Phenolics as carbonyl scavengers: an additional protection against lipid oxidative damage in foods.

Phenolics as carbonyl scavengers: an additional protection against lipid oxidative damage in foods. Mª ISABEL AGUILAR TÉLLEZ Department of Organic Chemistry Faculty of Chemistry Department of Lipids Characterization and Quality Instituto de la Grasa Rosario Zamora Corchero, Profesora de Investigación del Consejo Superior de Investigaciones Científicas (CSIC), y Francisco Javier Hidalgo García, Profesor de Investigación del Consejo Superior de Investigaciones Científicas (CSIC), INFORMAN que el presente trabajo titulado “Phenolics as carbonyl scavengers: an additional protection against lipid oxidative damage in foods” ha sido realizado bajo nuestra dirección y asesoramiento en los laboratorios del Departamento de Caracterización y Calidad de Lípidos del Instituto de la Grasa (CSIC) de Sevilla durante los cursos académicos 2013-2014 al 2017- 2018, constituyendo la Memoria que presenta la Lcda. Mª Isabel Aguilar Téllez para aspirar al grado de Doctor con la mención “Doctor Internacional” por la Universidad de Sevilla. Consideramos que el trabajo realizado es original y ha dado lugar a cinco artículos científicos en revistas con amplio reconocimiento internacional. La excelente labor realizada por la Lcda. Mª Isabel Aguilar Téllez, tanto en tareas experimentales como en las discusiones de los resultados, acredita una formación adecuada para la obtención del título de Doctor con la mención “Doctor Internacional” por la Universidad de Sevilla. Y para que conste a efectos de tramitación de la referida Tesis, firmamos el presente informe en Sevilla a ocho de Febrero de dos mil dieciocho. Prof. Rosario Zamora Corchero Prof. Francisco J. Hidalgo García VºBº Tutora de la Tesis Prof. Mª Carmen Ortíz Mellet. Catedrática de la Universidad de Sevilla Phenolics as carbonyl scavengers: an additional protection against lipid oxidative damage in foods. Doctoral thesis for the degree of Philosophiae Doctor presented by Mª Isabel Aguilar Téllez Seville, 2018. Department of Organic Chemistry Faculty of Chemistry Department of Lipids Characterization and Quality Instituto de la Grasa Thesis Supervisor Approval Thesis Supervisor Approval Department of Organic Chemistry Faculty of Chemistry Prof. Francisco J. Hidalgo García Prof. Rosario Zamora Corchero Instituto de la Grasa Instituto de la Grasa Spanish National Research Council (CSIC) Spanish National Research Council (CSIC) Seville, Spain Seville, Spain Department of Lipids Characterization and Quality Instituto de la Grasa Tutor: Professor Mª Carmen Ortiz Mellet Department of Organic Chemistry Faculty of Chemistry University of Seville Seville, Spain Supervisors: Professor Francisco Javier Hidalgo García Instituto de la Grasa Spanish National Research Council (CSIC) Seville, Spain Professor Rosario Zamora Corchero Instituto de la Grasa Spanish National Research Council (CSIC) Seville, Spain Reviewers: Dr. Stephen Elmore Department of Food and Nutritional Sciences University of Reading Whiteknights, Reading, UK. Dr. Pedro Jesús García Moreno National Food Institute Technical University of Denmark Lyngby, Denmark University of Seville Seville, 2018. Indexes IV 1.3.5. The hypothesis of the triple defensive barrier of phenolic compounds against the lipid oxidative damage produced in foods ............................................................................................. 53 2. AIMS OF THE STUDY ........................................................................... 57 3. MATERIALS AND METHODS ............................................................. 61 3.1. Materials ............................................................................................. 63 3.1.1. Commercial products ................................................................. 63 3.1.2. Syntheses of epoxyalkenals ........................................................ 64 3.1.3. Syntheses of oxoalkenals ............................................................ 66 3.2. Analytical instrumentation ................................................................ 67 3.2.1. Gas chromatography-mass spectrometry (GC-MS) ............... 67 3.2.2. Liquid chromatographytandem mass spectrometry (LCMS/MS) ....................................................................................... 67 3.2.3. Liquid chromatography-high resolution mass spectrometry (LC-HRMS) ......................................................... 67 3.2.4. Nuclear magnetic resonance spectroscopy (NMR) .................. 68 3.2.5. Other equipment ........................................................................ 68 3.3. Studied systems ................................................................................... 69 3.3.1. Formation of carbonyl-phenol adducts in the reaction of alkanals and phenolic compounds ............................................ 69 3.3.1.1. Synthesis and characterization of alkanal-phenol adducts ... 69 3.3.1.2. Effect of aldehyde chain length on the formation of alkanal-phenol adducts ........................................................ 70 3.3.1.3. Effect of aldehyde branching on the formation of alkanalphenol adducts ...................................................................... 70 Indexes V 3.3.2. Formation of carbonyl-phenol adducts in the reaction of epoxyalkenals and phenolic compounds ..................................... 71 3.3.3. Formation of carbonyl-phenol adducts in the reaction of oxoalkenals and phenolic compounds ......................................... 72 3.3.4. Formation of carbonyl-phenol adducts in the reaction of alkenals and quercetin .................................................................. 73 3.3.5. Thermal degradation of unsaturated aldehydes ........................ 74 3.3.6. Deep-frying experiments .............................................................. 75 3.4. Analytical determinations .................................................................. 76 3.4.1. Determination of aldehydes ......................................................... 76 3.4.1.1. Aldehydes produced by thermal degradation of alkadienals and 2-alkenals ................................................... 76 3.4.1.2. Aldehydes in heated oils ........................................................ 79 3.4.2. Determination of carbonyl-phenol adducts ............................... 80 3.4.2.1. Adducts formed with alkanals ............................................... 80 3.4.2.2. Adducts formed with epoxyalkenals ...................................... 80 3.4.2.3. Adducts formed with oxoalkenals .......................................... 81 3.4.2.4. Adducts formed with alkenals in model systems.................... 82 3.4.2.5. Adducts formed with alkenals in fried onions ....................... 83 3.5. Statistical analyses .............................................................................. 83 4. RESULTS .................................................................................................. 85 4.1. Alkanal-trapping ability of phenolic compounds ............................ 87 4.1.1. Characterization of the adducts produced in the reaction between alkanals and phenolic compounds ................................ 88 Indexes VI 4.1.1.1. Compounds isolated and characterized in the reaction of pentanal and 2-methylresorcinol after acetylation of the reaction mixture .................................................................... 88 4.1.1.2. Compounds isolated and characterized in the different alkanal/phenolic compound mixtures assayed ..................... 92 4.1.2. The structure-activity relationships (SAR) of aldehydes and phenolic compounds in the formation of carbonyl-phenol adducts ......................................................................................... 102 4.2. Epoxyalkenal-trapping ability of phenolic compounds ................ 108 4.2.1. Characterization of the adducts produced in the reaction between epoxyalkenals and phenolic compounds .................... 108 4.2.1.1. Compounds isolated and characterized in the reaction of 4,5-epoxy-2-heptenal and 2-methylresorcinol .................... 114 4.2.1.2. Compounds isolated and characterized in the reaction of 4,5-epoxy-2-decenal and 2,5-dimethylresorcinol ............... 116 4.2.1.3. Compounds isolated and characterized in the reaction of 4,5-epoxy-2-hexenal and 2-methylresorcinol ..................... 117 4.2.1.4. Compounds isolated and characterized in the reaction of 4,5-epoxy-2-heptenal and 2,5-dimethylresorcinol .............. 118 4.2.2. Effect of reaction conditions on the formation of epoxyalkenal-phenol adducts in the reaction of 4,5-epoxy-2- heptenal and 2-methylresorcinol ............................................... 119 4.3. Oxoalkenal-trapping ability of phenolic compounds .................... 128 4.3.1. Characterization of the adducts produced in the reaction between oxoalkenals and phenolic compounds ........................ 128 4.3.1.1. Compounds isolated and characterized in the reaction of 4-oxo-2-hexenal and 2-methylresorcinol ............................ 134 Indexes VII 4.3.1.2. Compounds isolated and characterized in the reaction of 4-oxo-2-nonenal and 2-methylresorcinol ........................... 137 4.3.1.3. Compounds isolated and characterized in the reaction of 4-oxo-2-nonenal and resorcinol ......................................... 140 4.3.1.4. Compounds isolated and characterized in the reaction of fumaraldehyde and 2-methylresorcinol .............................. 141 4.3.1.5. Compounds isolated and characterized in the reaction of fumaraldehyde and resorcinol ............................................ 142 4.3.2. Effect of reaction conditions on the formation of oxoalkenal-phenol adducts in the reaction of 4-oxo-2- hexenal and 2-methylresorcinol .............................................. 142 4.4. Thermal degradation of unsaturated aldehydes ............................ 150 4.4.1. Thermal degradation of 2-alkenals and 2,4-alkadienals ......... 151 4.4.2. Thermal degradation of 2-pentenal .......................................... 154 4.4.3. Thermal degradation of 2-octenal............................................. 159 4.4.4. Thermal degradation of 2,4-heptadienal .................................. 162 4.4.5. Thermal degradation of 2,4-decadienal.................................... 167 4.5. Carbonyl-trapping by phenols during the frying process ............ 172 4.5.1. Formation of carbonyl-phenol adducts by reaction of acrolein, crotonaldehyde, or (E)-2-pentenal with quercetin . 173 4.5.2. Fate of toxicologically relevant carbonyls during thermal heating of oils in the presence and in the absence of added food ............................................................................................ 181 4.5.3. Formation of aldehyde-phenol adducts in fried onions .......... 182 4.5.4. Identification of aldehyde-phenol adducts in commercially crispy fried onions .................................................................... 184 5. DISCUSSION ......................................................................................... 187 Indexes VIII 5.1. The phenolic trapping of alkanals .................................................. 189 5.2. The phenolic trapping of epoxyalkenals ......................................... 195 5.3. The phenolic trapping of oxoalkenals ............................................. 200 5.4. Lipid-derived aldehyde degradation under thermal conditions .. 203 5.5. Lipid-derived carbonyls are trapped by phenolics under common food processing conditions ............................................... 205 5.6. Lipid-derived carbonyl trapping, an additional protective mechanism of food phenolics .......................................................... 206 6. CONCLUSIONS .................................................................................... 211 7. REFERENCES ....................................................................................... 217 Indexes IX LIST OF FIGURES Figure 1. Chemical structures of some fatty acids corresponding to main PUFAs families ........................................................................................... 5 Figure 2. Lipid oxidation pathway .................................................................. 6 Figure 3. The initiation step of lipid oxidation for linoleic acid...................... 7 Figure 4. Primary lipid oxidation products of linoleic acid ........................... 10 Figure 5. Degradation of linoleic acid hydroperoxides to hydroxy-, epoxy- and oxo-fatty acids ........................................................................ 11 Figure 6. Some typical scission products of 9-hydroperoxy linoleic acid. .... 13 Figure 7. Proposed mechanism for 2-pentenal and propanal formation from linolenic acid .................................................................................... 15 Figure 8. Structures of the tertiary lipid oxidation products and analogues employed in this study. ............................................................................. 16 Figure 9. Formation of pyridinium salts by carbonyl-amine reactions involving alkanals ..................................................................................... 19 Figure 10. Formation of 2-alkylpyridines by carbonyl-amine reactions involving different lipid-derived reactive carbonyls. ................................ 20 Figure 11. Formation of dihydropyridines by oligomerization of malondialdehyde in the presence of alkanals and later reaction with amino compounds. .................................................................................... 20 Figure 12. Formation of 1-alkylpyrroles and 1-alkyl-2-(1’- hydroxyalkyl)pyrroles by reaction of 4,5-epoxy-2-alkenals with amino compounds ................................................................................................ 21 Figure 13. Formation of 1,2-dialkylpyrroles by carbonyl-amine reactions involving different lipid-derived reactive carbonyls. ................................ 22 Figure 14. Formation of imidazolium salts by carbonyl-amine reactions involving glyoxal ...................................................................................... 23 Indexes X Figure 15. Formation of Michael adducts by addition of an imidazole ring, such as the present in the side chain of histidine, to 4,5-epoxy-2- alkenals...................................................................................................... 23 Figure 16. General scheme for amino acid degradations induced by lipidderived reactive carbonyls......................................................................... 25 Figure 17. Proposed pathway for the conversion of amino acids into Strecker aldehydes in the presence of lipid-derived reactive carbonyls ... 26 Figure 18. Proposed pathway for the conversion of amino acids into - keto acids in the presence of lipid-derived reactive carbonyls .................. 28 Figure 19. Proposed pathway for the conversion of amino acids into amines and Strecker aldehydes in the presence of lipid-derived reactive carbonyls ..................................................................................... 29 Figure 20. Proposed pathway for the conversion of amines into Strecker aldehydes in the presence of lipid-derived reactive carbonyls .................. 29 Figure 21. Proposed pathway for the conversion of amines into olefins in the presence of lipid-derived reactive carbonyls ....................................... 30 Figure 22. The formation of carbonyl-amine adducts (advanced lipoxidation end-products, ALEs) in the course of lipid oxidation. .......... 32 Figure 23. Proposed mechanisms for the formation of brown pigments by carbonyl–amine reactions.......................................................................... 33 Figure 24. Carbonyl-amine reactions as a broadcasting of lipid oxidative damage to food macromolecules containing amino groups ...................... 34 Figure 25. Structures of benzoic and hydroxycinnamic acids, flavonoids and stilbenes .............................................................................................. 36 Figure 26. Chemical structures of model phenolic compounds used in this study. .................................................................................................. 37 Indexes XI Figure 27. Carbon skeleton of flavane. ......................................................... 38 Figure 28. Chemical structure of quercetin. .................................................. 38 Figure 29. Mechanism for the transition metal chelation by flavonoids. ...... 41 Figure 30. Resonance delocalization of phenoxyl radical. ............................ 43 Figure 31. A termination reaction between a phenoxyl radical and a lipid peroxyl radical ........................................................................................... 44 Figure 32. Comparative electronic delocalization produced in m-, o-, and p-diphenols after proton loss ..................................................................... 50 Figure 33. Alkanal addition to resorcinol. ..................................................... 51 Figure 34. Reaction of 2-alkenals with resorcinol......................................... 52 Figure 35. Structure of diACR-conjugated phloretin adduct. ....................... 52 Figure 36. Presence of chelating, free radical-scavenging, and carbonylscavenging regions in selected flavonoids ................................................ 54 Figure 37. Chemical structures of some of the aldehydes employed in this study ................................................................................................... 64 Figure 38. Synthesis of epoxyalkenals used in this study. ............................ 65 Figure 39. Syntheses of 4-oxo-2-nonenal and 4-oxo-2-hexenal. ................... 66 Figure 40. Total ion chromatogram obtained by GC-MS of the reaction between pentanal and 2-methylresorcinol after acetylation ...................... 89 Figure 41. Chemical structures of the compounds isolated and characterized by NMR and MS in the reaction between pentanal and 2-methylresorcinol. ................................................................................... 90 Figure 42. Chemical structures of the carbonyl-phenol adducts isolated and characterized by NMR and MS in the different assayed reactions between alkanals and phenolics ................................................................ 94 Figure 43. Time-course of carbonyl-phenol adduct formation in the reactions of: (A), resorcinol; (B), 2-methylresorcinol; (C), 2,5- Indexes XII dimethylresorcinol; and (D), orcinol; with propanal, butanal, pentanal, or hexanal ................................................................................................ 104 Figure 44. Time-course of carbonyl-phenol adduct formation in the reactions of butanal and 2-methylpropanal with 2-methylresorcinol. ..... 106 Figure 45. Time-course of carbonyl-phenol adduct formation in the reactions of pentanal, 2-methylbutanal, and 3-methylbutanal with 2- methylresorcinol.………………………………………………………..107 Figure 46. Total ion chromatogram obtained by GC-MS of the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol after acetylation ............................................................................................... 109 Figure 47. Chemical structures of the carbonyl-phenol adducts isolated and characterized by NMR and MS in the different assayed reactions between epoxyalkenals and phenolics .................................................... 110 Figure 48. HMBC couplings exhibited by compounds 20a and 25a .......... 111 Figure 49. Chemical structures of adducts produced in the reaction between epoxyalkenals and phenolics .................................................... 113 Figure 50. Effect of reaction pH on: (A), phenol disappearance; and (B), formation of carbonyl-phenol adducts in the reaction between 4,5- epoxy-2-heptenal and 2-methylresorcinol ............................................... 120 Figure 51. Effect of epoxyalkenal concentration on: (A), phenol disappearance; and (B), formation of carbonyl-phenol adducts in the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol ............ 121 Figure 52. Effect of phenol concentration on: (A), remaining phenol; and (B), formation of carbonyl-phenol adducts in the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol ........................................ 122 Figure 53. Effect of reaction time and temperature on the phenol disappearance in the reaction between 4,5-epoxy-2-heptenal and 2- methylresorcinol ...................................................................................... 123 Indexes XIII Figure 54. Effect of reaction time and temperature on the formation of compound 21 ........................................................................................... 124 Figure 55. Effect of reaction time and temperature on the formation of compound 25c ......................................................................................... 125 Figure 56. Effect of reaction time and temperature on the formation of compound 25a ......................................................................................... 126 Figure 57. Effect of reaction time and temperature on the formation of compound 25b......................................................................................... 127 Figure 58. Effect of reaction time and temperature on the formation of compound 25d......................................................................................... 128 Figure 59. Total ion chromatograms obtained by GC-MS of the reaction between 2-methylresorcinol and: (A), 4-oxo-2-hexenal, and (B), 4- oxo-2-nonenal, after acetylation .............................................................. 130 Figure 60. Chemical structures of the carbonyl-phenol adducts isolated and characterized by NMR and MS in the different assayed reactions between oxoalkenals and phenolics ........................................................ 131 Figure 61. Chemical structures of adducts produced in the reaction between oxoalkenals and phenolics ........................................................ 134 Figure 62. Effect of reaction pH on the formation of carbonyl-phenol adducts in the reaction between 4-oxo-2-hexenal and 2- methylresorcinol ...................................................................................... 143 Figure 63. Effect of oxoalkenal concentration on the formation of carbonyl-phenol adducts in the reaction between 4-oxo-2-hexenal and 2-methylresorcinol .................................................................................. 144 Figure 64. Effect of phenol concentration on the formation of carbonylphenol adducts in the reaction between 4-oxo-2-hexenal and 2- methylresorcinol. ..................................................................................... 145 Indexes XX ROS. Reactive oxygen species SAR. Structure-activity relationships TAG. Triacylglycerol. TOF. Time-of-flight Indexes XXI LIST OF ISOLATED AND CHARACTERIZED COMPOUNDS Compound 2. 4-(1-Methoxypentyl)-2-methyl-1,3-phenylene diacetate Compound 3. 2-Methyl-4-(pent-1-en-1-yl)-1,3-phenylene diacetate Compound 4. 4-(1-Acetoxypentyl)-2-methyl-1,3-phenylene diacetate Compound 5. 4-Butyl-8-methyl-3-propylchroman-2,7-diyl diacetate Compound 6. 4-(1-Methoxypropyl)benzene-1,3-diol Compound 7. 4-(1-Methoxypentyl)benzene-1,3-diol Compound 8. 2-(1-Hydroxypentyl)benzene-1,3-diol Compound 9a and 9b 4-(1-Methoxy-2-methylbutyl)benzene-1,3-diol Compound 10. 4-(1-Methoxyhexyl)benzene-1,3-diol Compound 11. 4-(1-Methoxypropyl)-2-methylbenzene-1,3-diol Compound 12. 4-(1-Methoxypentyl)-2-methylbenzene-1,3-diol Compound 13. 2-Methyl-4-(pent-1-en-1-yl)benzene-1,3-diol Compound 14a and 14b. 4-(1-Methoxy-2-methylbutyl)-2-methylbenzene- 1,3-diol Compound 15. 4-(1-Methoxy-3-methylbutyl)-2-methylbenzene-1,3-diol Compound 16. 4-(Hex-1-en-1-yl)-2-methylbenzene-1,3-diol Compound 17. 1-(2,4-Dihydroxyphenyl)-2-hydroxyethan-1-one Compound 18. 1-(2,4-dihydroxy-3-methylphenyl)-2-hydroxyethan-1-one Compound 19. 2-Ethyl-3-methyl-2H-chromen-7-ol Compound 20a and 20b. 4-Ethyl-2-hydroxy-6-methyl-1,3a,4,9b-tetrahydro- 2H-furo[2,3-c]chromen-7-yl acetate Compound 21. 2-Ethyl-8-methyl-4-(2-oxoethyl)-2H-chromen-7-yl acetate Indexes XXII Compound 22. 2-Hydroxy-6,9-dimethyl-4-pentyl-1,3a,4,9b-tetrahydro-2H- furo[2,3-c]chromen-7-yl acetate Compound 23. 8-Methyl-4-(2-oxoethyl)-2-pentyl-2H-chromen-7-yl acetate Compound 24. 1,8-Dimethyl-3,4,4a,9a-tetrahydro-1H-pyrano[3,4- b]benzofuran-3,7-diyl diacetate Compound 25a and 25b. 1-Ethyl-8-methyl-3,4,4a,9a-tetrahydro-1H- pyrano[3,4-b]benzofuran-3,7-diyl diacetate Compound 26. 1-Ethyl-5,8-dimethyl-3,4,4a,9a-tetrahydro-1H-pyrano[3,4- b]benzofuran-3,7-diyl diacetate Compound 29. 7-Methyl-2-(2-oxobutyl)benzofuran-6-yl acetate Compound 30a and 30b. 8a-Ethyl-7-methyl-2,3,3a,8a-tetrahydrofuro[2,3- b]benzofuran-2,6-diyl diacetate Compound 31a and 31b. 8-Methyl-4-propionylchromane-2,7-diyl diacetate Compound 32. 7-Methyl-2-(2-oxoheptyl)benzofuran-6-yl acetate Compound 33a and 33b. 7-Methyl-8a-pentyl-2,3,3a,8a-tetrahydrofuro[2,3- b]benzofuran-2,6-diyl diacetate Compound 34a and 34b. 4-Hexanoyl-8-methylchromane-2,7-diyl diacetate Compound 35. 2-(2-Hydroxyheptyl)-7-methylbenzofuran-6-ol Compound 36. 2-(2-Hydroxyheptyl)benzofuran-6-ol Compound 37. 2-(2-Hydroxyethyl)-7-methylbenzofuran-6-ol Compound 38. 2-(2-Hydroxyethyl)-7-methyl-2,3-dihydrobenzofuran-3,6-diol Compound 39. 2-(2-Hydroxyethyl)benzofuran-6-ol Compound 56. 2-(3,4-Dihydroxyphenyl)-3,5,8-trihydroxy-9,10-dihydro- 4H,8H-pyrano[2,3-f ]chromen-4-one Compound 57. 2-(3,4-Dihydroxyphenyl)-3,5,8-trihydroxy-10-methyl-9,10- dihydro- 4H,8H-pyrano[2,3-f ]chromen-4-one Indexes XXIII Compound 61. 2-(3,4-Dihydroxyphenyl)-3,5-dihydroxy-8-methyl-4H,8H- pyrano-[2,3-f ]chromen-4-one Compound 62. 2-(3,4-Dihydroxyphenyl)- 3,5-dihydroxy-8-ethyl-4H,8H- pyrano-[2,3-f ]chromen-4-one Indexes XXIV Indexes XXV LIST OF ORIGINAL PUBLICATIONS Paper 1. Zamora, R., Navarro, J. L., Aguilar, I., & Hidalgo, F. J. (2015). Lipid-derived aldehyde degradation under thermal conditions. Food Chemistry, 174, 89-96. Paper 2. Zamora, R., Aguilar, I., Granvogl, M., & Hidalgo, F. J. (2016). Toxicologically relevant aldehydes produced during frying process are trapped by food phenolics. Journal of Agricultural and Food Chemistry, 64, 5583-5589. Paper 3. Zamora, R., Aguilar, I., & Hidalgo, F. J. (2017). Epoxyalkenaltrapping ability of phenolic compounds. Food Chemistry, 237, 444-452. Paper 4. Hidalgo, F. J., Aguilar, I., & Zamora, R. (2017). Model studies on the effect of aldehyde structure on their selective trapping by phenolic compounds. Journal of Agricultural and Food Chemistry, 65, 4736-4743. Paper 5. Hidalgo, F. J., Aguilar, I., & Zamora, R. (2017). Phenolic trapping of lipid oxidation products 4-oxo-2-alkenals. Food Chemistry, 240, 822-830. Indexes XXVI Introduction 1 1. INTRODUCTION Introduction 2 Introduction 3 1.1. Lipids as a source of reactive carbonyls in foods. 1.1.1. Lipids. Lipids are a complex group of organic compounds that are insoluble in water but soluble in organic solvents (e.g., chloroform, ether), and are present or are derived from living organisms (Nichols & Sanderson, 2003; Coultate, 2016). Therefore, this broad definition includes many chemically heterogeneous substances, such as acylglycerols, waxes, phospholipids, sphingolipids, steroids, tocopherols, carotenoids, terpenes and polycyclic hydrocarbons. Lipids can be classified on the basis of their physical properties at room temperature (oils are liquid and fats are solid), their polarity [neutral (e.g., triacylglycerols) or polar (e.g., phospholipids)], or their essentiality for humans (essential or nonessential fatty acids). Besides, based on structure, lipids can be classified as simple (acylglycerols, waxes and esters of sterols), complex (glycerophospholipids, glyceroglycolipids and sphingolipids) and derived (fatty alcohols, fatty acids, steroids, fat-soluble vitamins and hydrocarbons) (Mara Block & Barrera-Arellano, 2012; O´Keefe, 2008). Lipids are necessary components in our diet because of their relation to our health. Thus, they are sources of fat-soluble vitamins (A, D, E, and K) and other functional compounds such as essential fatty acids, and provide energy for organic functions. In addition, foods lipids play an important role in food quality by contributing to sensory attributes such as color, rheological properties, and flavor. In general, lipids have a desirable impact on the sensory properties of foods by affecting the mouth-feel. On the other hand, this desirable contribution may be easily lost as a consequence of deteriorative reactions, mainly oxidation and hydrolysis (Kolakowska & Sikorski, 2003; McClements & Decker, 2017). Introduction 10 complex mixture of LOPs is produced (McClements, & Decker, 2017; Bastos et al., 2017). Because of their complexity, LOPs are usually classified depending on their oxidation level into primary, secondary, and tertiary LOPs. A. Primary LOPs. Hydroperoxides, which are produced in the propagation and termination steps, are the primary LOPs. These products are odorless and tasteless since they are not volatiles. However, they are degraded into compounds that are responsible for off-flavors (Erickson, 2008; Márquez- Ruiz et al., 2014; Belitz et al., 2009). They are produced at the initial step of oxidation and their concentration decreases afterwards. Nevertheless, they are an important fraction of LOPs at low and moderate temperatures. Figure 4 shows the hydroperoxides formed as a consequence of linoleic acid oxidation. Figure 4. Primary lipid oxidation products of linoleic acid. R1 = C5H11 and R2 = C8H15O2. Introduction 11 B. Secondary LOPs. Secondary LOPs are formed from hydroperoxides by different transformations and cleavages (Figure 5). This cascade of reactions, which includes abstraction of hydrogens, cyclization, addition to double bonds and scissions, generates hundreds of different compounds, including volatile and non-volatile products (such as aldehydes, ketones, acids, esters, alcohols, and short-chain hydrocarbons) (McClements, & Decker, 2017; Schaich, 2013). Figure 5. Degradation of linoleic acid hydroperoxides to hydroxy-, epoxy- and oxo-fatty acids. R1 = C5H11 and R2 = C8H15O2 (Belitz et al., 2009). Introduction 12 These compounds can be classified according to their molecular weights within three groups: compounds with molecular weights lower than the non-oxidized lipid, oxidized monomers or compounds with molecular weights similar to those of the non-oxidized lipid, and dimeric and polymeric compounds with molecular weights higher than the non-oxidized lipid. The compounds with molecular weights lower than the non-oxidized lipid are produced by alkoxyl radical breakdown (Márquez-Ruiz et al., 2014). Alkoxyl radicals can suffer scission of the C-C bond on either side of the RO group producing an oxo compound and an alkyl or alkenyl radical on the aliphatic chain (McClements, & Decker, 2017; Schaich, 2013; Kim & Min, 2008). The formed radical can then either abstract one hydrogen atom to form a stable compound (alkanes, alkenes, fatty acids) and propagate the radical chain, or react with a hydroxyl radical to form an alcohol or with oxygen to form a hydroperoxide, which might be further decomposed. Figure 6 shows some of the scission products produced during linoleic acid oxidation. It includes the formation of 2,4-decadienal, a secondary LOP used in the experiments described in this thesis. 2,4- Heptadienal, also used in this study, is produced similarly from linolenic acid. Introduction 13 Figure 6. Some typical scission products of 9-hydroperoxy linoleic acid. R1 = C5H11 and R2 = C8H15O2. (Adapted from Schaich, 2013). Introduction 14 Fragments containing the oxygen atom (mainly aldehydes) become the volatile and flavor compounds most strongly associated with rancidity, and they have an impact even at extremely low concentrations (Belitz et al., 2009; Schaich, 2013). The main volatile compounds produced by oleic acid, linoleic acid and linolenic acid are summarized in Table 1. Table 1. Volatile compounds formed by lipid oxidation of unsaturated fatty acids. Oleic acid Linoleic acid Linolenic acid Heptanal Pentane Propanal Octanal Pentanal 1-Penten-3-one Nonanal Hexanal 2-Butenal Decanal Heptanal 2-Pentenal 2-Decenal 2-Heptenal 2-Hexenal 2-Undecenal Octanal 3-Hexenal 1-Octen-3-one 2-Heptenal 1-Octen-3-hydroperoxide 2,4-Heptadienal 2-Octenal 2,4-Octadienal 3-Nonenal 3,5-Octadien-2-one 2-Nonenal 1,5-Octadien-3-one 2-Decenal 1,5-Octadien-3-hydroperoxide 2,4-Nonadienal 2,6-Nonadienal 2,4-Decadienal 2,4,7-Decatrienal 4,5-Epoxy-2-decenal Adapted from Belitz et al. (2009). This table includes some of the carbonyl compounds used in the experiments described in this work, such as propanal, pentanal, hexanal, 2- pentenal, 2-octenal, 2,4- heptadienal and 2,4-decadienal. The mechanism for alkadienal formation is shown in Figure 6. The mechanism proposed for the formation of 2-alkenals and alkanals from linolenic acid is summarized in Figure 7. Other alkenals and alkanals are produced similarly. Introduction 15 Figure 7. Proposed mechanism for 2-pentenal and propanal formation from linolenic acid. R1 = C8H15O2 (Frankel, 2005). When these reactions occur without scission of the carbon chain, compounds with molecular weights similar to those of non-oxidized lipids are produced (Figure 5). They have an oxygenated function (usually epoxy, hydroxyl or oxo group). These compounds are the most abundant LOPs formed at ambient or moderate temperatures (Márquez-Ruiz et al., 2014; Gardner, 1989). Finally, the reaction of two or more lipid radicals produce dimeric and polymeric compounds (Márquez-Ruiz et al., 2014). C. Tertiary LOPs. Secondary LOPs can suffer further oxidations to produce new products with two oxygenated functions, which are named tertiary LOPs. For example, an alcoxyl radical can form either secondary stable products or tertiary products if it suffers a new oxidation. These compounds usually have two of the following functional groups: hydroperoxy, epoxy, hydroxyl or oxo, Introduction 16 and they are very reactive. These products play a major role in the color and flavor changes produced in food as a consequence of lipid oxidation (Hidalgo, & Zamora, 2004). Figure 8 collects the structures of the tertiary LOPs used in this investigation. 4,5-Epoxy-2-heptenal and 4-oxo-2-hexenal are both oxidation products of ω3 fatty acyl chains, while 4,5-epoxy-2-decenal and 4- oxo-2-nonenal are derived from ω6 fatty acyl chains. In addition, fumaraldehyde and 4,5-epoxy-2-hexenal were also used as models of 4-oxo-2- alkenals and 4,5-epoxy-2-alkenals, respectively, in some experiments. The mechanism of formation of 4,5-epoxy-2-alkenals is shown in Figure 5. Figure 8. Structures of the tertiary LOPs and analogues employed in this study. 1.1.5. Consequences of lipid oxidation. Lipid oxidation strongly impacts on food properties and quality (Schaich, 2013; Brewer, 2014; Steel et al., 2012; Correddu et al., 2015; Liu et al., 2016; Sainsbury et al., 2016). Thus, LOPs affect food sensory properties generating flavors, off-odors, browning, texture alterations and color loss. In general, volatile compounds produced during lipid oxidation are detrimental to food quality although there are some food products, such as fried foods, dried cereals, and cheeses, for which small amounts of LOPs are important positive components for their flavor profile (McClements & Decker, 2017). Introduction 17 Aldehydes are main responsible for the formation of off-flavors as a consequence of lipid oxidation. Undesirable odors resulting from oxidation are commonly described as stale, wet cardboard, painty, grassy, or rancid (Brewer, 2014). In addition, lipid oxidation causes a reduction in the nutritional value of the food, and it can seriously impact toxicity and safety due to the formation of potentially toxic compounds, including , -unsaturated aldehydes (Guillén & Goicoechea, 2008a; Guillén & Uriarte, 2012; Liu et al., 2003; Zarkovic, 2003; Guillén & Goicoechea, 2008b; Picklo et al., 2011; Kawai et al., 2006; Kasai & Kawai, 2008; Demir et al., 2013; Lee et al., 2002; Van Hecke et al., 2015; Zeng et al., 2015). Some , -unsaturated aldehydes, such as acrolein or crotonaldehyde, have been related to an adverse effect on human health. These relevant aldehydes have been found in both frying oils and fried foods (Ewert et al., 2011; Granvogl, 2014; Ewert et al., 2014). They have been employed in this study. Other toxic aldehydes employed in this research have been 4,5-epoxy-2-alkenals and 4-oxo-2-alkenals. 1.2. The broadcasting of lipid oxidative damage. The formation of LOPs is neither the only consequence of lipid oxidation nor the final step of the lipid oxidative process. In fact, this is the initial step of a very complex cascade of reactions. Thus, many of the produced LOPs are not stable, and they react with the surrounding food components, broadcasting the oxidative damage from lipids to all kind of molecules. Among the different produced reactions, carbonyl-amine reactions are particularly important because they have been shown to produce significant changes in food with both positive and negative consequences (Zamora, & Hidalgo, 2016). Introduction 18 1.2.1. Carbonyl-amine reactions produced by LOPs. Analogously to the reactive carbonyl compounds derived from carbohydrates, the lipid-derived reactive carbonyls are also able to react with the amino groups of amines, amino acids, aminophospholipids, and proteins, among other food macromolecules, producing a complex cascade of reactions and numerous compounds. Many of these compounds have been traditionally considered exclusive of Maillard reaction (Zamora & Hidalgo, 2011). These reactions are the final step of the oxidation pathway when lipid oxidation occurs in the presence of amino compounds. Thus, lipid-derived carbonyls react with amino compounds producing nonenzymatic browning and the formation of advanced lipoxidation end-products (ALEs), which can be defined as carbonyl-amine adducts originated from lipids. 1.2.2. Formation of carbonyl-amine adducts. The initial step of carbonyl-amine reactions is the formation of the corresponding imine. The stabilization of this imine can occur without the loss of any portion of either the lipid or the amino compound, and the formation of low molecular weight monomeric ALEs is produced (Hidalgo & Zamora, 2000a; Hidalgo & Zamora, 2017). Particularly, the formation of diverse heterocyclic aromatic structures having five and six atoms has been described. One of the first carbonyl-amine adducts described was a pyridinium salt (A and B in Figure 9) formed between amino compounds and aldehydes (Suyama & Adachi, 1979). The formation is shown in Figure 9. Pyridinium salts are the products formed at low temperature. Nevertheless, upon heating, pyridinium salts suffer a Strecker elimination and the corresponding pyridines (C and D) are formed. Introduction 19 Figure 9. Formation of pyridinium salts A and B by carbonylamine reactions involving alkanals. Pyridinium salts A and B are converted into pyridines C and D, respectively, by heat treatment. Formation of other pyridines has also been described in these reactions. Thus, 2-alkylpyridines (E) are also produced at low temperatures as secondary products in the Strecker degradation of amino acids initiated by different carbonyls including 2,4-alkadienals, 4,5-epoxy-2-alkenals, and nonvolatile long-chain LOPs having epoxy and oxo functions (Hidalgo & Zamora, 2004; Zamora et al., 2007; Zamora et al., 2005). The reaction takes place in several steps as indicated in Figure 10 (Zamora et al., 2015). In addition, 2-alkylpyridines (E) can also be produced at high temperatures when starting from 2,4-alkadienals as indicated in Figure 10 (Zamora et al., 2009). Two 2-alkylpyridines are always produced in reactions collected in Figure 10. 2-Pentylpyridine is the product derived from ω6 fatty acyl chains and 2-ethylpyridine is the product derived from ω3 fatty acyl chains. Introduction 26 The Strecker degradation of amino acids produced by secondary and tertiary LOPs was firstly described in 2004 for the formation of phenylacetaldehyde by phenylalanine degradation in the presence of epoxyalkenals (Hidalgo & Zamora, 2004), and later extended to other lipidderived reactive carbonyls (Hidalgo et al., 2005; Zamora et al., 2005; Zamora et al., 2007; Zamora et al., 2008). However, the reaction is produced to different extents depending on the LOPs involved. This is likely a consequence of both the influence of the lipid-derived carbonyl on the reaction pathway and the existence of parallel reactions in which these oxidized lipids may be involved and that compete with the Strecker reaction (Hidalgo & Zamora, 2016; Zamora et al., 2008). A general mechanism for the Strecker degradation of amino acids produced by lipid-derived reactive carbonyls is shown in Figure 17. The reaction starts with the formation of the corresponding imine (M), followed by a decarboxylation and hydrolysis of the new imine (N) produced. Figure 17. Proposed pathway for the conversion of amino acids into Strecker aldehydes in the presence of lipid-derived reactive carbonyls (Hidalgo & Zamora, 2016). Introduction 27 In addition to aldehydes, some LOPs, which are ketones, can also produce the Strecker degradation of amino acids (Zamora et al., 2005; Zamora et al., 2007). The obtained results showed that the ability of aldehydes and ketones for degrading amino acids was similar in most experiments and differences found might be more related to the different solubility of the LOPs assayed than to a different reactivity. Although hydroperoxides do not have a carbonyl group, they also produce the Strecker degradation of amino acids, because they are easily decomposed and converted into free radicals and carbonyl compounds that react as described above (Hidalgo & Zamora, 2016). Under aerobic conditions, the formation of Strecker aldehydes is favored. However, they can suffer further reactions. Thus, they can also be oxidized to shorter aldehydes having two carbons less than the original amino acid or, under appropriate conditions, they can react with other food components. As an example, the Strecker aldehydes are involved in the production of heterocyclic aromatic amines (HAAs) such as 2-amino-1- methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). This compound is produced by reaction of phenylacetaldehyde, the Strecker aldehyde of phenylalanine, and creati(ni)ne in the presence of the in situ formed formaldehyde and ammonia (Zamora & Hidalgo, 2015). If this degradation of amino acids takes place without decarboxylation, the produced compounds are -keto acids (Zamora et al., 2006a; Zamora & Hidalgo, 2011; Hidalgo & Zamora, 2016). The pathway for this reaction is shown in Figure 18. These compounds are also an important source of flavor compounds in foods. -Keto acids are not thermally stable compounds, and they produce new flavors under mild conditions. Among them, their degradation to Strecker aldehydes and other aldehydes having two fewer carbon atoms than the original amino acid has been described (Hidalgo & Zamora, 2016). Introduction 28 Figure 18. Proposed pathway for the conversion of amino acids into a-keto acids in the presence of lipid-derived reactive carbonyls (Hidalgo & Zamora, 2016). Amino acids can also be converted into amines by action of reactive carbonyls. These amines are usually known as biogenic amines because they have been traditionally considered to be produced as a consequence of the action of microorganisms (Hidalgo et al., 2010; Hidalgo & Zamora, 2016; Zamora & Hidalgo, 2011). The reaction pathway proposed for this reaction is shown in Figure 19. Introduction 29 Figure 19. Proposed pathway for the conversion of amino acids into amines and Strecker aldehydes in the presence of lipidderived reactive carbonyls (Hidalgo & Zamora, 2016). These amines can suffer later reactions in the presence of lipidderived reactive carbonyls. One of them is their conversion into the corresponding Strecker aldehydes, which seems to takes place according to the reaction pathway shown in Figure 20 (Zamora et al., 2012). Figure 20. Proposed pathway for the conversion of amines into Strecker aldehydes in the presence of lipid-derived reactive carbonyls (Hidalgo & Zamora, 2016). Introduction 30 Another reaction suffered by amines is an elimination reaction that converts them into olefins: the corresponding vinylogous derivatives of the amino acids. The reaction takes place as shown in Figure 21. In the presence of the lipid-derived reactive carbonyl, the amine produces first the corresponding imine (P), which may be then converted into an iminium ion that suffer a elimination. This conversion is favored by the lipid because a reactive carbon in the lipid chain may react with the nitrogen of the imine to produce the corresponding iminium ion (Hidalgo & Zamora, 2016). Figure 21. Proposed pathway for the conversion of amines into olefins in the presence of lipid-derived reactive carbonyls (Hidalgo & Zamora, 2016). This reaction has lately received a considerable attention because of its potential implication in the formation of acrylamide as a consequence of asparagine degradation in the presence of reactive carbonyl compounds, including lipid-derived reactive carbonyls (Zamora & Hidalgo, 2011; Zamora & Hidalgo, 2008; Xu et al., 2016; Muttucumaru et al., 2017). Although only deamination reactions of phenylethylamine and 3-aminopropionamide by lipid carbonyls have been studied to date (Zamora et al., 2009), other biogenic amines are likely to suffer similar eliminations. Introduction 31 These vinylogous derivatives of amino acids are not stable compounds and can suffer the addition of nucleophiles. This reaction has recently received a considerable attention because of its possible use in the elimination of acrylamide (Zamora et al., 2010; Zamora et al., 2011; Salazar et al., 2012). Compounds having either amino or sulfhydryl groups are added to these vinylogous derivatives, and there are significant differences between the addition of either amino or sulfhydryl groups. These differences have been described by Hidalgo & Zamora (2016). 1.2.4. Formation of polymers. The imine produced between lipid-derived reactive carbonyls and amino compounds can also evolve into the formation of polymers, which are responsible (at least to a significant extent) for the browning developed in these reactions. Different kinds of colored polymers with a variable nitrogen content are produced in these reactions by different pathways as shown in Figure 22 (Hidalgo & Zamora, 2000a). One of these pathways is the reaction of proteins with lipid-derived reactive carbonyls. This reaction produces the formation of carbonyl-amine adducts in the terminal amino acid residues of the protein, and some of these adducts can also suffer later polymerizations (Hidalgo & Zamora, 2000b). Introduction 32 Figure 22. The formation of carbonyl-amine adducts (advanced lipoxidation end-products, ALEs) in the course of lipid oxidation. The other pathway is the polymerization of some of the low molecular weight ALEs described in previous subsections. Within this last pathway, two main mechanisms have been described for the formation of high molecular weight ALEs. The first one is a repeated aldol condensation which is shown in Figure 23 (Hidalgo & Zamora, 2000a; Hidalgo & Zamora, 2017). The produced polymeric brown materials are unstable and generate volatiles by scission or dehydration, which affect the flavor characteristics of foods during cooking and processing. The second one is based on the polymerization of N- substituted hydroxyalkylpyrroles (Figure 12 and 23) (Hidalgo & Zamora, 1995). These compounds polymerize spontaneously forming dimers, trimers, tetramers, and higher polymers as shown in Figure 23 (Hidalgo & Zamora, 1993). Introduction 33 Figure 23. Proposed mechanisms for the formation of brown pigments by carbonyl–amine reactions. Lipid carbonyls react with amino compounds either forming imines or hydroxyalkylpyrroles depending on the lipid carbonyl involved (Zamora & Hidalgo, 2011). All reactions described in this section show that Figure 2 should be completed by including the broadcasting of the lipid oxidative damage by means of the carbonyl compounds produced in the course of lipid oxidation. This has been collected in Figure 24. As observed in this figure, the initial product formed between the carbonyl compound and the amino compound is the corresponding imine which can suffer then three kinds of reactions: stabilization reactions to produce the corresponding carbonyl-amine adducts of a molecular weight similar to the addition of the molecular weights of the reagents; scission reactions to produce volatile compounds; and polymerization reactions to produce the polymers responsible for the browning formation in these reactions (Figure 24). Introduction 34 Figure 24. Carbonyl-amine reactions as a broadcasting of lipid oxidative damage to food macromolecules containing amino groups (Adapted from Zamora & Hidalgo, 2016). Introduction 35 1.3. The control of lipid oxidative damage by phenolics. 1.3.1. Phenolic compounds. Phenols are a large group of compounds (more than 8000 phenolic compounds have been reported (Crozier et al., 2009)) characterized by having at least one aromatic ring with one or more hydroxyl groups attached. They are plant secondary metabolites commonly found in many herbs, fruits, vegetables, grains and cereals, green and black teas, coffee beans, propolis, and red and white wines (Leopoldini et al., 2011). Phenolics can be classified by the number and arrangement of their phenol subunits into simple phenols and polyphenols. Simple phenols can have only hydroxyl groups or also a carboxylic group such as in phenolic acids. In addition, polyphenols, which possess at least two phenol subunits, include flavonoids and stilbenes among others (Figure 25). Polyphenols possessing three or more phenol subunits are known as tannins (Leopoldini et al., 2011). Introduction 42 such as temperature and pH (Selvaraj et al., 2013; Malesev & Kuntic, 2007). In addition, the formation of complexes of various stoichiometries is possible. 1.3.3. The free radical scavenging ability of phenolic compounds and the control of lipid oxidative damage. Phenolic compounds also delay the lipid oxidation process by scavenging the free radicals that either initiate the lipid oxidation or take part in the propagation of the free radical chain (Zamora & Hidalgo, 2016). This is consequence of their well-known ability of scavenging a wide range of free radical species, including reactive oxygen (ROS), nitrogen (RNS), and chlorine species, such as superoxide, hydroxyl, peroxyl and alkoxyl radicals, and peroxynitrous and hypochlorous acids (Maqsood et al., 2014; Abeywickrama et al., 2016; Rouphael et al., 2016). Phenolic compounds interact with peroxyl (ROO) or alkoxyl (RO) radicals, and also directly with lipid radicals (R), to convert them to more stable, nonradical products by the following reactions where AH is the phenolic compound (Reische et al., 2008; McClements & Decker, 2017): The reaction is produced by donation of a hydrogen atom to the radical at the same time that the phenolic compound is converted into a phenolic free radical (phenoxyl radical). Phenolic antioxidants are thought to interact mainly with peroxyl radicals because they have a longer lifetime (they are less reactive). On the contrary, hydroxyl radicals (OH) are so reactive that they interact with the molecules closest to their site of production. Since Introduction 43 phenolic compounds are generally found at low concentrations they would be less likely to react with these highly reactive free radicals. Antioxidant efficiency is dependent on the ability of the phenolic compound to donate hydrogen to a free radical. As bond energy of a hydrogen in a free radical scavengers (FRS) decreases, the transfer of the hydrogen to a free radical is more energetically favorable and thus more rapid. The efficiency of the FRS is also dependent on the energy of the resulting FRS radical. Effective FRS, such as phenolic compounds, form low energy radicals (phenoxyl radicals in this case) due to resonance delocalization. The phenoxyl radical is stabilized by electron delocalization around a phenolic ring(s) to form stable resonance hybrids (Figure 30). Figure 30. Resonance delocalization of phenoxyl radical. (Adapted from McClements & Decker, 2017). Effective FRS also produce radicals that do not react rapidly with oxygen to form hydroperoxides. Therefore, the phenoxyl radical produced by hydrogen donation has a very low reactivity with oxygen and lipids. Thus, they reduce the rate of propagation of the free radical chain. Besides, phenoxyl radicals may participate in termination reactions with other phenoxyl radicals or lipid radicals to form nonradical species by the following reactions: Introduction 44 Therefore, each FRS is capable of inactivating at least two free radicals. The first radical is inactivated when the FRS interacts with peroxyl, alkoxyl, or lipid radicals. The second radical is inactivated when the FRS takes part in termination reactions with another FRS radical or lipid radical (Figure 31). Figure 31. A termination reaction between a phenoxyl radical and a lipid peroxyl radical (ROO) (Adapted from McClements & Decker, 2017). In foods, the efficiency of phenolic FRS depends on additional factors other than chemical reactivity. Thus, factors such as volatility, pH, sensitivity, and polarity can influence the activity of the FRS in stored and processed foods (Decker, 2008). The free radical-scavenging potential of natural polyphenolic compounds appears to depend on the pattern (both number and location) of free hydroxyl groups on the flavonoid skeleton (Brewer, 2011). This distribution seems to be more important than the flavan backbone alone (Heim et al., 2002). Besides, other types of substituents can influence the free radical scavenging activity. The hydroxyl configuration in the B-ring plays the main role in the scavenging of ROS and RNS (Heim et al., 2002). Hydroxyl groups in the B- ring donate hydrogen and an electron to free radicals, stabilizing them and giving rise to a relatively stable flavonoid radical. A 3’,4’-catechol structure in the B-ring strongly enhances lipid oxidation inhibition because it yields a fairly stable o-semiquinone radical stabilized by electron delocalization. Introduction 45 Furthermore, phenolics containing three adjacent hydroxyl groups are more effective FRS than their dihydroxyl counterparts. Flavonoids lacking catechol or o-trihydroxyl (pyrogallol) systems form relatively unstable radicals and are weak scavengers (Pannala et al., 2001). The significance of other hydroxyl configurations is less clear. It seems that A-ring substitution correlates little with antioxidant activity and compared to the B-ring hydroxylation pattern, the impact of the A-ring arrangement on antioxidant activity is of questionable significance. Regarding O-methylation, the differences in antioxidant activity between polyhydroxylated and polymethoxylated flavonoids are most likely due to differences in both hydrophobicity and molecular planarity. Suppression of antioxidant activity by O-methylation may reflect steric effects that perturb planarity (Heim et al., 2002). In the presence of glycosides, it has been shown that aglycones are more potent antioxidants than their corresponding glycosides. Like methylation, O-glycosylation interferes with the coplanarity of the B-ring with the rest of the flavonoid and its ability to delocalize electrons (Heim et al., 2002). 1.3.4. The carbonyl scavenging ability of phenolic compounds and the control of lipid oxidative damage. Free radical scavenging and chelating properties of phenolics do not explain the ability of these compounds to inhibit some processes in which free radicals are not involved. Thus, the formation of the heterocyclic aromatic amine (HAA) PhIP is a consequence of carbonyl chemistry. This reaction is inhibited in the presence of phenolic compounds (Murkovic et al., 1998; Janoszka, 2010; Persson et al., 2003; Salazar et al., 2014), but this inhibition did not correlate with the antioxidant/free radical scavenging capacity of the phenolic compounds responsible for its inhibition. For example, Cheng et al. Introduction 46 (2007) evaluated the relative activities of 12 food-derived dietary phenolic compounds (including naringenin, quercetin, and epigallocatechin gallate, among others) in inhibiting HAAs formation in model systems and beef patties. The data clearly showed the inhibitory activity of a few phenolic compounds on PhIP formation in model system and all of them in beef patties. In fact, naringenin, a flavonoid found in many citrus fruits, was found to be the most promising inhibitor in both chemical model system and beef patties, suggesting its great potential for practical application in daily cuisine. However, a good positive correlation with their free radical scavenging activities could not be established suggesting that radical scavenging activity may not be the principal mechanism of intervention of these phytochemicals. In fact, recent studies have shown that this effect seems to be a consequence of the ability of phenolic compounds to scavenge lipid-derived reactive carbonyls (Salazar et al., 2014; Cheng et al., 2008). 1.3.4.1. The carbonyl compounds scavenged by phenolics. The ability of phenolics to react with carbonyl compounds has been long known (Lo et al., 2011; Peng et al., 2008; Totlani & Peterson, 2005). In fact, recent studies have shown that they can act as natural inhibitors against the formation of advanced glycation end-products (AGEs) by trapping or scavenging reactive carbonyls (Wu et al., 2011; Peng et al., 2011; Mesías et al., 2014; Liu et al., 2014). The reactive carbonyls mostly studied in this sense have been glyoxal and methylglyoxal. Thus, the in vitro capacity of numerous phenolics to scavenge these aldehydes has been shown, including hydroxytyrosol (Navarro & Morales, 2015), a mixture of quercitrin, chlorogenic acid and rutin (Yoon & Shim, 2015), genistein (Lv et al., 2011); phlorotannins extracted from brown algae Fucus Vesiculosus (Liu & Gu, 2012), and tea polyphenols, particularly epicatechins in green tea and theaflavins in black tea (Lo et al., 2006; Totlani & Peterson, 2006). Introduction 47 In addition, the reaction between phenylacetaldehyde and phenols has also been studied, mostly in relation to the inhibition of PhIP formation (Cheng et al., 2009, 2008; Delgado et al., 2016). Thus, Cheng et al. (2008) demonstrated that naringenin exhibited a dose-dependent reduction of phenylacetaldehyde, a key intermediate in PhIP formation. The capability of naringenin to form adducts with phenylacetaldehyde was confirmed by isolating and characterizing two phenylacetaldehyde-naringenin adducts. Although lesser studied than the highly reactive short dicarbonyl compounds glyoxal and methylglyoxal, some reports have also shown that lipid-derived reactive carbonyls are potential compounds to be trapped by phenolic compounds. Thus, Delgado et al. (2016) studied the Strecker-type degradation of phenylalanine to produce phenylacetaldehyde in the presence of 2-pentenal and phenolic compounds, and they observed an antagonism when the aldehyde and the phenolics were simultaneously present, which suggested an ability of phenolics to react with both 2-pentenal and phenylacetaldehyde. In fact, Hidalgo and Zamora (2014) investigated the reaction between m-diphenols (resorcinol, 2-methylresorcinol, 2,5- dimethylresorcinol, 3-methylphenol, orcinol, and phloroglucinol) and 2- alkenals (2-pentenal and 2-octenal) and isolated and characterized the formed carbonyl-phenol adducts. Zhu et al. (2009a) also evaluated the carbonyl trapping capacity of 21 natural polyphenols against two cytotoxic lipidderived ,-unsaturated aldehydes, acrolein and 4-hydroxy-trans-2-nonenal (HNE). They found that flavan-3-ols, theaflavins, cyanomaclurins, and dihydrochalcones effectively trapped acrolein and/or HNE, and the most effective one was phloretin. In addition, Zhu et al. (2009b) employed green and black tea extracts to stabilize seal blubber oil and they showed that with the addition of green/black tea extracts, the contents of acetaldehyde, acrolein, malondialdehyde, and propanal, four major lipid peroxidation products, were reduced. All these results have pointed out that some lipid-derived carbonyls, including alkanals (acetaldehyde and propanal), dialdehydes Introduction 48 (malondialdehyde), alkenals (acrolein, 2-pentenal, 2-octenal), and HNE can be trapped by phenolics. However, unequivocal structures for the produced adducts were not determined in most cases. Moreover, the ability of phenolic compounds to trap many other carbonyl compounds produced in the lipid oxidation, such as the very reactive and toxic epoxyalkenals or oxoalkenals, remains to be investigated. 1.3.4.2. The structure-activity relationships (SAR) for the carbonylscavenging function of phenolics. Analogously to the observed chelating and free radical scavenging activities, the carbonyl trapping potential of phenolic compounds depends on the pattern (both number and location) of free hydroxyl groups. In a study investigating the SAR of phenolic compounds for the inhibition of PhIP formation, Salazar et al. (2014) analyzed the role of twentyfive phenolic compounds on the PhIP produced in phenylalanine/creatinine/oxidized lipid reaction mixtures. They found that phenols having two hydroxyl groups at meta positions of the aromatic ring were the most efficient inhibitors. Besides, they observed that the presence of alkyl or carboxylic groups as additional substituents in the aromatic ring slightly reduced the inhibitory effect. On the other hand, the introduction of additional hydroxyl and amino groups mostly cancelled the inhibitory effect, which was also mostly absent in o- and p-dihydroxy derivatives. In complex phenols, the presence of several rings with opposite effects produced a reduced inhibitory effect. Nevertheless, although reduced with respect to resorcinol and its simple derivatives, the assayed complex phenols still exhibited a significant PhIP inhibitory effect. This ability of m-diphenols to reduce PhIP formation was a consequence of the formation of carbonyl-phenol adducts as shown by Hidalgo and Zamora (2014) when studying the reaction between 2-alkenals Introduction 49 and m-diphenols. Thus, any substituent or substitution pattern that favors the nucleophilicity of some phenolic carbons, will favor the carbonyl-trapping ability of these compounds. This reaction is mainly favored in m-diphenols and not in o- or pdiphenols because m-diphenols have aromatic carbons with a higher nucleophilicity than o- and p-diphenols due to the electronic delocalization shown in Figure 32. The position of the two hydroxyl groups of m-diphenols in the aromatic ring favors the loss of the aromatic proton at positions 2, 4 or 6. These positions are activated by the two hydroxyl groups. On the other hand, each hydroxyl group in o- and p-diphenols activates different aromatic carbons, which reduces their nucleophilicity. By using the same reasoning, the lower electronic density in the carbonyl compound will facilitate the reaction. This is the reason of the high reactivity of glyoxal and methylglyoxal. Introduction 50 Figure 32. Comparative electronic delocalization produced in m- (top), o- (centre), and p-diphenols (bottom) after proton loss (Hidalgo & Zamora, 2014). 1.3.4.3. The chemical structures of produced carbonyl-phenol adducts. Carbonyl-phenol reactions are produced as a consequence of the high nucleophilicity of some phenolic carbons and the low electronic density in the carbonyl carbon. Therefore, the reaction should start with the addition of the aromatic carbon or hydroxyl group of the phenol to the carbonyl carbon, or the conjugated carbon-carbon double bond in the case of , -unsaturated carbonyls. The formed products should be produced by stabilization of this initial addition product. This has been observed in the reactions studied until now (Sang et al., 2007; Lv et al., 2011; Shao et al., 2008; Hidalgo & Zamora, 2014; Zhu et al., 2009a; Beretta et al., 2008). Introduction 51 Thus, saturated carbonyl compounds suffer the addition of the phenolic compound and the corresponding alcohol is produced (Figure 33). Depending on the reaction conditions and both the carbonyl and the phenolic compound involved, the formed adduct can suffer then a dehydration reaction to produce a more stable conjugated compound. An example of this dehydration has been observed in the reaction of phenylacetaldehyde with different flavonoids (Cheng et al., 2008; Zheng et al., 2016). Figure 33. Alkanal addition to resorcinol. In the case of the highly reactive short dicarbonyl compounds glyoxal and methylglyoxal, the reaction occurs similarly (Wang & Ho, 2012; Lv et al., 2011; Sang et al., 2007; Lo et al., 2006; Yoon & Shim, 2015; Shao et al., 2008; Totlani & Peterson, 2006; Navarro & Morales, 2015). Although much more complex, the reaction of 2-alkenals with phenolic compounds also occurs similarly, as shown in Figure 34. Thus, the reaction takes place by initial addition of either the aromatic carbon or hydroxyl group of the phenolic to the carbon-carbon double bond of the carbonyl compound and later stabilization of these initial addition products. The adduct produced by the initial addition of the aromatic carbon to the aldehyde is relatively stable after formation of a hemiacetal structure. However, the adduct produced by the initial addition of the hydroxyl group to the aldehyde is more unstable and suffers a dehydration. These compounds have been isolated and characterized by nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry (MS) for a significant number of Aims of the Study 58 Aims of the Study 59 As described in the Introduction section, phenolic compounds might, in addition to chelating metals and scavenging free radicals, also be limiting the produced lipid oxidative damage by trapping the lipid-derived reactive carbonyls formed in the course of lipid oxidation. Furthermore, this trapping would constitute a third protective barrier of phenolic compounds against the broadcasting of the consequences of lipid oxidation to other food macromolecules. However, this barrier is still poorly understood because neither the carbonyl-trapping ability of all LOPs has been studied nor the possibility that this carbonyl-trapping might be produced under usual cooking conditions has been investigated. Therefore, the general objective of this thesis is to determine whether the carbonyl-trapping ability of phenolics is general for main LOPs, and to study the presence of carbonyl-phenol adducts in processed food products. To carry out this general objective, different specific objectives have been proposed: 1. To study the reaction among phenolic compounds and the main groups of carbonyl compounds produced as a consequence of lipid oxidation, including alkanals, alkadienals, 4,5-epoxy-2-alkenals, and 4-oxo-2-alkenals. 2. To investigate the reaction conditions that favor the formation of the corresponding carbonyl-phenol adducts. 3. To understand the structure-activity relationships (SAR) of both phenolics and carbonyl compounds that favor the formation of carbonyl-phenol adducts. 4. To determine the reaction pathways by which carbonyl-phenol reactions are produced. 5. To study the lipid-derived carbonyl-trapping ability of phenolic compounds under common food processing conditions in order to Aims of the Study 60 know whether carbonyl-phenol adducts might be usual food components in processed foods. Materials and Methods 61 3. MATERIALS AND METHODS Materials and Methods 62 Materials and Methods 63 3.1. Materials. 3.1.1. Commercial products. A series of aldehydes having different number of carbons and unsaturations were employed in these studies. These aldehydes included alkanals such as propanal, butanal, 2-methylpropanal, pentanal, 2- methylbutanal, 3-methylbutanal, hexanal, and glyoxal; 2-alkenals such as acrolein, crotonaldehyde, 2-pentenal, 2-octenal, and 2-methyl-2-pentenal; and alkadienals, such as 2,4-heptadienal and 2,4-decadienal. Fumaraldehyde was also used in some experiments. It was prepared from commercial fumaraldehyde bis(dimethyl acetal) by treating an aqueous solution of the acetal with acid resin (Dowex 50WX8) for 1 h. The chemical structures of all these compounds are given in Figure 37. As model phenolic compounds, resorcinol, 2-methylresorcinol, 2,5- dimethylresorcinol, orcinol, and quercetin were employed. The chemical structures of these phenolic compounds are given in Figure 26 and Figure 28 in the Introduction section 1.3.1. All these compounds were purchased from Sigma-Aldrich (St. Louis, MO), Fluka (Buchs, Switzerland), Alfa Aesar (HaverHill, MA), or TCI Chemicals (Zwijndrecht, Belgium), and were of either analytical grade or the highest available quality. Citric acid (99.5%), boric acid, sodium hydroxide (99%) and the sodium phosphate used to prepare the different buffers were obtained from Merck (Darmstadt, Germany). Rapeseed oil, red onions (Allium cepa L.), and commercially crispy fried onions were purchased at local supermarkets. Sephadex LH-20 was obtained from GE Healthcare Europe (Freiburg, Germany). Silica gel, alumina, and thin-layer chromatography plates (thickness of 0.2 mm) were purchased from Macherey Nagel (Gmbh, Düren, Germany). Materials and Methods 64 All solvents used in these studies (methanol, hexane, diethyl ether, acetonitrile, chloroform, acetone) were analytical grade and were purchased from Merck. Figure 37. Chemical structures of some of the aldehydes employed in this study. 3.1.2. Syntheses of epoxyalkenals. Three 4,5-epoxy-2-alkenals were employed in this study. They were 4,5-epoxy-2-hexenal, 4,5-epoxy-2-heptenal, and 4,5-epoxy-2-decenal. Their syntheses were carried out from the corresponding 2,4-alkadienals (2,4- Materials and Methods 65 hexadienal, 2,4-heptadienal, and 2,4-decadienal, respectively) by epoxidation with 3-chloroperoxybenzoic acid (Zamora et al., 2006b) (Figure 38). Figure 38. Synthesis of epoxyalkenals used in this study. In short, 3-chloroperoxybenzoic acid (69.5 mmol) was dissolved in chloroform (175 ml), washed three times with 100 ml of buffer (0.2 M Na2HPO4·12H2O adjusted to pH 7.5 with 0.1 M citric acid) followed by three 100 ml portions of water, and dried with anhydrous sodium sulfate. This solution was added slowly to a solution of 2,4-alkadienal (39.9 mmol) in chloroform (30 ml), which was stirred at room temperature overnight and, finally, washed three times with 100 ml of buffer (0.2 M Na2HPO4·12H2O adjusted to pH 7.5 with 0.1 M citric acid monohydrate) followed by three extractions with 100 ml of water. The organic solution was dried over anhydrous sodium sulfate and concentrated under vacuum. The residue was fractionated by column chromatography on silica gel 60 (230-400 mesh; Macherey-Nagel) using mixtures of hexane and acetone as eluent. The separation was controlled by thin-layer chromatography. 4,5-Epoxy-2- alkenals were obtained chromatographically pure. Confirmations of identity and purity were obtained by gas chromatography-mass spectrometry (GCMS) by comparing retention times and mass spectra with authentic standards prepared previously in the laboratory. Materials and Methods 66 3.1.3. Syntheses of oxoalkenals. Two 4-oxo-2-alkenals were employed in this study. They were 4-oxo- 2-hexenal and 4-oxo-2-nonenal, which were prepared by ring opening of the corresponding 2-alkylfuran: 2-ethylfuran and 2-pentylfuran, respectively (Shimozu et al., 2009) (Figure 39). Briefly, N-bromosuccinimide (10.7 mmol) and pyridine (10 ml) were added to a solution of 2-alkylfuran (16.6 mmol) in 11 ml of tetrahydrofuran/acetone/water (5:4:2) in an ice bath. The reaction mixture was stirred for 1 h at this temperature and, then, kept on stirring at room temperature for 2 h. After this time, the mixture was diluted with 10 ml of water and extracted three times with 50 ml of chloroform. The combined chloroformic extracts were washed succesively with hydrochloric acid and water in order to remove pyridine, then dried over anhydrous sodium sulfate, and finally concentrated under vacuum. The residue was fractionated by column chromatography on silica gel 60 (230-400 mesh; Macherey-Nagel) using mixtures of hexane and ethyl acetate as eluent. The separation was controlled by thin-layer chromatography. 4-Oxo-2-alkenals were obtained chromatographically pure. Confirmations of identity and purity were obtained by GC-MS as described above. Figure 39. Syntheses of 4-oxo-2-nonenal and 4-oxo-2-hexenal. Materials and Methods 67 3.2. Analytical instrumentation. 3.2.1. Gas chromatography-mass spectrometry (GC-MS). GC-MS analyses were conducted with two instruments. Some GCMS analyses were carried out with a Hewlett-Packard 6890 GC Plus coupled with an Agilent 5973 MSD (mass selective detector, quadrupole type). A fused-silica HP5-MS capillary column (30 m × 0.25 μm i.d.; coating thickness, 0.25 μm) was used, and 1 μl of sample was injected in the pulsed splitless mode. Other GC-MS analyses were conducted with an Agilent 7820 gas chromatograph coupled with an Agilent 5977B MSD (mass selective detector, quadrupole type). In this case, the employed column was a fusedsilica DB-5MS UI capillary column (30 m 0.25 i.d.; coating thickness, 0.25 µm), and 1 µl of sample was injected in the pulsed splitless mode. 3.2.2. Liquid chromatographytandem mass spectrometry (LC-MS/MS). Samples were analyzed by using an Agilent liquid chromatography system (1200 Series) consisting of binary pump (G1312A), degasser (G1379B), and autosampler (G1329A), connected to a triple quadrupole API 2000 mass spectrometer (Applied Biosystems, Foster City, CA) using an electrospray ionisation interface in positive ionization mode (ESI+), or in negative ionization mode (ESI−) in some cases. Compounds were separated on a Zorbax Eclipse XDB-C18 (150 mm 4.6 mm, 5 μm) column from Agilent (Santa Clara, CA). 3.2.3. Liquid chromatography-high resolution mass spectrometry (LCHRMS). The LC-ESI-MS system consisted of a Dionex Ultimate 3000RS U- HPLC (Thermo Fisher Scientific, Waltham, MA) coupled to a micrOTOF-QII Materials and Methods 74 (crotonaldehyde), or 96 h ((E)-2-pentenal), respectively. At the end of the heating time, the reaction mixture was fractionated on a Sephadex LH-20 column using methanol/water (80/20, v/v) as eluent at a flow rate of 15 ml/h. Eluted products were detected by MS/MS using direct injection and the analytical conditions described in section 3.4.2.4. The compounds formed in the different assayed reactions (see section 4.5) were isolated and characterized by 1D and 2D NMR and LC-HRMS, using the analytical conditions described in sections 3.2.4 and 3.4.2.4, respectively. 3.3.5. Thermal degradation of unsaturated aldehydes. Two different procedures were followed depending on whether the formed compounds were going to be either identified or quantified. The identification of thermal degradation products of the studied aldehydes was carried out by GC-MS after derivatization with O-(2,3,4,5,6- pentafluorobenzyl)hydroxylamine hydrochloride, according to a previously described procedure (Zamora et al., 2006a), which was modified. Briefly, the aldehyde (4 mol) was either heated alone or in the presence of 200 l of 50 mM buffer pH 8, for 1 h at 200 ºC in a closed test tube under either nitrogen or air. Two buffers were employed: sodium phosphate and sodium borate. At the end of the heating process, samples were cooled (5 min at room temperature and 10 min at 20 ºC) and derivatized with 400 l of a freshly prepared solution of O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine hydrochloride (10 mg/ml in methanol). The resulting solution was stirred and incubated for 1 h at 37 ºC. Finally, reactions were studied by GC-MS, using the analytical conditions described in section 3.4.1.1. Quantification of the compounds produced was carried out by LCMS/MS after derivatization with dansylhydrazine according to a previously described procedure (Zamora et al., 2014), which was also modified. In short, a solution of the aldehyde (0–80 mol) in tetrahydrofuran (80 l) was treated with 0.2 M buffer (420 l) and, then, heated for the indicated time and Materials and Methods 75 temperature in a closed test tube under air. The solvent employed in most of experiments was 0.2 M sodium phosphate buffer, pH 8. In addition, three buffers were employed when studying the effect of reaction pH: sodium citrate, pH 2.15-5; sodium phosphate, pH 6–8; and sodium borate, pH 9–12. At the end of the heating process, samples were cooled (5 min at room temperature and 10 min at 20 ºC). Fifty microliters of these cooled samples were diluted with 350 l of methanol, and treated with 50 l of the internal standard (a solution of 88 mol of formaldehyde-d2 in 2 ml of methanol), 150 l of trifluoromethanesulfonic acid solution (3% in methanol), and 200 l of dansylhydrazine solution (4 mg/ml in methanol). The resulting solution was incubated for 15 min at 100 ºC, then maintained for 1 h at 25 ºC, and finally diluted with 200 l of eluent A (a 30:70 mixture of 0.2% formic acid in acetonitrile and 4 mM ammonium acetate), and analyzed by LC-MS/MS using the analytical conditions described in section 3.4.1.1. 3.3.6. Deep-frying experiments. Deep-frying experiments were carried out in rapeseed oil (100 g) containing acrolein, crotonaldehyde, and (E)-2-pentenal (2.7 μmol/g of oil of each aldehyde), which were added before heating the oil. The oil was first preheated for 9 min to achieve 160 °C and, then, heated in the presence, or absence, of thin slices of onions (5 g) for further 3 min. At the end of the heating time, the heated oils were analyzed for aldehyde contents by 1H NMR using the procedure described in section 3.4.1.2. The fried onions were also analyzed to identify aldehyde-quercetin adducts. Aldehyde-quercetin adducts were firstly extracted and concentrated according to the following procedure: samples of onions (5 g each) were homogenized in water (20 ml) and hexane (50 ml). The resulting mixture was centrifuged at 7500g for 10 min at room temperature. The organic layer was discarded and other 50 ml of hexane were added. The mixture was homogenized again and centrifuged at 7500g for 10 min. The organic layer Materials and Methods 76 was discarded and methanol (80 ml) was added. The mixture was newly homogenized and centrifuged at 7500g for 10 min. The supernatant was collected and the solid was extracted with 100 ml of methanol/water (80/20, v/v). The resulting mixture was centrifuged at 7500g for 10 min and the supernatant was collected. Both supernatants were combined, concentrated to about 4 ml using a rotatory evaporator (30 °C, 16 mbar) and fractionated on a Sephadex LH-20 column using methanol/water (80/20, v/v) as eluent at a flow rate of 15 mL/h. Fractions obtained were analyzed by LC-HRMS, using the analytical conditions described in section 3.4.2.5. The different carbonylphenol adducts were identified by comparison with adducts synthesized as described in section 3.3.4, which were used as reference compounds. The same procedure was used to analyze commercially prepared crispy fried onions. 3.4. Analytical determinations. 3.4.1. Determination of aldehydes. 3.4.1.1. Aldehydes produced by thermal degradation of alkadienals amd 2- alkenals. Thermal degradation products of the studied aldehydes (2-pentenal, 2- octenal, 2,4-heptadienal and 2,4-decadienal) were determined using the samples prepared in section 3.3.5. Samples for identification purposes were analyzed by GC-MS. In this case, working conditions were as follows: carrier gas, helium (1 ml/min at constant flow); injector temperature, 250 ºC; oven temperature, from 70 (1 min) to 240 ºC at 5 ºC/min and, then, to 325 ºC at 10 ºC/min; transfer line to MSD, 280 ºC; ionisation EI, 70 eV. Reaction products were identified by Materials and Methods 77 comparison of mass spectra and retention times to those of authentic standards. On the other hand, samples for quantification purposes were analyzed by LC-MS/MS. Working conditions were as follow: as eluent A, a 30:70 mixture of 0.2% formic acid in acetonitrile and 4 mM ammonium acetate was used. As eluent B, a 0.2% formic acid solution in acetonitrile was employed. The mobile phase was delivered at 0.5 ml/min using the following gradient: for 0–13 min, the content of mobile phase B was 7%; for 13–20 min, the content of mobile phase B was increased linearly from 7% to 60%; for 20–30 min, the content of mobile phase B was 60%; for 30–32 min, the content of mobile phase B was increased linearly from 60% to 90%; for 32–42 min, the content of mobile phase B was 90%; and for 42–45 min, the content of mobile phase B was decreased linearly from 90% to 7%. Mass spectrometric acquisition was performed using multiple reaction monitoring (MRM). The nebulizer gas (synthetic air), the curtain gas (nitrogen), and the heater gas (synthetic air) were set at 40, 25, and 50 (arbitrary units), respectively. The collision gas (nitrogen) was set at 3 (arbitrary units). The heater gas temperature was set at 500 ºC and the electrospray capillary voltage to 5.5 kV. The fragment ions in MRM mode were produced by collision-activated dissociation of selected precursor ions in the collision cell of the triple quadrupole and the selected products analyzed with the second analyzer of the instrument. Three transitions were acquired for the identification of each dansylhydrazone derivative. To establish the appropriate MRM conditions for the individual compounds, the mass spectrometric conditions were optimized using infusion with a syringe pump to select the most suitable ion transitions for the target analytes. Precursor and product ions used for quantification and confirmation purposes, and operating conditions are summarized in Table 3. Materials and Methods 78 Table 3. Optimization of MRM transitions for detection of aldehydes. Aldehyde Monitored transition DP FP EP CEP CE CXP Formaldehyde-d2 280.0 156.1 26 360 8 18 47 6 280.0 115.2 26 360 8 18 67 4 280.0 171.1 26 360 8 18 31 6 Formaldehyde 278.1 170.1 26 370 10 18 35 6 278.1 128.1 26 370 10 18 71 4 278.1 115.2 26 370 10 18 77 4 Propanal 306.2 156.0 26 370 10.5 14 53 6 306.2 115.1 26 370 10.5 14 71 4 306.2 171.1 26 370 10.5 14 31 6 2-Pentenal 332.2 156.1 26 370 10 14 55 6 332.2 171.1 26 370 10 14 37 6 332.2 115.1 26 370 10 14 79 4 2-Methyl-2-pentenal 346.2 156.1 21 370 6.5 26 57 6 346.2 171.1 21 370 6.5 26 39 8 346.2 115.1 21 370 6.5 26 79 6 2,4-Heptadienal 358.1 170.1 21 370 8 16 29 6 358.1 171.1 21 370 8 16 35 6 358.1 115.2 21 370 8 16 79 4 Hexanal 348.1 156.1 26 370 10.5 16 61 6 348.1 115.1 26 370 10.5 16 83 4 348.1 171.2 26 370 10.5 16 41 6 2-Octenal 374.1 156.1 26 350 11.5 14 61 6 374.1 171.1 26 350 11.5 14 39 6 374.1 115.1 26 350 11.5 14 83 4 2,4-Decadienal 400.1 170.0 21 370 10.5 16 33 6 400.1 171.1 21 370 10.5 16 37 6 400.1 95.1 21 370 10.5 16 37 6 Abbreviations: DP, declustering potential; FP, focusing potential; EP, entrance potential; CEP, collision cell entrance potential; EC, collision energy; CXP, collision cell exit potential. Materials and Methods 79 Quantification of the different aldehydes was carried out by preparing five standard curves of aldehyde mixtures in 500 l of the mixture tetrahydrofuran/sodium phosphate buffer, pH 8, and following the whole procedure described above. For each curve, seven different concentration levels of each aldehyde (0–2 nmol) were used. Aldehyde contents were directly proportional to aldehyde/IS area ratios (r > 0.99, p < 0.0001). All data given are the mean of, at least, three independent experiments. 3.4.1.2. Aldehydes in heated oils. Aldehydes content in oil samples subjected to deep-frying experiments described in section 3.3.6 was determined by 1H NMR analogously to the analysis of oil components developed by Sopelana et al. (2013). Briefly, oil samples (200 mg) were diluted with CDCl3 (400 μl), which contained 0.2% of nondeuterated chloroform, and their spectra were obtained by 1H NMR, using the analytical conditions described in section 3.2.4. Quantification of the three aldehydes (acrolein, crotonaldehyde, and (E)-2-pentenal) was carried out by considering the area of the CHCl3 signal (at δ 7.29 ppm) as internal standard. The proton of the aldehyde group appeared as a doublet that was independent for each one of the three analyzed aldehydes and appeared far from other signals in the spectrum. The signals of the aldehydic proton appeared at δ 9.55 (d, J = 7.4 Hz) for acrolein, 9.47 (d, J = 7.9 Hz) for crotonaldehyde, and 9.49 (d, J = 7.7 Hz) for (E)-2-pentenal. Four samples were analyzed for each heated oil and each sample was measured three times. Therefore, the aldehyde content for each analyzed oil was the mean value of 12 determinations. Materials and Methods 80 3.4.2. Determination of carbonyl-phenol adducts. 3.4.2.1. Adducts formed with alkanals. Alkanal-phenol adducts were determined in the samples prepared according to the procedure described in section 3.3.1 by GC-MS. Working conditions were as follows: carrier gas, helium (1 mL/min at constant flow); injector temperature, 250 °C; transfer line to MSD, 280 °C; ionization EI, 70 eV; ion source temperature, 230 °C. Two oven temperatures were programmed. For acetylated derivatives, the oven temperature was programmed from 100 °C (1 min) to 300 °C at 15 °C/min and, then, 5 min at 300 °C. For nonacetylated derivatives, the oven temperature was programmed from 40 °C (1 min) to 240 °C at 12 °C/min, then to 300 °C at 20 °C/min, and, finally, 5 min at 300 °C. The flow velocities employed were 30 cm/s for acetylated compounds and 37 cm/s for nonacetylated compounds. Quantification of alkanal-phenol adducts was carried out by preparing standard curves of the isolated adducts in the 515 µl of solution employed in samples prepared in section 3.3.1. Seven different concentration levels were used. Adduct content was directly proportional to the aldehyde/internal standard area ratio (r > 0.98, p < 0.001). The coefficients of variation were less than 10 %. When one adduct was not available, such as the adducts derived from either orcinol or butanal, the calibration curve used was that of the adduct with the closest chemical structure. 3.4.2.2. Adducts formed with epoxyalkenals. Epoxyalkenal-phenol adducts were determined in the samples prepared according to the procedure described in section 3.3.2 by GC-MS. The following working conditions were employed: carried gas, helium (0.8 ml/min at constant pressure); injector, 250 C; oven temperature programmed from 100 C (1 min) to 300 C at 15 C/min and then 5 min at 300 C; Materials and Methods 81 transfer line to MSD, 280 C; ionization EI, 70 eV; ion source temperature, 230 C; mass range 50–550 amu. The quantification of aldehyde-phenol adducts was carried out by preparing standard curves of the isolated adducts (20a, 20b, 25a, 25b) in the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol. For each curve, six different concentration levels were used. Adduct content was directly proportional to the adduct/internal standard area ratio (r > 0.99, p < 0.001). The coefficients of variation were <10%. Because the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol produced four isomers of 1-ethyl-8-methyl-3,4,4a,9a-tetrahydro-1H-pyrano[3,4-b]benzofuran-3,7- diyl diacetates (25a, 25b, 25c and 25d) and only two of them could be isolated and employed for preparing the corresponding standard curves (25a and 25b), the quantification of the other two isomers (25c, 25d) was carried out by employing a calibration curve that was the media of the other two. This approximation should be considered satisfactory because the calibration curves of the adducts 25a and 25b were very similar between them. 3.4.2.3. Adducts formed with oxoalkenals. Oxoalkenal-phenol adducts were determined in the samples prepared according to the procedure described in section 3.3.3 by GC-MS. Working conditions employed were similar to those described in section 3.4.2.2. Adducts were quantified by preparing standard curves of the isolated adducts (29-31) in the reaction between 4-oxo-2-hexenal and 2- methylresorcinol. Six different concentration levels were used. Adduct content was directly proportional to the adduct/internal standard area ratio (r > 0.99, p < 0.001). The coefficients of variation were <10%. Materials and Methods 82 3.4.2.4. Adducts formed with alkenals in model systems. Two different methods were used in the determination of aldehydequercetin adducts in model systems, prepared according to the procedure described in section 3.3.4. On the one hand, these adducts were analyzed by MS/MS using direct injection. It was used in order to detect the eluted products in the gel filtration separation carried out using a Sephadex LH-20 column, which was described in section 3.3.4. Working conditions for MS/MS analyses were as follows: the nebulizer gas and the curtain gas were set at 19 and 10 (arbitrary units), respectively. The electrospray capillary voltage was set to −4.5 kV, the declustering potential was −50 V, the focusing potential was −400 V and the entrance potential was −10 V. In addition, aldehyde-quercetin adducts were analyzed by LC-HRMS, in order to characterize the adducts synthesized as reference compounds. Working conditions were as follows: as eluent A, a mixture of acetonitrile containing 0.2% formic acid and 4 mM ammonium formiate (30/70, v/v) was used. As eluent B, acetonitrile containing 0.2% formic acid was employed. The flow rate was 0.5 mL/min in linear gradient mode: 0−13 min 7% B, 13−20 min from 7 to 60% B, 20−30 min 60% B, 30− 32 min from 60 to 90% B, 32−42 min 90% B, 42−45 min from 90 to 7% B. A split postcolumn with a flow rate of 0.25 mL/min was inserted directly into the mass spectrometer ESI source. The scan range applied was m/z 50−1500 and mass resolving power was always over 18,000 (m/Δm). The instrument was operated in the negative ion mode. Mass spectra and data were obtained by broadband Collision Induced Dissociation (bbCID) mode, providing MS and MS/MS spectra simultaneously. Collision energy was estimated dynamically based on appropriate values for the mass and stepped across a ± 10% magnitude range to ensure good quality fragmentation spectra. The instrument control was performed using Compass 1.3 for micrOTOF-QII + Focus Option Version 3.0 (Bruker Daltonics). Materials and Methods 83 3.4.2.5. Adducts formed with alkenals in fried onions. Aldehyde-quercetin adducts were analyzed by LC-HRMS, in order to detect their presence in onions fried in the laboratory and commercially prepared crispy fried onions. Fried onion samples were prepared for their analysis according to the procedure described in section 3.3.6. Working conditions for LC-HRMS analyses were similar to those described in section 3.4.2.4. 3.5. Statistical analyses. All quantitative data given are mean ± standard deviation (SD) values of, at least, three independent experiments. Statistical comparisons among different groups were made using analysis of variance. When significant F values were obtained, group differences were evaluated by the Tukey test (Snedecor & Cochran, 1980). Statistical differences among the amounts of aldehydes remaining in the oils after heating in the presence or in the absence of onions were evaluated by the Student t-test. Statistical comparisons were carried out using Origin® v. 7.0 (OriginLab Corporation, Northampton, MA). The significance level is p < 0.05 unless otherwise indicated. Results 90 Figure 41. Chemical structures of the compounds isolated and characterized by NMR and MS in the reaction between pentanal and 2-methylresorcinol. The chromatographic, spectroscopic, and spectrometric data of the isolated and characterized compounds are the following: 4-(1-Methoxypentyl)-2-methyl-1,3-phenylene diacetate (2). Yield: 13.8%. Retention time: 11.21 min. 1H NMR (CDCl3): (ppm) 0.90t (3H, J = 7.1 Hz, H5’), 1.29m and 1.41m (2H, H3’), 1.32m (2H, H4’), 1.62m and 1.75m (2H, H2’), 1.98s (3H, CH3C2), 2.33s (3H, CH3CO), 2.33s (3H, CH3CO), 3.17s (3H, CH3O), 4.15dd (1H, J = 5.0 Hz, J = 8.0 Hz, H1’), 7.00d (1H, J = 8.1 Hz, H6), and 7.26d (1H, J = 8.1 Hz, H5). 13C NMR (CDCl3): (ppm) 10.11 (CH3C2), 13.97 (C5’), 20.74 (CH3CO), 20.79 (CH3CO), 22.52 (C4’), 28.07 (C3’), 36.63 (C2’), 56.97 (CH3O), 79.05 (C1’), 119.93 (C6), 123.66 (C2), 124.97 (C5), 132.36 (C4), 148.03 (C3), 148.89 (C1), 168.89 (CO), and 168.92 (CO). MS, m/z (%, ion structure): 308 (0.02, M+), 277 (0.02, Results 91 M+ – CH3O), 265 (3, M+ – CH3CO), 251 (50, M+ – CH3CH2CH2), 234 (19, C13H14O4), 209 (76, 251 – CH2CO), 192 (70, 234 – CH2CO), and 167 (100, 209 – CH2CO). 2-Methyl-4-(pent-1-en-1-yl)-1,3-phenylene diacetate (3). Yield: 5.2%. Retention time: 11.46 min. 1H NMR (CDCl3): (ppm) 0.96t (3H, J = 7.3 Hz, H5’), 1.50sx (2H, J = 7.3 Hz, H4’), 1.99s (3H, CH3C2), 2.20qd (2H, J = 7.3 Hz, J = 1.2 Hz, H3’), 2.33s (3H, CH3CO), 2.37s (3H, CH3CO), 6.18dt (1H, J = 15.8 Hz, J = 6.8 Hz, H2’), 6.31d (1H, J = 15.8 Hz, H1’), 6.94d (1H, J = 8.5 Hz, H6), and 7.37d (1H, J = 8.5 Hz, H5). 13C NMR (CDCl3): (ppm) 10.11 (CH3C2), 13.64 (C5’), 20.49 (CH3CO), 20.80 (CH3CO), 22.38 (C4’), 35.37 (C3’), 119.83 (C6), 123.10 (C1’), 123.55 (C2), 123.94 (C5), 128.83 (C4), 133.85 (C2’), 146.94 (C3), 148.50 (C1), 168.58 (CO), and 169.03 (CO). MS, m/z (%, ion structure): 276 (17, M+), 234 (38, M+ – CH2CO), 192 (100, 234 – CH2CO), and 163 (97, 192 – CH3CH2). 4-(1-Acetoxypentyl)-2-methyl-1,3-phenylene diacetate (4). Yield: 3.2%. Retention time: 12.22 min. 1H NMR (CDCl3): (ppm) 0.90t (3H, J = 7.1 Hz, H5’), 1.24m and 1.34m (2H, H3’), 1.34m (2H, H4’), 1.79m and 1.92m (2H, H2’), 1.98s (3H, CH3C2), 2.04s (3H, CH3COOC1’), 2.34s (3H, CH3CO), 2.38s (3H, CH3CO), 5.87br (1H, H1’), 6.99d (1H, J = 8.4 Hz, H6), and 7.29d (1H, J = 8.4 Hz, H5). 13C NMR (CDCl3): (ppm) 10.24 (CH3C2), 13.90 (C5’), 20.59 (CH3COOC1), 20.82 (CH3CO), 21.06 (CH3CO), 22.34 (C4’), 27.78 (C3’), 34.44 (C2’), 70.95 (C1’), 119.92 (C6), 124.17 (C2), 125.43 (C5), 130.46 (C4), 147.77 (C3), 149.41 (C1), 168.60 (CO), and 168.86 (CO). MS, m/z (%, ion structure): 336 (0.2, M+), 277 (1, M+ – CH3COO), 276 (0.4, M+ – CH3COOH), 234 (16, 276 – CH2CO), 192 (100, 234 – CH2CO), and 163 (62, 192 – CH3CH2). Results 92 4-Butyl-8-methyl-3-propylchroman-2,7-diyl diacetate (5). Yield: 2.9%. Retention time: 13.65 min. 1H NMR (CDCl3): (ppm) 0.92t (3H, J = 7.2 Hz, CH3CH2), 0.95t (3H, J = 7.0 Hz, CH3CH2), 1.4m (8H, CH3CH2, CH3CH2, CH3CH2CH2, and CH3CH2CH2), 2.02s (3H, CH3C8), 2.06s (3H, CH3CO), 2.33s (3H, CH3CO), 2.55dd (1H, J = 5.5 Hz and J = 9.9 Hz, H3), 2.80dt (2H, J = 5.3 Hz and J = 7.7 Hz, CH2C4), 2.95dt (1H, J = 4.8 Hz and J = 7.5 Hz, H4), 6.37dd (1H, J = 1.3 Hz and J = 1.9 Hz, H2), 6.67d (1H, J = 8.3 Hz, H6), and 6.98d (1H, J = 8.3 Hz, H7). 13C NMR (CDCl3): (ppm) 9.23 (CH3C8), 14.01 (CH3CH2), 14.23 (CH3CH2), 20.13 (CH3CH2), 20.83 (CH3CO), 21.19 (CH3CO), 23.01 (CH3CH2), 27.48 (CH3CH2CH2), 29.97 (CH3CH2CH2), 33.83 (C4), 36.72 (CH2C4), 37.45 (C3), 93.90 (C2), 114.86 (C6), 118.60 (C8), 122.55 (C4a), 127.31 (C5), 148.54 (C7), 150.45 (C8a), 169.35 (CO), and 169.64 (CO). MS, m/z (%, ion structure): 362 (7, M+), 320 (8, M+ – CH2CO), 303 (3, M+ – CH2COO), 302 (1, M+ – CH2COOH), 260 (9, 302 – propene), 259 (22, 302 – CH3CH2CH2), 245 (19, 302 – CH3CH2CH2CH2), 217 (32, 260 – CH3CH2CH2), and 203 (100, 260 – CH3CH2CH2CH2). 4.1.1.2. Compounds isolated and characterized in the different alkanal/phenolic compound reaction mixtures assayed. In order to prepare the adducts needed for the SAR studies, a number of reactions involving different saturated aldehydes and phenolic compounds were carried out and the corresponding carbonyl-phenol adducts were analyzed and characterized. Thus, the adducts produced in the reactions between propanal, 2-methylbutanal, pentanal and hexanal with resorcinol were isolated and characterized. In the same way, the adducts produced in the reactions between propanal, 2-methylbutanal, 3-metilbutanal, pentanal and hexanal with 2-methylresorcinol were isolated and characterized as well. Results 93 Furthermore, analogous reactions between glyoxal and resorcinol or 2- methylresorcinol, and the reaction between resorcinol and 2-methyl-2- pentenal were also studied for comparison purposes. Reactions were carried out analogously to the reaction between pentanal and 2-methylresorcinol discussed in subsection 4.1.1.1, but formed compounds were isolated without acetylation. In all studied reactions, only major reaction products were isolated and characterized. Figure 42 shows the adducts isolated in the different assayed reactions. The chromatographic, spectroscopic, and spectrometric data of the isolated and characterized compounds are the following: The reaction between resorcinol and propanal produced 4-(1- methoxypropyl)benzene-1,3-diol (6). The reaction was heated for 48 h at 60 ºC and compound 6 was isolated using hexane-diethyl ether (7:3) as eluent. Retention time: 18.18 min. 1H NMR (CD3OD): (ppm) 0.88t (3H, J = 7.5 Hz, H3’), 1.67m and 1.77m (2H, H2’), 3.22s (3H, CH3O), 4.42t (1H, J = 6.7 Hz, H1’), 6.29d (1H, J = 2.3 Hz, H2), 6.31dd (1H, J = 2.3 Hz, J = 8.2 Hz, H6), and 6.96d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD): (ppm) 9.15 (C3’), 28.98 (C2’), 55.24 (CH3O), 80.31 (C1’), 102.07 (C2), 106.39 (C6), 118.25 (C4), 127.61 (C5), 156.17 (C3), and 157.21 (C1). MS, m/z (%, ion structure): 150 (100, M+ – CH3OH), 133 (19, 150 – OH), and 123 (36, C7H7O2). Results 94 Figure 42. Chemical structures of the carbonyl-phenol adducts isolated and characterized by NMR and MS in the different assayed reactions between alkanals and phenolics. The aldehyde and the phenolic compound involved in the formation of each adduct is described in the text. The reaction between resorcinol and pentanal produced 4-(1- methoxypentyl)benzene-1,3-diol (7) and 2-(1-hydroxypentyl)benzene-1,3-diol Results 95 (8). The reaction was heated for 72 h at 60 ºC and the compounds were isolated using hexane-diethyl ether (3:1) as eluent. Retention time of compound 7: 20.04 min. Retention time of compound 8: 18.87 min. 1H NMR (CD3OD) of compound 7: (ppm) 0.90t (3H, J = 7.2 Hz, H5’), 1.35m (4H, H3’ and H4’), 1.65m and 1.75m (2H, H2’), 3.21s (3H, CH3O), 4.50t (1H, J = 6.6 Hz, H1’), 6.28d (1H, J = 2.4 Hz, H2), 6.31dd (1H, J = 2.4 Hz, J = 8.2 Hz, H6), and 6.96d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD) of compound 7: (ppm) 13.02 (C5’), 22.26 (C4’), 27.73 (C3’), 35.92 (C2’), 55.14 (CH3O), 78.31 (C1’), 102.03 (C2), 106.40 (C6), 118.55 (C4), 127.52 (C5), 156.14 (C3), and 157.20 (C1). MS of compound 7, m/z (%, ion structure): 178 (38, M+ – CH3OH), and 149 (100, 178 – CH3CH2). 1H NMR (CD3OD) of compound 8: (ppm) 0.92t (3H, J = 7.2 Hz, H5’), 1.36m (2H, H4’), 1.36m and 1.49m (2H, H3’), 1.78m and 1.81m (2H, H2’), 5.22dd (1H, J = 5.4 Hz, J = 7.7 Hz, H1’), 6.27d (2H, J = 8.1 Hz, H4 and H6), and 6.87t (1H, J = 8.1 Hz, H5). 13C NMR (CD3OD) of compound 8: (ppm) 13.03 (C5’), 22.27 (C4’), 27.48 (C3’), 36.17 (C2’), 68.95 (C1’), 106.76 (C4 and C6), 115.54 (C2), 127.54 (C5), and 155.72 (C1 and C3). MS of the dehydrated product formed from compound 8, m/z (%, ion structure): 178 (48, M+), 149 (100, M+ – CH3CH2), and 123 (56, C7H7O2). The reaction between resorcinol and 2-methylbutanal produced 4-(1- methoxy-2-methylbutyl)benzene-1,3-diol (9a and 9b). This compound has two chiral centers at C1’ and C2’. Therefore, it was produced as two pairs of diastereomers, which have been named 9a and 9b. Their retention times were 18.64 and 18.86 min. The reaction was heated for 48 h at 60 ºC and the compounds were isolated using hexane-diethyl ether (7:3) as eluent. 1H NMR (CD3OD) of compound 9a: (ppm) 0.73d (3H, J = 6.9 Hz, CH3C2’), 0.92t (3H, J = 7.5 Hz, C4’), 1.21m and 1.76m (2H, H3’), 1.73m (2H, H2’), 3.19s (3H, CH3O), 4.24d (1H, J = 7.8 Hz, H1’), 6.28d (1H, J = 2.4 Hz, H2), 6.32dd (1H, J = 2.4 Hz, J = 8.2 Hz, H6), and 6.92d (1H, J = 8.2 Hz, H5). 13C NMR Results 96 (CD3OD) of compound 9a: (ppm) 10.19 (C4’), 14.26 (CH3C2’), 24.98 (C3’), 40.21 (C2’), 55.44 (CH3O), 83.13 (C1’), 101.93 (C2), 106.16 (C6), 117.37 (C4), 128.46 (C5), 156.60 (C3), and 157.09 (C1). MS of compound 9a, m/z (%, ion structure): 178 (100, M+ – CH3OH), 163 (71, 178 – CH3), 149 (84, 178 – CH3CH2), and 123 (75, C7H7O2). 1H NMR (CD3OD) of compound 9b: (ppm) 0.86t (3H, J = 7.5 Hz, C4’), 0.95d (3H, J = 6.7 Hz, CH3C2’), 1.10m and 1.37m (2H, H3’), 1.73m (2H, H2’), 3.20s (3H, CH3O), 4.28d (1H, J = 6.9 Hz, H1’), 6.28d (1H, J = 2.4 Hz, H2), 6.32dd (1H, J = 2.4 Hz, J = 8.2 Hz, H6), and 6.93d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD) of compound 9b: (ppm) 10.51 (C4’), 13.90 (CH3C2’), 25.53 (C3’), 40.18 (C2’), 55.60 (CH3O), 82.98 (C1’), 101.96 (C2), 106.26 (C6), 117.54 (C4), 128.26 (C5), 156.40 (C3), and 157.17 (C1). MS of compound 9b, m/z (%, ion structure): 178 (100, M+ – CH3OH), 163 (80, 178 – CH3), 149 (95, 178 – CH3CH2), and 123 (81, C7H7O2). The reaction between resorcinol and hexanal produced 4-(1- methoxyhexyl)benzene-1,3-diol (10). The reaction was heated for 3 h at 100 ºC and compound 10 was isolated using hexane-diethyl ether (75:25) as eluent. Retention time: 12.08 min. 1H NMR (CD3OD): (ppm) 0.88t (3H, J = 6.9 Hz, H6’), 1.27m (6H, H3’, H4’, and H5’), 1.66m (2H, H2’), 3.21s (3H, CH3O), 4.48t (1H, J = 6.7 Hz, H1’), 6.26dd (1H, J = 2.5 Hz, J = 8.1 Hz, H6), 6.30d (1H, J = 2.5 Hz, H2), and 6.94d (1H, J = 8.1 Hz, H5). 13C NMR (CD3OD): (ppm) 14.40 (C6’), 23.70 (C5’), 26.58 (C4’), 32.94 (C3’), 37.62 (C2’), 56.51 (CH3O), 79.93 (C1’), 103.33 (C6), 107.74 (C2), 119.93 (C4), 128.82 (C5), 157.55 (C3), and 158.60 (C1). MS, m/z (%, ion structure): 192 (34, M+ – CH3OH), 163 (2, 192 – CH3CH2), and 149 (100, 192 – CH3CH2CH2). The reaction between 2-methylresorcinol and propanal produced 4-(1- methoxypropyl)-2-methylbenzene-1,3-diol (11). The reaction was heated for 72 h at 60 ºC and compound 11 was isolated using hexane-diethyl ether (9:1) Results 97 as eluent. Retention time: 18.52 min. 1H NMR (CD3OD): (ppm) 0.88t (3H, J = 7.5 Hz, H3’), 1.68m and 1.83m (2H, H2’), 2.07s (3H, CH3C2), 3.29s (3H, CH3O), 4.23t (1H, J = 6.9 Hz, H1’), 6.33d (1H, J = 8.2 Hz, H6), and 6.96d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD): (ppm) 7.08 (CH3C2), 9.32 (C3’), 28.84 (C2’), 55.60 (CH3O), 80.81 (C1’), 106.04 (C6), 111.36 (C2), 116.85 (C4), 125.02 (C5), 153.92 (C3), and 155.41 (C1). MS, m/z (%, ion structure): 164 (100, M+ – CH3OH), 149 (42, 164 – CH3), 147 (17, 164 – OH), and 137 (20, C8H9O2). The reaction between 2-methylresorcinol and pentanal produced 4-(1- methoxypentyl)-2-methylbenzene-1,3-diol (12) and 2-methyl-4-(pent-1-en-1- yl)benzene-1,3-diol (13). The reaction was heated for 72 h at 60 ºC and the compounds were isolated using hexane-diethyl ether (85:15) as eluent. Retention time of compound 12: 20.28 min. Retention time of compound 13: 20.38 min. 1H NMR (CD3OD) of compound 12: (ppm) 0.89t (3H, J = 7.2 Hz, H5’), 1.23m and 1.38m (2H, H3’), 1.31m (2H, H4’), 1.65m and 1.82m (2H, H2’), 2.08s (3H, CH3C2), 3.28s (3H, CH3O), 4.31t (1H, J = 6.9 Hz, H1’), 6.33d (1H, J = 8.2 Hz, H6), and 6.69d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD) of compound 12: (ppm) 7.09 (CH3C2), 13.04 (C5’), 22.23 (C4’), 27.84 (C3’), 35.72 (C2’), 55.52 (CH3O), 83.31 (C1’), 106.05 (C6), 111.36 (C2), 117.15 (C4), 124.89 (C5), 153.89 (C3), and 155.41 (C1). MS of compound 12, m/z (%, ion structure): 192 (36, M+ – CH3OH), and 163 (100, 192 – CH3CH2). 1H NMR (CD3OD) of compound 13: (ppm) 0.95t (3H, J = 7.4 Hz, H5’), 1.50sx (2H, J = 7.4 Hz, H4’), 2.08s (3H, CH3C2), 2.18qd (2H, J = 1.4 Hz, J = 7.2 Hz, H3’), 5.98dt (1H, J = 7.2 Hz, J = 15.7 Hz, H2’), 6.33d (1H, J = 8.4 Hz, H6), 6.63d (1H, J = 15.7 Hz, H1’), and 7.03d (1H, J = 8.4 Hz, H5). 13C NMR (CD3OD) of compound 13: (ppm) 7.48 (CH3C2), 12.74 (C5’), 22.62 (C4’), 35.27 (C3’), 106.98 (C6), 111.33 (C2), 117.90 (C4), 123.04 (C5), 125.02 (C1’), 127.45 (C2’), 152.39 (C3), and 154.86 (C1). MS Results 98 of compound 13, m/z (%, ion structure): 192 (35, M+), and 163 (100, M+ – CH3CH2). The reaction between 2-methylresorcinol and 2-methylbutanal produced 4-(1-methoxy-2-methylbutyl)-2-methylbenzene-1,3-diol (14a and 14b). Analogously to compounds 9a and 9b, this compound also has two chiral centers at C1’ and C2’. Therefore, it was produced as two pairs of diastereomers, which have been named 14a and 14b. Their retention times were 19.05 and 19.27 min. The reaction was heated for 72 h at 60 ºC and the compounds were isolated using hexane-diethyl ether (85:15) as eluent. 1H NMR (CD3OD) of compound 14a: (ppm) 0.71d (3H, J = 6.9 Hz, CH3C2’), 0.84t (3H, J = 7.4 Hz, C4’), 1.20m and 1.80m (2H, H3’), 1.79m (1H, H2’), 2.07s (3H, CH3C2), 3.28s (3H, CH3O), 4.01d (1H, J = 8.1 Hz, H1’), 6.32d (1H, J = 8.2 Hz, H6), and 6.65d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD) of compound 14a: (ppm) 7.09 (CH3C2), 10.52 (C4’), 14.60 (CH3C2’), 25.10 (C3’), 39.78 (C2’), 55.99 (CH3O), 88.17 (C1’), 105.79 (C6), 111.23 (C2), 115.60 (C4), 126.11 (C5), 154.22 (C3), and 154.33 (C1). MS of compound 14a, m/z (%, ion structure): 192 (82, M+ – CH3OH), 177 (48, 192 – CH3), 163 (100, 192 – CH3CH2), and 137 (49, C8H9O2). 1H NMR (CD3OD) of compound 14b: (ppm) 0.92t (t, 3H, J = 6.9 Hz, C4’), 0.99d (d, 3H, J = 6.7 Hz, CH3C2’), 1.05m and 1.30m (2H, H3’), 1.72m (1H, H2’), 2.07s (3H, CH3C2), 3.29s (3H, CH3O), 4.07d (1H, J = 7.5 Hz, H1’), 6.33d (1H, J = 8.2 Hz, H6), and 6.66d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD) of compound 14b: (ppm) 7.09 (CH3C2), 10.16 (C4’), 14.15 (CH3C2’), 25.65 (C3’), 40.39 (C2’), 56.12 (CH3O), 87.98 (C1’), 105.87 (C6), 111.23 (C6), 115.83 (C4), 125.88 (C5), 154.17 (C3), and 155.41 (C1). MS of compound 14b, m/z (%, ion structure): 192 (90, M+ – CH3OH), 177 (45, 192 – CH3), 163 (100, 192 – CH3CH2), and 137 (49, C8H9O2). The reaction between 2-methylresorcinol and 3-methylbutanal produced 4-(1-methoxy-3-methylbutyl)-2-methylbenzene-1,3-diol (15). The Results 99 reaction was heated for 48 h at 60 ºC and compound 15 was isolated using hexane-diethyl ether (85:15) as eluent. Retention time: 19.97 min. 1H NMR (CD3OD): (ppm) 0.92d (3H, J = 6.6 Hz, C4’), 0.94d (3H, J = 6.6 Hz, CH3C3’), 1.46m and 1.76m (2H, H3’), 1.67hp (1H, J = 6.6 Hz, H3’), 2.08s (3H, CH3C2), 3.27s (3H, CH3O), 4.42dd (1H, J = 6.0 Hz, J = 8.2 Hz, H1’), 6.34d (1H, J = 8.2 Hz, H6), and 6.71d (1H, J = 8.2 Hz, H5). 13C NMR (CD3OD): (ppm) 7.09 (CH3C2), 21.42 (CH3C3’), 22.07 (C4’), 24.53 (C3’), 45.26 (C2’), 55.43 (CH3O), 81.33 (C1’), 106.14 (C6), 111.40 (C2), 117.45 (C4), 124.70 (C5), 153.92 (C3), and 155.41 (C1). MS, m/z (%, ion structure): 192 (67, M+ – CH3OH), 177 (83, 192 – CH3), 175 (19, 192 – OH), and 137 (100, C8H9O2). The reaction between 2-methylresorcinol and hexanal produced 4- (hex-1-en-1-yl)-2-methylbenzene-1,3-diol (16). The reaction was heated for 24 h at 100 ºC and compound 16 was isolated using hexane-diethyl ether (80:20) as eluent. Retention time: 13.55 min. 1H NMR (CD3OD): (ppm) 0.93t (3H, J = 7.3 Hz, H6’), 1.40m (4H, H4’ and H5’), 2.06s (3H, CH3C2), 2.18qd (2H, J = 1.5 Hz, J = 7.1 Hz, H3’), 5.96dt (1H, J = 7.1 Hz, J = 15.7 Hz, H2’), 6.32d (1H, J = 8.3 Hz, H6), 6.60dt (1H, J = 1.5 Hz, 15.7 Hz, H1’), and 7.01d (1H, J = 8.3 Hz, H5). 13C NMR (CD3OD): (ppm) 8.87 (CH3C2), 14.39 (C6’), 23.36 (C5’), 33.16 (C4’), 34.21 (C3’), 108.29 (C6), 112.61 (C2), 119.26 (C4), 124.36 (C5), 126.15 (C1’), 128.93 (C2’), 153.72 (C3), and 156.19 (C1). MS, m/z (%, ion structure): 206 (39, M+), 191 (1, M+ – CH3), 177 (100, M+ – CH3CH2), 163 (100, M+ – CH3CH2CH2), and 137 (19, C8H9O2). The reaction between resorcinol and glyoxal produced 1-(2,4- dihydroxyphenyl)-2-hydroxyethan-1-one (17). The reaction, using a glyoxalresorcinol ratio of 4:1, was heated for 3 h at 100 ºC and compound 17 was isolated using hexane-diethyl ether (1:3) as eluent. Retention time: 12.18 min. 1H NMR (CD3OD): (ppm) 4.80s (2H, CH2OH), 6.28d (1H, J = 2.3 Hz, H2), 6.35dd (1H, J = 2.3 Hz, J = 8.7 Hz, H6), and 7.59d (1H, J = 8.7 Hz, H5). 13C Results 106 than of 2-methylpropanal/2-methylresorcinol adduct was produced when aldehydes having four carbon atoms were studied (Figure 44). 0 120 240 360 040 80 150 175 Adduct (nmol/ mol aldehyde) Time (h) Figure 44. Time-course of carbonyl-phenol adduct formation in the reactions of butanal () and 2-methylpropanal () with 2- methylresorcinol. Reaction was carried out by heating a mixture of 2-methylresorcinol (40 mol) and the two alkanals (20 µmol of each) in 500 l of methanol containing 20 l of triethylamine at 60 ºC under nitrogen for the indicated times. Aldehydes having five carbon atoms exhibited two different behaviors (Figure 45). Thus, pentanal produced a higher amount of adducts than 2- methylbutanal. On the other hand, similar amounts of adducts were produced by both pentanal and 3-methylbutanal. Results 107 040 80 150 175 0 120 240 360 Time (h) Adduct (nmol/ mol aldehyde) Figure 45. Time-course of carbonyl-phenol adduct formation in the reactions of pentanal (), 2-methylbutanal (), and 3- methylbutanal () with 2-methylresorcinol. Reaction was carried out by heating a mixture of 2-methylresorcinol (60 mol) and the three alkanals (20 µmol of each) in 500 l of methanol containing 20 l of triethylamine at 60 ºC under nitrogen for the indicated times. Reaction rates for adduct formation between branched aldehydes and phenolic compounds were calculated as described above and they are collected in Table 6. Results 108 Table 6. Effect of Aldehyde Branching on the Rate Constants of Carbonyl-Phenol Adduct Formation Phenol Aldehyde Rate constant [(mmol adduct)·(mol aldehyde)–1·h–1] 2-Methylresorcinol Butanal 3.44 0.20 2-Methylpropanal 1.87 0.03 2-Methylresorcinol Pentanal 3.68 0.11 2-Methylbutanal 1.02 0.05 3-Methylbutanal 3.34 0.11 These results are a consequence of the steric hindrance introduced by the methyl group at position 2 of the aldehyde. In addition, this methyl group inhibited the formation of the 2-alkenal produced by aldol condensation and no carbonyl-phenol adducts involving two molecules of 2-methylbutanal were observed. 4.2. Epoxyalkenal-trapping ability of phenolic compounds. This section describes the isolation and characterization of the carbonyl-phenol adducts produced in the reaction of 4,5-epoxy-2-alkenals and phenolic compounds. In addition, the reaction conditions that promote the formation of the corresponding adducts are also reported. 4.2.1. Characterization of the adducts produced in the reaction between epoxyalkenals and phenolic compounds. The reaction between epoxyalkenals and phenolic compounds is very fast and rapidly produces a complex mixture of compounds (Hidalgo et al., 2017). Different attempts were carried out to isolate the produced compounds, Results 109 but all of them were unsuccessful because of the high reactivity of the formed adducts, which spontaneously suffered further reactions, including polymerizations. Only after their stabilization, they could be isolated and characterized. This stabilization was achieved by acetylation of the hydroxyl groups present in the formed adducts. As observed in Figure 46, acetylation allowed that carbonyl-phenol adducts could be easily identified by GC-MS. This technique was employed to control the chromatographic separation of the formed adducts. The compounds isolated and characterized in these reactions are described in subsections 4.2.1.1, 4.2.1.2, and 4.2.1.3. 11.5 12.0 12.5 13.0 13.5 14.0 25d 25a 25b 25c Time (min) 21 Figure 46. Total ion chromatogram obtained by GC-MS of the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol after acetylation. Only the portion corresponding to epoxyalkenal/phenol adducts is shown. Chemical structures for the identified compounds are given in Figure 47. Two different kinds of adducts were isolated by column chromatography and characterized by 1D and 2D NMR and MS (Figure 47). Results 110 Figure 47. Chemical structures of the carbonyl-phenol adducts isolated and characterized by NMR and MS in the different assayed reactions between epoxyalkenals and phenolics. The aldehyde and the phenolic compound involved in the formation of each adduct is described in the text. Differently to the adducts produced between alkanals and phenolics, which were the result of a single addition of the phenolic compound to the alkanal, the adducts produced with epoxyalkenals resulted more complex, and structures could only be determined by using 2D NMR. Thus, COSY, HMQC and HMBC experiments were carried out. Main HMBC couplings exhibited by compounds 20a and 25a are shown in Figure 48. HMBC couplings of compounds 20b, 22 were identical to those of compound 20a. HMBC couplings of compounds 24, 25b, and 26 were identical to those of compound 25a. Results 111 Figure 48. HMBC couplings exhibited by compounds 20a and 25a. The first type of adducts isolated in reactions between epoxyalkenals and phenolic compounds were 1,3a,4,9b-tetrahydro-2H-furo[2,3-c]chromen- 7-yl acetates (compounds 20a, 20b and 22). This cyclic form was the only observed in solution by NMR. However, because the hemiacetalic hydroxyl group was not acetylated, when these compounds were studied by GC-MS, the cyclic hemiacetal was opened and the free hydroxyl group suffered a dehydration to produce the corresponding 4-(2-oxoethyl)-2H-chromen-7-yl acetate (compounds 21 and 23). Compounds 20a, 20b and 22 had only one acetyl group and two phenolic protons (or one in the case of compound 22), indicating that one phenolic hydroxyl group and one of the phenolic protons were involved in the reaction with the epoxyalkenal. The epoxyalkenal chain could be easily identified in the adduct by means of a COSY experiment. The study of its chain showed that its epoxy, double bond and carbonyl groups were also involved in the reaction. The final characterization of their structures was carried out by using HMQC and HMBC experiments. In particular, HMBC showed long distance couplings, which were employed for the unequivocal identification of the formed structures. Two isomers (20a and 20b) were isolated in the reaction between 4,5- epoxy-2-heptenal and 2-methylresorcinol. This is consequence of the different chiral centers present in the molecule (C4, C3a, and C9b). Two of them (C3a Results 112 and C4) corresponded to the epoxy carbons of the epoxyalkenal (C4 and C5) and were already present in the aldehyde. Because the initial alkadienal (2,4- heptadienal) was trans,trans and the epoxidation was syn, the obtained epoxyalkenal should be a mixture of the isomer 4R,5R and its enantiomer. Supposing that the formation of the adduct did not imply a racemization, the creation of the new chiral center at C9b would explain the existence of the two isomers 20a and 20b. In addition, C2 is also chiral, although it is not fixed and the interconversion between the two absolute configurations of this carbon might be produced. The 1H NMR spectra of both isomers were quite similar. The most significant differences, as expected, corresponded to the protons belonging to chiral carbons C4, C3a, C9b, and C2. One difference was the chemical shift of protons H4 and H9b, which were interchanged in compounds 20a and 20b as confirmed by HMQC experiments. Other differences were the coupling constants, which were particularly different for H2. After dehydration, both compounds produced the same compound 21. The second type of adducts produced in these reactions were 3,4,4a,9a-tetrahydro-1H-pyrano[3,4-b]benzofuran-3,7-diyl diacetates (24-26). Compounds 24-26 had two acetyl groups and two phenolic protons (or one in the case of compound 26), indicating that, at least, one of the phenolic protons was involved in the reaction with the epoxyalkenal. The epoxyalkenal chain could also be easily identified in the adduct by means of a COSY experiment. The study of its chain showed that epoxy, double bond and carbonyl groups of the epoxyalkenal were also involved in the reaction. The final characterization of the structures was carried out by using HMQC and HMBC experiments. Analogously to the above described for compounds 20a and 20b, the reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol produced several compounds of this type (25a, 25b, 25c and 25d). In this case, four isomers were detected by GC-MS (Figure 46), two of which (25a and 25b) could be isolated and characterized. The existence of these 4 isomers is a consequence of the presence in the molecule of 4 chiral carbons: C1, C9a, C4a, and C3. Two of them (C1 and C9a) were already present in the initial Results 113 epoxyalkenal, but C4a and C3 were created in the reaction with the phenol and C3 was fixed in the later acetylation. The 1H NMR spectra of stereoisomers 25a and 25b were similar between them and the main differences were again the chemical shifts and coupling constants of protons close to chiral carbons. Thus, for example, the acetyl group of C3 was different for the two isomers. On the other hand, the acetyl group of C7, which is quite far from the chiral centers, was identical for the two isomers. These differences were also observed in the mass spectra. Thus, although the fragmentation pattern was identical for the two isomers, the intensities of the produced fragments were different. Nevertheless, all compounds 24-26 had very similar fragmentation patterns by MS. Although only a limited number of compounds could be isolated and characterized in the studied reactions, no other compounds having different mass fragmentation patterns were formed in the studied reactions to a significant extent. Therefore, it can be concluded that the reaction between epoxyalkenals and phenolic compounds mainly produces in a first step 1,3a,4,9b-tetrahydro-2H-furo[2,3-c]chromene-2,7-diols (27) and 3,4,4a,9a- tetrahydro-1H-pyrano[3,4-b]benzofuran-3,7-diols (28) (Figure 49). These compounds have reactive hydroxyl and carbonyl groups that can be involved in further reactions, which is in agreement with the reactivity observed for the formed adducts. Figure 49. Chemical structures of adducts produced in the reaction between epoxyalkenals and phenolics. Results 114 The spectroscopic and spectrometric data of the carbonyl-phenol adducts isolated and characterized in the different epoxyalkenal-phenol reactions assayed are collected in the next subsections. 4.2.1.1. Compounds isolated and characterized in the reaction of 4,5-epoxy-2- heptenal and 2-methylresorcinol. The reaction between 4,5-epoxy-2-heptenal and 2-methylresorcinol produced five carbonyl-phenol adducts (Figure 46). Four of them could be isolated and characterized. Their spectroscopic and spectrometric data are the following: 4-Ethyl-2-hydroxy-6-methyl-1,3a,4,9b-tetrahydro-2H-furo[2,3- c]chromen-7-yl acetate (20a and 20b). Two diastereomers of this compound could be isolated and characterized. 1H NMR (CDCl3) of compound 20a: (ppm) 1.04t (3H, J = 7.4 Hz, CH3CH2), 1.65qu (2H, J = 7.4 Hz, CH3CH2), 2.02s (3H, CH3C6), 2.21m and 2.29m (2H, H1), 2.33s (3H, CH3CO), 3.52m (1H, H4), 3.89q,br (1H, J = 7.5 Hz, H9b), 4.65t (1H, J = 9.3 Hz, H3a), 6.07dd (1H, J = 2.7 Hz and 5.2 Hz, H2), 6.59d (1H, J = 8.0 Hz, H8), and 6.97d (1H, J = 8.0 Hz, H9). 13C NMR (CDCl3) of compound 20a: (ppm) 9.19 (CH3C6), 9.45 (CH3CH2C4), 20.76 (CH3CO), 26.22 (CH3CH2C4), 29.58 (C1), 36.56 (C9b), 74.33 (C4), 82.31 (C3a), 92.17 (C2), 113.51 (C6), 114.39 (C8), 120.90 (C9), 125.77 (C9a), 149.71 (C7), 158.60 (C5a), 169.39 (CH3CO). 1H NMR (CDCl3) of compound 20b: (ppm) 1.06t (3H, J = 7.4 Hz, CH3CH2), 1.64qu (2H, J = 7.4 Hz, CH3CH2), 2.02s (3H, CH3C6), 1.95m and 2.30m (2H, H1), 2.32s (3H, CH3CO), 3.51m (1H, H9b), 3.84dt (1H, J = 3.9 Hz, J = 8.7 Hz, J = 8.7 Hz, H4), 4.58t (1H, J = 9.3 Hz, H3a), 6.15dd (1H, J = 5.9 Hz and 8.5 Hz, H2), 6.55d (1H, J = 8.0 Hz, H8), and 6.96d (1H, J = 8.0 Hz, H9). 13C NMR (CDCl3) of compound 20b: (ppm) 9.24 (CH3C6), 9.55 (CH3CH2C4), 20.74 (CH3CO), 26.36 (CH3CH2C4), 30.56 (C1), 37.42 (C9b), 71.81 (C4), 84.16 (C3a), 91.92 (C2), 113.39 (C6), 114.13 (C8), 121.30 (C9), 126.16 (C9a), Results 115 149.65 (C7), 159.16 (C5a), 169.35 (CH3CO). These compounds suffered a dehydration when studied by GC-MS and were converted into 2-ethyl-8- methyl-4-(2-oxoethyl)-2H-chromen-7-yl acetate (21). MS, m/z (%, ion structure): 274 (19, M+), 232 (33, M+ – CH2CO), 204 (15, 232 – ethene), 203 (100, 232 – CH3CH2), 189 (11, 204 – CH3), 175 (19, 204 – CHO), 174 (31, 204 – HCHO) and 161 (13, 204 – CH2CHO). 1-Ethyl-8-methyl-3,4,4a,9a-tetrahydro-1H-pyrano[3,4-b]benzofuran- 3,7-diyl diacetate (25a and 25b). Two diastereomers of this compound could be isolated and characterized. 1H NMR (CDCl3) of compound 25a: (ppm) 0.91t (3H, J = 7.4 Hz, CH3CH2), 1.57m and 1.78m (2H, CH2C1), 1.99s (3H, CH3C8), 2.15s (3H, CH3CO), 2.29m (2H, H4), 2.32s (3H, CH3CO), 3.05d (1H, J = 3.9 Hz, H4a), 4.29d (1H, J = 8.3 Hz, H9a), 4.75dt (1H, J = 3.8 Hz, J = 8.1 Hz, H1), 5.83d (1H, J = 3.3 Hz, H3), 6.56d (1H, J = 8.1 Hz, H6), and 6.83d (1H, J = 8.1 Hz, H5). 13C NMR (CDCl3) of compound 25a: (ppm) 8.99 (CH3C8), 9.78 (CH3CH2), 20.77 (CH3CO), 21.11 (CH3CO), 23.93 (CH2C1), 30.19 (C4), 37.72 (C4a), 73.76 (C1), 89.85 (C9a), 99.34 (C3), 113.74 (C6), 118.61 (C8), 123.52 (C5), 126.15 (C4b), 149.18 (C7), 150.59 (C8a), 169.31 (CH3CO), and 170.62 (CH3CO). MS of compound 25a, m/z (%, ion structure): 334 (4, M+), 292 (1, M+ – CH2CO), 274 (2, 278 – H2O), 245 (55, 274 – CHO), 232 (16, 274 – CH2CO), 217 (2, 232 – CH3), 203 (100, 232 – CH3CH2), 191 (18, 232 – C3H5), 190 (5, 232 – CH2CO), 189 (15, 232 – CH2CHO), 175 (8, 190 – CH3), and 161 (39, 190 – CH3CH2). 1H NMR (CDCl3) of compound 25b: (ppm) 0.85t (3H, J = 7.5 Hz, CH3CH2), 1.65m and 1.76m (2H, CH2C1), 1.98s (3H, CH3C8), 2.03s (3H, CH3CO), 2.25dd (2H, J = 2.4 Hz, J = 3.2 Hz, H4), 2.31s (3H, CH3CO), 3.20dd (1H, J = 3.0 Hz, J = 5.1 Hz, H4a), 4.22dd (1H, J = 3.2 Hz, J = 8.5 Hz, H9a), 4.47ddd (1H, J = 4.1 Hz, J = 6.5 Hz, J = 8.5 Hz, H1), 5.80br (1H, H3), 6.55d (1H, J = 8.1 Hz, H6), and 6.68d (1H, J = 8.1 Hz, H5). 13C NMR (CDCl3) of compound 25b: (ppm) 8.62 (CH3CH2), 9.07 (CH3C8), 20.78 (CH3CO), 21.00 (CH3CO), 24.37 (CH2C1), 32.87 (C4), 38.40 (C4a), 73.94 (C1), 87.95 (C9a), 99.13 (C3), References 218 References 219 Abeywickrama, G., Debnath, S. C., Ambigaipalan, P., & Shahidi, F. (2016). Phenolics of selected cranberry genotypes (Vaccinium macrocarpon Ait.) and their antioxidant efficacy. Journal of Agricultural and Food Chemistry, 64, 9342−9351. Andjelkovic, M., Van Camp, J., De Meulenaer, B., Depaemelaere, G., Socaciu, C., Verloo, M., & Verhe, R. (2006). Iron-chelation properties of phenolic acids bearing catechol and galloyl groups. Food Chemistry, 98, 23-31. Bastos, L. C. S., Costa, E. A. A., & Pereira, P. A. P. (2017). Development, validation and application of an UFLC-DAD-ESI-MS method for determination of carbonyl compounds in soybean oil during continuous heating. Food Chemistry, 218, 518–524. Belitz, H. D., Grosch, W., & Schieberle, P. (2009). Food Chemistry. 4th revised and extended edition, Springer-Verlag, Berlin, Germany. Beretta, G., Furlanettob, S., Regazzoni, L., Zarrella, M., & Facino, R. M. (2008). Quenching of , -unsaturated aldehydes by green tea polyphenols: HPLC-ESI-MS/MS studies. Journal of Pharmaceutical and Biomedical Analysis, 48, 606-611. Brewer, M. S. (2011). Natural antioxidants: sources, compounds, mechanisms of action, and potential applications. Comprehensive Reviews in Food Science and Food Safety, 10, 221-247. Brewer, S. (2014). Effects of oxidation on sensory characteristics of food components during processing and storage. In: Food Oxidants and Antioxidants: Chemical, Biological, and Functional Properties. Bartosz, G., ed., CRC Press, Boca Raton, 159-186. References 220 Bukhari, S. B., Memon, S., Mahroof-Tahir, M., & Bhanger, M. I. (2009). Synthesis, characterization and antioxidant activity copper-quercetin complex. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 71, 1901-1906. Cai, Y. Z., Sun, M., Xing, J., Luo, Q., & Corke, H. (2006). Structure-radical scavenging activity relationships of phenolic compounds from traiditional Chinese medicinal plants. Life Sciences, 78, 2872-2888. Cheng, K. W., Chen, F., & Wang, M. (2007). Inhibitory activities of dietary phenolic compounds on heterocyclic amine formation in both chemical model system and beef patties. Molecular Nutrition & Food Research, 51, 969-976. Cheng, K. W., Wong, C. C., Chao, J., Lo, C., Chen, F., Chu, I. K., Che, C. M., Ho, C. T. & Wang, M. (2009). Inhibition of mutagenic PhIP formation by epigallocatechin gallate via scavenging of phenylacetaldehyde. Molecular Nutrition & Food Research, 53, 716-725. Cheng, K. W., Wong, C. C., Cho, C. K., Chu, I. K., Sze, H. K., Lo, C., Chen, F., & Wang, M. (2008). Trapping of phenylacetaldehyde as a key mechanism responsible for naringenin´s inhibitory activity in mutagenic 2- amino-1-methyl-6-phenylimidazo[4,5-b]pyridine formation. Chemical Research in Toxicology, 21, 2026-2034. Correddu, F., Nuda, A., Manca, M. G., Pulina, G., & Dalsgaard, T. K. (2015). Light-induced lipid oxidation in sheep milk: effects of dietary grap seed and linseed, alone or in combination, on milk oxidative stability. Journal of Agricultural and Food Chemistry, 63, 3980−3986. Coultate, T. (2016). Lipids. In: Food. The Chemistry of its components. Coultate, T., eds., 6th ed., The Royal Society of Chemistry, Cambridge, 107-171. References 221 Crozier, A., Jaganath, I. B., & Clifford, M. N. (2009). Dietary phenolics: chemistry, bioavailability and effects on health. Natural Products Reports, 26, 1001-1043. Crozier, A., Lean, M. E. J., Mc Donald, M. S., & Black, C. (1997). Quantitative analysis of the flavonoid content of commercial tomatoes, onions, lettuce, and celery. Journal of Agricultural and Food Chemistry, 45, 590-595. Da Silva, T. O., & de Paula Pereira, P. A. (2008). Influence of time, surface- to-volume ratio, and heating process (continuous or intermittent) on the emission rates of selected carbonyl compounds during thermal oxidation of palm and soybean oils. Journal of Agricultural and Food Chemistry, 56, 3129-3135. Decker, E. A. (2008). Antioxidant mechanisms. In: Food Lipids: Chemistry, Nutrition, and Biotechnology. Akoh, C. C., & Min, D. B., eds., 3rd ed., CRC Press, Boca Raton, 409-433. Delgado, R. M., Hidalgo, F. J., Zamora, R. (2016). Antagonism between lipidderived reactive carbonyls and phenolic compounds in the Strecker degradation of amino acids. Food Chemistry, 194, 1143-1148. Demir, E., Turna, F., Kaya, B., Creus, A., & Marcos, R. (2013). Mutagenic/recombinogenic effects of four lipid peroxidation products in Drosophila. Food Chemistry, 53, 221–227. Erickson, M. C. (2008). Lipid oxidation of muscle foods. In: Food Lipids: Chemistry, Nutrition, and Biotechnology. Akoh, C. C., & Min, D. B., eds., 3rd ed., CRC Press, Boca Raton, 321-347. References 222 Ewert, A., Granvogl, M., & Schieberle, P. (2011). Development of two stable isotope dilution assays for the quantitation of acrolein in heat-processed fats. Journal of Agricultural and Food Chemistry, 59, 3582–3589. Ewert, A., Granvogl, M., & Schieberle, P. (2014). Isotope-labeling studies on the formation pathway of acrolein during heat processing of oils. Journal of Agricultural and Food Chemistry, 62, 8524-8529. Frankel, E. N. (2005). Lipid Oxidation. 2nd ed., The Oily Press, Bridgwater. Gardner, H. W. (1989). Oxygen radical chemistry of polyunsaturated fatty acids. Free Radical Biology & Medicine, 7, 65-86. Gliszczynska-Swiglo, A., & Oszmianski, J. (2014). Antioxidant and prooxidant activity of food components. In: Food Oxidants and Antioxidants: Chemical, Biological, and Functional Properties. Bartosz, G., ed., CRC Press, Boca Raton, 375-432. Granvogl, M. (2014). Development of three stable isotope dilution assays for the quantitation of (E)-2-butenal (crotonaldehyde) in heat-processed edible fats and oils as well as in food. Journal of Agricultural and Food Chemistry, 62, 1272–1282. Guillen, M. D., & Goicoechea, E. (2008a). Formation of oxygenated , - unsaturated aldehydes and other toxic compounds in sunflower oil oxidation at room temperature in closed receptacles. Food Chemistry, 111, 157–164. Guillen, M. D., & Goicoechea, E. (2008b). Toxic oxygenated , -unsaturated aldehydes and their study in foods: A review. Critical Reviews in Food Science and Nutrition, 48, 119-136. Guillén, M. D., & Uriarte, P. S. (2012). Aldehydes contained in edible oils of a very different nature after prolonged heating at frying temperature: References 223 Presence of toxic oxygenated , unsaturated aldehydes. Food Chemistry, 131, 915-926. Heim, K. E., Tagliaferro, A. R., & Bolbiya, D. J. (2002). Flavonoid antioxidants: Chemistry, metabolism and structure-activity relationships. Journal of Nutritional Biochemistry, 13, 572-584. Hermund, D. B., Karadag, A., Andersen, U., Jonsdottir, R., Kristinsson, H. G., Alasalvar, C., & Jacobsen, C. (2016). Oxidative stability of granola bars enriched with multilayered fish oil emulsion in the presence of novel brown seaweed based antioxidants. Journal of Agricultural and Food Chemistry, 64, 8359−8368. Hidalgo, F. J., Delgado, R. M., Navarro, J. L., & Zamora, R. (2010). Asparagine decarboxylation by lipid oxidation products in model systems. Journal of Agricultural and Food Chemistry, 58, 10512–10517. Hidalgo, F. J., Delgado, R. M., & Zamora, R. (2017). Protective effect of phenolic compounds on carbonyl-amine reactions produced by lipidderived reactive carbonyls. Food Chemistry, 229, 388-395. Hidalgo, F. J., Gallardo, E., & Zamora, R. (2005). Strecker type degradation of phenylalanine by 4-hydroxy-2-nonenal in model systems. Journal of Agricultural and Food Chemistry, 53, 10254-10259. Hidalgo, F. J., León, M. M., & Zamora, R. (2016). Amino acid decarboxylations produced by lipid-derived reactive carbonyls in amino acid mixtures. Food Chemistry, 209, 256–261. Hidalgo, F. J., & Zamora, R. (1993). Fluorescent pyrrole products from carbonyl-amine reactions. Journal of Biological Chemistry, 268, 16190– 16197. References 224 Hidalgo, F. J., & Zamora, R. (1995). In vitro production of long chain pyrrole fatty esters from carbonyl-amine reactions. Journal of Lipid Research, 36, 725–735. Hidalgo, F. J., & Zamora, R. (2000a). The role of lipids in nonenzymatic browning. Grasas y Aceites, 51(1-2), 35-49. Hidalgo, F. J., & Zamora, R. (2000b). Modification of bovine serum albumin structure following reaction with 4,5(E)-epoxy-2(E)-heptenal. Chemical Research in Toxicology, 13, 501–508. Hidalgo, F. J., & Zamora, R. (2004). Strecker-type degradation produced by the lipid oxidation products 4,5-epoxy-2-alkenals. Journal of Agricultural and Food Chemistry, 52, 7126-7131. Hidalgo, F. J., & Zamora, R. (2014). 2-Alkenal scavenging ability of mdiphenols. Food Chemistry, 160, 118-126. Hidalgo, F. J., & Zamora, R. (2016). Amino acid degradations produced by lipid oxidation products. Critical Reviews in Food Science and Nutrition, 56, 1242–1252. Hidalgo, F. J., & Zamora, R (2017). Food processing antioxidants. Advances in Food and Nutrition Research, 81, 31-64. Janoszka, B. (2010). Heterocyclic amines and azaarenes in pan-fried meat and its gravy fried without additives and in the presence of onion and garlic. Food Chemistry, 120, 463-473. Kasai, H., & Kawai, K. (2008). 4-Oxo-2-hexenal, a mutagen formed by -3 fat peroxidation: Occurrence, detection, and adduct formation. Mutation Research, 659, 56–59. References 225 Kawai, K., Matsuno, K., & Kasai, H. (2006). Detection of 4-oxo-hexenal, a novel mutagenic product of lipid peroxidation, in human diet and cooking vapor. Mutation Research, 603, 186–192. Khokhar, S., & Owusu Apenten, R. K. (2003). Iron binding characteristics of phenolic compounds: Some tentative structure-activity relations. Food Chemistry, 81, 133-140. Kikugawa, K., Ido, Y., & Mikami, A. (1984). Studies on peroxidized lipids. VI. Fluorescent products derived from the reaction of primary amines, malondialdehyde and monofunctional aldehydes. Journal of the American Oil Chemists’ Society, 61, 1574–1581. Kim, S. Y., Li, J. L., Lim, N. R., Kang, B. S., & Park, H. J. (2016). Prediction of warmed-over flavor development in cooked chicken by colorimetric sensor array. Food Chemistry, 211, 440–447. Kim, H. J., & Min, D. B. (2008). Chemistry of Lipid Oxidation. In: Food Lipids: Chemistry, Nutrition, and Biotechnology. Akoh, C. C., & Min, D. B., eds., 3rd ed., CRC Press, Boca Raton, 299-318. Kitryte, V., Adams, A., Venskutonis, P. R., & De Kimpe, N. (2012). Impact of lipid oxidation-derived aldehydes and ascorbic acid on the antioxidant activity of model melanoidins. Food Chemistry, 135, 1273–1283. Kokkinidou, S., & Peterson, D. G. (2014). Control of Maillard-type off-flavor development in ultrahigh-temperature-processed bovine milk by phenolic chemistry. Journal of Agricultural and Food Chemistry, 62, 8023−8033. Kolakowska, A. (2003). Lipid Oxidation in Food Systems. In: Chemical and Functional Properties of Food Lipids. Sikorski, Z. E., & Kolakowska, A., eds., CRC Press, Boca Raton, 133-166. References 226 Kolakowska, A., & Sikorski, Z. E. (2003). The role of lipids in food quality. In: Chemical and Functional Properties of Food Lipids. Sikorski, Z. E., & Kolakowska, A., eds., CRC Press, Boca Raton, 1-8. Lee, J., Xiao, L., Zhang, G., Ebeler, S. E., & Mitchell, A. E. (2014). Influence of storage on volatile profiles in roasted almonds (Prunus dulcis). Journal of Agricultural and Food Chemistry, 62, 11236−11245. Lee, S. H., Oe, T., & Blair, I.A. (2002). 4,5-Epoxy-2(E)-decenal-induced formation of 1, N2-etheno-2’-deoxyguanosine adducts. Chemical Research in Toxicology, 15, 300–304. Leopoldini, M., Russo, N., Chiodo, S., & Toscano, M. (2006). Iron chelation by the powerful antioxidant flavonoid quercetin. Journal of Agricultural and Food Chemistry, 54, 6343-6352. Leopoldini, M., Russo, N., & Toscano, M. (2011). The molecular basis of working mechanism of natural polyphenolic antioxidants. Food Chemistry, 125, 288-306. Lindsay, R. C. (2017). Food additives. In: Fennema´s Food Chemistry. Damodaran, S., & Parkin, K. L., eds., 5th ed., CRC Press, Boca Raton, 803- 864. Liu, H., & Gu, L. (2012). Phlorotannins from brown algae (Fucus versiculosus) inhibited the formation of advanced glycation endproducts by scavenging reactive carbonyls. Journal of Agricultural and Food Chemistry, 60, 1326-1334. Liu, Q., Raina, A.K., Smith, M.A., Sayre, L.M., & Perry, G. (2003). Hydroxynonenal, toxic carbonyls and Alzheimer disease. Molecular Aspects of Medicine, 24, 305–313. References 227 Liu, S., Liu, F., Xue, Y., & Gao, Y. (2016). Evaluation on oxidative stability of walnut beverage emulsions. Food Chemistry, 203, 409–416. Liu, W., Ma, H., Frost, L., Yuan, T., Dain, J. A., & Seeram, N. P. (2014). Pomegranate phenolics inhibit formation of advanced glycation endproducts by scavenging reactive carbonyl species. Food & Function, 5, 2996-3004. Lo, C. Y., Hsiao, W. T., & Chen, X. Y. (2011). Efficiency of trapping methylglyoxal by phenols and phenolic acids. Journal of Food Science, 76, H90-H96. Lo, C. Y., Li, S., Tan, D., Pan, M. H., Sang, S., & Ho, C. T. (2006). Trapping reactions of reactive carbonyl species with tea polyphenols in simulated physiological conditions. Molecular Nutrition and Food Research, 50, 118-1128. Lv, L., Shao, X., Chen, H., Ho, C. T., & Sang, S. (2011). Genistein inhibits advanced glycation end product formation by trapping methylglyoxal. Chemical Research in Toxicology, 24, 579-586. Malesev, D., & Kuntic, V. (2007). Investigation of metal-flavonoid chelates and the determination of flavonoids via metal-flavonoid complexing reactions. Journal of the Serbian Chemical Society, 72, 921-939. Maqsood, S., Benjakul, S., Abushelaibi, A., & Alam, A. (2014). Phenolic compounds and plant phenolic extracts as natural antioxidants in prevention of lipid oxidation in seafood: A detailed review. Comprehensive Reviews in Food Science and Food Safety, 13, 1125-1140. Mara Block, J., & Barrera-Arellano, D. (2012). Introducción a la química de lípidos. In: Temas selectos en aceites y grasas. Volumen 2-Química. Mara Block J., & Barrera-Arellano D., eds., Blucher, Sao Paulo, 1-36. References 234 Umano, K., & Shibamoto, T. (1987). Analysis of acrolein from heated cooking oils and beef fat. Journal of Agricultural and Food Chemistry, 35, 909-912. Van Hecke, T., Vossen, E., Hemeryck, L. Y., Vanden Bussche, J., Vanhaecke, L., & De Smet, S. (2015). Increased oxidative and nitrosative reactions during digestion could contribute to the association between well-done red meat consumption and colorectal cancer. Food Chemistry, 187, 29–36. Wang. Y, & Ho, C. T. (2012). Flavour chemistry of methylglyoxal and glyoxal. Chemical Society Reviews, 41, 4140-4149. Wells-Knecht, K. J., Brinkmann, E., & Baynes, J. W. (1995). Characterization of an imidazolium salt formed from glyoxal and Nα-hippuryllysine: A model for Maillard reaction crosslinks in proteins. Journal of Organic Chemistry, 60, 6246–6247. Wright, J. S., Johnson, E. R., DiLabio, G. A. (2001). Predicting the activity of phenolic antioxidants: Theoretical method, analysis of substituent effects, and application to major families of antioxidants. Journal of the American Chemical Society, 123, 1173–1183. Wu, C. H., Huang, S. M., Lin, J. A., & Yen G. C. (2011). Inhibition of advanced glycation endproduct formation by foodstuffs. Food & Function, 2, 224-234. Xu, F., Oruna-Concha, M. J., & Elmore, J. S. (2016). The use of asparaginase to reduce acrylamide levels in cooked food. Food Chemistry, 210, 163- 171. Yoon, S. R., & Shim, S. M. (2015). Inhibitory effect of polyphenols in Houttuynia cordata on advanced glycation end-products (AGEs) by References 235 trapping methylglyoxal. LWT – Food Science and Technology, 61, 158- 163. Zamora, R., Alaiz, M., & Hidalgo, F. J. (1999). Modification of histidine residues by 4,5-epoxy-2-alkenals. Chemical Research in Toxicology, 12, 654–660. Zamora, R., Alcón, E., & Hidalgo, F. J. (2014). Ammonia and formaldehyde participate in the formation of 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) in addition to create(ni)ne and phenylacetaldehyde. Food Chemistry, 155, 74-80. Zamora, R., Delgado, R. M., & Hidalgo, F. J. (2009). Conversion of 3- aminopropionamide and 3-alkylaminopropionamides into acrylamide in model systems. Molecular Nutrition and Food Research, 53, 1512–1520. Zamora, R., Delgado, R. M., & Hidalgo, F. J. (2010). Model reactions of acrylamide with selected amino compounds. Journal of Agricultural and Food Chemistry, 58, 1708–1713. Zamora, R., Delgado, R. M., & Hidalgo, F. J. (2011). Amino phospholipids and lecithins as mitigating agents for acrylamide in asparagine/glucose and asparagine/2,4-decadienal model systems. Food Chemistry, 126, 104-108. Zamora, R., Delgado, R. M., & Hidalgo, F. J. (2012). Chemical conversion of phenylethylamine into phenylacetaldehyde by carbonyl−amine reactions in model systems. Journal of Agricultural and Food Chemistry, 60, 5491−5496. Zamora, R., Gallardo, E., & Hidalgo, F. J. (2006b). Amine degradation by 4,5-epoxy-2-decenal in model systems. Journal of Agricultural and Food Chemistry, 54, 2398-2404. References 236 Zamora, R., Gallardo, E., & Hidalgo, F. J. (2007). Strecker degradation of phenylalanine initiated by 2,4-decadienal or methyl 13-oxooctadeca-9,11- dienoate in model systems. Journal of Agricultural and Food Chemistry, 55, 1308-1314. Zamora, R., Gallardo, E., & Hidalgo, F. J. (2008). Model studies on the degradation of phenylalanine initiated by lipid hydroperoxides and their secondary and tertiary oxidation products. Journal of Agricultural and Food Chemistry, 56, 7970–7975. Zamora, R., Gallardo, E., Navarro, J. L., & Hidalgo, F. J. (2005). Streckertype degradation of phenylalanine by methyl 9,10-epoxy-13-oxo-11- octadecenoate and methyl 12,13-epoxy-9-oxo-11-octadecenoate. Journal of Agricultural and Food Chemistry, 53, 4583-4588. Zamora, R., & Hidalgo, F. J. (2005). 2-Alkylpyrrole formation from 4,5- epoxy-2-alkenals.Chemical Research in Toxicology, 18, 342–348. Zamora, R., & Hidalgo, F. J. (2008). Contribution of lipid oxidation products to acrylamide formation in model systems. Journal of Agricultural and Food Chemistry, 56, 6075–6080. Zamora, R., & Hidalgo, F. J. (2011). The Maillard reaction and lipid oxidation. Lipid Technology, 23(3), 59-62. Zamora, R., & Hidalgo, F. J. (2015). 2-Amino-1-methyl-6- phenylimidazo[4,5-b]pyridine (PhIP) formation and fate: an example of the coordinate contribution of lipid oxidation and Maillard reaction to the production and elimination of processing-related food toxicants. RSC Advances, 5, 9709–9721. References 237 Zamora, R., & Hidalgo, F. J. (2016). The triple defensive barrier of phenolic compounds against the lipid oxidation-induced damage in food products. Trends in Food Science & Technology, 54, 165-174. Zamora, R., León, M. M., & Hidalgo, F. J. (2015). Oxidative versus nonoxidative decarboxylation of amino acids: Conditions for the preferential formation of either Strecker aldehydes or amines in amino acid/lipidderived reactive carbonyl model systems. Journal of Agricultural and Food Chemistry, 63, 8037−8043. Zamora, R., Navarro, J. L., Gallardo, E., & Hidalgo, F. J. (2006a). Chemical conversion of -amino acids into -keto acids by 4,5-epoxy-2-decenal. Journal of Agricultural and Food Chemistry, 54, 6101-6105. Zarkovic, N. (2003). 4-Hydroxynonenal as a bioactive marker of pathophysiological processes. Molecular Aspects of Medicine, 24, 281– 291. Zeng, Z., Ma, J., Liu, B., & Jiang, H. (2015). Amino acid-catalyzed formation of 2-vinylfuran from lipid-derived 4-oxo-2-hexenal. Food Chemistry, 188, 591–595. Zheng, Z. P., Yan, Y., Xia, J., Zhang, S., Wang, M., Chen, J., & Xu, Y. (2016). A phenylacetaldehyde-flavonoid adduct, 8-C-(E-phenylethenyl)- norartocarpetin, exhibits intrinsic apoptosis and MAPK pathways-related anticancer potential on HepG2, SMMC-7721 and QGY-7703. Food Chemistry, 197, 1085-1092. Zhu, Q., Liang, C. P., Cheng, K. W., Peng, X., Lo, C. Y., Shahidi, F., Chen, F., Ho, C. T., & Wang, M. (2009b). Trapping effects of green and black tea extracts on peroxidation-derived carbonyl substances of seal blubber oil. Journal of Agricultural and Food Chemistry, 57, 1065-1069. References 238 Zhu, Q., Zheng, Z. P., Cheng, K. W., Wu, J. J., Zhang, S., Tang, Y. S., Sze, K. H., Chen, J., Chen, F., & Wang, M. (2009a). Natural polyphenols as direct trapping agents of lipid peroxidation-derived acrolein and 4-hydroxy- trans-2-nonenal. Chemical Research in Toxicology, 22, 1721-1727.