T. Koch et al. / Tetrahedron 58 (2002) 3271±3274
3273
4. Experimental
removed in vacuo and compound 7 was puri®ed by chroma-
tography on silica gel using the same solvent as before.
Yield: 0.029 g (53%). Spectroscopic data of the products
obtained were in agreement with the literature.16
4.1. General
All reactions were run in ¯ame-dried glassware under argon.
1
Solvents and reagents were dried prior to use. H and 13C
4.1.3. (9Z,11E,15Z)-13-Oxooctadeca-9,11,15-trienoic acid
(4). A well stirred solution of the hydroxy acid resulting
from reduction of 2 (23 mg, 0.08 mM) in CH2Cl2 (300
mL) was treated at room temperature with Bobbitt's reagent
(0.027 g, 0.09 mM) and silica gel (5 mg) to catalyse the
oxidation. Stirring was continued for 30 min followed by
removal of solvent in a gentle stream of argon. The crude
residue was taken up in hexane/ethyl acetate/acetic acid
(1:1:0.1%, v/v/v) and puri®ed by column chromatography
on silica gel using the same solvent. Yield: 13.3 mg (58%).
1H NMR (500 MHz, CDCl3) d: 0.92 (t, J7.50 Hz, 3H),
1.15±1.4 (m, 8H), 1.50±1.62 (m, 2H), 2.02 (m, 2H),
2.20±2.33 (m, 4H), 3.24 (d, J6.91 Hz, 2H), 5.43±5.59
(m, 2H), 5.75±5.90 (m, 1H), 6.05 (t, J11.3 Hz, 1H),
6.13 (d, J15.3 Hz, 1H), 7.42±7.5 (dd, J11.6, 15.3 Hz,
1H). 13C NMR (125 MHz, CDCl3) d: 13.9, 20.9, 24.6,
27.2, 28.3, 28.6, 29.0, 29.4, 33.9, 40.1, 120.5, 126.9,
128.6, 135.3, 137.7, 142.8, 179.3, 198.7. MS (EI, 70 eV)
m/z: 292 (M1z, 7), 224 (16), 223 (100), 149 (29), 135 (10),
123 (12), 109 (15), 107 (17), 97 (14), 95 (27), 83 (17), 81
(64), 79 (19), 69 (24), 67 (30), 57 (23), 55 (45). IR (KBr,
neat): 3019, 2931, 2856, 1706, 1665, 1627, 1586, 1457,
1412, 1272, 1189, 996 cm21. HR-MS m/z calcd for
C18H28O3 292.2038, found 292.2039.
NMR: chemical shifts of H (500 MHz) and 13C NMR
1
(125 MHz) are given in ppm (d) down®eld relative to
TMS as internal standard. Thin layer chromatography was
performed with silica gel plates. Column chromatographic
separations were performed on silica gel (Merck, Kieselgel
60; 230±400 mesh). Soybean lipoxygenase was obtained
from Sigma±Aldrich, D-82024 Taufkirchen, Germany.
The con®gurational purity of the products was established
by gas chromatography on a capillary column, Econo-capq
EC-5 (SE 54), 15 m£0.25 mm (Alltech, Deer®eld), under
temperature programmed conditions. Free acids and
alcohols were derivatised with diazomethane and MSTFA
prior to analysis.
4.1.1. (13S,9Z,11E,15Z-Hydroxyoctadeca-9,11,15-trienoic
acid (6). Linolenic acid (1.0 g, 3.6 mM) was emulsi®ed by
sonication in a borate buffer (600 mL, 0.1 M, pH 9.0). To
the rapidly stirred emulsion (208C) was added lipoxygenase
from soybean (0.1 g, 5.3 U). After a few minutes the cloudy
solution became clear and stirring was continued for 1 h
followed by addition of dil. HCl (10 mL, 2 M). The crude
hydroperoxide was extracted with CHCl3 (3£500 mL) and
dried using anhydrous Na2SO4, followed by removal of
solvent at 208C. Puri®cation was achieved by chroma-
tography on silica gel using hexane/ethyl acetate/acetic
acid (600:400:1) for elution. Yield: 0.27 g (39%). An
aliquot of the hydroperoxide (60 mg, 0.20 mM) was reduced
to the alcohol 6 by stirring for 10 min at room temperature
with P(OCH3)3 (100 mL, 0.816 mM) in CH2Cl2 (8 mL).
Solvents and excess reagent were removed in vacuo and
the alcohol was puri®ed by chromatography on silica gel
using the same solvent as before. Yield: 0.029 g (53%). The
spectroscopic data of the compounds were in agreement
with the literature.9
4.1.4.
(10E,12Z,15Z)-9-Oxooctadeca-10,12,15-trienoic
acid (5). A well stirred solution of the hydroxy acid result-
ing from reduction of 3 (8.7 mg, 29.7 mM) in CH2Cl2
(300 mL) was treated at room temperature with Bobbitt's
reagent (10 mg) and silica gel (5 mg) to catalyse the oxida-
tion. Stirring was continued for 30 min followed by removal
of solvent in a gentle stream of argon. The crude residue was
taken up in hexane/ethyl acetate/acetic acid (1:1:0.1%,
v/v/v) and puri®ed by column chromatography on silica
1
gel. Yield: 6.0 mg (69%). H NMR (500 MHz, CDCl3) d:
0.92 (t, J7.48 Hz, 3H), 1.15±1.35 (m, 6H), 1.47±1.61 (m,
4H), 2.03 (m, 2H), 2.28 (t, J7.48 Hz, 2H), 2.5 (t, J
7.48 Hz, 2H), 3.0 (t, J7.48 Hz, 2H), 5.25 (m, 1H), 5.4
(m, 1H), 5.79 (m, 1H), 6.06 (t, J11.4 Hz, 1H), 6.12 (d,
J15.21 Hz, 1H), 7.45 (dd, J11.6, 15.21 Hz, 1H). 13C
NMR (125 MHz, CDCl3) d: 14.2, 20.6, 24.2, 24.5, 24.6,
26.5, 29.0, 29.1, 34.0, 41.0, 125.2, 126.8, 129.6, 133.3,
136.8, 140.2, 179.9, 201.1. IR (KBr, neat): 3015, 2962,
2928, 2856, 1695, 1620, 1589, 1468, 1408, 1306, 1226,
1196, 1113, 1068, 992, 966 cm21. MS (EI, 70 eV) m/z:
292 (M1z, 75), 274 (17), 223 (64), 171 (74), 149 (55), 135
(17), 125 (46), 121 (51), 107 (50), 97 (38), 95 (46), 93 (46),
83 (49), 81 (54), 79 (77), 77 (32), 69 (23), 67 (36), 55 (100).
HR-MS m/z calcd for C18H28O3 292.2038, found 292.2037.
4.1.2. (9S,10E,12Z,15Z-Hydroxyoctadeca-10,12,15-tri-
enoic acid (7). Tomatoes from the local market (1 kg)
were homogenized in a phosphate buffer (600 mL, 0.1 M,
pH 5.7) with simultaneous aeration. Then, linolenic acid
1 (1.0 g, 3.6 mM), previously emulsi®ed by sonication
(30 min) in the same buffer (300 mL) containing Tween
20 (1 mL), was added. Ef®cient stirring with simultaneous
aeration was maintained for 4 h and then the reaction was
stopped by addition of 0.2 M HCl (20 mL). The crude
hydroperoxide was extracted with ether (3£800 mL).
Separation of the phases was forced by centrifugation at
2500 rpm. Solid material accumulating at the interphase
was extracted with ether (3£100 mL) and the combined
extracts dried (anhydrous Na2SO4). Removal of solvent at
208C afforded 0.73 g (65%) of the crude hydroperoxide 3.
An aliquot of 0.10 g was puri®ed by preparative thin layer
chromatography using a ternary solvent mixture of hexane/
ethyl acetate/acetic acid (600:400:1) for elution. Yield:
0.058 g (58%). The hydroperoxide (58 mg, 0.19 mM)
was reduced to the alcohol 7 by stirring for 10 min at
room temperature with P(OCH3)3 (100 mL, 0.816 mM) in
CH2Cl2 (8 mL). Solvents and excess of reagent were
Acknowledgements
Support by the Fonds der Chemischen Industrie, Frankfurt,
and a fellowship from the Alexander von Humboldt Society
(M. Hoskovec) are gratefully acknowledged. We thank
R. Lauchli for proof reading the manuscript.