Improved synthesis of the epoxy isoprostane phospholipid PEIPC and its reactivity with amines

Article abstract of DOI:10.1021/ol8014804

(Chemical Equation Presented) An improved synthesis of the naturally occurring hydroxy ketone 1-palmitoyl-2-(5,6)-epoxyisoprostane E ₂-sn-glycero-3-phosphocholine (PEIPC) 1, a compound that plays a role in endothelial activation in atherosclerosis, has been carried out using a PMB ether as the key protecting group. Opening of an intermediate with pentylamine shows that the allylic epoxide is the position of attack by nucleophiles.

Full text of DOI:10.1021/ol8014804

ORGANIC
LETTERS
2008
Vol. 10, No. 19
4207-4209
Improved Synthesis of the Epoxy
Isoprostane Phospholipid PEIPC and its
Reactivity with Amines
Michael E. Jung,*,† Judith A. Berliner,‡,§ Lukasz Koroniak,†
B. Gabriel Gugiu,§ and Andrew D. Watson‡
Departments of Chemistry and Biochemistry, Medicine/Cardiology, and Pathology,
UniVersity of California, Los Angeles, California 90095
jung@chem.ucla.edu
Received June 30, 2008
ABSTRACT
An improved synthesis of the naturally occurring hydroxy ketone 1-palmitoyl-2-(5,6)-epoxyisoprostane E2-sn-glycero-3-phosphocholine (PEIPC)
1, a compound that plays a role in endothelial activation in atherosclerosis, has been carried out using a PMB ether as the key protecting
group. Opening of an intermediate with pentylamine shows that the allylic epoxide is the position of attack by nucleophiles.
The in vitro oxidation product of arachidonoyl phosphatidyl-
choline, 1-palmitoyl-2-(5,6)-epoxyisoprostane E2-sn-glycero-3-
phosphocholine (PEIPC), has been previously isolated, and its
biological activity described.1 It is present in atherosclerotic
lesions, oxidized lipoproteins, the membranes of cells exposed
to oxidative stress, and in apoptotic and necrotic cells.2 We have
demonstrated that at least five HPLC-separable isomers of
PEIPC were formed from the oxidation of 1-palmitoyl-2-
arachidonoyl-sn-glycero-3-phosphocholine (PAPC). The most
active isomer 1 was shown to activate several important
inflammatory responses that contribute to atherosclerosis,
including endothelial-monocyte interaction, and synthesis of
monocyte activators.3,4 On the basis of mass spectrometry of
the natural compound and the proton NMR spectra of the
dehydration product, the compound was tentatively assigned
the structure 1.5 On the basis of our earlier synthetic work
in this area,6 we recently reported the development of a triply
convergent7 coupling strategy, beginning with the diacetate
3 that permitted the total synthesis of PEIPC 1 (Scheme 1).8
Thus, conversion of 3 to the bromo alkene 4, coupling with
the epoxyaldehyde 5, and further transformations gave the
silyl-protected PEIPC 6. However, fluoride-promoted depro-
tection gave significant amounts of the dehydration product
PECPC 2 at the expense of 1. We report herein an improved
total synthesis of PEIPC 1 which uses a different alcohol
protecting group and therefore allows a higher-yielding
deprotection as the final step to give more and purer material
with less of the dehydration product.
(4) Cole, A. L.; Subbanagounder, G.; Mukhopadhyay, S.; Berliner, J. A.;
Vora, D. K. Arterioscler. Thromb. Vasc. Biol. 2003, 23, 1384–1390.
(5) Watson, A. D.; Subbanagounder, G.; Welsbie, D. S.; Faull, K. F.;
Navab, M.; Jung, M. E.; Fogelman, A. M.; Berliner, J. A. J. Biol. Chem.
1999, 274, 24787–98.
† Departments of Chemistry and Biochemistry.
‡ Department of Medicine/Cardiology.
§ Department of Pathology.
(6) Jung, M. E.; Kers, A.; Subbanagounder, G.; Berliner, J. A. J. Chem.
Soc., Chem. Commun. 2003, 167.
(1) Watson, A. D.; Leitinger, N.; Navab, M.; Faull, K. F.; Horkko, S.;
Witztum, J. L.; Palinski, W.; Schwenke, D.; Salomon, R. G.; Sha, W.;
Subbanagounder, G.; Fogelman, A. M.; Berliner, J. A. J. Biol. Chem. 1997,
272, 13597–607.
(7) (a) Suzuki, M.; Kawagishi, T.; Suzuki, T.; Noyori, R. Tetrahedron
Lett. 1982, 4057. (b) Noyori, R.; Suzuki, M. Chemtracts: Org. Chem. 1990,
173. (c) Noyori, R.; Suzuki, M. Angew. Chem., Int. Ed. Engl. 1984, 847.
(d) Snider, B. B.; Yang, K. J. Org. Chem. 1992, 57, 3615.
(8) Jung, M. E.; Berliner, J. A.; Angst, D.; Yue, D.; Koroniak, L.;
Watson, A. D.; Li, R. Org. Lett. 2005, 7, 3933.
(2) Berliner, J. A.; Watson, A. D. N. Engl. J. Med. 2005, 353, 8–11.
(3) Subbanagounder, G.; Wong, J. W.; Faull, K. F.; Miller, E.; Witztum,
J. L.; Berliner, J. A. J. Biol. Chem. 2002, 277, 7271–7281
.
10.1021/ol8014804 CCC: $40.75
Published on Web 08/28/2008
2008 American Chemical Society

Scheme 1
Scheme 3
hydroboration-oxidation of the vinyl group at -78 °C gave
a 2:1 mixture of the separable primary alcohols in 86% yield
(57% of the trans and 29% of the cis). Dess-Martin
periodinane oxidation gave in 66% yield the aldehyde which
on Wittig reaction with the ylide from hexyl bromide
afforded the desired Z-alkene 11 in 72% yield. In a like manner,
the cis-alcohol could be converted into the epimer 11′.
The coupling of the bromide 11 with the aldehyde 5 was
carried out by first forming the vinyllithium species by treatment
of 11 (dried with molecular sieves) with tert-butyllithium
followed by addition of the epoxyaldehyde 5 (also dried with
molecular sieves) to give the allylic alcohol 12 in 66% yield
(Scheme 4). Hydrolysis of the silyl enol ether, dehydration, and
The components for the key coupling were prepared as
follows. The epoxy aldehyde 5 was prepared by a slight
modification of our earlier route,5 namely, initial silyl protection
of pentane-1,5-diol followed by Swern oxidation, Horner-
Emmons reaction, and reduction to give the E-allylic alcohol 7
which was subjected to Sharpless asymmetric epoxidation and
finally oxidized (Dess-Martin periodinane) to give the epoxide
5 (Scheme 2). The second key component was prepared in the
Scheme 4
Scheme 2
same manner as described earlier,6,8 namely, conversion of
the achiral diacetate 3 via three steps into the monopivalate
8 using a lipase resolution as the key step. Protection of the
alcohol as the p-methoxybenzyl (PMB) ether followed by
reductive removal of the pivalate, oxidation, and bromination
gave the 2-bromo-4-arylmethoxycyclopentenone 9 (Scheme
3). 1,4-Addition of vinylcopper to the enone 9 in the presence
of tert-butyldimethylsilyl (TBS) chloride9 gave an inseparable
2:1 mixture of the trans and cis disubstituted bromoenol
ethers 10 in 94% yield. Although these compounds could
not be separated at this stage, they were easily separated at
the next. As we had done in our earlier synthesis, selective
deprotection of the silyl ether were effected in 75% yield with
formic acid in aq. THF to give the enone 13. The final two
steps (Dess-Martin periodinane and then chlorite oxidation)
converted the primary alcohol into the desired acid 14, namely,
the PMB ether of EI in 63% yield over the two steps. To test
the ability to oxidatively remove the PMB ether in the real
system, we decided to convert this protected EI into EI itself.
(9) (a) Lipshutz, B. H.; Ellsworth, E. L.; Dimock, S. H.; Smith, R. A. J.
J. Org. Chem. 1989, 54, 4977. (b) Lipshutz, B. H.; Crow, R.; Ellsworth,
E. L.; Dimock, S. H.; Smith, R. A.; Behling, J. R. J. Am. Chem. Soc. 1990,
112, 4063.
4208
Org. Lett., Vol. 10, No. 19, 2008

This was done by the oxidation of 14 with DDQ in dichlo-
romethane/pH 7 buffer followed by filtration through Celite and
then silica gel chromatography to give EI, 15, in 54% yield.
Thus, the oxidative removal of the PMB ether does not result
in dehydration to give EC as did our earlier attempts at
desilylation of the corresponding silyl ethers.8
The final steps in the synthesis of PEIPC involved coupling
of the PMB protected EI 14 to commercially available lyso-
PC 16 and final deprotection (Scheme 5).
dehydration of the ꢀ-hydroxyketone to give PECPC 2, to the
oxidatively removable PMB group allowed for the easy
purification and isolation of PEIPC 1. The synthetic PEIPC 1
had an HPLC retention time and mass spectrometric profile
identical to the naturally occurring compound1,3,5 and also
exhibited the biological activity expected of PEIPC.11
Finally, it is likely that PEIPC may form a covalent adduct
with various cellular proteins since treatment of proteins with
PEIPC and then extensive washing does not free the material
from the peptide.12 Therefore, it is very important to know what
part of the PEIPC molecule is responsible for its covalent
attachment to proteins, since then one may be better able to
develop inhibitors of this covalent linking. A priori, either the
enone or the allylic epoxide could be the required electrophile
(or perhaps the cyclopentenone PECPC could be formed in cells
and serve as the electrophile). To obtain some information on
the reactivity of this system, we treated the protected enone
epoxide 18 (prepared by a route similar to that described here
for 13)8 with pentyl amine as a surrogate for lysine-containing
peptides in cells (Scheme 6). We isolated a large amount of
Scheme 5
Scheme 6
the starting material 18 but also the product 19 in which the
allylic epoxide has been opened by the amine.13 The structure
was determined by extensive NMR experiments which showed
that the proton R to the amine (Hc: δ 3.37) was a dd and coupled
to the proton (Ha: δ 6.17) on the exocyclic double bond of the
enone while the proton R to the oxygen (Hb: δ 3.66) was a multiplet
and did not couple to Ha. Thus, it is likely that other biological
nucleophiles may also react with PEIPC in a similar fashion.
Thus, we have synthesized the important naturally occurring
material PEIPC 1 by a modified route that allowed for the
production of larger quantities and purer material. We have also
determined that the allylic epoxide functionality is the most
likely center for the attack of nucleophiles on this system; e.g.,
18 gives 19. Further work in this area is underway.
We again used the method we had developed in our earlier
work,8 namely, reaction of 14 with 10 equiv of the Yamaguchi
reagent (2,4,6-trichlorobenzoyl chloride) and DMAP at 23 °C
for 1 h in the presence of fresh lyso-PC 16, to afford a 70%
yield of the coupled product 17.10 The key removal of the PMB
ether protecting group was effected as above using DDQ in
dichloromethane/pH 7 buffer at 22 °C for 2 h followed by
filtration through Celite to give the crude PEIPC. This material
could be purified by silica gel chromatography followed by
passage through a short C18 column to give the desired
compound PEIPC 1 as a pure material in 42% isolated yield.
Thus, the change of the protecting group strategy from the base-
labile TBDPS group, the removal of which caused extensive
(10) Other methods for such couplings have been developed. See: (a)
Gu, X.; Sun, M.; Gugiu, B.; Hazen, S.; Crabb, J. W.; Salomon, R. G. J.
Org. Chem. 2003, 68, 3749. (b) Acharya, H. P.; Kobayashi, Y. Angew.
Chem., Int. Ed. 2005, 44, 3481.
Acknowledgment. We thank the National Institutes of
Health (HL64731) for generous support.
(11) See Supporting Information for more details about the biological
activity.
Supporting Information Available: Experimental pro-
cedures and spectral data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(12) Gugiu, B. G.; Mouillesseaux, K.; Duong, V.; Herzog, T.; Hekimian,
A.; Koroniak, L.; Vondriska, T. M.; Watson, A. D. J. Lipid Res. 2008, 49,
510.
(13) The silyloxy group in the cyclopentanone ring underwent ꢀ-elim-
ination under these mildly basic conditions to give the cyclopentenone
product as we had seen earlier in the synthesis of PECPC.8
OL8014804
Org. Lett., Vol. 10, No. 19, 2008
4209