First synthesis of phosphatidylcholine and cholesteryl derivatives bearing an unsaturated aldehyde residue

Article abstract of DOI:10.1039/b206231n

The first synthesis of phosphatidylcholine and cholesteryl esters bearing a shortened acyl chain with a 2,4-dienal terminal is described.

Full text of DOI:10.1039/b206231n

View Article Online / Journal Homepage / Table of Contents for this issue
First synthesis of phosphatidylcholine and cholesteryl derivatives
bearing an unsaturated aldehyde residue†
Arnold N. Onyango,a Teruhiko Nitoda,a Takao Kaneko,b Mitsuyoshi Matsuo,c
Shuhei Nakajimaa and Naomichi Baba*a
1
a Department of Bioresources Chemistry, Faculty of Agriculture, Okayama University,
Tsushimanaka, Okayama 700-8530, Japan. E-mail: babanaom@cc.okayama-u.ac.jp
b Tokyo Metropolitan Institute of Gerontology, 35-2 Sakaecho, Itabashi-ku, Tokyo 173-0015,
Japan
c Faculty of Science, Konan University, Higashinada-ku, Kobe 658, Japan
Received (in Cambridge, UK) 27th June 2002, Accepted 25th July 2002
First published as an Advance Article on the web 5th August 2002
The first synthesis of phosphatidylcholine and cholesteryl
esters bearing a shortened acyl chain with a 2,4-dienal
terminal is described.
Much attention has recently been paid to lipids containing an
oxidatively shortened fatty acyl chain with a carbonyl terminal,
the so-called core aldehydes, since their reactivity with bio-
logical nucleophiles such as the amino group gives them
important biological implications. For example, phospholipid
and cholesteryl ester core aldehydes modify membrane proteins
and lipoproteins, and contribute in various ways to athero-
genesis.1 Various platelet activating factor (PAF)-like activities
of phosphatidylcholine (PC) core aldehydes have also been
documented.1,2 The core aldehydes that have so far been
employed in previous studies are esters of saturated simple
aldehyde acids such as 5-oxopentanoic and 9-oxononanoic
acids. In biological systems, however, the oxidative degradation
of polyunsaturated acyl groups in lipids should also produce
unsaturated core aldehydes. Tokumura et al. recently identified
a 13 carbon dienal-containing PC as a major product of the
15-lipoxygenase catalyzed oxidation of PCs under hypoxic
conditions, and this reaction is likely to occur in vivo.2 The bio-
logical activities of this kind of PC and cholesteryl core alde-
hydes should therefore be examined, especially in view of the
fact that the role of 12/15 lipoxygenase in atherogenesis has
Although fatty acid hydroperoxides decompose extremely
fast at 50 ЊC,6 we could not achieve considerable decomposition
of 13-hydroperoxy-octadecatrienoic acid (HPOT) under a
nitrogen atmosphere even at 60 ЊC, prompting us to use Fe2ϩ to
effect the formation of the requisite alkoxy radicals according
to the Fenton reaction (eqn. (1)).
recently attracted much attention.3 It has also been suggested
that the difficulty in detecting such unsaturated core aldehydes
from biological systems might be due to their stronger inter-
actions with protein side chains than carbonyl groups alone.4
Indeed, their Schiff base adducts should be stabilized by the
conjugation of their diene system with the imine double bond.
Besides, they might also form Michael type adducts with thiol
and amino groups. Thus, the characterization of such adducts
will be important for understanding the fate and effects of these
aldehydes in biological systems. For these studies, the avail-
ability of large amounts of unsaturated core aldehydes in a pure
state is vital. Here we report the preparation of such PC and
cholesteryl esters, compounds 1–4.
During the homolytic decomposition of PC hydroperoxides,
the formation of PC core aldehyde instead of other products
seems to be greatly favoured by the accompanying formation
of an allyl radical upon β-scission of the alkoxy radical.5 We
thus anticipated that in the absence of oxygen, the homo-
lytic decomposition of ω-6 hydroperoxides of n-3 fatty acids
(Scheme 1) might give preparatively useful amounts of ω-oxo
fatty acids, which could then be esterified to lysophosphatidyl-
choline (lysoPC) or cholesterol.
(1)
This reaction afforded a mixture of some decomposition
products including the desired 13-oxotrideca-9,11-dienoic acid
(OTDA) as the major product, and 13-hydroxyoctadecatrienoic
acid. Unfortunately, these latter two compounds could not be
separated from each other by silica gel chromatography. There-
fore as shown in Scheme 2, HPOT obtained from α-linolenic
acid by the action of soybean lipoxygenase was first converted
to methyl ester 5, which was then cleaved by FeSO4ؒ7H2O at 60
ЊC to give a mixture of methyl OTDA isomers 6 and 7 in 51%
isolated yield. This yield was impressive since the decom-
position of fatty acid hydroperoxides has not generally been
associated with such a level of selectivity for any of the possible
products. The structures of compounds 6 and 7 were confirmed
by EI MS and 1H NMR.7 Stereochemical assignment was based
on the coupling constants of 11 and 15 Hz for the cis and trans
double bonds respectively in compound 6. For compound 7,
however, the coupling constant for H-9 and H-10 could not be
† Electronic Supplementary Information (ESI) available: experimental
data for compounds 1–9. See http://www.rsc.org/suppdata/p1/b2/
b206231n/
DOI: 10.1039/b206231n
J. Chem. Soc., Perkin Trans. 1, 2002, 1941–1943
This journal is © The Royal Society of Chemistry 2002
1941

View Article Online
Table 1 Dependence of the total yield and the ratio of methyl OTDA
isomers 6 and 7 on reaction time at 60 ЊCa
of the isomerization was found to depend on the amount of
DCC used.
Not surprisingly, free oxo acids 8 and 9 could not be
obtained from their esters 6 and 7 by either basic or acidic
hydrolysis, owing to the reactivity of the unsaturated carbonyl
groups under such conditions. Fortunately, lipase PS (from
Pseudomonas sp., Amano Seiyaku Co., Ltd.) was found to
effectively perform the hydrolysis, affording 8 and 9 in 80%
yield and with retention of the isomeric ratio.
Time/min
Yield (%)
6 : 7 ratio
10b
30
40
1 : 0
3 : 1
1 : 1
1 : 4
51
46
38
60
a 60 ЊC gave better results than lower temperatures. b 1H NMR revealed
that a little of compound 6 had formed, but it could not be separated
from the remaining hydroperoxide.
Although all previously reported syntheses of core aldehydes
involved protecting the aldehyde function of the oxo acid as a
dimethyl acetal prior to the DCC–DMAP mediated esterifi-
cation,8 we found that such a protection–deprotection sequence
could be avoided. Thus in the presence of DCC and a catalytic
amount of DMAP, a mixture of compounds 8 and 9 were
esterified to lysoPC prepared as previously described,9 affording
core aldehydes 1 and 2 in 90% yield. As mentioned above, some
double bond isomerization occurred during this reaction, and
it was found to depend on the amount of DCC used. For
example, with 1 equivalent of DCC, a 2 : 1 mixture of acids 8
and 9 gave a 3 : 2 mixture of core aldehydes 1 and 2, which ratio
remarkably changed to 1 : 9 when 1.2 equivalents of DCC was
used.‡ Purification by chromatography on silica gel (CHCl3–
MeOH–H2O, 60 : 30 : 1) and subsequent reverse phase HPLC
on an ODS column (MeOH–CHCl3–H2O, 100 : 4.5 : 5) gave 1
and 2 with high purity as resinous solids whose structural
integrity was confirmed by MS and 1H NMR.10 However, their
separation was not achieved under these conditions.§
Scheme 1 Formation of ω-oxo fatty acid and allylic radical by the
homolytic decomposition of ω-6 hydroperoxides of n-3 fatty acids.
Similarly, the esterification of oxo acids 8 and 9 to cholesterol
afforded core aldehydes 3 and 4,11 which were obtained as
iridescent solids in 87% yield after careful column chrom-
atography on silica gel (hexane–EtOAc, 95 : 5). Fortunately,
their separation was also achieved thereby. Their purity was
confirmed by analytical scale HPLC on an inertsil column
eluted with hexane–chloroform (8 : 2).
The first successful synthesis of these compounds described
herein sets the stage for studies on their biological activities and
biologically relevant chemistry, and provides reference com-
pounds for their possible identification in biological systems.
This method should also be applicable to the synthesis of other
core aldehydes having a longer ω-oxo acyl group, whose facile
formation in vitro has previously been reported.5
Scheme 2 Reagents and conditions: i. soybean lipoxygenase, O2, borate
buffer (pH 9), 6 h; ii. diazomethane; iii. FeSO4ؒ7H2O, EtOH, 60 ЊC, N2;
iv. lipase PS (Amano), H2O, rt, 3 h; v. DCC, DMAP, EtOH-free CHCl3,
rt, 20 h.
calculated because their signals overlapped. Nevertheless, the
coupling constant for its H-11 and H-12 was 15 Hz like in
compound 6, indicating that its H-9 and H-10 must be in a trans
relationship.
Experimental
General
As shown in Table 1, both the total yield and the isomeric
ratio were found to depend on the reaction time. From these
results it is clear that compound 7 was formed from 6. We pro-
pose that this isomerization was due to electrophilic attack on
the carbonyl oxygen of 6 as shown in Scheme 3, since both Fe2ϩ
Apart from the preparation of compounds 6 and 7 from com-
pound 5, all reactions in the present synthesis were conducted
in the presence of a trace amount of the free-radical scavenging
antioxidant 2,6-di-tert-butyl-p-hydroxytoluene (BHT). All reac-
tions were conducted under a nitrogen atmosphere. Column
chromatography was done on silica gel 60.
Preparation of aldehyde esters 6 and 7
To ethanol (200 ml) maintained at 60 ЊC were added methyl
ester 5 (600 mg, 1.8 mmol), and FeSO4ؒ7H2O (495 mg,
1.8 mmol). This mixture was stirred vigorously for 30 minutes,
after which a trace of BHT was added, followed by solvent
evaporation in vacuo and column chromatography (hexane–
EtOAc, 9.3 : 0.7), affording a mixture of aldehyde esters 6 and 7
(218 mg, 51%). Characterization data are given in ref. 7 below.
Preparation of oxo acids 8 and 9
Scheme 3 Proposed mechanism for double bond isomerization in the
unsaturated oxo fatty acid derivatives.
A mixture of esters 6 and 7 (200 mg, 0.84 mmol) and lipase PS
(200 mg, 6000 units) in water (4 ml) was stirred at rt for 4 hours,
followed by extraction with ethyl acetate. The organic phase
was washed with water and dried over Na2SO4. Subsequent
solvent evaporation in vacuo and column chromatography
(hexane–EtOAc, 4 : 1) afforded oxo acids 8 and 9 (150 mg,
80%). 1H NMR was similar to that for esters 6 and 7, but lack-
ing the signal at 3.66 ppm. HR-MS (EI): Found, 244.1400 m/z
(Mϩ, C13H20O3 requires 244.1412).
and Fe3ϩ which are present in the reaction mixture (eqn. (1)) are
Lewis acids. That the conversion of 7 back to 6 was unfavour-
able should be due to greater stability of the former than the
latter. Additional isomerization was observed during the DCC-
mediated esterification reaction (vide infra). This might also be
explained, at least in part, by the mechanism presented in
Scheme 3, with DCC acting as the electrophile, since the extent
1942
J. Chem. Soc., Perkin Trans. 1, 2002, 1941–1943

View Article Online
MS (EI): Found, 238.1567 m/z (Mϩ, C14H22O3 requires 238.1569);
Acknowledgements
Compound 7: Rf = 0.27 (hexane–EtOAc, 9 : 1); 1H NMR (500 MHz,
CDCl3): δ 1.31 (6H, m, 4-, 5-, 6-H), 1.45 (2H, m, 7-H), 1.59 (2H, m,
3-H), 2.20–2.30 (4H, m, 2-, 8-H), 3.66 (3H, s, -OCH3), 6.07 (1H, dd,
J 7.9 and 15.3, 12-H), 6.16–6.34 (2H, m, 9-, 10-H), 7.08 (1H, dd,
J 10.1 and 15.3, 11-H), 9.52 (1H, d, J 7.9, 13-H); HR-MS (EI):
Found, 238.1567 m/z (Mϩ, C14H22O3 requires 238.1569).
We are grateful for using the Mass and NMR spectrometers
of the MS laboratory (Faculty of Agriculture), the API and
SC-NMR laboratories of Okayama University.
8 (a) H. Boechzelt, B. Karten, P. M. Abuja, W. Sattler and M. Mittel-
bach, J. Lipid Res., 1998, 39, 1503–1507; (b) A. D. Watson, N.
Leitinger, M. Navab, K. F. Faull, S. Horrko, J. L. Witzum, W. Palin-
ski, D. Schwenke, R. G. Salomon, W. Sha, G. Subbanagounder, A.
M. Fogelman and J. A. Berliner, J. Biol. Chem., 1997, 272, 13597–
13607 and the references therein.
References
‡ Protection of the aldehyde group is not expected to eliminate double
bond isomerization because the deprotection step involves acidic
conditions.
§ The composition of such core aldehydes in vivo is not known, but
besides possible isomerization of the (Z,E)-isomer as presented above,
both isomers are expected to form directly from the corresponding
nonenzymatically-produced (Z,E) and (E,E)-hydroperoxides.
9 N. Baba, K. Yoneda, S. Tahara, J. Iwasa, T. Kaneko and M. Matsuo,
J. Chem. Soc., Chem. Commun., 1990, 18, 1281–1282.
1
10 Compound 1: H NMR (500 MHz, CDCl3): δ 0.87 (t, 3H, ω-CH3),
1.18–1.35 (34H, m, CH2 × 17), 1.45 (2H, m, 7Ј-H), 1.60 (4H, m,
OCOCH2 CH2 × 2), 2.20–2.30 (6H, m, 8Ј-H, OCOCH2 × 2), 3.35
(9H, s, NMe3), 3.79 (2H, m, OCH2CH2N), 3.94 (2H, m, CH2OP),
4.11 (1 proton of OCH2CH(OR)CH2OP), 4.31(2H, m, OCH2-
CH2N), 4.38 (1H, d, 1 proton of OCH2CH(OR)CH2OP), 5.19 (1H,
br s, OCH2CH(OR)CH2OP), 6.00 (1H, m, 9Ј-H), 6.14 (1H, dd, J 7.9
and 15.2, 12Ј-H), 6.26 (1H, t, J 11.4, 10Ј-H), 7.44 (1H, dd, J 11.4 and
15.2, 11Ј-H), 9.61 (1H, d, J 7.9, 13Ј-H); ES MS: Found, 730.6 m/z;
HR-FAB MS: Found, 730.5032 m/z (C39H72NO9P ϩ Hϩ requires
730.5023); Compound 2: same as for 1 except: 6.07 (1H, dd, J 7.9,
15.3, 12Ј-H), 6.16–6.34 (2H, m, 9Ј, 10Ј-H), 7.08 (1H, dd, J 10.1 and
15.3, 11Ј-H), 9.52 (1H, d, J 7.9, 13Ј-H).
1 For a recent review, see: A. Kuksis, inform (AOCS), 2000, 11,
746–752.
2 A. Tokumura, T. Sumida, M. Toujima, K. Kogure, K. Fukuzawa,
Y. Takahashi and S. Yamamoto, J. Lipid Res., 2000, 41, 953–962.
3 For recent reviews, see: (a) C. D. Funk and T. Cyrus, Trends in
Cardiovascular Medicine, 2001, 11, 116–124; (b) M. K. Cathcart and
V. A. Folcik, Free Rad. Biol. Medicine, 2000, 28, 1726–1734.
4 G. Hoppe, A. Ravandi, D. Herrera, A. Kuksis and H. F. Hoff,
J. Lipid Res., 1997, 38, 1347–1360.
5 A. N. Onyango, T. Inoue, S. Nakajima, N. Baba, T. Kaneko,
M. Matsuo and S. Shimizu, Angew. Chem., Int. Ed. Engl., 2001, 40,
1755–1757.
11 Compound 3: Rf = 0.44 (hexane–EtOAc, 9 : 1); 1H NMR (500 MHz,
CDCl3): δ 0.5–2.6 (57H, CH3 × 5, CH2 × 18, CH × 6), 4.59 (1H, m,
3-H), 5.38 (1H, m, 6-H), 6.00 (1H, m, 9Ј-H), 6.14 (1H, dd, J 7.9 and
15.2, 12Ј-H), 6.26 (1H, t, J 11.4, 10Ј-H), 7.44 (1H, dd, J 11.4 and
15.2, 11Ј-H), 9.61 (1H, d, J 7.9, 13Ј-H); HR-FAB MS: Found,
593.4932 (C40H64O3 ϩ Hϩ requires 593.4933); Compound 4: Rf = 0.35
6 T. Nishike, S. Kondo, T. Yamamoto, A. Shigeeda, Y. Yamamoto,
H. Takamura and T. Matoba, Biosci. Biotechnol. Biochem., 1997, 61,
1973–1976.
7 Compound 6: Rf = 0.31 (hexane–EtOAc, 9 : 1); 1H NMR (500 MHz,
CDCl3): δ 1.31 (6H, m, 4-, 5-, 6-H), 1.45 (2H, m, 7-H), 1.59 (2H, m,
3-H), 2.20–2.30 (4H, m, 2-, 8-H), 3.66 (3H, s, -OCH3), 6.00 (1H, m,
9-H), 6.14 (1H, dd, J 7.9 and 15.2, 12-H), 6.25 (1H, t, J 11.4, 10-H),
7.44 (1H, dd, J 11.4 and 15.2, 11-H), 9.61 (1H, d, J 7.9, 13-H); HR-
1
(hexane–EtOAc, 9 : 1); H NMR same as for 3 except: δ 6.07 (1H,
dd, J 7.9 and 15.3, 12Ј-H), 6.16–6.34 (2H, m, 9Ј-, 10Ј-H), 7.08 (1H,
dd, J 10.1 and 15.3, 11Ј-H), 9.52 (1H, d, J 7.9, 13Ј-H).
J. Chem. Soc., Perkin Trans. 1, 2002, 1941–1943
1943