alde hide

Upload: veronica-drg

Post on 14-Apr-2018

219 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/27/2019 Alde Hide

    1/21

    Aldehydic lipid peroxidation products derived from linoleic acid

    Peter Spiteller, Werner Kern, Josef Reiner, Gerhard Spiteller *

    Lehrstuhl Organische Chemie I, Universitat Bayreuth, Universitatsstrasse 30, 95440 Bayreuth, Germany

    Received 20 October 2000; received in revised form 12 February 2001; accepted 13 February 2001

    Abstract

    Lipid peroxidation (LPO) processes observed in diseases connected with inflammation involve mainly linoleic acid. Its

    primary LPO products, 9-hydroperoxy-10,12-octadecadienoic acid (9-HPODE) and 13-hydroperoxy-9,11-octadecadienoic

    acid (13-HPODE), decompose in multistep degradation reactions. These reactions were investigated in model studies:

    decomposition of either 9-HPODE or 13-HPODE by Fe2 catalyzed air oxidation generates (with the exception of

    corresponding hydroxy and oxo derivatives) identical products in often nearly equal amounts, pointing to a common

    intermediate. Pairs of carbonyl compounds were recognized by reacting the oxidation mixtures with pentafluorobenzylhy-

    droxylamine. Even if a pure lipid hydroperoxide is subjected to decomposition a great variety of products is generated, since

    primary products suffer further transformations. Therefore pure primarily decomposition products of HPODEs were

    exposed to stirring in air with or without addition of iron ions. Thus we observed that primary products containing the

    structural element R-CH= CH-CH = CH-CH = O add water and then they are cleaved by retroaldol reactions. 2,4-

    Decadienal is degraded in the absence of iron ions to 2-butenal, hexanal and 5-oxodecanal. Small amounts of buten-1,4-dial

    were also detected. Addition of m-chloroperbenzoic acid transforms 2,4-decadienal to 4-hydroxy-2-nonenal. 4,5-Epoxy-2-decenal, synthetically available by treatment of 2,4-decadienal with dimethyldioxirane, is hydrolyzed to 4,5-dihydroxy-

    2-decenal. 2001 Elsevier Science B.V. All rights reserved.

    Keywords: Lipid peroxidation; 2,4-Decadienal; 4-Hydroxy-2-nonenal; 2-Butenal; 5-Oxodecanal; 4,5-Epoxy-2-decenal; 4-Oxo-2-nonenal;

    4,5-Dihydroxy-2-decenal; Buten-1,4-dial; 9-Hydroxy-12-oxo-10-dodecenoic acid

    1388-1981 / 01 / $ ^ see front matter 2001 Elsevier Science B.V. All rights reserved.

    PII: S 1 3 8 8 - 1 9 8 1 ( 0 1 ) 0 0 1 0 0 - 7

    Abbreviations: EA, ethyl acetate; CH, cyclohexane; GC, gas chromatography; GC/MS, gas chromatography-mass spectrometry;

    4-HNE, 4-hydroxy-2-nonenal; 9-HODE, 9-hydroxy-10,12-octadecadienoic acid; 9-HPODE, 9-hydroperoxy-10,12-octadecadienoic acid;13S-HODE, (13S,9Z,11E)-13-hydroxy-9,11-octadecadienoic acid; 13S-HPODE, (13S,9Z,11E)-13-hydroperoxy-9,11-octadecadienoic acid.

    Expressions without stereochemical specications ^ e.g. 13-HODE ^ are used for racemic mixtures. This is the case for products generated

    in nonenzymic LPO processes; HPLC, high performance liquid chromatography; KODE, oxodecadienoic acid; Lc, radical of an unsat-

    urated fatty acid; LA, linoleic acid; LDL, low density lipoprotein; LH, polyunsaturated fatty acid; LOc, alkoxy radical of a polyunsat-

    urated fatty acid; LOH, hydroxy derivative of a polyunsaturated fatty acid; LOOc, peroxy radical of a polyunsaturated fatty acid;

    LOOH, hydroperoxide of a polyunsaturated fatty acid; LPO, lipid peroxidation; MCPBA, meta-chloroperbenzoic acid; MSTFA, N-

    methyl-N-(trimethylsilyl)triuoroacetamide; NP-HPLC, normal phase high performance liquid chromatography; m/v, molar mass vol-

    ume; PFBHA, pentauorobenzylhydroxylamine; PFBO, pentauorobenzyloxime; PUFA, polyunsaturated fatty acid; RI, retention

    index; TIC, total ion current; TLC, thin layer chromatography; TMS, trimethylsilyl

    * Corresponding author. Fax: +49-921-552-671; E-mail: [email protected]

    Biochimica et Biophysica Acta 1531 (2001) 188^208

    www.bba-direct.com

  • 7/27/2019 Alde Hide

    2/21

    1. Introduction

    Injury of plant [1,2] or mammalian tissue [3] re-

    sults in fast activation of enzymes: for instance in

    mammalian cells phospholipase A2 is activated [4].Phospholipase A2 liberates fatty acids situated in

    position 2 of phospholipids. In this position there

    are localized preferentially polyunsaturated fatty

    acids (PUFAs), characterized by the presence of

    -CH = CH-CH2-CH = CH groups [5]. Thus liberated

    PUFAs (in mammals mainly linoleic and arachidonic

    acid) are substrates for lipoxygenases, also activated

    by cell injury. Lipoxygenases transform PUFAs to

    lipid hydroperoxides (LOOHs) [6,7]. When, after

    cell injury, high amounts of PUFAs are oered to

    lipoxygenases [8], apparently due to depletion of oxy-

    gen, Lc radicals are not fast enough transformed to

    LOOHs. Radicals react probably under these condi-

    tions with histidine residues surrounding the active

    center: the altered histidine residues are no longer

    able to keep the iron ion in the active center of

    lipoxygenase in its complex bond. As a consequence

    iron ions are released and react with LOOHs to gen-

    erate LOc and LOOc radicals [9,10]. These in turn

    induce autocatalytic lipid peroxidation (LPO) reac-

    tions. Products formed in these autocatalytic process-

    es are observed after tissue injury and inammation,

    they are characteristic indicators of cell degradation.Increased amounts of LPO products, for instance of

    malondialdehyde (a degradation product of arachi-

    donic acid), but especially of hydroxyoctadecadienoic

    acids (HODEs) ^ derived from linoleic acid ^ are

    observed in many diseases connected with inamma-

    tion [11], for instance in atherosclerosis [12] or rheu-

    matic arthritis [13].

    Autocatalytic LPO processes are distinguished

    from those introduced by lipoxygenases by the

    ability of radicals to attack any -CH = CH-CH2-

    CH = CH group with about equal probability. Thus

    radicals attack not only free acids, but also conju-

    gates (most lipoxygenases require as substrates free

    fatty acids). Substrates for radical attack are there-

    fore phospholipids containing PUFAs, or cholesterol

    linoleate, a main constituent of low density lipopro-

    tein (LDL). Radicals do not dierentiate between

    PUFAs. In tissue and blood linoleic acid conjugates

    outweigh in number by far conjugates of arachidonic

    acid. Therefore the main amount of LPO products

    obtained after oxidation of LDL originates from li-

    noleic acid and not from arachidonic acid. Linoleic

    acid oxidation products were detected, for instance,

    in minimally modied low density lipoprotein

    (MMLDL) [14] which is still recognized by theLDL cell receptor [15]. Thus oxidized linoleic acid

    products are introduced into endothelial cells with

    the consequence that some of these products might

    contribute to damage these cells from inside [14].

    Identication of LPO products in biological uids

    is very dicult, since they are generated in low

    amounts only, compared to the huge amounts of

    other tissue constituents. Their detection is facili-

    tated, if their chemical properties are known. These

    properties have been investigated by model reac-

    tions: when pure PUFAs are subjected to LPO

    much higher amounts of LPO products are obtained

    than by biological reactions. The reaction products

    are separated after derivatization by gas chromatog-

    raphy (GC). Identication is achieved by mass spec-

    trometry (MS). Such investigations revealed genera-

    tion of a great variety of products [16^23].

    LPO processes are induced in germinating seeds

    [24^26], and by injury of plant material [27]. LPO

    products are involved in phytoalexin production

    [7,28] and plant aging [29]. LOOH are generated in

    mammals by injury [3,11,30], and are apparently in-

    volved in aging [31,32]. Thus LPO processes seem tobe the consequence of a change in the cell membrane

    structure.

    In addition LPO processes are involved in the de-

    composition of fat [33^37].

    Minor degradation products of LPO processes are

    carbonyl compounds. Some of these develop potent

    physiological properties, for instance, prostaglandins

    [38], derived from arachidonic acid in mammals, and

    jasmonic acid, derived from linolenic acid, in plants

    [39^42].

    In contrast the generation and properties of car-

    bonyl containing LPO products derived from linoleic

    acid have not yet been investigated so well. Linoleic

    acid is the most abundant polyunsaturated fatty acid

    in plants and mammals. Its LPO products are gen-

    erated compared to LPO products of other PUFAs

    in much larger amounts. Some of the carbonyl com-

    pounds derived from linoleic acid develop important

    physiological properties. 4-Hydroxy-2-nonenal (4-

    HNE) and 2,4-decadienal [43^45] are highly toxic

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 189

  • 7/27/2019 Alde Hide

    3/21

    to mammalian cells, in plants 4-HNE inhibits growth

    of pathogens [46], 2-alkenals are also genotoxic [47]

    and inhibit growth of fungal cultures [46]. Hexanal

    and other aldehydes decrease seed germination [48].

    Thus a re-examination of degradation processes of

    linoleic acid hydroperoxides seemed of interest.

    Precursor molecules of linoleic acid LPO products

    are 13-hydroperoxy-9,11-octadecadienoic acid (13-

    HPODE) and 9-hydroperoxy-10,12-octadecadienoic

    acid (9-HPODE), generated either by enzymic or

    nonenzymic processes (Scheme 1).

    In order to investigate the decomposition products

    of HPODEs we subjected puried 9S-HPODE (not

    so easily available as 13S-HPODE) and 13S-HPODE

    to air oxidation catalyzed by iron ions and investi-gated the products after trapping with pentauoro-

    benzylhydroxylamine (PFBHA) [49,50]. Moreover

    we subjected primary generated LPO products of

    13S-HPODEs, 2,4-decadienal and 4,5-epoxy-2-dece-

    nal, to further degradation in order to facilitate the

    recognition of their fate in the course of multistep

    degradation reactions. The results of these investiga-

    tions are reported in this paper.

    2. Materials and methods

    2.1. Materials

    2E,4E-Decadienal, 2-butenal, m-chloroperbenzoic

    acid, hexadecanol, octadecanol, linoleic acid, tri-

    methyl phosphite, dimethyl sulde and soybean

    lipoxygenase (lipoxidase) were purchased from Fluka

    (Neu Ulm, Germany); formic acid p.a., hydrochloric

    acid 32%, sodium hydroxide p.a., sodium sulfate p.a.

    and sodium hydrogen carbonate were bought from

    Merck (Darmstadt, Germany). N-Methyl-N-(tri-

    methylsilyl)triuoroacetamide (MSTFA) was ob-

    tained from Macherey-Nagel (Du ren, Germany).

    Solvents (Jackle Chemie, Nu rnberg, Germany) weredistilled before use.

    2.2. Gas chromatography

    Gas chromatography was carried out with a Carlo

    Erba Instruments HRGC 5160 Mega Series gas chro-

    matograph equipped with a ame ionization detector

    using a DB-1 fused silica capillary column (30

    mU0.32 mm i.d.; JpW Scientic, Germany), cov-

    ered with a 0.1 Wm layer of liquid phase. The temper-

    ature of the detector was kept at 290C, the injector

    temperature was 280C. Injection volume: 0.2^0.7 Wl

    of a 1% (m/v) solution of the residue obtained after

    Fe2 catalyzed air oxidation of 9- and 13-HPODE

    and appropriate derivatization (see later). Tempera-

    ture program: 5 min isotherm at 60C, 3C/min 60^

    280C, then 15 min isotherm at 280C. The carrier

    gas was hydrogen and the splitting ratio 1:30. Peak

    integration and data recording was done with a

    Merck Hitachi Chromatointegrator D-2500.

    Kovats indices were determined by co-injection of

    a standard mixture of saturated straight chain alka-

    nes (C10-C36) [51].

    2.3. Gas chromatography/mass spectrometry

    GC/MS was performed on a Finnigan MAT 95,

    equipped with an EI ion source operating at 70 eV.

    A Hewlett-Packard 5890 Series II gas chromato-

    graph with a fused silica DB-5 glass capillary column

    (30 mU0.25 mm i.d., covered with a layer of 0.1 Wm

    of liquid phase; JpW Scientic) was used for sample

    separation. The injector temperature was kept at

    280C, injection volumes were 0.6^1.5W

    l of a 1^2%

    (m/v) solution of the residue obtained after air oxi-

    dation of linoleic acid and appropriate derivatization

    (see later). Temperature program: 3 min isotherm

    50C, within 2 min increase to 100C, then 3C/min

    to 300C, nally 10 min isotherm at 300C.

    2.4. NMR

    NMR measurements were carried out with a

    Scheme 1. Generation of 9-HPODE and 13-HPODE.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208190

  • 7/27/2019 Alde Hide

    4/21

    Bruker AMX 600 and a Bruker ARX 300 instrument

    with solvent peak as internal reference (CDCl3 :

    NH7.26, NC77.7).

    2.5. HPLC

    Analytical normal phase HPLC (NP-HPLC):

    pump system Beckman System Gold with program-

    mable solvent module 126, detector: Beckman Sys-

    tem Gold diode array detector 168. Column: Bis-

    cho, Nucleosil 3 Wm (25 cmU4.6 mm i.d.).

    Precolumn: Bischo spherisorb 5 Wm Si/NP (2

    cmU0.4 mm).

    Preparative NP-HPLC: Beckman System Gold

    with programmable solvent module 125, program-

    mable detector module 166. Column: Bischo Ultra-

    sep FS 100 (6 Wm), 25 cmU20 mm i.d., precolumn:

    Bischo Ultrasep (2 cmU20 mm i.d.).

    2.6. Preparation of 9S-hydroperoxy-10,12-

    octadecadienoic acid (9S-HPODE) and

    13S-hydroperoxy-9,11-octadecadienoic acid

    (13S-HPODE)

    9S-HPODE and 13S-HPODE were prepared ac-

    cording to Gardner [52] by reacting linoleic acid at

    pH 6 with soybean lipoxygenase. Purication of 9S-

    HPODE and 13S-HPODE was achieved by NP-HPLC [53,54]. The mixture (1014 mg) consisting of

    linoleic acid and 9- and 13-HPODE ^ obtained by

    soybean lipoxygenase (lipoxidase) catalyzed O2 oxi-

    dation of 1090.8 mg (3.9 mmol) linoleic acid ^ was

    dissolved in 35 ml n-hexane and 7 ml 2-propanol.

    1 ml of this mixture, containing 24.1 mg of the mix-

    ture, was brought onto a preparative silica gel HPLC

    column (Bischo, Ultrasep FS 100, d=2 cm, l= 25

    cm, partial diameter 6 Wm; precolumn: same pack-

    age, d=2 cm, l= 5 cm. Solvent mixture A : n-hexane:

    acetic acid 1000:1; mixture B: n-hexane:acetic acid:

    isopropanol 950:1 :50; relation A:B = 76:24. Flow

    rate 15 ml/min; UV absorbtion 234 nm. Four

    fractions were obtained eluting between 22 min

    and 35 min: fraction 1 (Rt = 22 min), 13S-hydro-

    peroxy-9Z,11E-octadecadienoic acid; fraction 2

    (Rt = 25 min), 13S-hydroperoxy-9E,11E-octadecadi-

    enoic acid; fraction 3 (Rt = 30 min), 9S-hydroper-

    oxy-10E,12Z-octadecadienoic acid; fraction 4

    (Rt = 35 min), 9S-hydroperoxy-10E,12E-octadecadi-

    enoic acid. The separation was repeated ten times.

    After ten runs the column was rinsed with 100%solvent A and equilibrated. Yield (ten collections):

    97.3 mg 13S,9Z,11E-HPODE (34.4%), 41.4 mg

    9S,10E,12Z-HPODE (14.6%).

    2.7. Incubation of 13S-9Z,11E-HPODE and

    9S-10E,12Z-HPODE by Fe2+ catalysis

    Decomposition of HPODEs was carried out as

    described recently [55].

    Briey 5 mg (18 Wmol) 13S-HPODE (respectively

    9S-HPODE) were emulsied in 7.5 ml of 0.1 M

    phosphate buer (pH 7.4) and 15 ml 0.2 M KCl

    solution at 37C. The reaction mixture was stirred

    vigorously. Then decomposition was started by addi-

    tion of 5 mg (18 Wmol) FeSO4W7H2O. The solution

    was stirred at room temperature in an air atmos-

    phere. In other experiments the amount of Fe-

    SO4W7H2O was reduced to 1/10 respectively 1/100 of

    the amount of HPODEs.

    After 30 min a 2 ml aliquot was withdrawn. In

    order to reduce hydroperoxides and to stop radical

    induced reactions 100 Wl dimethyl sulde (respec-

    tively trimethyl phosphite) were added. After shakingfor 5 min 200 Wl 0.05 M PFBHAWHCl in methanol

    were admixed followed by stirring in a N2 atmos-

    phere for 3 h. Two drops of HCl (5%, p.a.) were

    added, then the solution was extracted twice each

    with 1 ml of chloroform. The CHCl3 layer was sep-

    arated and dried with Na2SO4. Na2SO4 was removed

    by decantation, the solvent was blown o by a ow

    of nitrogen at room temperature.

    The residue was treated with an excess of etheral

    diazomethane solution for 30 s [56] (according to

    [56], addition of about 10% methanol greatly im-

    proves the eciency of diazomethane methylation).

    The residue obtained after removal of solvent and

    excess of diazomethane in a stream of N2 was re-

    acted with 50 Wl MSTFA at room temperature for

    30 min. 1.0 Wl of this solution was used for GC

    analysis, 1.5 Wl for GC/MS analysis.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 191

  • 7/27/2019 Alde Hide

    5/21

    2.8. Generation of 1-(trimethylsiloxyhexyl)-

    phosphonic acid dimethyl ester by addition of

    trimethyl phosphite to hexanal followed by

    trimethylsilylation with MSTFA

    100 mg (1 mmol) hexanal, 124 mg (1 mmol) tri-

    methyl phosphite and 18 mg (1 mmol) H2O were

    mixed and allowed to stand at room temperature.

    After 10 min 5 ml of H2O were added. The reaction

    product was extracted twice with each 5 ml of ethyl

    acetate (EA), dried over Na2SO4 and the organic

    solvent was removed by evaporation in vacuo. Yield:

    184 mg (87.6%).

    The reaction product was puried on a TLC plate

    (20U20 cm covered with 1 mm of silica gel 60 PF254,

    Merck), solvent EA, Rf = 0.25. Yield of (1-hydroxy-

    hexyl)phosphonic acid dimethyl ester: 110 mg

    (52.4%).1H-NMR (600 MHz, CDCl3) : N= 0.88 (t, J=7.2

    Hz, 3H, 6-CH3), 1.26^1.36 (m, 4H, 4-CH2, 5-CH2),

    1.38 (m, 1H, 3-CH2), 1.61 (m, 1H, 3-CH2), 1.70 (m,

    1H, 2-CH2), 1.75 (m, 1H, 2-CH2), 3.81 (d,3J

    (P,H) = 10.2 Hz, 6H), 3.90 (ddd, 2J (P,H) = 10.2 Hz,

    J=4.8 Hz, J= 4.8 Hz, 1H, 1-CH). 13C-NMR (75.6

    MHz, CDCl3) : N= 14.66 (6-C), 23.13 (5-C), 25.94

    (d, 3J (C,P) = 13.6 Hz, 3-C), 31.98 (2-C or 4-C),

    32.09 (2-C or 4-C), 53.98 (d, 2J (C,P) = 7.6 Hz,

    1-C), 68.45 (d, 1J (C,P) = 160.3 Hz, 1-C).Trimethylsilylation of this product was achieved

    by adding 50 Wl of MSTFA. The solution was kept

    at room temperature for 15 min and was then inves-

    tigated without removal of excess reagent by GC/

    MS.

    If the reaction is carried out without addition of

    water even after 2 days standing at room tempera-

    ture besides the expected addition product great

    amounts of the MSTFA adduct of hexanal were ob-

    tained, indicating that water catalysis is required for

    fast reaction.

    2.9. Synthesis of 4-hydroxy-2-nonenal by oxidation of

    2E,4E-decadienal with m-chlorobenzoic acid

    1064 mg (7.0 mmol) of 2E,4E-decadienal and 1330

    mg (7.7 mmol of m-chloroperbenzoic acid (100%)

    were dissolved in 10 ml dry methylene chloride.

    The solution was stirred for 3 h at room tempera-

    ture. The main amount of m-chlorobenzoic acid was

    removed by extraction with a saturated NaHCO3 so-

    lution. The remaining organic solvent was removed

    applying vacuum. The residue was dissolved in

    CH2Cl2 and separated by thin layer chromatography

    (solvent CH:EA, 3 :1; Rf = 0.23). Yield: 399.4 mg(37.5%).

    2.10. Oxidation of 2E,4E-decadienal with

    m-chloroperbenzoic acid (MCPBA) and

    trapping of formic acid

    17.3 mg (0.114 mmol) 2,4-decadienal dissolved in

    9 ml dry methylene chloride and 15.7 mg (0.091

    mmol) dry MCPBA dissolved in 1 ml dry methylene

    chloride were stirred for 3 h at room temperature.

    Then 27.6 mg (0.114 mmol) hexadecanol were added.

    The solution was kept at room temperature for

    5 days. 10 Wl MSTFA were added to a 1/100 aliquot

    of the solution (100 Wl). The reaction mixture was

    kept overnight at room temperature, then it was in-

    troduced into the GC/MS without removal of the

    solvent. The GC/MS investigation revealed besides

    the presence of the trimethylsilyl (TMS) ester of

    m-chlorobenzoic acid (retention index (RI) = 1376)

    small amounts of trimethylsilylated 4-hydroxy-2-

    nonenal (RI = 1392); RI and mass spectrum turned

    out to be identical with literature data [50]. Rather

    large peaks indicated trimethylsilylated adducts of4-hydroxy-2-nonenal (RI1 = 1754 and RI2 = 1759).

    RI and mass spectra were identical with published

    data [50]. A peak at RI = 1918 was recognized to

    represent according to its mass spectrum hexadecyl

    formate. A peak at RI = 1970 corresponded to the

    TMS ether of unreacted hexadecanol.

    Exchange of hexadecanol versus octadecanol in

    the reaction sequence provided the corresponding

    octadecyl formate (RI = 2121).

    Structures of hexadecyl formate respectively octa-

    decyl formate were proven by reaction of formic acid

    with hexadecanol (respectively octadecanol) and run-

    ning GC/MS. MS and RI proved identical with the

    products obtained after reacting 2,4-decadienal with

    MCPBA.

    2.11. Decomposition of 2E,4E-decadienal by stirring

    at room temperature

    30 Wl 2E,4E-decadienal were suspended in 30 ml

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208192

  • 7/27/2019 Alde Hide

    6/21

    0.05 M phosphate buer (pH 7.4). The suspension

    was stirred at room temperature in an atmosphere of

    air; seven 1 ml samples were withdrawn in 1 h time

    intervals. One sample was withdrawn after 24 h.

    The solution of the removed samples was acidiedwith two drops of 5% aqueous HCl to pH 2. The

    aqueous solution was extracted with 1 ml ether, the

    etheral layer was dried with Na2SO4 and ltered.

    Ether was blown o by a stream of N2. 200 Wl of a

    0.05 M solution of PFBHAWHCl in Tris buer and 50

    Wl of CH3OH were added. The solution was kept

    overnight at room temperature, then 1 ml water

    was added. The aqueous solution was extracted

    with 1 ml of a mixture of CH:AE, 4:1 and dried.

    50 Wl of etheral CH2N2 solution were added and

    solvent was removed under N2. After adding of 50

    Wl EA the mixture was investigated by GC/MS.

    Already after 2 h stirring pentauorobenzyloxime

    (PFBO) derivatives of 2E,4E-decadienal, 2-butenal

    and n-hexanal were recognized by corresponding

    small peaks. After 24 h 2-butenal and n-hexanal cor-

    responded to the main peaks. One double peak

    (RI1 = 2480, RI2 = 2485) was ascribed to the di-

    PFBO derivative of 5-oxodecanal. Small peaks indi-

    cated the presence of formaldehyde and ethylmethyl-

    ketone (apparently impurities), and also of acetalde-

    hyde, which might be generated in the course of a

    retroaldol reaction of butenal or of 2E,4E-decadie-nal.

    If the reaction mixture was stirred at 37C and if

    removed samples were reduced and PFBHAWHCl was

    added, we were unable to detect 2-butenal. We rec-

    ognized that stirring 2-butenal in an air atmosphere

    or nitrogen generates 3-hydroxybutanal (water addi-

    tion). Thus butenal is further decomposed.

    2.12. Synthesis of 4,5-epoxy-2-decenal

    18.1 mg (0.12 mmol) 2,4-decadienal were treated

    with a 0.07 molar solution of 2,2-dimethyldioxirane

    in acetone at 4C [57]. The solution was kept for 4 h

    at 4C. After removal of the solvent at 30C the

    residue was puried by thin layer chromatography

    (ve plates, Macherey-Nagel, Polygram SIL

    6/UV254) using a mixture of CH:EA, 3:1. Rf = 0.57,

    RI = 1336. Yield: 8.1 mg (40.5%). RI and MS were

    identical with published data [50].

    2.13. Saponication of 4,5-epoxy-2-decenal and

    trimethylsilylation of products

    11.8 mg (0.07 mmol) 4,5-epoxy-2-decenal sus-

    pended in 1 ml HCl p.a. (5%) were stirred for 5 minat room temperature in a N2 atmosphere. The solu-

    tion was extracted twice with 1 ml EA, then the

    organic solution was dried over Na2SO4. After re-

    moval of the solvent in a stream of nitrogen 50 Wl

    MSTFA were added. The solution was kept for 1 h

    at room temperature, then a sample of the MSTFA

    solution was injected without removal of the reagent

    in the GC/MS combination.

    The gas chromatogram showed several reaction

    products; the main product (RI = 1693) turned out

    to be 4,5-ditrimethylsiloxy-2E-decenal. In addition a

    small peak (RI = 1687) corresponded to its Z-isomer.

    A second main product was identied to be 5-tri-

    methylsilyloxy-4-chloro-2-decenal (RI = 1787) gener-

    ated by addition of Cl3 instead of water. MS:

    m/z =348 (M, 8), 333 (5), 313 (22), 279 (45), 277

    (100), 259 (2), 225 (2), 189 (4), 169 (3), 167 (4), 147

    (13), 115 (5), 93 (7), 73 (65), 45 (6).

    Some starting material was also detected due to

    incomplete reaction in the form of its MSTFA ad-

    duct (RI = 1807) [50].

    2.14. Oxidation of 4,5-epoxy-2-decenal withm-chloroperbenzoic acid

    26.0 mg (0.155 mmol) of puried 4,5-epoxy-2-de-

    cenal and 26.7 mg (0.155 mmol) of MCPBA 100%

    (from 48.5 mg MCPBA 55%) were dissolved in 10 ml

    of dry methylene chloride and stirred for 3 h under

    N2 atmosphere. After this time 1/10 aliquot (1 ml)

    was withdrawn and solvent was removed under N2.

    Then 20 Wl of MSTFA were added and kept at room

    temperature for 24 h. Then the sample was investi-

    gated by GC/MS. The GC/MS investigation revealed

    besides the presence ofm-chlorobenzoic acid trimeth-

    ylsilyl ester (M = 286, RI = 1376) and the MSTFA

    adduct of the starting material (M = 367, RI = 1807)

    generation of 4,5-epoxy-2-decenoic acid trimethylsilyl

    ester (RI = 1639, M = 256, m/z =256 (M, 1), 241

    (4), 225 (4), 173 (2), 169 (8), 156 (47), 139 (5), 129

    (4), 113 (6), 103 (6), 81 (7), 75 (44), 73 (100), 55 (9),

    43 (9), 41 (18)). 4-HNE was produced in the form of

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 193

  • 7/27/2019 Alde Hide

    7/21

  • 7/27/2019 Alde Hide

    8/21

    Table 1

    Identied products after reaction of 9S-HPODE and 13S-HPODE with Fe2, followed by reduction with CH3SCH3

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 195

  • 7/27/2019 Alde Hide

    9/21

    Table 1. (continued).

    *The structure of this compound has been elucidated by Hamberg [62] after oxidation of linoleic acid by soybean lipoxygenase. We

    [55] obtained the same compound by iron ion induced air oxidation of linoleic acid, but ascribed to it falsely the structure of a tri-

    methylsilylated enol ether of a 13-hydroxy-9-oxo-11-octadecenoic acid [63]. We are indebted to Prof. Hamberg for his comments con-

    cerning this error.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208196

  • 7/27/2019 Alde Hide

    10/21

    besides 13S-HPODE also 9S-HPODE is obtained

    [52].

    Since we needed both hydroperoxides the mixture

    of 9S- and 13S-HPODE ^ prepared by incubation of

    soybean lipoxygenase with linoleic acid at pH 6.0 ^

    was separated by normal phase HPLC [53,54]. 9S-

    HPODE was thus obtained much faster and easier

    than by application of tomato lipoxygenase.

    HPODEs were then subjected to decomposition in

    an air atmosphere by iron ion catalysis. Progress of

    the reaction was checked by removal of samples in

    time intervals and after appropriate derivatization by

    GC and MS: when iron ions were added in molar

    amounts in respect to HPODEs oxidation started

    much faster than by reduction of the iron ion

    amount to 1/10 molar ratio and less (1/100 m). Since

    independently of the molar ratio of added iron ions

    always identical oxidation products were detectable

    in comparable ratios, we concluded that the amount

    of iron ions does not change the product pattern and

    therefore we felt allowed to reduce the reaction time

    by applying equimolar amounts of iron ions to start

    Fig. 3. GC run of the decomposition products of 13-HPODE after reduction with P(OCH3)3, short reaction with diazomethane and

    MSTFA. 1, H11C5-CH(OTMS)-CH= CH-OTMS; 2, H3COOC-(CH2)7-CHO; 3, H3COOC-(CH2)7-COOCH3 ; 4, H3COOC-(CH2)6-

    CH= CH-CHO; 5, H11C5-[CH(OTMS)]2-CH= CH-CHO; 6, H3COOC-(CH2)7-CH(OTMS)-N(CH3)-COCF3 ; 7, H3COOC-(CH2)7-

    CH= CH-CH = CH-CHO; 8, H3COOC-(CH2)7-CH(OTMS)-CH= CH-CHO; 9, H3COOC-(CH2)7-CH(OTMS)-PO(OCH3)2.

    Fig. 2. Mass spectrum of the bis-PFBO derivative obtained from 4-oxo-2-nonenal.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 197

  • 7/27/2019 Alde Hide

    11/21

  • 7/27/2019 Alde Hide

    12/21

    are reduced by iron ions before addition of dimethyl

    sulde.

    Further major dierences concern epoxyhydroxy

    acids in which the hydroxy group is anked by the

    epoxy group and the double bond: 9-HPODE gen-

    erates preferentially 9,10-epoxy-11-hydroxy-12-octa-

    decenoic acid (peak 16) while 13-HPODE is decom-

    posed to 12,13-epoxy-11-hydroxy-9-octadecenoicacid (peak 15).

    Epoxyhydroxy acids in which the epoxy and the

    hydroxy group are separated by a double bond were

    not detected: according to Gardner [9,67] these ep-

    oxyhydroxy acids are readily hydrolyzed due to the

    allylic hydroxylic group to a mesomeric carbocation,

    thus giving a mixture of 9,12,13-trihydroxy-10-octa-

    decadienoic acids (peak 23 and 24) and 9,10,13-tri-

    hydroxy-11-octadecadienoic acids (peak 21 and 22),

    independent of whether 9- or 13-HPODEs are de-

    composed. Therefore these trihydroxy acids (peaks21^24) were generated in comparable amounts.

    Additional (minor) dierences in the chromato-

    grams (Fig. 1) concern methylates of oxodienoic

    Fig. 4. Gas chromatogram of the compounds obtained after stirring 2,4-decadienal in the absence iron ions in an air atmosphere. The

    aldehydic products were trapped by reaction with pentauorobenzylhydroxylamine. 1, H3C-CH = CH-CH = N-O-CH2-C6F5 ; 2, H3C-

    (CH2)4-CH= N-O-CH2-C6F5 ; 3, H3C-(CH2)4-CH = CH-CH = CH-CH = N-O-CH2-C6F5 ; 4, H3C-(CH2)4-C(N-OCH2-C6F5)-(CH2)3-

    CH = N-O-CH2-C6F5.

    Fig. 5. Mass spectrum of bis-PFBO derivative of 5-oxodecanal.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 199

  • 7/27/2019 Alde Hide

    13/21

    acids. In the chromatogram obtained from 9S-

    HPODE peaks are recognized which correspond to

    methyl esters of E/Z-isomeric 9-oxo-10,12-octadeca-

    dienoic acids (9-KODEs, peaks 26 and 27), peaks

    corresponding to 13-oxo-9,11-octadecadienoic acids

    (13-KODEs, peaks 28 and 29) were detectable only

    in traces, while in the chromatogram obtained from

    13S-HPODE nearly exclusively methylates of 13-

    KODEs are recognized, and 9-KODEs are nearly

    absent. In addition 2-hydroxyaldehydes 2-hydroxy-

    heptanal (peak 4, mainly generated from 13-

    HPODE) and 9-hydroxy-10-oxodecanoic acid (peak12, mainly generated from 9-HPODE) are distin-

    guished considerably in quantitative respect in both

    GC runs.

    Thus the main degradation products, HODEs and

    epoxy-11-hydroxy acids, generated by introduction

    of oxygen at the carbon adjacent to the epoxy group,

    as well as some minor carboxylic products (KODEs

    and K-hydroxyaldehydes) have largely preservedtheir regiospecicity.

    The main aldehydic compounds hexanal (peak 3)

    and 9-oxononanoic acid (peak 10) were produced in

    comparable amounts. Also 2,4-decadienal (peak 8) or

    13-oxotridecadienoic acid, primary decomposition

    products of HPODEs, were obtained in nearly equal

    yield.

    In addition we detected dicarbonyl compounds:

    bis-PFBO derivatives of E,Z-isomeric 9,12-dioxo-

    10-dodecenoic acids, recently identied in lentils

    [68], and their analogues, 4-oxo-2-nonenals (peak

    18) [69]. The mass spectrum of the bis-PFBO deriv-

    ative of 4-oxo-2-nonenal is reproduced in Fig. 2.

    The mass spectra (Fig. 1) of all mentioned aldehy-

    dic PFBO derivatives have been published elsewhere

    [50].

    The main decomposition products of HPODEs are

    trihydroxyenoic acids. They are derived by hydrolysis

    of corresponding epoxyhydroxy acids, well known

    since a long time ago [23,37]. The GCs indicate

    that E/Z-isomers of 9,10,13-trihydroxy-11-octadece-

    noic acid (peaks 20 and 23) and 9,12,13-trihydroxy-

    10-octadecenoic acid (peaks 21 and 22) are generatedfrom 9- and 13-HPODE in nearly equal amounts.

    When HPODEs were stirred for an extended peri-

    od of time (24 h) in an air atmosphere peaks not

    Scheme 4. Generation of the characteristic degradation products

    in the mass spectrum of the bis-PFBO derivative of 5-oxodeca-

    nal.

    Fig. 6. Mass spectrum of the bis-PFBO derivative obtained from 2-butenedial.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208200

  • 7/27/2019 Alde Hide

    14/21

  • 7/27/2019 Alde Hide

    15/21

    3.1. Synthesis of 4,5-epoxy-2-decenal

    The structure deduced by the mass spectrum of

    4,5-bistrimethylsilyloxy-2-decenal required proof by

    synthesis. 2E,4E-Decadienal, an easily commercially

    available compound, was selected as starting material

    for synthesis of 4,5-epoxy-2-decenal which should be

    then hydrolyzed to 4,5-dihydroxy-2-decenal.

    Since 4,5-epoxy-2-decenal was obtained with the

    common epoxidation reagent, m-chloroperbenzoic

    acid, in poor yield only, 2E,4E-decadienal 8 was re-

    acted with Adams reagent, dimethyldioxiran [57] (seeScheme 2).

    The obtained 4,5-epoxy-2-decenal was hydrolyzed

    with aqueous hydrochloric acid, then the reaction

    mixture was treated with MSTFA. Short treatment

    produced mainly 4,5-bistrimethylsilyloxy-2-decenal

    besides 4-chloro-5-hydroxy-2-decenal (see Scheme

    2), while prolonged treatment resulted in an increase

    in the adducts as deduced by GC/MS measurements.

    Small amounts of 4,5-epoxy-2-decenal have been

    detected by Grein et al. after stirring 2,4-decadienal

    in an atmosphere of pure oxygen [71]. In addition

    generation of 4-HNE was also reported [71].

    4-HNE has one carbon less than 2,4-decadienal.

    Considering that the carbonyl group of an aldehyde

    might undergo a Criegee rearrangement [70,72], we

    supposed that 4-HNE might have been generated by

    an epoxidation reaction of 2,4-decadienal to 4,5-ep-

    oxy-2-decenal, followed by addition of oxygen to the

    aldehydic group and formation of a peroxyl inter-

    mediate. Hydroperoxides are well known to suer a

    Criegee rearrangement [73]. The reaction product

    was expected to be hydrolyzed to 4-HNE (see

    Scheme 3).

    Stirring 2,4-decadienal in an atmosphere of air

    (not oxygen) generated 4,5-epoxydecenal only intraces. In contrast rather large quantities of 4-HNE

    were detected by GC/MS after addition of m-chloro-

    perbenzoic acid.

    In the course of the supposed reaction sequence

    (see Scheme 3) formic acid should be generated.

    The identication of small amounts of formic acid

    is dicult by GC or GC/MS due to its low molecular

    weight and high volatility. Therefore we tried to trap

    the expected formic acid by transforming it into a

    less volatile, easily detectable derivative ^ hexadecyl

    formate ^ by addition of hexadecanol. When 2,4-

    decadienal was treated with MCPBA in the presence

    of hexadecanol and the reaction mixture was sub-

    jected after trimethylsilylation to a GC/MS investi-

    gation we identied hexadecyl formate besides 4-

    HNE and its MSTFA adduct. When instead of hex-

    adecanol octadecanol was added octadecyl formate

    was obtained, corroborating the proposed degrada-

    tion mechanism.

    In order to prove that 4,5-epoxy-2-decenal is an

    intermediate on the reaction path from 2,4-decadie-

    nal to 4-HNE we subjected 4,5-epoxy-2-nonenal to

    oxidation with MCPBA. After 3 h reaction at roomtemperature, besides starting material only 4,5-ep-

    oxy-2-nonenoic acid was obtained, thus proving

    that generation of 4-HNE starts with a rearrange-

    ment which is followed by epoxidation (see Scheme

    6) and not by a primary epoxidation as indicated in

    Scheme 3.

    4,5-Epoxy-2-decenal acid turned out to be rather

    stable to air oxidation in the presence of iron ions.

    After 2 days standing at room temperature only a

    small amount of products of high molecular weight

    (not yet identied) were detected.

    3.2. Decomposition of 2,4-decadienal to butenal,

    hexanal, 5-oxodecanal and buten-1,4-dial

    When 2,4-decadienal was stirred overnight in an

    air atmosphere without addition of m-chloroperben-

    zoic acid three main aldehydic products were recog-

    nized besides still present starting material by GC/

    MS after trapping the aldehydic products by addi-

    Scheme 5. Generation of the dimethyl phosphite addition prod-

    ucts to hexanal, identied after reaction with MSTFA.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208202

  • 7/27/2019 Alde Hide

    16/21

    tion of pentauorobenzylhydroxylamine: hexanal,

    butenal and 5-oxodecanal. In addition buten-1,4-

    dial was generated in small amounts (Fig. 4).

    5-Oxodecanal was identied according to the char-

    acteristic mass spectrum of its PFBO derivative (Fig.5). The molecular ion of even mass (m/z 560) ^ cor-

    roborated by ions at M-212 (m/z 348) and M-197 (m/

    z 363) ^ indicated the presence of two nitrogen atoms

    corresponding to derivatization of the original com-

    pound with two molecules of pentauorobenzyl-

    amine. A peak at m/z 321 (560^239) was explained

    by assuming a McLaerty rearrangement while the

    ion m/z 238 indicated cleavage of the allylic bond

    (see Scheme 4).

    2-Buten-1,4-dial was identied by the mass spec-

    trum (Fig. 6) of its bispentauorobenzyloxime deriv-

    ative: again the even molecular weight indicated the

    reaction with two molecules of pentauorobenzylhy-

    droxylamine. When the mass of two CH = N-O-CH2-

    C6H5 groups (448) is subtracted from the molecular

    ions (474) the remaining mass (26) corresponds to a

    CH = CH group, thus indicating the structure OHC-

    CH = CH-CHO for the original molecule.

    3.3. Decomposition of 2-butenal by stirring in an air

    atmosphere

    2-Butenal contains an K,L-unsaturated doublebond; therefore it was expected to be able to under-

    go similar reactions as 2E,4E-decadienal. We de-

    tected, indeed, after stirring 2-butenal for 2 h at

    room temperature, followed by reaction with

    PFBHA and trimethylsilylation, after GC/MS analy-

    sis 3-trimethylsilyloxybutanal.

    3.4. Generation of artifacts by reduction with

    P(OCH3)3

    In GCs obtained by reduction with P(OCH3

    )3

    (but

    not with (CH3)2S) additional peaks were observed

    (see Fig. 3). After treatment with pentauorobenzyl-

    hydroxylamine for instance a pair of products with

    peaks at mass 167 and 182 (Figs. 7 and 8) was de-

    tected.

    The molecular weights of the compounds were de-

    duced by M-15 ions to be 282, respectively 368. The

    presence of a fragment with mass 173 (A) in the one

    and 259 (B) in the other spectrum indicated the pres-

    ence of a H11C5-CH=OTMSk (A) respectively a

    H3COOC-(CH2)7-CH = OTMSk (B) residue. Gener-

    ation of the peaks at m/z 182 in both spectra was

    visualized as loss of C5H11CHO, respectively of

    H3COOC-(CH2)7-CHO, interpreted by migration ofthe TMS residue at a CH-OTMS group to a carbon-

    yl function in the rest of the molecule [74]. After

    subtraction of 73 mass units (SiCH3)3) from 182, a

    fragment of 109 (182373) mass units remained. In

    addition in one of the mass spectra (Fig. 8) also a

    positively charged ion of mass 109 was obtained.

    High resolution measurements allowed to deduce

    its molecular formula to be C2H6PO3. Thus the com-

    pounds with m/z 282 respectively m/z 368 were de-

    duced to have been generated by reaction of

    P(OCH3)3 with the carbonyl group of the aldehyde

    followed by reaction with MSTFA (Scheme 5).

    Therefore the compound with molecular weight

    282 was deduced to be a reaction product of

    P(OCH3)3 with hexanal, already described in the lit-

    erature [75], and that of molecular weight 368 with 9-

    oxononanoic acid. Similar small peaks corresponded

    to other aldehydes which represented analogous ad-

    dition products.

    Scheme 6. Generation of 4-HNE and pentylfuran from 2,4-

    decadienal.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 203

  • 7/27/2019 Alde Hide

    17/21

    4. Discussion

    Chan et al. [76] have observed the generation of

    pairs of products when they thermally decomposed

    either 9- or 13-HPODE in the inlet line of a GC.

    They explained this fact by a fast conversion of the

    9- to the 13-hydroperoxide and vice versa. In con-

    trast to Chan we reduced the hydroperoxides prior to

    gas chromatography excluding rearrangement of the

    peroxyl group. In addition we protected the func-

    tional groups by derivatization. Therefore isomeriza-

    tion in the GC is excluded.

    Using HPODEs, labeled with 18O in the peroxyl

    group [77], Chan proved a rearrangement of 9-

    HPODE to 13-HPODE and vice versa. Our investi-gations using also 18HPODEs conrmed these re-

    sults. We detected both labeled oxygen atoms in

    the resulting epoxyhydroxy acids in amounts of

    about 70%. Such retention of both oxygens in epox-

    yhydroxy acids was reported previously in enzymic

    reactions [78] or in reactions catalyzed by hematin

    [79]. The retention of both labeled oxygens was ex-

    plained by Dix and Marnett [79] by assuming a `re-

    bound' mechanism, in the course of which the peroxybond is cleaved, and one oxygen is transferred to

    hematin, followed by retransfer of a hydroxy group

    to the acid. Our experiments demonstrate that the

    oxygen transfer neither needs an enzyme nor requires

    hematin but just the presence of iron ions (or prob-

    ably other bivalent heavy metal ions). We interpret

    the identication of pairs of reaction products (these

    represent only minor LPO products) ^ independently

    of whether 9S-HPODE or 13S-HPODE is used as

    starting material ^ by formation of an identical in-

    termediate from 9- and 13-HPODE. We speculate

    that this intermediate is a carbocation, generated

    by the formation of a complex bond between the

    oxygen of the peroxyl group and the iron ion induc-

    ing loss of the peroxyl groups. This intermediate

    should resemble the radical which Chan et al. have

    proposed to explain the conversion of 9- to 13-

    HPODE and vice versa [77]. Isomeric 12,13-epoxy-

    9-hydroxy-10-octadecenoic acids and 9,10-epoxy-13-

    hydroxy-11-octadecenoic acids are the main LPO

    products of linoleic acid. Their genesis has been

    studied in detail [9,67,80,81]. They are assumed to

    be derived by reduction of the corresponding hydro-peroxides to alkoxyl radicals followed by rearrange-

    ment to epoxyallylic radicals [16,82]. In a second, less

    prominent degradation path, isomeric 9,10-epoxy-13-

    hydroperoxy-11-octadecenoic acids are degraded to

    Scheme 8. Generation of 2-butenal and hexanal by water addition to 2,4-decadienal followed by a retroaldol reaction. An alternative

    reaction induces generation of 5-oxodecanal.

    Scheme 7. Thermal decomposition of 12,13-epoxy-9-hydroper-

    oxy-10-octadecenoic acid generates nally 4,5-dihydroxy-2-dece-

    nal.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208204

  • 7/27/2019 Alde Hide

    18/21

    4,5-epoxy-2-decenal [70]. This compound ^ also a

    main volatile obtained by thermal decomposition of

    butter fat [83] ^ turned out to be a potent odorant

    [84]. Detection of 4,5-dihydroxy-2-decenal by Fe2

    catalyzed decomposition of HPODEs indicates that4,5-epoxy-2-decenal is not only generated by heating

    of 9,10-epoxy-13-hydroperoxy-11-octadecenoic acid

    but also by catalysis with iron ions.

    In order to prove that 4,5-epoxy-2-decenal repre-

    sents an intermediate in LPO processes it was syn-

    thesized by epoxidation of 2,4-decadienal: epoxida-

    tion with m-chlorobenzoic acid provided 4,5-epoxy-

    2-decenal in poor yield only. A much better yield was

    obtained by applying Adams catalyst, dimethyldiox-

    irane [57]. Hydrolysis of 4,5-epoxy-2-decenal revealed

    the identity of the obtained products with com-

    pounds identied in HPODE samples after stirring

    in an atmosphere of air by iron catalysis. This proves

    that 4,5-epoxy-2-decenal is generated in the course of

    nonenzymic LPO processes, and in fact we were able

    to trap 4,5-epoxy-2-decanal by reaction with PFBHO

    in the product mixture generated by stirring HPODE

    in air by Fe2 catalysis: The presence of 4,5-epoxy-2-

    decenal by air treatment of HPODEs in air probably

    escaped previously detection, since mass spectra of

    4,5-epoxy-2-decenal and also of its derivatives, for

    instance of its pentauorobenzyloximes, are not

    very typical [50]. In contrast, mass spectra of hydro-lysis products derived from 4,5-epoxy-2-decenal,

    show very characteristic spectra after appropriate de-

    rivatization. This enabled, for instance, detection of

    4,5-dihydroxy-2-decenal in homogenates of liver mi-

    crosomes [85] and indicates that 4,5-epoxy-2-decenal

    is generated in living matter too.

    When 2,4-decadienal was subjected to stirring in

    an air atmosphere 4-HNE was identied as trace

    product. Addition of m-chloroperbenzoic acid in-

    creased the yield of 4-HNE considerably. This obser-

    vation indicated involvement of a Criegee rearrange-

    ment in the generation of 4-HNE, as already

    suggested by Pryor [72]. Therefore we subjected

    2E,4E-decadienal to oxidation with m-chloroperben-

    zoic acid and obtained 4-HNE in good yield. This

    procedure represents the simplest method to generate

    4-HNE. Previously used methods started from 3Z-

    nonen-1-ol: 3Z-nonen-1-ol was epoxidized, then the

    alcoholic group was oxidized. Treatment with base

    aorded nally 4-HNE [86]. The involvement of a

    rearrangement reaction in the conversion of 2,4-deca-

    dienal to 4-HNE was proven by trapping formic

    acid, produced during the hydrolysis of the generated

    formic acid ester as intermediate.

    4-HNE is a toxic product identied in humanblood and tissue samples of atherosclerotic patients

    [44]. 4-HNE was found to be generated in plants by

    enzymic decomposition of 9-HPODE: 9-HPODE is

    assumed to be decomposed to 3-nonenal which is

    transformed by an alkenal oxidase to 2E-4-hydroper-

    oxy-2-nonenal. The latter is reduced by a peroxygen-

    ase to 4-HNE [87,88]. Similarly 13-HPODE is de-

    graded in plants enzymically to 9-hydroxy-12-oxo-

    10-dodecenoic acid (9-hydroxytraumatin) [89]. Here

    we observed that generation of 4-HNE and 9-hy-

    droxytraumatin does not require enzymes. Moreover

    the GCs indicate that 9-HPODE is the main precur-

    sor of 4-HNE and 13-HPODE of 9-hydroxytrauma-

    tin in the iron catalyzed decomposition of HPODEs.

    Double bonds react with peroxyl radicals by

    epoxidation [90]. Therefore we considered that 4-

    HNE might be generated via epoxidation of 2,4-dec-

    adienal to 4,5-epoxy-2-decenal by a Criegee reaction.

    This possibility is excluded, since we failed to trans-

    form 4,5-epoxy-2-decenal by treatment with hydro-

    peroxides, H2O2 or even MCPBA to 4-HNE: we

    obtained mainly 2,4-decadienoic acid. Thus 4-HNE

    is only generated starting from 2,4-decadienal, whenthe rearrangement of 2,4-decadienal precedes the ep-

    oxidation (Scheme 6). It remains an open question

    how much the reaction outlined in Scheme 6 contrib-

    utes to the generation of 4-HNE, since we detected

    2,4-decadienal after iron ion induced decomposition

    of 13-HPODE only in traces.

    4-HNE is easily cyclized to pentylfuran in acidic

    medium [14] or by heat. The latter compound is a

    common constituent in the volatiles of cooked meat

    [91] and plant oils [92]. It was also found after ther-

    mal degradation of methyl-12,13-epoxy-9-hydroper-

    oxy-10-octadecenoate [70]. 4,5-Epoxy-2-decenal is

    not transferred to 4-HNE but suers hydrolysis to

    4,5-dihydroxy-2-decenal (Scheme 7).

    When commercially available 2E,4E-decadienal

    was stirred in an air atmosphere in the absence of

    iron ions, dierent aldehydic products were recog-

    nized: 2,4-decadienal was decomposed mainly to

    2-butenal and hexanal (Fig. 4), as already reported

    by Swoboda [97] (Scheme 8). These authors inter-

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 205

  • 7/27/2019 Alde Hide

    19/21

    preted the generation of 2-butenal as retroaldol rear-

    rangement.

    2-Butenal was trapped after short stirring by reac-

    tion with PFBHA, but disappeared after stirring for

    several hours. In order to recognize the fate of2-butenal we subjected pure 2-butenal to stirring in

    an air atmosphere. Already after 2 h stirring at room

    temperature we were able to trap 3-hydroxybutanal,

    the expected intermediate in a retroaldol reaction.

    Thus stirring in air without addition of iron ions

    decomposes K,L-unsaturated aldehydes as well as

    2,4-dienals. Similar results were obtained by stirring

    2-butenal in a nitrogen atmosphere (excluding reac-

    tion with O2).

    The identication of 2-butenal as decomposition

    product of 2E,4E-decadienal by trapping is remark-

    able, since recently acrolein was reported as LPO

    product. Acrolein was identied by immunostaining

    [93]. Although generation of acrolein is possible by

    thermal decomposition of glycerols (derived from the

    glycerol part of the molecule), its formation by LPO

    of a PUFA is hardly deducible. Since aldehydes are

    trapped with high sensitivity in the form of penta-

    uorobenzyloximes and since we have not been able

    to recognize even a trace of acrolein, we assume that

    the immunostaining reaction of Uchida [93] concerns

    generally 2-unsaturated aldehydes and that 2-butenal

    or other 2-alkenals have been responsible for thepositive staining reaction.

    A 1,6-water addition to carbon 5 and to the car-

    bonyl oxygen of 2,4-decadienal generates 5-hydroxy-

    3-decenal, which can isomerize to either 5-hydroxy-2-

    decenal (which is decomposed to 2-butenal and hexa-

    nal via retroaldol reaction) or 5-hydroxy-4-decenal,

    which is enolized to 5-oxodecanal (Scheme 8).

    2-Hydroxyaldehydes have been detected in this

    and previous investigations [94,95]. They are not gen-

    erated by water addition to 4-HNE ^ as might be

    deduced considering the generation of butenal from

    2,4-decadienal ^ since 2-hydroxyheptanal obtained

    from 18O labeled HPODEs contained the labeled

    oxygen atom in the hydroxy group. 2-Hydroxyalde-

    hydes suer easily dehydration, especially if not de-

    rivatized at the OH groups. Therefore, by applying

    GC for separation of LPO products without protec-

    tion of OH groups [96,97], 2-heptene is formed from

    2-hydroxyheptanal as artifact and mimics the pres-

    ence of 2-heptene as LPO product.

    In addition to hexanal, butenal, 5-oxodecanal and

    the starting material (2,4-decadienal), a small double

    peak is recognized in GCs obtained after exposing

    2,4-decadienal to stirring in an air atmosphere (Fig.

    4). This double peak corresponds according to themass spectra to isomeric 2-butenedials (in the form

    of their PBFO derivatives).

    Their genesis might be visualized by attack of sin-

    glet oxygen at the 4,5-double bond of 2,4-decadienal

    by formation of a dioxetane. Its decomposition in-

    duces formation of 2-butenedial and hexanal.

    Recently Gardner [69] reported the isolation of

    4-oxo-2-nonenal after lipoxygenase oxidation of 3Z-

    nonenal. We detected the analogous compound 9,12-

    dioxo-10-dodecenoic acid in lentils [68]. Since we did

    not use enzymes in the experiments described in this

    paper the dioxo compounds might be derived as out-

    lined above by singlet oxygen attack to the 4,5-dou-

    ble bond of 9-oxo-10,12-octadecadienoic acid fol-

    lowed by cleavage of the oxetane bond. Likewise

    generation of 4-oxo-2-nonenal might be interpreted

    as addition of singlet oxygen to the 9,10-double bond

    in 13-KODE, followed by cleavage of the peroxyl

    bond and the C-C bond between carbon 9 and 10.

    5. Conclusion

    Evaluation of natural induced lipid peroxidation

    processes requires the investigation of model com-

    pounds. Detected primary intermediates should be

    subjected to further oxidation in order to elucidate

    the sequence of reactions.

    The concentration of the dierent products in

    LPO is strongly dependent on reaction conditions

    (the presence of heavy metal ions) and the time

    when samples are withdrawn. Many of the products

    described in this paper have been detected after arti-

    cial oxidation of LDL, but also after storage of

    LDL [14]. Since nearly all products have also been

    found after a myocardial infarction [98], we are

    forced to assume that all oxidation products ^ ob-

    served after articial LPO of linoleic acid ^ are also

    generated in all diseases connected with cell injury or

    inammation. As shown in this paper the K,L,Q,N-

    carbon system is prone to water addition followed

    by a retroaldol reaction. Biological material contains

    glutathione, a much more ecient nucleophile than

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208206

  • 7/27/2019 Alde Hide

    20/21

    water. We therefore assume that the greatest amount

    of dienals is captured immediately after genesis by

    this biological reagent.

    Some aldehydic LPO products have been shown to

    develop profound physiological properties, for in-stance 2,4-decadienal which induces apoptosis [99]

    and is of high cytotoxicity [45]. Since plant and

    mammalian cells respond to any change in the cell

    membrane structure [100] caused by inside (prolifer-

    ation) or outside (injury) events by generation of

    aldehydic LPO products, these might represent a

    part of the cell to cell communication system [101].

    Consequently it would be of great importance to

    study the physiological properties of aldehydic LPO

    products in more detail. Moreover epoxy hydroxy

    acids are also of high chemical reactivity. Since

    they are the main LPO products they may also in-

    uence biological processes.

    LPO processes are deeply involved in the genera-

    tion of chronical diseases, for instance in atheroscle-

    rosis. The inuence of linoleic acid oxidation prod-

    ucts on these diseases has only been recognized

    recently. Therefore the investigation of linoleic acid

    LPO products promises to contribute to the elucida-

    tion of the genesis of these diseases and the processes

    which are connected with aging and death.

    References

    [1] T. Galliard, in: G. Kahl (Ed.), Biochemistry of Wounded

    Plant Tissue, de Gruyter Verlag, Berlin, 1978, pp. 155^201.

    [2] H. Spreitzer, J. Schmidt, G. Spiteller, Fat Sci. Technol. 91

    (1989) 108^113.

    [3] E.D. Wills, Biochem. J. 99 (1966) 667^676.

    [4] S. Parthasarathy, U.P. Steinbrecher, J. Barnett, J.L. Witz-

    tum, D. Steinberg, Proc. Natl. Acad. Sci. USA 82 (1985)

    3000^3004.

    [5] G.B. Ansell, S. Spanner, New Compr. Biochem. 4 (1982) 1^

    49.

    [6] S. Yamamoto, Biochim. Biophys. Acta 1128 (1992) 117^131.[7] S. Rosahl, Z. Naturforsch. Sect. C J. Biosci. 51 (1996) 123^

    138.

    [8] C. Fuchs, G. Spiteller, Z. Naturforsch. 55C (2000) 643^648.

    [9] H.W. Gardner, Free Radic. Biol. Med. 7 (1989) 65^86.

    [10] B. Halliwell, S. Afr. J. Sci. 87 (1991) 576^584.

    [11] G. Spiteller, Free Radic. Biol. Med. 25 (1996) 1003^1009.

    [12] W. Jira, W. Carson, A. Schramm, G. Spiteller, Chem. Phys.

    Lipids 91 (1998) 1^11.

    [13] W. Jira, A. Richter, G. Spiteller, Chem. Phys. Lipids 87

    (1997) 81^89.

    [14] D. Spiteller, G. Spiteller, Angew. Chem. 39 (2000) 585^

    589.

    [15] J.A. Berliner, M.C. Territo, A. Sevanian, S. Ramin, J.A.

    Kim, B. Bamshad, M. Esterson, A.M. Fogelman, J. Clin.

    Invest. 85 (1990) 1260^1266.

    [16] N.A. Porter, Methods Enzymol. 105 (1984) 273^282.[17] H.W. Gardner, R. Kleiman, D. Weisleder, Lipids 9 (1974)

    696^706.

    [18] H.W. Gardner, R. Kleiman, Biochim. Biophys. Acta 665

    (1981) 113^125.

    [19] H.W. Gardner, J. Agric. Food Chem. 23 (1975) 129^136.

    [20] W. Grosch, in : H.W.S. Chan (Ed.), Autoxidation Unsatu-

    rated Lipids, Academic Press, London, 1987, pp. 95^139.

    [21] H.W. Gardner, Biochim. Biophys. Acta 1084 (1991) 221^

    239.

    [22] N.A. Porter, S.E. Caldwell, K. Mills, Lipids 30 (1995) 277^

    290.

    [23] H.W. Gardner, in: W.W. Christie (Ed.), Advances in Lipid

    Methodology-Four, Oily Press, Dundee, 1997, Ch. 1, pp. 1^

    43, and refs. cited therein.

    [24] H.W. Gardner, J. Lipid Res. 11 (1970) 311^321.

    [25] A. Graveland, Lipids 8 (1973) 599^605.

    [26] I. Feussner, C. Wasternack, H. Kindl, H. Kuehn, Proc. Natl.

    Acad. Sci. USA 92 (1995) 11849^11853.

    [27] T. Galliard, Annu. Proc. Phytochem. Soc. (Recent Adv.

    Chem. Biochem. Plant Lipids Proc. Symp.) 12 (1975) 319^

    357.

    [28] C. Rusterucci, V. Stallart, M.-L. Milat, A. Pugin, P. Ricci,

    J.-P. Blein, Plant Physiol. 111 (1996) 885^891.

    [29] J.N. Siedow, Annu. Rev. Plant Physiol. Plant Mol. Biol. 42

    (1991) 145^188.

    [30] P.A. Sandstrom, P.W. Tebbey, S. Van Cleave, T.M. Buttke,

    J. Biol. Chem. 269 (1994) 798^801.[31] D.M. Hockenbery, BioEssays 17 (1995) 631^638.

    [32] J. Sastre, F.V. Pallardo, J. Vina, Age 19 (1996) 129^139.

    [33] S.S. Chang, R.J. Peterson, C.T. Ho, J. Am. Oil Chem. Soc.

    55 (1978) 718^727.

    [34] E. Selke, W.K. Rohwedder, H.J. Dutton, J. Am. Oil Chem.

    Soc. 57 (1980) 25^30.

    [35] E.N. Frankel, Chem. Phys. Lipids 44 (1987) 73^85.

    [36] W. Grosch, Fat Sci. Technol. 91 (1989) 1^6.

    [37] E.N. Frankel, Lipid Oxidation, Oily Press, Dundee, 1989,

    Ch. 11, pp. 227^248, and refs. cited therein.

    [38] F. Marks, G. Fu rstenberger (Eds.), Prostaglandins, Leuko-

    trienes, and other Eicosanoids. From Biogenesis to Clinical

    Applications, vol. 1, Wiley-VCH, Weinheim, 1999.

    [39] G. Sembdner, C. Klose, Biol. Rundsch. 23 (1985) 29^40.

    [40] E.E. Farmer, C.A. Ryan, Proc. Natl. Acad. Sci. USA 87

    (1990) 7713^7716.

    [41] M.J. Mu ller, W. Brodschelm, E. Spannagl, M.H. Zenk,

    Proc. Natl. Acad. Sci. USA 90 (1993) 7490^7494.

    [42] S. Blechert, W. Brodschelm, S. Hoelder, L. Kammerer, T.M.

    Kutchan, M.J. Mueller, Z.-Q. Xia, Proc. Natl. Acad. Sci.

    USA 92 (1995) 4099^4105.

    [43] T. Kaneko, S. Honda, S.I. Nakano, M. Matsuo, Chem.-Biol.

    Interact. 63 (1987) 127^137.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208 207

  • 7/27/2019 Alde Hide

    21/21

    [44] H. Esterbauer, R.J. Schaur, H. Zollner, Free Radic. Biol.

    Med. 11 (1991) 81^128.

    [45] C. Nappez, S. Battu, J.L. Beneytout, Cancer Lett. 99 (1996)

    115^119.

    [46] S.F. Vaughn, H.W. Gardner, J. Chem. Ecol. 19 (1993) 2337^

    2345.[47] U. Dittberner, G. Eisenbrand, H. Zankl, Mutat. Res. 335

    (1995) 259^265.

    [48] H.W. Gardner, D.L. Dornbos Jr., A.E. Desjardins, J. Agric.

    Food Chem. 38 (1990) 1316^1320.

    [49] F.J.G.M. van Kuijk, A.N. Siakotos, L.G. Fong, R.J. Ste-

    phens, D.W. Thomas, Anal. Biochem. 224 (1995) 420^424.

    [50] G. Spiteller, W. Kern, P. Spiteller, J. Chromatogr. A 843

    (1999) 29^98.

    [51] E. Kovats, Helv. Chim. Acta 41 (1958) 1915^1932.

    [52] H.W. Gardner, Biochim. Biophys. Acta 1001 (1989) 274^

    281.

    [53] H. Hughes, C.V. Smith, E.C. Horning, J.R. Mitchell, Anal.

    Biochem. 130 (1983) 431^436.

    [54] Z. Wu, D.S. Robinson, C. Domoney, R. Casey, J. Agric.

    Food Chem. 43 (1995) 337^342.

    [55] P. Spiteller, G. Spiteller, Biochim. Biophys. Acta 1392 (1998)

    23^40.

    [56] M. Hamberg, P. Fahlstadius, Arch. Biochem. Biophys. 276

    (1990) 518^526.

    [57] W. Adam, L. Hadjiarapoglou, A. Smerz, Chem. Ber. 124

    (1991) 227^232.

    [58] J.A. Matthew, H.W.S. Chan, T. Galliard, Lipids 12 (1977)

    324^326.

    [59] J.I. Shulman, J. Org. Chem. 42 (1977) 3970^3972.

    [60] G. Lakshminarayana, D.G. Cornwell, Lipids 21 (1986) 175^

    177.

    [61] S. Srivastava, A. Chandra, L.-F. Wang, W.E. Seifert Jr.,B.B. DaGue, N.H. Ansari, S.K. Srivastava, A. Bhatnagar,

    J. Biol. Chem. 273 (1998) 10893^10900.

    [62] M. Hamberg, Biochim. Biophys. Acta 752 (1983) 191^197.

    [63] G. Spiteller, Chem. Phys. Lipids 95 (1998) 105^162.

    [64] H.W. Gardner, C.G. Crawford, Biochim. Biophys. Acta 665

    (1981) 126^133.

    [65] G. Streckert, H.J. Stan, Lipids 10 (1975) 847^854.

    [66] E.N. Frankel, W.E. Ne, W.K. Rohwedder, B.P.S. Kham-

    bay, R.F. Garwood, B.C.L. Weedon, Lipids 12 (1977) 908^

    913.

    [67] H.W. Gardner, D. Weisleder, E.C. Nelson, J. Org. Chem. 49

    (1984) 508^515.

    [68] B. Gallasch, G. Spiteller, Lipids 35 (2000) 953^960.

    [69] H.W. Gardner, M.J. Grove, Plant Physiol. 116 (1998) 1359^

    1366.

    [70] H.W. Gardner, E. Selke, Lipids 19 (1984) 375^380.

    [71] B. Grein, M. Huer, G. Scheller, P. Schreier, J. Agric. Food

    Chem. 41 (1993) 2385^2390.

    [72] W.A. Pryor, N.A. Porter, Free Radic. Biol. Med. 8 (1990)

    541^543.

    [73] G. Ohlo, H. Strickler, B. Willhalm, C. Borer, M. Hinder,

    Helv. Chim. Acta 53 (1970) 623^637.

    [74] G. Eglinton, D.H. Hunneman, A. McCormick, Org. Mass

    Spectrom. 1 (1968) 593^611.

    [75] H. Sasai, M. Bougauchi, T. Arai, M. Shibasaki, Tetrahe-

    dron Lett. 38 (1997) 2117^2120.

    [76] H.W.S. Chan, F.A.A. Prescott, P.A.T. Swoboda, J. Am.

    Oil Chem. Soc. 53 (1976) 572^576.[77] H.W.S. Chan, G. Levett, J.A. Matthew, Chem. Phys. Lip-

    ids 24 (1979) 245^256.

    [78] G.J. Garssen, J.F.G. Vliegenthart, J. Boldingh, Biochem.

    J. 122 (1971) 327^332.

    [79] T.A. Dix, L.J. Marnett, J. Am. Chem. Soc. 105 (1983)

    7001^7002.

    [80] H.W. Gardner, E.C. Nelson, L.W. Tjarks, R.E. England,

    Chem. Phys. Lipids 35 (1984) 87^101.

    [81] M. Hamberg, Chem. Phys. Lipids 43 (1987) 55^67.

    [82] T.D. Lindstrom, S.D. Aust, Arch. Biochem. Biophys. 233

    (1984) 80^87.

    [83] P.A.T. Swoboda, K.E. Peers, J. Sci. Food Agric. 29 (1978)

    803^807.

    [84] K. Gassenmeier, P. Schieberle, J. Am. Oil Chem. Soc. 71

    (1994) 1315^1319.

    [85] A. Benedetti, M. Comporti, R. Fulceri, H. Esterbauer, Bio-

    chim. Biophys. Acta 792 (1984) 172^181.

    [86] H.W. Gardner, R.J. Bartelt, D. Weisleder, Lipids 27 (1992)

    686^689.

    [87] H. Takamura, H.W. Gardner, Biochim. Biophys. Acta 1303

    (1996) 83^91.

    [88] H.W. Gardner, M. Hamberg, J. Biol. Chem. 268 (1993)

    6971^6977.

    [89] H.W. Gardner, Lipids 33 (1998) 745^749.

    [90] P.R. Ortiz de Montellano, L.A. Grab, Biochemistry 26

    (1987) 5310^5314.

    [91] K.W. Um, M.E. Bailey, A.D. Clarke, R.R. Chao, J. Agric.Food Chem. 40 (1992) 1641^1646.

    [92] N.J. Engeseth, E.G. Perkins, B.P. Klein, J. Am. Oil Chem.

    Soc. 69 (1992) 841^846.

    [93] K. Uchida, M. Kanematsu, Y. Morimitsu, T. Osawa, N.

    Noguchi, E. Niki, J. Biol. Chem. 273 (1998) 16058^

    16066.

    [94] A. Loidl, G. Spiteller, Org. Mass Spectrom. 28 (1993) 153^

    156.

    [95] A. Loidl-Stahlhofen, G. Spiteller, Biochim. Biophys. Acta

    1211 (1994) 156^160.

    [96] A.M. Gaddis, R. Ellis, G.T. Currie, J. Am. Oil Chem. Soc.

    38 (1961) 371^375.

    [97] P.A.T. Swoboda, C.H. Lea, J. Sci. Food Agric. 16 (1965)

    680^688.

    [98] A. Dudda, G. Spiteller, F. Kobelt, Chem. Phys. Lipids 82

    (1996) 39^51.

    [99] A. Miralto, G. Barone, G. Romano, S.A. Poulet, A. Ia-

    nora, G.L. Russo, I. Buttino, G. Mazzarella, M. Laabir,

    M. Cabrini, M. Giacobbe, Nature 402 (1999) 173^176.

    [100] G. Spiteller, Stud. Nat. Prod. Chem. (2001) submitted.

    [101] M.C. Ra, Nature 356 (1992) 397^400.

    P. Spiteller et al. / Biochimica et Biophysica Acta 1531 (2001) 188^208208