine and enzymatic hydrolysis for the acid-labile amino acid cysteine. Amino acids were separated by cationic ion exchange chromatography, derivatized by ninhydrin, and detected using a wavelength of 570 nm.22 Valine was taken as internal standard and amino acids were expressed as valine equivalents. Incubation of trans-2-Heptenal with Nα-Acetyl-L-lysine. For identification of major amino acid side chain reaction products between trans-2-heptenal and Nα-acetyl-L-lysine, 10 mM trans-2heptenal and 10 mM Nα-acetyl-L-lysine were heated for 4 h at 75 °C in 5 mL methanol in a 10 mL round-bottom flask under reflux. 1.0 mL was taken for direct HPLC-DAD-ESI-MS/MS analysis and 3.0 mL were evaporated under a gentle stream of nitrogen. After hydrolysis with 3 mL of 6 M hydrochloric acid for 23 h at 110 °C under nitrogen, 1.0 mL of the sample was dried at 37 °C in vacuo by using a SpeedVac vacuum concentrator (Thermo Fisher Scientific, Waltham, U.S.A.), redissolved in 2.0 mL methanol and analyzed by LC-DAD-ESI-MS/ MS. Preparation of Modified Peanut Proteins. Previously shelled and skinned raw peanuts were crushed and defatted by Soxhleth extraction using petroleum ether for 6 h at 78 °C. For protein extraction, 21 mL 0.1 M sodium bicarbonate (pH 10.4) were added to 3.0 g defatted peanut meal.23 Samples were homogenized using an ultraturrax at 11.200 rpm for 2 min and proteins extracted for 1 h at 6 °C while stirring constantly. After centrifugation at 8643g for 15 min at 4 °C, the supernatant was collected and the residue was extracted analogously. Following dialysis (14 kDa MWCO) against deionized water for 48 h at 6 °C, the samples were lyophilized. Protein content was analyzed by the Kjeldahl method using the factor 5.3 for oilseeds.24 Amino acid composition was analyzed by amino acid analysis. Amino acid contents of cysteine, histidine, lysine, and arginine were 3.83 ± 0.79, 19.27 ± 0.06, 32.58 ± 1.39, and 70.58 ± 3.22 mmol/ 100 g protein, respectively. Approximately 100 mg of peanut proteins were dissolved in 0.1 M phosphate buffer (pH 7.4) and trans-2heptenal was added in 1.0 mL methanol in following molar ratios related to the sum of relevant reactive amino acids histidine, lysine, and arginine: 0.1:1, 0.2:1, 0.5:1, 1:1, 5:1, and 10:1. Finally, the samples were made up to 25 mL with phosphate buffer (pH 7.4). Additionally, a blank sample consisting of peanut proteins without trans-2-heptenal was prepared. The incubation was carried out under a nitrogen atmosphere for 24 h at 37 °C while stirring. After dialysis (14 kDa MWCO) against deionized water for 24 h at 6 °C, the samples were lyophilized and protein contents were analyzed by the Kjeldahl method. For analysis of the amino acid modification rate (lysine, histidine, and arginine), the samples were hydrolyzed by hydrochloric acid and subjected to amino acid analysis. For cysteine analysis, samples were hydrolyzed enzymatically. For analysis of BPP-lysine, 12 mg of samples were hydrolyzed by 4 mL 6 M hydrochloric acid for 23 h at 110 °C. 1.0 mL of samples were dried in vacuo using a SpeedVac vacuum concentrator (Thermo Fisher Scientific, Waltham, U.S.A.) at 37 °C, redissolved in 0.25 mL deionized water, 0.2 mL methanol and 0.05 mL of internal standard N-benzoylglycyl-L-phenylalanine in methanol:water (50:50, v/v) with a final concentration of 1 μg/mL. After filtration (0.45 μm), samples were subjected to HPLC-ESI-MS/ MS analysis. Isolation of (Z)-1-(5-Amino-5-carboxypentyl)-4-butyl-3(pent-1-en-1-yl)pyridin-1-ium (cis-BPP-lysine) and (E)-1-(5Amino-5-carboxypentyl)-4-butyl-3-(pent-1-en-1-yl)pyridin-1ium (trans-BPP-lysine). In a 500 mL round-bottom flask 376.5 mg (2.0 mmol) Nα-acetyl-L-lysine and 827.0 μL (6.0 mmol) trans-2heptenal were heated for 4 h at 75 °C in 200 mL methanol under reflux. After evaporation of methanol in vacuo at 40 °C, 150 mL 6 M 8502

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

respectively. For quantitation of BPP-lysine in modified peanut proteins 1 to 10 μL of samples were separated using a Zorbax SBC18 column (50 × 2.1 mm) from Agilent Technologies (Böblingen, Germany) at 30 °C and a flow rate of 0.25 mL/min. A gradient was used with solvent A (water) and B (acetonitrile), each containing 2 mM heptafluorobutyric acid and 11 mM acetic acid. Gradient was as follows: 2% B was held for 5 min, increased to 51% B within 17 min, to 90% B within 2 min, held at 90% B for 6 min, decreased to 2% B within 1 min and held at 2% B for 6 min. Measurements were performed using the multi reaction monitoring (MRM) mode in positive mode, gas temperature was set to 300 °C, gas flow to 11 mL/ min and nebulizer pressure to 15 psi. Internal standard Nbenzoylglycyl-L-phenylalanine was analyzed from 18 to 20.1 min with a mass transition from m/z 327.1 to 166.1 for quantitation using fragmentor and collision voltages of 80 and 4 V, respectively and a dwell time of 300 ms. Mass transition from m/z 327.1 to 105.1 was used for qualification using fragmentor and collision voltages of 80 and 32 V, respectively and a dwell time of 300 ms. cis-BPP-lysine was analyzed from 20.1 to 21.0 min with a mass transition from m/z 333.3 to 204.2 for quantitation using fragmentor and collision voltages of 133 and 21 V, respectively, and a dwell time of 100 ms. Mass transition from m/z 333.3 to 84.2 was used for qualification using fragmentor and collision voltages of 133 and 33 V, respectively, and a dwell time of 100 ms. trans-BPP-lysine was analyzed from 21.0 to 23.5 min with a mass transition from m/z 333.3 to 204.2 for quantitation using fragmentor and collision voltages of 134 and 21 V, respectively, and a dwell time of 100 ms. Mass transition from m/z 333.3 to 84.2 was used for qualification using fragmentor and collision voltages of 134 and 30 V, respectively and a dwell time of 100 ms. Quantification was realized using authentic cis-and trans-BPP-lysine in a native peanut protein hydrolyzate matrix calibration after addition of internal standard Nbenzoylglycyl-L-phenylalanine with a final concentration of 1 μg/mL. Statistical Treatment. Results are expressed as mean values ± standard deviations of two separate measurements.

nonanal are the most important secondary products of lipid peroxidation. Hexanal and 2-heptenal are mainly formed by the autoxidation of linoleic acid and nonanal is mainly formed from oleic acid.14 Being a rich source of unsaturated fatty acids, peanut oil consists of about 55% oleic acid and 22% linolenic acid.1 Considering the high reactivity of the 2-alkenals,15−17 it is likely that in the presence of peanut proteins, these compounds can easily react with nucleophilic amino acid side chains for example the ε-amino group of lysine. Kinetic Studies on the Reaction of trans-2-Heptenal and Nα-Benzoylglycyl-L-lysine. Focusing on 2-heptenal, kinetic studies were performed to get an insight into the reaction behavior with the ε-amino-group of lysine using Nαbenzoylglycyl-L-lysine as a model for protein bound lysine. trans-2-Heptenal and Nα-benzoylglycyl-L-lysine were incubated in equimolar amounts under mild reaction conditions at 37 °C for 24 h in methanol. Analogously, blank samples consisting only of trans-2-heptenal or Nα-benzoylglycyl-L-lysine, respectively, were incubated. Figure 2 shows that the amounts of

■

RESULTS AND DISCUSSION Identification of Volatile Secondary Lipid Peroxidation Products of Heated Peanut Oil. To identify potential precursors for amino acid side chain modifications formed in the course of lipid peroxidation during roasting of peanuts, refined peanut oil was heated under roasting conditions for 20 min at 170 °C. Figure 1 shows selected volatile carbonyl compounds in the headspace analyzed by GC-MS (EI). Identification was achieved by comparing retention times and mass spectra to corresponding reference material. Compounds were semiquantitated by evaluating the peak areas of the analytes related to the peak areas of the internal standards pxylene and 1,3-diethylbenzene. Hexanal, 2-heptenal, and

Figure 2. Kinetic studies on the reaction of equimolar amounts of trans-2-heptenal and Nα-benzoylglycyl-L-lysine at 37 °C for 24 h in methanol. Open squares, blank samples of Nα-benzoylglycyl-L-lysine; open circles, blank samples of trans-2-heptenal; filled squares, Nαbenzoylglycyl-L-lysine in equimolar reaction mixture; filled circles, trans-2-heptenal in equimolar reaction mixture.

trans-2-heptenal and Nα-benzoylglycyl-L-lysine in the blank samples were stable over the reaction time, whereas both the amounts of trans-2-heptenal and Nα-benzoylglycyl-L-lysine in the equimolar reaction mixture were decreasing. The observation that trans-2-heptenal decreased more than 2-fold faster than Nα-benzoylglycyl-L-lysine may indicate trans-2heptenal was decomposed or polymerized in the presence of the ε-amino-group of lysine. A proposed reaction pathway for polymerization initially starts with the formation of Schiff base adducts which polymerize by aldol condensation resulting in polymers from which the amino compound even may cleave.25,26 Another possibility is that he majority of the newly formed adducts arised by the reaction of two or more molecules of trans-2-heptenal and one molecule Nα-benzoylglycyl-L-lysine. Possible reaction products are pyridinium derivatives as previously described for other 2-alkenales and the ε-amino-group of lysine (see the Introduction).16,17,27 Identification of Major Reaction Products of trans-2Heptenal and Nα-Acetyl-L-lysine. For the identification of major reaction products of trans-2-heptenal and the ε-aminogroup of lysine, trans-2-heptenal and Nα-acetyl-L-lysine were

Figure 1. Semiquantitated amounts of selected volatile lipid peroxidation products of peanut oil heated under roasting conditions for 20 min at 170 °C. 8503

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

Figure 3. Identification of cis- and trans-BPP-lysine by HPLC-DAD-ESI-MS/MS after incubation of Nα-acetyl-L-lysine with trans-2-heptenal in equimolar amounts in methanol for 4 h at 75 °C. (A, B) HPLC-ESI-MS-scan-chromatograms and HPLC-UV-chromatograms (gray) at 227 nm before and after acid hydrolysis, respectively. Arrows indicate peaks resulting from Nα-acetyl-L-lysine (m/z 189.2) and the two pyridinium derivatives (m/z 375.3 and m/z 333.3). (C, D) product ion patterns after fragmentation of the hydrolyzed pyridinium derivatives with m/z 333.3 (first and second peak) and individual ions resulting from cis- and trans-pyridinium derivatives with m/z 333.3, respectively.

incubated in equimolar amounts in methanol for 4 h at 75 °C. Newly formed products were analyzed before and after acid hydrolysis by HPLC-DAD-ESI-MS/MS. Figure 3A shows that under these conditions, two UV-active main products with monoisotopic masses of m/z 375.3 were detectable. These masses are indicative for pyridinium derivatives resulting from one molecule Nα-acetyl-L-lysine, two molecules of trans-2heptenal, the loss of two water molecules and an oxidation step. As previously described, cis- and trans-derivatives may be formed, being the cis-derivative of the compound eluting earlier in the chromatogram.16,17 Even when the reaction partners were incubated in equimolar ratio, some amounts of Nα-acetyl+ L-lysine (M+H peak at m/z 189.2) were remaining. This fits with the previous results of the kinetic studies, underlining that the formed adducts consist of more than one molecule of trans-

2-heptenal. After acid hydrolysis, two UV-active main adducts with m/z 333.3 were detectable, showing that the deacetylation (mass difference, −42) was successful (Figure 3B). The stability of pyridinium derivatives against acid hydrolysis was analogously described for lysine-2-nonenal-pyridinium derivatives.17 To ensure these findings, product ion scans of the two hydrolyzed adducts were performed. Figure 3C,D indicates that both substances lead to the same product ions. The fragmentation patterns show typical lysine fragments with m/ z 84.2 and m/z 130.1 as well as a decarboxylation step with m/z 288.3. The ions with m/z 204.2 result from a fragmentation between the δ-carbon of lysine and the ε-nitrogen being incorporated in the pyridinium structure and therefore are strong hints for the existence of pyridinium derivatives. Due to the fact that both structures gave the same fragmentation 8504

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

derivative and J = 15.8 Hz for the trans-derivative and by means of 2D 1H/1H NOESY experiments. Strong NOEs were observed between proton 3 and protons 14, 15, and 16 (weak) being indicative for the cis-product. Additionally, no NOE was observed between protons 3 and 13 and only a weak NOE was observed between protons 12 and 14. In contrast, no NOEs were observed between proton 3 and protons 14, 15, 16 by analyzing the trans-product, whereas strong NOEs between protons 3 and 13 and protons 12 and 14 were observed. To the best of our knowledge, this is the first profound structural characterization of these two lysine derivatives. Analogously to the reaction of the ε-amino-group of lysine with trans-2hexenal,16 an exemplarily proposed reaction mechanism for the formation of trans-BPP-lysine following the reaction of trans-2heptenal with the ε-amino-group of lysine is given in Figure 5. In the first step, trans-2-heptenal probably reacts with the εamino-group of lysine, resulting in a Schiff’s base adduct, followed by a Michael addition of a second molecule of trans-2heptenal, cyclization, and oxidation reaction to the final pyridinium derivative. Quantitation of BPP-Lysine Derivatives in Modified Peanut Proteins. To investigate whether the pyridinium derivatives can be formed in food proteins as well, extracted proteins of native peanuts were incubated with trans-2-heptenal under mild physiological conditions at 37 °C for 24 h in phosphate buffer (pH 7.4) as previously described for other 2alkenals.17,28,29 Incubations were performed in different molar ratios between trans-2-heptenal and the sum of relevant reactive amino acids histidine, lysine and arginine. The amounts of these amino acids in the native peanut protein extract were 19.27 ± 0.06, 32.58 ± 1.39 and 70.58 ± 3.22 mmol/100 g protein, respectively. Cysteine was not regarded as a relevant amino acid because of its low amount in peanut proteins (3.83 ± 0.79 mmol/100 g protein) and its ability to form disulfide bonds. After acid hydrolysis, both BPP-lysine derivatives were quantitated by LC-ESI-MS/MS (MRM mode). Figure 6 shows a representative LC-ESI-MS/MS-chromatogram (m/z 333.3 > 204.2) of the hydrolyzed peanut proteins modified in the 0.1:1 ratio. It can be seen that the separation of the cis- and trans-derivatives can be achieved using an RP-18 column and addition of heptafluorobutyric acid to the regular ion pair reagent acetic acid. The quantitated amounts of both derivatives are shown in Figure 7. These results show clearly that a modification of peanut proteins by trans-2-heptenal is possible and that the amounts of formed BPP-lysine derivatives depend on the amount of trans-2-heptenal within the incubation samples. In the blank sample, consisting of peanut proteins treated in the same way but without trans-2-heptenal,

patterns, the presence of the cis- and trans-pyridinium derivatives can be assumed. The formation of the cis-derivative with trans-2-heptenal as reactant may be due to an isomerization of the trans-2-heptenal in the presence of the ε-aminogroup of lysine.17 On the basis of the UV-chromatograms before and after acid hydrolysis, it can be seen that the transderivative is the major product. This may be due to its thermodynamically preferred structure or due to the fact that the isomerization only leads to lower amounts of the cisderivative. This is confirmed by a recent literature report, in which cis- and trans-nonenal-pyridinium derivatives were quantitated in mice kidneys in comparable amounts after exposing the animals to ferric nitrilotriacetate.14 It is, therefore, likely that in real food samples, the cis- and trans-2-heptenal derivatives of lysine will also appear in equimolar amounts. For the isolation and unequivocal structure elucidation of the formed reaction products, Nα-acetyl-L-lysine was incubated with a 3-fold excess of trans-2-heptenal to induce the formation of the pyridinium derivatives. After acid hydrolysis, the products were separated and isolated by semipreparative HPLC with UV-detection, using an RP-18 column. Identification was realized by LC-ESI-MS/MS measurements and one and twodimensional NMR spectroscopy. Following collision-induced fragmentation, both substances showed the same identical fragmentation patterns as previously described. Figure 4 shows

Figure 4. Structure of trans-BPP-lysine showing the observed relevant heteronuclear multiple bond correlations (HMBC).

exemplarily the structure of (E)-1-(5-amino-5-carboxypentyl)4-butyl-3-(pent-1-en-1-yl)pyridin-1-ium (trans-BPP-lysine) and relevant heteronuclear multiple bond correlations (HMBC). The couplings in the HMBC spectrum of (Z)-1-(5-amino-5carboxypentyl)-4-butyl-3-(pent-1-en-1-yl)pyridin-1-ium (cisBPP-lysine) were analogous. Differentiation of cis- and transBPP-lysine was possible on the basis of the 1H vinyl proton (H12 and H-13) coupling constants J = 11.3 Hz for the cis-

Figure 5. Proposed formation pathway for the formation of trans-BPP-lysine. 8505

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

Article

only one molecule of trans-2-heptenal. Further cross-linking may occur due to a Michael addition of a pyridinium derivative to a Schiff base-adduct.16 Taking into account that the amounts of 2-heptenal formed in the course of lipid peroxidation as a result of peanut roasting will be probably lower than the amounts of trans-2-heptenal added to the peanut protein extract samples, it can be shown that especially in the presence of small amounts of trans-2-heptenal, nearly the whole lysine modification can be explained by the formation of the BPPlysine derivatives. So it is possible to cover most of the adducts formed by the reaction of the ε-amino group of lysine and 2heptenal by measuring the two BPP-lysine derivatives. In conclusion, we were able to show that trans-2-heptenal, a well-known secondary product formed in the course of the autoxidation of linoleic acid,14 is one of the most important volatile carbonyl compounds formed during heating peanut oil under roasting conditions and is able to modify the ε-amino group of lysine, leading to the formation of two isomeric pyridinium derivatives. Considering the whole peanut, it is likely that the volatile compounds formed in the course of the lipid peroxidation during the roasting process are able to react with the amino acid side chains of peanut proteins. Hence, these two derivatives represent good markers for a protein modification by 2-heptenal. Taking into account that other 2alkenals, like 2-hexenal, 2-octenal, or 2-nonenal, are able to form analogous adducts16,17,27 and that “mixed” derivatives consisting for example of one molecule each of lysine, 2hexenal, and 2-heptenal, respectively, are possible, a large variety of adducts can be expected. Furthermore, other secondary products of the lipid peroxidation like alkanals, 2,4alkadienals or 4-hydroxyalkenals are also able to form adducts with amino acid side chains.13,30,31 Due to the broad spectrum of possible precursors, we conclude that in food, especially as a result of roasting lipid-rich peanuts, a complex spectrum of adducts between amino acid side chains and secondary products of lipid peroxidation can be formed accounting for the large modification rate of lysine. Analogous to glycation reactions between reducing carbohydrates and amino compounds (also referred to a Maillard reaction or nonenzymatic glycosylation), the complex reactions involving carbonyl compounds from lipid peroxidation and amino acids could be called “lipation”. Due to the formation of these posttranslationally formed lipation products, the binding ability of immunoglobulines E against peanut proteins may be altered. This could have an influence on the allergenic potential of roasted peanuts. Further studies are needed to investigate whether lipation products are formed in roasted peanuts as well and whether these modifications have an influence on the allergenic potential of roasted peanuts.

Figure 6. Representative LC-ESI-MS/MS-chromatogram (MRM mode, transition m/z 333.3 > 204.2) of the acid hydrolyzate of the peanut protein extract modified by trans-2-heptenal in the ratio 0.1 to 1 related to the sum of the potentially reactive amino acids histidine, lysine and arginine.

Figure 7. Quantitated amounts of cis- and trans-BPP-lysine in modified peanut protein extracts by LC-ESI-MS/MS and decrease of lysine analyzed by amino acid analysis.

no derivatives could be measured. As previously discussed, the trans-derivative is the major product formed. Additionally, a direct relationship between the formation of the BPP-lysine derivatives and the decrease in lysine can be seen (Figure 8). In the slightly modified samples, up to 80% of the decrease in lysine can be explained by the formed BPP-lysine derivatives. At higher levels of trans-2-heptenal further reactions will become relevant. Having two electrophilic centers, possible reactions might be the formation of Michael- and Schiff’s base-adducts. These can lead for example to lysine cross-links by involving

■

AUTHOR INFORMATION

Corresponding Author

* Phone: +49-351-463-34647; fax: +49-351-463-34138; e-mail: [email protected]. Notes

Parts of this study were presented as poster contributions on the “11th International Symposium on the Maillard Reaction”, September 16−20, 2012, Nancy, France, and at the conference “Chemical Reactions in Foods VII”, November 14−16, 2012, Prague, Czech Republic. The authors declare no competing financial interest.

Figure 8. Explainable decreases of lysine due to the formation of cisand trans-BPP-lysine. 8506

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507

Journal of Agricultural and Food Chemistry

■

Article

peanuts. I. Major qualitative and some quantitative differences. J. Agr. Food Chem. 1972, 20, 700−706. (13) Esterbauer, H.; Schaur, R. J.; Zollner, H. Chemistry and biochemistry of 4-hydroxynonenal, malondialdehyde and related aldehydes. Free Radic. Biol. Med. 1991, 11, 81−128. (14) Esterbauer, H.; Zollner, H.; Schaur, R. J. Aldehydes Formed by Lipid Peroxidation: Mechanisms of Formation, Occurence, And Determination. In Membrane Lipid Oxidation; Vigo-Pelfrey, C., Ed.; CRC Press: Boca Raton, FL, 1990; Vol. 1, pp 239−283. (15) Uchida, K. Role of reactive aldehyde in cardiovascular diseases. Free Radic. Biol. Med. 2000, 28, 1685−1696. (16) Baker, A.; Zidek, L.; Wiesler, D.; Chmelik, J.; Pagel, M.; Novotny, M. V. Reaction of N-acetylglycyllysine methyl ester with 2alkenals: An alternative model for covalent modification of proteins. Chem. Res. Toxicol. 1998, 11, 730−740. (17) Ishino, K.; Wakita, C.; Shibata, T.; Toyokuni, S.; Machida, S.; Matsuda, S.; Matsuda, T.; Uchida, K. Lipid peroxidation generates body odor component trans-2-nonenal covalently bound to protein in vivo. J. Biol. Chem. 2010, 285, 15302−15313. (18) Lynch, M. P.; Faustman, C.; Silbart, L. K.; Rood, D.; Furr, H. C. Detection of lipid-derived aldehydes and aldehyde:protein adducts in vitro and in beef. J. Food Sci. 2001, 66, 1093−1099. (19) Zamora, R.; Hidalgo, F. J. Modification of lysine amino groups by the lipid-peroxidation product 4,5(E)-epoxy-2(E)-heptenal. Lipids 1994, 29, 243−249. (20) Zamora, R.; Alaiz, M.; Hidalgo, F. J. Determination of ϵ-Npyrrolylnorleucine in fresh food products. J. Agric. Food Chem. 1999, 47, 1942−1947. (21) Henle, T.; Walter, H.; Klostermeyer, H. Evaluation of the extent of the early Maillard-reaction in milk-products by direct measurement of the Amadori-product lactuloselysine. Z. Lebensm. Unters. Forsch. 1991, 193, 119−122. (22) Henle, T.; Walter, H.; Krause, I.; Klostermeyer, H. Efficient determination of individual Maillard compounds in heat-treated milk products by amino acid analysis. Int. Dairy J. 1991, 1, 125−135. (23) Chung, S. Y.; Champagne, E. T. Allergenicity of Maillard reaction products from peanut proteins. J. Agric. Food Chem. 1999, 47, 5227−5231. (24) Matissek, R.; Steiner, G.; Fischer, M. Lebensmittelanalytik, 4th ed.; Springer-Verlag: Berlin Heidelberg, 2010. (25) Hidalgo, F. J.; Zamora, R. The role of lipids in nonenzymatic browning. Grasas y Aceites 2000, 51, 35−49. (26) Gardner, H. W. Lipid hydroperoxide reactivity with proteins and amino-acids: a review. J. Agric. Food Chem. 1979, 27, 220−229. (27) Alaiz, M.; Barragán, S. Reaction of a lysyl residue analogue with E-2-octenal. Chem. Phys. Lipids 1995, 75, 43−49. (28) Furuhata, A.; Ishii, T.; Kumazawa, S.; Yamada, T.; Nakayama, T.; Uchida, K. N-epsilon-(3-methylpyridinium)lysine, a major antigenic adduct generated in acrolein-modified protein. J. Biol. Chem. 2003, 278, 48658−48665. (29) Ichihashi, K.; Osawa, T.; Toyokuni, S.; Uchida, K. Endogenous formation of protein adducts with carcinogenic aldehydes Implications for oxidative stress. J. Biol. Chem. 2001, 276, 23903− 23913. (30) Fenaille, F.; Tabet, J.-C.; Guy, P. A. Study of peptides containing modified lysine residues by tandem mass spectrometry: precursor ion scanning of hexanal-modified peptides. Rapid Commun. Mass Spectrom. 2004, 18, 67−76. (31) Kato, H.; Kaneko, S.; Okitani, A.; Ito, M.; Hayase, F. Modification of lysine residues to alkyl-substituted pyridiniums on exposure of proteins to vaporized hexanal. Agric. Biol. Chem. 1986, 50, 1223−1228.

ACKNOWLEDGMENTS We thank Karla Schlosser, Institute of Food Chemistry, for performing the amino acid analysis, Anke Peritz, Institute of Organic Chemistry, for performing the elemental analysis and Dr. Margit Gruner, Institute of Organic Chemistry, for performing the NMR analysis.

■

ABBREVIATIONS USED: AGE, advanced glycation endproduct; IgE, immunoglobulin E; cis-BPP-lysine, (Z)-1-(5-amino-5-carboxypentyl)-4-butyl-3(pent-1-en-1-yl)pyridin-1-ium; trans-BPP-lysine, (E)-1-(5amino-5-carboxypentyl)-4-butyl-3-(pent-1-en-1-yl)pyridin-1ium; CML, Nε-carboxymethyllysine; MWCO, molecular weight cut off; GC-MS (EI), gas chromatography with mass spectrometry after electron-impact ionization; LC-DAD-ESIMS/MS, liquid chromatography-diode array detector-electrospray ionization-tandem mass spectrometry

■

REFERENCES

(1) Souci, S. W.; Fachmann, W.; Kraut, H. Food Composition and Nutrition Tables, 7th ed.; MedPharm Scientific Publishers: Stuttgart, Germany, 2008; p 397. (2) Beyer, K.; Morrow, E.; Li, X. M.; Bardina, L.; Bannon, G. A.; Burks, A. W.; Sampson, H. A. Effects of cooking methods on peanut allergenicity. J. Allergy Clin. Immunol. 2001, 107, 1077−1081. (3) Cong, Y.-J.; Lou, F.; Li, L.-F.; Xue, W.-T.; Wang, J.; Zhang, H.; Fan, J.-F. Identification and quantification of major peanut allergens (in China) with IgE-binding property. J. Food Biochem. 2008, 32, 353− 367. (4) Mondoulet, L.; Drumare, M.-F.; Ah-Leung, S.; Paty, E.; Scheinmann, P.; Wal, J.-M.; Bernard, H. Influence of thermal processing on the IgE binding capacity of peanut allergens. Rev. Fr. Allergol. 2003, 43, 486−491. (5) Blanc, F.; Vissers, Y. M.; Adel-Patient, K.; Rigby, N. M.; Mackie, A. R.; Gunning, A. P.; Wellner, N. K.; Skov, P. S.; Przybylski-Nicaise, L.; Ballmer-Weber, B.; Zuidmeer-Jongejan, L.; Szépfalusi, Z.; Ruinemans-Koerts, J.; Jansen, A. P. H.; Bernard, H.; Wal, J.-M.; Savelkoul, H. F. J.; Wichers, H. J.; Mills, E. N. C. Boiling peanut Ara h 1 results in the formation of aggregates with reduced allergenicity. Mol. Nutr. Food Res. 2011, 55, 1887−1894. (6) Maleki, S. J.; Chung, S.-Y.; Champagne, E. T.; Raufman, J.-P. The effects of roasting on the allergenic properties of peanut proteins. J. Allergy Clin. Immunol. 2000, 106, 763−768. (7) Mueller, G. A.; Maleki, S. J.; Johnson, K.; Hurlburt, B. K.; Cheng, H.; Ruan, S.; Nesbit, J. B.; Pomés, A.; Edwards, L. L.; Schorzman, A.; Deterding, L. J.; Park, H.; Tomer, K. B.; London, R. E.; Williams, J. G. Identification of Maillard reaction products on peanut allergens that influence binding to the receptor for advanced glycation end products. Allergy 2013, 68, 1546−1554. (8) Vissers, Y. M.; Iwan, M.; Adel-Patient, K.; Skov, P. S.; Rigby, N. M.; Johnson, P. E.; Müller, P. M.; Przybylski-Nicaise, L.; Schaap, M.; Ruinemans-Koerts, J.; Jansen, A. P. H.; Mills, E. N. C.; Savelkoul, H. F. J.; Wichers, H. J. Effect of roasting on the allergenicity of major peanut allergens Ara h 1 and Ara h 2/6: the necessity of degranulation assays. Clin. Exp. Allergy 2011, 41, 1631−1642. (9) Wellner, A.; Nußpickel, L.; Henle, T. Glycation compounds in peanuts. Eur. Food Res. Technol. 2012, 234, 423−429. (10) Maguire, L. S.; O’Sullivan, S. M.; Galvin, K.; O’Connor, T. P.; O’Brien, N. M. Fatty acid profile, tocopherol, squalene and phytosterol content of walnuts, almonds, peanuts, hazelnuts and the macadamia nut. Int. J. Food Sci. Nutr. 2004, 55, 171−178. (11) Liu, X.; Jin, Q.; Liu, Y.; Huang, J.; Wang, X.; Mao, W.; Wang, S. Changes in volatile compounds of peanut oil during the roasting process for production of aromatic roasted peanut oil. J. Food Sci. 2011, 76, 404−412. (12) Brown, D. F.; Senn, V. J.; Stanley, J. B.; Dollear, F. G. Comparison of carbonyl compounds in raw and roasted runner 8507

dx.doi.org/10.1021/jf502501f | J. Agric. Food Chem. 2014, 62, 8500−8507