The conjugate addition-Peterson olefination reaction for the preparation of cross-conjugated cyclopentenone, PPAR-γ ligands

Article abstract of DOI:10.1039/b814619e

5-Alkylidenecyclopent-2-enones 15a-q may be prepared via a conjugate addition-Peterson olefination sequence, best achieved in one-pot, using exo-2-trimethylsilyl-3a,4,7,7a-tetrahydro-4,7-methanoinden-1-one 12, followed by a retro-Diels-Alder reaction. The geometry of the exocyclic alkene may be controlled according to the use of organometallic species in the conjugate addition step; organocuprate reagents are found to selectively lead to the formation of E-exocyclic alkene adducts, whereas Grignard reagents favour the formation of Z-alkenyl isomers. The use of enantiomerically enriched 12, accessed from an asymmetric Pauson-Khand reaction, affords the corresponding enantioenriched 5-alkylidenecyclopent-2-enones and this approach is exemplified by the short, stereoselective total syntheses of two cyclopentenone phytoprostanes 51 and 13,14-dehydrophytoprostane J₁65. The ability of this family of synthetic compounds to activate the peroxisome proliferator activated receptor-γ is reported.

Full text of DOI:10.1039/b814619e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The conjugate addition–Peterson olefination reaction for the preparation of
cross-conjugated cyclopentenone, PPAR-c ligands†
Mazhar Iqbal,‡a Patricia Duffy,b Paul Evans,*b George Cloughley,c Bernard Allan,§c Agust´ı Lledo´,d
Xavier Verdaguerd and Antoni Rierad
Received 21st August 2008, Accepted 23rd September 2008
First published as an Advance Article on the web 4th November 2008
DOI: 10.1039/b814619e
5-Alkylidenecyclopent-2-enones 15a–q may be prepared via a conjugate addition–Peterson olefination
sequence, best achieved in one-pot, using exo-2-trimethylsilyl-3a,4,7,7a-tetrahydro-4,7-methanoinden-
1-one 12, followed by a retro-Diels–Alder reaction. The geometry of the exocyclic alkene may be
controlled according to the use of organometallic species in the conjugate addition step; organocuprate
reagents are found to selectively lead to the formation of E-exocyclic alkene adducts, whereas Grignard
reagents favour the formation of Z-alkenyl isomers. The use of enantiomerically enriched 12, accessed
from an asymmetric Pauson–Khand reaction, affords the corresponding enantioenriched
5-alkylidenecyclopent-2-enones and this approach is exemplified by the short, stereoselective total
syntheses of two cyclopentenone phytoprostanes 51 and 13,14-dehydrophytoprostane J1 65. The ability
of this family of synthetic compounds to activate the peroxisome proliferator activated receptor-g is
reported.
Another facet to this story concerns the recent discovery that
Introduction
structurally similar, albeit racemic, prostanoid natural products
Natural products possessing the cyclopentane structure are widely
are also formed, particularly during events of cellular (oxidative)
distributed throughout the animal, bacterial and plant kingdoms.
stress. Such compounds, termed isoprostanes to distinguish them
A subset of this broader class is the family of compounds derived
from the enzymatically derived prostaglandins, are thought to be
from fatty acids that possess a cyclopentenone group (see Fig. 1).
derived from the non-enzymatic oxidation of membrane-bound
polyunsaturated fatty acids.5 These natural products provide a
link to the popular current concept that dietary manipulation of
the levels of polyunsaturated fatty acids might protect cells from
the destructive influence of reactive oxygen species.6
These compounds are intimately involved in several important and
related cellular processes, including mediation of the inflammatory
response and involvement in cellular defence pathways.1 Evidence
indicates that these biological effects may directly result from
the covalent modification of cysteinyl groups present in proteins
following conjugate addition to the electrophilic a,b-unsaturated
ring.2 Biosynthetically, however, the origins of these structurally
related compounds appear to be different. PGA2 2 and J2 3, for
example, may be formed in mammalian systems following the
oxidation and subsequent dehydration of PGF2a 1.1 PGF2a 1 itself
being derived from the arachidonic acid-cyclooxygenase cascade.3
Further allylic dehydration of PGJ2 3, which has been reported
to occur in serum albumin, generates D12,14-15-deoxy-PGJ2 4.4
The cross-conjugated juxtaposition of the dienone unit present
in 4 is also found in several other, structurally related, naturally oc-
curring compounds. Again these compounds, from different bio-
logical sources, appear to possess interesting biological properties.1
In plants, 12-oxophytodienonic acid 5, the biosynthetic precursor
to jasmonic acid, is derived from linolenic acid via the allene
oxide synthase pathway.7 Dehydrophytodienoic acid 6 (13,14-
dehydrophytoprostane J1 originally also called chromomoric
acid)8 has been isolated from several plant sources and, despite
the structural similarity to 5, it is not known whether this
compound is derived enzymatically, or from the plant version of
the isoprostane pathway.9 The plant congeners of the mammalian
isoprostanes are termed phytoprostanes.9 Marine derived natural
products related to 4 and 6 are also known and clavulone 7 is
a representative example.10 These compounds typically display
greater oxidation than their terrestrial counterparts and often
include halogenation.10 Recently a new class of cross-conjugated
cyclopentenone has been discovered that possesses an allylic
epoxide in the alkylidene side chain. Such racemic compounds,
exemplified by 8, are, again believed to be derived via the
isoprostane pathway.11
aDepartment of Chemistry, University of Liverpool, Liverpool, UK L69 7ZD
bCentre for Synthesis and Chemical Biology, School of Chemistry and
Chemical Biology, University College Dublin, Dublin 4, Ireland. E-mail:
paul.evans@ucd.ie
cConway Institute, University College Dublin, Dublin 4, Ireland
dUnitat de Recerca en S´ıtesis Asime`trica, Institute for Research in
Biomedicine (IRB Barcelona), Departament de Qu´ımica Orga`nica, Uni-
versitat de Barcelona, Parc Cient´ıfic de Barcelona, c/Baldiri Rexiac 10,
08028, Barcelona, Spain
† Electronic supplementary information (ESI) available: Full experimental
details and spectroscopic data. CCDC reference number 219462. For ESI
and crystallographic data in CIF or other electronic format see DOI:
10.1039/b814619e
‡ Current address: National Institute for Biotechnology & Genetic Engi-
neering (NIBGE), PO Box 577, Jhang Road, Faisalabad, Pakistan.
Primary prostaglandins, such as the historic PGF2a, elicit
their potent biological activities via their direct interaction with
specific receptors on the surface of the cell plasma membrane.1
§ Current address: Dept. Molecular Biology, Genentech Inc. 1 DNA Way,
MS 37, South San Francisco, Ca 94080-4990, USA.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4649–4661 | 4649
©

View Article Online
Fig. 1 Representative cyclopentenone fatty acid derived natural products.
Typically such C20 arachidonic acid derivatives possess pro-
inflammatory and muscle contracting effects.1,3 In contrast, the
latter, cyclopentenone-containing members of the prostaglandin
family, such as 2, 3 and particularly 4, appear to counteract these
pro-inflammatory effects.1 Evidence indicates that they may do
this by interfering with the processes of gene transcription and
ultimately protein translation via their interaction with several
transcription factors.1,3
the peptide adiponectin, which controls formation of adipocytes
(fat cells). In 1995 it was demonstrated that the unsaturated
cyclopentenone, D12,14-15-deoxy-PGJ2 4, bound to and activated
PPAR-g and consequently the suggestion was made that 4 was the
natural ligand for this orphan receptor.18 This proposal, however,
has remained a contentious matter of debate, primarily since
the natural detectable levels of 4 have never been in the same
range as the levels needed to activate PPAR-g appreciably.19 One
plausible proposal explaining these detection problems and the
apparent low in vivo concentration of 4 concerns its reactivity
and the fact that free D12,14-15-deoxy-PGJ2 4 may be removed
from circulation via either conjugation to reduced glutathione
(GSH)1,2,20 and/or by association with serum albumin.20 The
reactivity of 4 with GSH and the stability of the corresponding
glutathione adduct has been investigated by Noyori and co-
workers.2 Furthermore, a relationship and interplay between
NFkB and PPAR-g has recently been uncovered which seems to
demonstrate that activation of PPAR-g leads directly to inhibition
of NFkB possibly by a nuclear export mechanism.21
Although the precise mechanisms are different, in plant species
the cyclopentenoid natural products, exemplified by 5 and 6,
appear to play important roles involving the defence and home-
ostasis of their host.9 For example, it has been reported that
either wounding of the plant, or pathogen attack, induces the
accumulation of cyclopentanoid compounds. These compounds
activate defence related genes and induce detoxification responses.
Even the exogenous application of phytoprostanes 5 and 6 initiates
the biosynthesis of secondary metabolites such as phytoalexins.22
Based on the interesting profile of biological activities demon-
strated by natural products possessing the cross-conjugated cy-
clopentenone structural motif we became interested in developing
a modular synthetic method that enabled the straightforward
preparation of analogues.23 The ultimate aim was to evaluate
these analogues in terms of their ability to activate PPAR-g in
the hope that a relationship between structure and activity could
be uncovered. We envisaged that the structural motif present in this
class of compounds could be efficiently accessed via a conjugate
addition–Peterson olefination sequence performed on a suitable
masked a-silyl cyclopentadienone synthon 9 (see Fig. 2). This
process would serve to install both the alkyl and alkylidene (C-8
Nuclear factor kappa B (NF-kB), discovered by Baltimore
in 1986,12 plays a pivotal role in the inflammatory response
and its downstream gene products include i-NOS, COX-2 and
various cytokines responsible for initiating and perpetuating the
effects of inflammation.13 For this reason NF-kB has emerged
as a promising target for the development of compounds aimed
at inhibiting the excessive inflammation associated with various
conditions.13 In 2000 it was shown by Karin and Santoro that
unsaturated prostanoids inhibit NF-kB by binding to a thiol
residue contained in the activation loop of the b-subunit of the
IkB kinase IKK.14 This kinase is responsible for the activation of
NF-kB in the cytoplasm of the cell, which, following activation,
relocates to the nucleus and initiates gene transcription.13,14
Cyclopentenone prostaglandins also interfere with the heat shock
response process, which is a protective mechanism that cells use
under stressful conditions including, but not exclusively, extremes
of temperature.1,15 This process is mediated by a transcription
factor (heat shock factor) the activation of which leads to the
accumulation of proteins termed heat shock proteins (HSP), in
particular HSP70, whose role as a molecular chaperone has been
documented.15
The final important cyclopentenone prostaglandin target is
another transcription factor; peroxisome proliferator activated
receptor g (PPAR-g). This, so-called, orphan nuclear receptor was
discovered in the late eighties16 and soon after was reported to be
the molecular target for a clinically useful class of drugs for the
treatment of the symptoms of type 2 diabetes.17 Several members of
this class of compounds, known as the thiazolidindiones (TZDs),
ultimately became medicines for the treatment of this steadily
growing worldwide condition (see later). PPAR-g’s role appears
to be in the control of lipid and glucose homeostasis and its
gene products include enzymes involved in lipid oxidation and
4650 | Org. Biomol. Chem., 2008, 6, 4649–4661
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 2 Retrosynthetic analysis of cross-conjugated cyclopentenones based on a conjugate addition–Peterson olefination approach (M = metal).
and C-12) side-chains, the latter being at the desired oxidation
level. The reaction partners for this proposed sequence would
be a functionalised organometallic species represented by 10 and
an a,b-unsaturated aldehyde 11 and 9. A cyclopentene unit was
chosen to mask the endocyclic carbon–carbon double bond in the
cyclopentenone moiety, since, based on seminal studies by Stork
and Roussec,24 this group may be readily removed by a retro-Diels–
Alder process either under Lewis acidic, or flash vacuum pyrolysis
conditions. Furthermore due to its rigid, conformationally locked
molecular architecture high levels of stereoselectivity have been
observed following related conjugate addition reactions.25 Obvious
advantages in this approach include expediency and the ability to
readily vary the structures of the pendant groups in a modular
fashion. We also felt that there was a high possibility that the
conjugate addition reaction and the Peterson olefination could be
carried out sequentially, in the same reaction vessel. This efficient
one-pot transformation has been virtually ignored following its
initial disclosure in 1984,26 mainly because its success remained
limited to certain substrates. We have reported preliminary results
in this area.27
lacetylene 13 and norbornadiene either under our microwave
conditions,28 or traditional thermal conditions.29 Traces of the cor-
responding endo-diastereomer were detected in small amounts but
this impurity may be removed following column chromatography,
or recrystallisation from hexane (Scheme 1).
Pleasingly compound 12 readily participated in conjugate addi-
tion reactions on treatment with either organocuprate reagents
or with Grignard reagents in the presence of copper(I) salts.
These conjugate adducts could be isolated and characterised on
protonation (see for example Scheme 3) but we were, however,
delighted to find that the resultant intermediate enolate efficiently
participated in a one-pot Peterson olefination reaction with a vari-
ety of structurally diverse aldehydes.¶ Table 1 summarises the one-
pot conjugate addition–Peterson olefination reactions performed.
Initial experiments focused on the addition of Gilman’s cuprate to
12 in ether (Entry 1), or THF. This reaction was found to proceed
◦
◦
to completion between -78◦ C and -5 C. The reaction mixture
was then re-cooled to -78 C and benzaldehyde was added. On
warming, the formation of a more polar, UV active spot was
detected which on isolation proved to be the hoped for exocyclic
dienone 14a. One geometric isomer and one diastereomer were
detected and the structure of this solid compound 14a was
Results and discussion
Crucially, in relation to this proposed sequence, ( )-exo-2-
trimethylsilyl-3a,4,7,7a-tetrahydro-4,7-methanoinden-1-one 12 is
readily available in high yield and on multi-gram scale via an
intermolecular Pauson–Khand reaction between trimethylsily-
¶ Classical Peterson olefination reactions may be performed following
initial addition of LDA to the conjugate adduct followed by benzalde-
hyde. However, in our hands the yields for this two-step process were
considerably lower than the reported, in situ, method.
Scheme 1 Proposed three-step sequence for the construction of cross-conjugated cyclopentenones.
Table 1 The one-pot conjugate addition–Peterson olefination reaction of exo-12
Entry RMa
R¢
Adduct Yieldb E : Zc
Entry RMa
R¢
Adduct
Yieldb
E : Zc
65 : 35
1
2
Me2CuLi
Ph
Ph
Ph
Ph
14a
14a
14b
14c
14d
14e
14f
14g
14h
14i
93%
86%
91%
84%
92%
45%
94%
45%
13%
83%
68%
88%
93%
>95 : 5
75 : 25
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
90 : 10
75 : 25
14
15
16
17
18
19
20
21
22
23
24
25
Me2CuLi
MeMgBr
i-Pr
i-Pr
n-Hex
n-Hex
Ph
Ph
i-Pr
MeO2C(CH2)5
MeO2C(CH2)5
t-Bu
PhCOMe
(CH2)5CO
14l
14l
61%
83%
82%
69%
81%
94%
88%
81%
57%
—
MeMgBr
n-Bu2CuLi
n-Oct2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
Me2CuLi
MeMgBr
30 : 70
>95 : 5
30 : 70
60 : 40
95 : 5
50 : 50
>95 : 5
35 : 65
—
3
4
5
6
7
8
9
n-Bu2CuLi
n-BuMgCl
i-PrMgCl
VinylMgBr
VinylMgBr
n-Oct2CuLi
n-OctMgBr
Me2CuLi
14m
14m
14n
14o
14p
14q
14q
14r
14s
14t
4-O2NC6H4
4-MeOC6H4
2-Furyl
2-Pyridyl
N-Me-3-indole
N-Ts-3-indole
10
11
12
13
=
E-PhCH CH
14j
Me2CuLi
Me2CuLi
—
—
—
—
=
=
E-MeCH2CH CH 14k
E-MeCH2CH CH 14k
a Conditions: 2 eq. RLi, 1 eq. CuI, Et2O; or: 1.5 eq. RMgBr–Cl, 10 mol% CuI, Et2O (for further details see Experimental section). b Yield following
purification by flash column chromatography. c Ratio determined by 1H-NMR spectroscopy of crude adducts 14.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4649–4661 | 4651
©

View Article Online
unambiguously confirmed by X-ray diffraction (Fig. 3).30 This
indicated that the formation of an E-alkene had occurred and that
the initial conjugate addition had taken place on the less sterically
encumbered, lower face. Based on isolated yields we found that the
use of ether as a solvent for this process was preferable and further
investigations focused on its use. Addition of a solution of 12 in
ether to a mixture of methylmagnesium bromide in the presence
of 10 mol% copper(I) iodide again resulted in smooth conjugate
addition between -78 C and -5 C (Entry 2). The intermediate
enolate was re-cooled to -78 ◦C and benzaldehyde was added.
In terms of conversion the reaction proceeded equally effectively
under these conditions, however, in this instance inspection of
the crude 1H-NMR spectrum indicated the presence of two
compounds in a ratio of 75 : 25. The minor compound was
separable and proved to be the corresponding Z-adduct 14a.
Alternative alkyl organocuprate reagents proceeded in a similar
fashion with benzaldehyde (Entries 3 and 4); in each case only the
E-alkenyl isomer was detected. The use of alternative aromatic
and heteroaromatic aldehydes was successfully investigated in
conjunction with Me2CuLi (Entries 5–10) and again only one
geometrical isomer was detected for the newly formed double
bond. Based on the yields encountered it appears that electron
rich and Lewis basic aldehydes (i.e. Entries 6, 8 and 9) do
not participate as effectively as their more reactive, electron
poorer counterparts (Entry 5 and 10). The process described also
proceeded efficiently with a,b-unsaturated aldehydes (Entries 11,
12 and 13), thereby providing a strategy for the installation of
the exocyclic dienone unit present in natural products 4, 6 and 7
(Fig. 1). The 1,4-addition of isopropylmagnesium chloride, and
vinylmagnesium bromide proceeded smoothly using catalytic CuI
(10 mol%) and in both cases the corresponding magnesium enolate
reacted efficiently with benzaldehyde, affording enones 14n and
14o as separable mixtures of E- and Z-stereoisomers (Entries 18
and 19).
At this stage it was of interest to ascertain whether aldehydes
possessing a-protons would participate in the process, or whether
intermediate 17 might prove to be incompatible with such,
potentially acidic, reaction partners. This concern proved to be
unfounded and freshly distilled isobutyraldehyde gave good yields
of the corresponding adducts 14l using either the Gilman cuprate,
or MeMgBr (Entries 14 and 15). Interestingly, the stereochemical
integrity of the double bond formed was much less well defined
in these instances and significant amounts of the readily separable
Z-stereoisomers were formed. As before (Entries 2) this lack of
stereoselectivity was most marked when the organomagnesium
reagent was employed. Finally, this reaction sequence described
was then applied to a rapid synthesis of ( )-TEI-9826 15q, an
analogue of PGA2 2 which has been evaluated in vivo as a potential
anticancer agent.31 Thus, addition of either (n-Oct)2CuLi, or n-
OctMgBr (10 mol% CuI) to 12, followed by addition of methyl
7-oxoheptanoate32 afforded the exocyclic enone 14q in good yield
(Entries 21 and 22). This latter reaction indicated that additional
carbonyl functionality in the aldehyde Peterson olefination partner
is tolerated under the reaction conditions. The use of pivaldehyde,
acetophenone and cyclohexanone (Entries 23–25) only afforded
the corresponding conjugate adduct and none of the hoped-for
product of Peterson olefination. Based on these failures it appears
that steric effects, in addition to the electronic reactivity of the
carbonyl species, play an important role in this process (see also
Scheme 5).
◦
◦
The retro-Diels–Alder reactions of the norbornadiene adducts
E-14a–14q were performed in dichloromethane at 40 ◦C using
MeAlCl2 and an excess of maleic anhydride as a cyclopentadiene
trap (Table 2).25,33 Generally, this method provided good yields of
the corresponding cross-conjugated cyclopentenones E-15a–15q
with only minimal alkylidene isomerisation. Notable exceptions
were compounds containing basic azo-functionality (Entries 8
and 9); in these examples none of the cyclopentenone products
15h and 15i were detected. It was also of interest that, following
Fig. 3 X-Ray crystal structure of adduct 14a (Ortep representation).
Table 2 The preparation of cross-conjugated prostanoid mimics 15 by retro-Diels–Alder cycloaddition
Entrya
R
R¢
Yieldb
E : Zc
Entry
R
R¢
Yieldb
E : Zc
90 : 10
33%g
92%h
82%
67%i
84%
33%j
87%
73%
73%
86%
=
=
1
2
3
4
5
6
7
8
9
Me
Me
n-Bu
n-Oct
Me
Me
Me
Me
Ph
Ph
Ph
Ph
E-15a
Z-15a
E-15b
E-15c
E-15d
E-15e
E-15f
E-15g
E-15h
E-15i
E-15j
84%
64%d
83%
76%
85%
68%
64%
NRe
NRe
28%f
78%
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
>95 : 5
—
12
13
14
15
16
17
18
19
20
21
Me
Me
Me
Me
n-Bu
n-Bu
i-Pr
Vinyl
Vinyl
n-Oct
E-MeCH2CH CH
E-MeCH2CH CH
i-Pr
i-Pr
n-Hex
n-Hex
Ph
Ph
E-15k
E-15k
E-15l
>95 : 5
90 : 10
90 : 10
>95 : 5
65 : 35
>95 : 5
>95 : 5
90 : 10
80 : 20
Z-15l
4-O2NC6H4
4-MeOC6H4
2-Furyl
2-Pyridyl
N-Me-3-indole
N-Ts-3-indole
E-15m
Z-15m
E-15n
E-15o
E-15p
E-15q
Me
Me
Me
—
75 : 25
90 : 10
i-Pr
MeO2C(CH2)5
10
11
=
E-PhCH CH
a Conditions unless otherwise stated: 1–1.5 eq. MeAlCl2, 5–10 eq. maleic anhydride, DCM, 40 ◦C (for further details see Experimental section). b Yield
following purification by flash column chromatography. c Ratio by 1H-NMR spectroscopy. d 84% based on recovered E-14a. e NR = no reaction. f 1 eq.
MeAlCl2, 5 eq. N-methyl maleimide, DCM, m-wave, 70 ◦C, 25 min, 85%. g 83% based on recovered 14k. h 1.5 eq. MeAlCl2, 15 eq. maleic anhydride, DCM,
m-wave, 120 ◦C, 70 sec. i 89% based on recovered E-14l. j 95% based on recovered E-14m.
4652 | Org. Biomol. Chem., 2008, 6, 4649–4661
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
employment of this Lewis acid based methodology, the purified
Z-exocyclic enones 14a, 14l and 14m (Entries 2, 15 and 17) under-
went significant alkene isomerisation, affording predominantly the
E-cross-conjugated cyclopentenone products 15a, 15l and 15m.
For example, treatment of Z-14l gave 15l in 67% yield (E : Z,
90 : 10) and 22% of isomerised starting material, E-14l. The
functionalised adduct 14q was smoothly converted to the target
dienone 15q in 86% yield.
In certain instances it proved advantageous to use alternative
reaction conditions. For example, the use of N-methyl maleimide
in conjunction with microwave heating gave E-15i in 85% yield,
whereas under standard conductive heating the isolated yield was
significantly lower (28%). The use of the alternative dienophile in
this case facilitated product purification (Entry 10). Microwave
irradiation for relatively short periods was also found to be
useful in the conversion of E,E-14k into the corresponding cross-
conjugated trienone E-15k, where both improved product yields
and reduced alkylidene isomerisation was observed (Entries 12
and 13).
In summary, the one-pot conjugate addition–Peterson olefina-
tion reaction was employed in order to efficiently install both the
2-alkylidene and 3-alkyl side-chains present in cross-conjugated
cyclopentenone natural products. The stereochemistry of the 2-
alkylidene side-chain depends on the type of organometallic
reagent employed in the conjugate addition reaction. Use of
organocuprate reagents afforded high levels of E-selectivity,
whereas copper-catalysed Grignard reagents gave significant
amounts of the corresponding Z-isomer. In all examples the
separation of geometrical alkylidene isomers proved possible by
flash column chromatography.
It seems reasonable to speculate that the observation concerning
the selective formation of E-Peterson olefination products with
organocuprate reagents versus the formation of Z-adducts (albeit
with variable selectivity) with Grignard reagents is due to a
change in counterion in the intermediate 17 (i.e. Li → MgBr)34
and that the identity of this enolate influences the path of the
subsequent Peterson reaction. Another observation concerning
the stereoselectivity was that the preponderance for Z-alkylidene
formation increases corresponding to the size of the aldehyde. For
this reason the one-pot conjugate addition–Peterson olefination
process using isobutyraldehyde was further investigated. Using
the Gilman cuprate to form 17 and subsequent addition of
isobutyraldehyde at -78 ◦C afforded the adduct 14l as a separable
mixture of E- and Z-isomers (E : Z ; 65 : 25 determined
1
by H-NMR spectroscopy of the crude reaction mixture). The
corresponding reaction with methylmagnesium bromide and a
catalytic amount of copper(I) iodide preferentially formed Z-14l
(E : Z; 30 : 70), which seemed to be relatively independent of the
temperature (i.e. -78 ◦C to 0 ◦C) at which the aldehyde was added.
Following addition of Gilman’s cuprate to 12 the intermediate 17
(M = Li) was treated with 2 equiv. of ZnCl2. Subsequent addition
of isobutyraldehyde, at -78 ◦C, afforded 14l in low yield (29%) but
predominantly as Z-14l (E : Z; 20 : 80). This serves to highlight,
again, that the counter ion plays a pivotal role in the stereochemical
outcome of the Peterson olefination process (Scheme 2).
The initial carbon–carbon bond forming reaction could dictate
the stereochemical outcome dependant on the identity of M.35
We feel that this initial bond formation is likely to occur on
the Si-face of 17 (for the diastereoselective protonation of 17 see
Scheme 3, compound 20); from this point the initial diastereomeric
Scheme 2 Mechanistic considerations in the conjugate addition–Peterson olefination reaction.
Scheme 3 Organocuprate and zinc conjugate addition reactions of 12.
The Royal Society of Chemistry 2008 Org. Biomol. Chem., 2008, 6, 4649–4661 | 4653
This journal is
©

View Article Online
adducts 18 and 19 would then generate the corresponding products
of Peterson olefination via a concerted silyloxy-elimination type
pathway. Although this seems to be a plausible explanation and
chelate (Zimmerman-Traxler–type) and open transition states
can be postulated, the situation is somewhat clouded by the
possibility that the initial diastereomeric adducts may undergo an
initial Brook-type rearrangement, enabling bond rotation prior
to elimination: it is also possible that the adducts may undergo
post-Peterson isomerisation to some extent.
At this stage we became interested in applying this one-
pot procedure for the preparation of more complex molecules.
Therefore, the use of organozinc reagents was considered since it is
well-documented that such species not only efficiently participate
in conjugate addition reactions but that they are also able to carry
functionality (such as carboxylic acid esters) not compatible with
organolithium and magnesium reagents.36 Initial investigations
were conducted to determine whether 12 underwent conjugate
addition with diethylzinc under standard conditions (Scheme 3).36
Whereas cyclopentenone 12 readily undergoes conjugate addition
under standard organocuprate conditions the analogous organoz-
inc reaction was sluggish under several literature conditions and
yields of only 10–40% of adduct 21 were obtained.36 Optimum
yields of 21 were achieved using Noyori’s sulfonamide promoted
conditions,37 however, due to the low conversion of 12 into
21 and the use of additional additives we felt that the use of
organozinc species was not attractive in terms of our one-pot
Peterson approach. The stereochemistry of the newly formed a-
keto centre was probed using NOE experiments. Irradiation of
the methyl silicon substituent gave an enhancement to one of the
diastereotopic methylene bridging protons thereby indicating that
protonation of the enolate 17 occurs on the Si face.
bond and elimination. Optimum yields for this process were
achieved following a one-pot reaction protocol in which all the
reagents were added simultaneously (Scheme 4).
It has been shown that lithiodithiane, in the presence of a
polar aprotic additive such as HMPA or DMPU,38 undergoes
1,4-conjugate addition to various enones. In our hands, although
we could successfully generate adduct 23, resulting from 1,4-
conjugate addition and protonation, attempts to affect in situ Pe-
terson olefination were unsuccessful. The use of malonate anions
were also studied and again, although under standard conditions39
the conjugate addition was successful, we were unable to link this
process to the Peterson olefination in order to prepare adducts
with a functional handle for further elaboration. It seems likely
that under these conditions the intermediate enolate (of the type
17) is formed reversibly and then undergoes protonation, which in
turn facilitates the protodesilylation process, thereby affording 24.
Interestingly, under similar reaction conditions, in the absence of
dimethylmalonate, 12 underwent efficient desilylation, a process
we feel is likely to proceed via reversible methoxide conjugate
addition.
At this stage all our experience of conducting the one-pot
conjugate addition–Peterson olefination reaction had been us-
ing the same compound, namely exo-2-trimethylsilyl-3a,4,7,7a-
tetrahydro-4,7-methanoinden-1-one 12. Consequently, it was of
interest to investigate whether alternative a-silicon bearing a,b-
unsaturated carbonyl compounds would also participate in this
tandem reaction. Therefore, compounds 26 and 27 possessing
different silicon substituents, were prepared using an inter-
molecular PKR (Scheme 5). Compound 28, in which a methyl
substituent is present in the b-position was also prepared and
these alternative cyclopentenone adducts were subjected to the
reaction conditions described in Table 1. Whereas, compound
12 efficiently generated 14k, the product of methyl conjugate
addition and Peterson olefination with E-pent-2-enal (Entries 1
and 2), substrate 26 gave solely the conjugate adduct 30 with
both organocuprate and Grignard reagents (Entries 3 and 4).
We attributed the recalcitrance of the intermediate enolate to
participate in the Peterson olefination reaction on the basis of
the increased steric bulk of the triisopropylsilyl unit. Use of
the less-bulky tert-butyldimethylsilyl group (Entry 5) led to a
The use of alternative nucleophilic species as reaction partners
in the conjugate addition–Peterson olefination reaction with the
same enone 12 were investigated. The results of these studies
are summarised in Scheme 4. It was found that a THF solution
of PhSLi and benzaldehyde, or isobutyraldehyde generated an
uncharacterised diastereomeric mixture of 22 (60 : 40) in 96%
and 90% yields respectively. The formation of this adduct appears
to result from an initial conjugate addition–Peterson olefination
process followed by nucleophilic attack at the exocyclic double
Scheme 4 Attempted conjugate addition–Peterson olefination using alternative nucleophilic reagents.
4654 | Org. Biomol. Chem., 2008, 6, 4649–4661 This journal is The Royal Society of Chemistry 2008
©

View Article Online
Scheme 5 Attempted conjugate addition–Peterson olefination using different PK adducts.
mixture of the conjugate adduct 31 (30%) and the Peterson
product 14k (41%) suggesting again that steric factors play a
significant role in the Peterson olefination reaction of this series of
compounds. We found that substrate 28 was particularly reluctant
to undergo conjugate addition under standard conditions, indeed
this compound only underwent conjugate addition in the presence
of TMSCl whereupon 32 was isolated in 44% yield following
silyl enol ether cleavage. Based on the stereoselective formation
of 20, we assume that the formation of 30 to 32 generated the
diastereomer resulting from protonation on the Si face of the
enolate although this was not fully investigated.
Alternative substrates do, however, effectively participate in
the process (Scheme 6). Monocyclic trimethylsilanes 35 and 36
were prepared in three steps according to a four-step literature
procedure40 and under the standard conditions, indicated in
Table 1, cyclopentenone 35 and cyclohexenone 36 afforded the
trans-enones 37, 38 and 39 stereoselectively albeit in variable,
unoptimised yields (32–94%).
The chemoselective alkylidene reduction of the products of
conjugate addition–Peterson olefination was investigated since
on reduction followed by retro-Diels–Alder cycloaddition, 4,5-
dialkyl substituted cyclopentenones of the type present in PGA2
2 and J2 3 would be accessed—such species would be of
potential interest in terms of biological comparison with their
cross-conjugated analogues. It has been reported that lithium
tri-tert-butoxyaluminium hydride preferentially participates in the
1,4-selective reduction of a,b-unsaturated ketones resembling
14.24b Therefore, we investigated whether we could utilise this
reagent for the preparation of dialkyl-substituted cyclopentenones
of the type 42. It was found that the use of THF as opposed
to Et2O was crucial in order to obtain reasonable 1,4-reduction
selectivity in the formation of 41b and 41m. However, significant
amounts of the corresponding (separable) 1,2-reduced products
40b and 40m were still obtained (Scheme 7). The stereochemistry
of the newly formed stereogenic centre was probed using NOE
experiments; irradiation of the benzylic protons enhanced one of
Scheme 6 The conjugate addition–Peterson reaction with cyclopent-2-enone 35 and cyclohex-2-enone 36.
Scheme 7 The conjugate reduction of exocyclic Peterson olefination adducts 14b and 14m.
The Royal Society of Chemistry 2008 Org. Biomol. Chem., 2008, 6, 4649–4661 | 4655
This journal is
©

View Article Online
the diastereotopic bridging methylene protons (see also Scheme 3).
Subsequent retro-Diels–Alder cycloaddition, under the conditions
described (Table 2), afforded the corresponding cyclopentenones
42b and 42m in good yield.
conditions41 no conversion of either compound could be achieved
(Scheme 8). This failure was attributed to steric congestion
around the reactive centre; therefore, alternative approaches were
explored.
Due to the rigid conformation of exo-12, the diastereoselec-
tivity of the conjugate addition reaction (and, for that matter,
protonation of the intermediate enolate) is high. Consequently, if
enantiomerically enriched forms of this compound were available,
the preparation of either enantiomer of cross-conjugated com-
pounds 15a–q could be envisaged using the chemistry described.
There are several reasons why a stereoselective synthesis of the
types of compound illustrated in Fig. 1 is of interest. Firstly,
from a biological perspective a comparison of the activities of
racemic with enantioenriched materials is important. From a
chemical perspective the configurational stability, or otherwise, of
the doubly allylic stereogenic centre is also of interest. Finally,
the question of stereogenicity and stereostability in this class
of compound is of interest more generally since, although the
prostanoids are synthesised following an enzymatically controlled
pathway, the isoprostanes are formed in racemic form via the
non-stereoselective oxidation of unsaturated fatty acids. The
stereochemistry of the plant congeners is currently not completely
known. Studies have shown that 12-oxophytodienoic acid 5 is
formed in optically active form via the jasmonic acid allene
oxide synthase–cyclase pathway. However, despite the structural
similarities, it has been proposed that compound 6 is not formed
following this pathway but is formed via the plant variant of
the isoprostane pathway and is therefore, racemic. Following the
original natural product isolation an optical rotation of +20 for
the isolated dPPJ1-I methyl ester 6 was reported by Bohlmann
and co-workers.8 Consequently, we felt that it was of interest
to investigate the synthesis of such compounds in non-racemic
form.
Recently we have reported several P,S chiral ligands for the
asymmetric intermolecular Pauson–Khand reaction. The chiral-
pool derived enantiopure ligand PuPHOS (45) and the PNSO
ligand 46 both provided excellent results in the cyclisation with
norbornadiene.42 In the case of 46 both enantiomers are readily
available in a multigram scale since they are prepared from the
commercially available tert-butylsulfinamide. Ligand 46 can be
stored either as its borane complex or as a borane-free ligand
(Scheme 8). Release of the borane protection group in situ with
DABCO and reaction of either 45 or 46 with the acetylene-
hexacarbonyl dicobalt complex afforded a biased mixture of
diastereomeric complexes (3 : 1 for 45, 12 : 1 for 46) from which
the major complex 47 can be isolated by crystallisation.
The subsequent Pauson–Khand cyclisation from 47 enabled the
preparation of the desired enantiomer of 12. Thus, starting from
the crystalline complex 47 the oxidative Pauson–Khand reaction
with norbornadiene afforded (+)-12 in high yield and 96–97% ee.42
Recrystallisation of this material from hexane gave (+)-12 in 99%
ee. In our case, subsequent synthetic studies focused on the use
of this enantiomer since conjugate addition–Peterson olefination
would generate compounds possessing the S-configuration present
in the naturally occurring cross-conjugated cyclopentenones (see
Fig. 1).
Part of the argument in favour of the non-enzymatic phyto-
prostane pathway is that regioisomeric compounds 51 are formed
in addition to compounds exhibiting the type of structure exem-
plified by 5 and 6.9 We employed a similar approach used in the
preparation of TEI-9826 14q (Table 1, Entry 21) for the synthesis of
ethyl phytoprostane 51 as the methyl ester (Scheme 9). Thus, both
geometrical isomers of aldehyde 49 were prepared stereoselectively
from 8-bromooctanoic acid and propargyl alcohol 48 using a
known route.43 Although it was found that Z-49 readily underwent
isomerisation both on prolonged storage in CDCl3 and on
silica gel during chromatographic purification, use of the crude
To this end the enzymatic kinetic resolution of racemic allylic
alcohol 43 and acetate 44 was investigated. This approach is
attractive from the perspective of an investigation into the
biological properties of this compound class since, in principle,
both enantiomers of 15 may be accessed. However, under standard
Scheme 8 Studies aimed towards the preparation of enantioenriched exo-12.
4656 | Org. Biomol. Chem., 2008, 6, 4649–4661 This journal is The Royal Society of Chemistry 2008
©

View Article Online
Scheme 9 Preparation of phytoprostanes 51.
Z-aldehyde directly in the conjugate addition–Peterson olefination
using ( )-12, gave the adduct 50 in good yield and stereoselectivity
(E,Z > 95%). Proton NMR spectroscopy confirmed retention
of the Z-alkene (J, 10.5 Hz). Finally, retro-Diels–Alder reaction
gave the phytoprostane natural product E,Z-51 (methyl ester)
as the 9-cis-geometrical isomer. It was found that the use of
short reaction periods and microwave irradiation was the key to
minimise isomerisation of the alkylidene bond formed following
the Peterson reaction. An identical synthetic approach using E-49
gave the corresponding 9-E stereoisomer (E,E-51). Following the
optimisation of this sequence in the racemic series an identical
approach was performed using (+)-12 and E-49. In this manner
(+)-E,E-51 was obtained in 56% overall yield for the two steps.
Analysis by chiral HPLC indicated that this material had an
enantiomeric excess of >99% thereby demonstrating that no
epimerisation of the doubly allylic, single stereogenic centre had
occurred during the retro-Diels–Alder reaction and purification.
Conversion of E,E-51 to the corresponding carboxylic acid E,E-
52 was efficiently achieved using hog liver esterase.44
Next we investigated the preparation of phytoprostane 6, and
its double bond stereoisomers employing analogous chemistry to
that reported by us previously for the synthesis of D12,14-15-deoxy-
PGJ1 61.45 Appropriately functionalised organometallic species
were prepared as illustrated in Scheme 10.
When the Grignard reagent 58 was employed in the tandem se-
quence, compound 62 was isolated (following Peterson olefination
with E-pent-2-enal) in 84% yield with an E : Z ratio of 65 : 35 for
the newly formed double bond (Scheme 11, Entry 2). In contrast,
use of corresponding cuprate 60 gave enone 62 in good yield as
essentially only the E-stereoisomer (E : Z; >95 : 5). As previously
observed the geometrical isomers of 62 were separable via flash
column chromatography. A standard deprotection–oxidation se-
quence was then employed to furnish the carboxylic acid in good
overall yields for the three-step sequence. Finally, unmasking of
the cyclopentenone group following an extremely rapid Lewis-acid
mediated retro-Diels–Alder reaction performed under microwave
irradiation gave ( )-6 in good yield. Alternatively, methylation
followed by the microwave-mediated retro-Diels–Alder process
afforded the methyl ester ( )-65 (39% overall). Saponification
of ( )-65 proceeded smoothly using hog liver esterase affording
carboxylic acid 6.
An identical sequence was employed using cuprate 60 and
(+)-12 in order to access enantioenriched enone (+)-65. All the
intermediates on this sequence demonstrated strongly dextroro-
tary optical rotations and chiral HPLC analysis of the product
(+)-65 (Chiralpak AS) indicated an enantiomeric excess of 94%,
again demonstrating that epimerisation of the doubly allylic
single asymmetric centre did not occur to any significant extent.
The dextrorotary reading obtained for 65 {[a]D +144 (c = 1.0,
CHCl3)} compares in sign with the literature value {[a]D +20
(c = 1.0, CHCl3)} recorded during the natural product isolation;8
however, the significant difference in magnitude suggests either
epimerisation during extraction from the natural source, or that
in nature, 6 is formed non-stereoselectively via the phytoprostane
pathway. Several geometric isomers of 6 have been synthesised
previously in racemic form by both the Bohlmann and Liu
groups8,46 and our synthetic compound corresponds with reported
literature values in terms of spectroscopic data.
Scheme 10 Preparation of silyl ether protected organometallic reagents.
The Royal Society of Chemistry 2008 Org. Biomol. Chem., 2008, 6, 4649–4661 | 4657
This journal is
©

View Article Online
Scheme 11 Use of the conjugate addition–Peterson olefination reaction in the preparation of cross-conjugated cyclopentenones possessing oxidised
alkyl substituents.
were compared with controls, including D12,14-15-deoxy-PGJ2 4
(Entry 15) and rosiglitazone 68 (Entry 16).
Activation of PPAR-c
Simple cross-conjugated compounds possessing either benzyli-
dene, or isopropylidene R¢ and methyl, or n-butyl side-chains
(Entries 1–3) did not significantly alter residual PPAR-g levels.
However, increasing the lengths of both side chains did lead to
compounds that activated this transcription factor. For example,
15m (R¢ = hexylidene and R = n-butyl) proved to be reasonably
effective in terms of PPAR-g activation. This effect was observed
particularly for the E-stereoisomer (Entry 4); Z-15m proved to be a
significantly less effective activator (Entry 5). Similarly, compound
42m, in which the exocyclic enone has undergone reduction (see
Scheme 7), also displayed diminished activation as compared to its
cross-conjugated analogue E-15m (Entry 6). It was subsequently
of interest to gauge whether the presence of an oxidised side chain
led to increased activity, particularly since the fatty acid derived
natural products possess this type of group (Fig. 1). Thus, E-15q
(TEI-9826) was studied and although this did increase the levels
of PPAR-g activation the additional functional group in the R¢
side-chain was found not to improve activation in comparison to
E-15m (Entry 7). The synthetic phytoprostane natural products
were then evaluated. Thus, it was found that both methyl ester
E-65 and carboxylic acid E-6 did indeed activate PPAR-g, in
the case of the methyl ester did this slightly more effectively,
possibly due to better cell permeation (Entries 8 and 9). Although
the levels of activation were higher than E-15m and E-15q the
EC50 value was not vastly improved. As before the corresponding
Z-alkylidene, Z-65 proved to be a much less effective agonist
compared to its E-isomer (Entry 10).49 Evaluation of (+)-65 (ee
94%) indicated that the presence of the 9-S stereogenic centre did
not improve activation compared to its racemic counterpart (Entry
11). The phytoprostane isomer 51 possessing a drastically different
orientation of the R and R¢ side-chains proved to be a poor
PPAR-g activator (Entry 12). Finally, our synthetic prostanoids E-
D12,14-15-deoxy-PGJ1 64 (Entry 13) and Z-D12,14-15-deoxy-PGJ1 64
(Entry 14) were evaluated and these compounds both proved to be
the most effective synthetic PPAR-g activators, particularly, again,
The cluster of so-called metabolic diseases, including type 2
diabetes hypertension and obesity represent a major global health
concern. For example, it is currently estimated that approximately
5% of the world’s population suffers from type 2 diabetes and
this value appears to be on the increase and consequently
has received widespread media coverage.17,47 All appear to be
closely linked to problems associated with our consumption
and ability to process fatty acids. Notably inflammation is a
characteristic additional problem associated with many disease
states including diabetes. Therefore, a dual acting compound
capable of inhibiting the development of inflammation via the
inhibition of NFkB and promotion of lipid metabolism and
insulin sensitivity via the activation of PPAR-g could be a useful
combination. The thiazolidindione (TZD) class of drugs for the
treatment of the symptoms of type 2 diabetes [such as rosiglitazone
66 (Avandia, GSK)] are currently widely prescribed to patients
suffering from the symptoms of this disease. These compounds
improve the patient prognosis by increasing the sensitivity of
cells towards insulin, possibly due to the release of a hor-
mone called adiponectin, indirectly by binding to and activating
PPAR-g.
Due to the structural similarity of our synthetic compounds to
the putative natural PPAR-g ligand, D12,14-15-deoxy-PGJ2 4, we
investigated whether our synthetic cross-conjugated compounds
possessed the ability to activate PPAR-g. In order to achieve
this a cell-based PPAR-g assay was employed based on the
Gal4-luciferase reporter gene.48 Synthetic compounds at different
doses were added to confluent human embryonic kidney cells
(HEK293T) transfected with murine PPAR-g and the luciferase
vector. Activation of PPAR-g was then determined based on
luminescence. Dose response curves were constructed and the
concentration at which 50% of the maximal activation (EC50) was
calculated. The level of PPAR-g activation was also determined
and was reported as a fold activation versus the baseline level
of activation (Table 3). Additionally, our levels of activation
4658 | Org. Biomol. Chem., 2008, 6, 4649–4661
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 3 Activation of PPAR-g by selected cross-conjugated D12,14-15-deoxy-PGJ2 mimics and phytoprostanes
Entry
Compound
R
R¢
PPAR-g
EC50a(SEM)b
Fold activationc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
E-15a
E-15l
E-15b
E-15m
Z-15m
42m
E-15q
E-65
E-6
Z-65
(+)-E-65
E-51
E-64
Z-64
4
Me
Me
Ph
i-Pr
Ph
ND
ND
ND
—
—
—
—
—
—
n-Bu
n-Bu
n-Hex
n-Hex
n-Hex
MeO2C(CH2)5
2.0 mM
7.0 mM
10.0 mM
2.0 mM
1.6 mM
2.2 mM
4.1 mM
2.4 mM
4.5 mM
0.5 mM
1.2 mM
0.8 mM
(0.09 mM)
(2.5 mM)
(4.0 mM)
(0.7 mM)
(0.09 mM)
(0.2 mM)
(1.2 mM)
(0.3 mM)
(2.2 mM)
(0.07 mM)
(0.11 mM)
(0.07 mM)
4.5
5.5
3.5
3.0
5.0
4.0
4.0
5.5
2.0
6.0
3.0
8.0
n-Bu
n-Bu
n-Oct
=
E-CH CHCH2Me
(CH2)7CO2Me
(CH2)7CO2H
(CH2)7CO2Me
(CH2)7CO2Me
Et
=
E-CH CHCH2Me
=
E-CH CHCH2Me
=
E-CH CHCH2Me
=
E-CH CH(CH2)7CO2Me
=
E-CH CH(CH2)4Me
(CH2)6CO2Me
(CH2)6CO2Me
R = Z-CH2CHCH(CH2)3CO2H;
=
E-CH CH(CH2)4Me
=
R¢ = E-CH CH(CH2)4Me
16
68
Rosiglitazone
1.4 mM
(0.05 mM)
40
a Based on an average of three experiments, each run in triplicate. Values calculated using PRISM software. b Standard error of the mean value. c Maximal
value versus the control run containing only the vehicle.
the E-stereoisomer. The levels of PPAR-g activation displayed by
E-64 mimic rather closely the putative ligand D12,14-15-deoxy-PGJ2
4 (Entry 15). Rosiglitazone 68 is a much more potent PPAR-
g ligand than any of these prostanoids and prostanoid mimics
(Entry 16).ꢀ However, several side-effects have been reported for
the TZD drug class, particularly related to patients’ weight gain
and odema.50 These side-effects have been linked to the high level
of PPAR-g activation, indicating that in the future less potent
PPAR-g compounds may have improved side-effect profiles.51
Experimental section
General procedure for the conjugate addition–Peterson
olefination reactions
3-Methyl-2-[1-phenylmeth-(E)-ylidene]-2,3,3a,4,7,7a-hexahydro-
4,7-methano-inden-1-one, E-14a. At -78 ◦C under nitrogen, a
slurry of CuI (509 mg, 2.67 mmol, 1.2 eq.) in Et2O (25 cm3)
was treated dropwise with a 1.6 M solution of MeLi in hexanes
(3.36 cm3, 5.37 mmol, 2.4 eq.). The reaction was warmed to -10 ◦C
over a period of 2 h. This solution was cooled to -20 ◦C before a
cooled (-20 ◦C) solution of the enone exo-12 (485 mg, 2.23 mmol,
1 eq.) in Et2O (25 cm3) was added in a dropwise fashion. The flask
containing exo-12 was washed with Et2O (5 cm3) and this was
also transferred to the reaction mixture. Stirring was continued
for 1.5 h during◦which time the temperature rose to -10 ◦C. Upon
cooling to -78 C, benzaldehyde (0.35 cm3, 3.44 mmol, 1.5 eq.)
was added. The reaction was stirred for 3 h and warmed from
Conclusion
In conclusion, we have developed an efficient and novel method
for the synthesis of cross-conjugated cyclopentadienones in two
high yielding steps. Significantly the conjugate addition reaction
proceeds with very high diastereoselectivity; the corresponding
enantioenriched Pauson–Khand adducts have been employed to
generate enantioenriched substrates. Furthermore, many of our
synthetic compounds mimic the behaviour of 4, the benchmark
PPAR-g activating fatty acid. In terms of the likely biological
mechanism by which these compounds elicit these agonistic effects,
it has been shown that a sulfahydryl, cysteine residue resides
in the PPAR-g–TZD binding pocket (Cys285)52 and one can
speculate that this residue forms a conjugate adduct with the
electrophilic, endocyclic alkene.2a In terms of a related biological
precedent for this type of protein modification, a link can be
made with the well-appreciated prenylation of cysteinyl containing
proteins.53
◦
-78 ◦C to 10 C. A saturated solution of NH4Cl (25 cm3) was
added and the resultant aqueous phase was further extracted
with Et2O (3 ¥ 25 cm3). The combined organic extracts were
dried over MgSO4, filtered and the solvent removed in vacuo. The
crude product was then purified by flash column chromatography
(Hex–Et2O; 19 : 1) affording E-14a as a colourless solid (520 mg,
93%). Recrystallisation from n-hexane gave crystals of E-14a
suitable for X-ray crystallography. Mp 82 ◦C (Hex); Rf 0.15
(Hex–Et2O; 19 : 1); (Found C, 85.95; H, 7.42%, C18H18O requires
C, 86.36; H, 7.25%); nmax/cm-1 2934, 1694, 1609, 1494, 1448,
1332, 1236, 1183; dH (400 MHz, CDCl3) 1.25 (3H, d, J 7.0 Hz,
CH3), 1.28 (1H, dt, J 1.5, 9.5 Hz, CH2), 1.36 (1H, d, J 9.5 Hz,
CH2), 1.93 (1H, d, J 7.5 Hz, CH), 2.48 (1H, d, J 7.5 Hz, CH),
2.86 (1H, s, CH), 3.12 (1H, s, CH), 3.19 (1H, q, J 7.0 Hz, CH),
6.18–6.26 (2H, m, CH), 7.28 (1H, d, J 2.0 Hz, CH), 7.34–7.44
ꢀ The EC50 value of 1.4 mM is somewhat misleading since 68 begins
to activate at a comparatively lower concentration and maintains and
increases activation at doses above 10 mM reaching approximately a 40-
fold activation over the control.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4649–4661 | 4659
©

View Article Online
(3H, m, ArH), 7.57 (2H, d, J 7.5 Hz, ArH); dC (100 MHz, CDCl3)
21.2, 38.9, 43.2, 48.5, 49.2, 49.5, 53.3, 128.8, 129.4, 130.7, 133.4,
135.9, 137.6, 139.0, 145.1, 209.0; m/z (EI) 250 (M+, 25%), 183
(100%), 156 (50%), 141 (50%), 128 (40%), 115 (70%), 91 (50%),
66 (90%).
and L. J. Roberts, J. Biol. Chem., 1999, 274, 10863; (c) for a recent
review see: U. Jahn, J.-M. Galano and T. Durand, Angew. Chem. Int.
Ed., 2008, 47, 5894.
6 (a) J. L. Quiles, G. Barja, M. Battino, J. Mataix and V. Solfrizzi,
Nutrition Rev., 2006, 64, S31; (b) M. H. Horani, M. J. Haas and A. D.
Mooradian, Nutrition, 2006, 22, 123; (c) G. Liu, D. M. Bibus, A. M.
Bode, W.-Y. Ma, R. T. Holman and Z. Dong, Proc. Natl. Acad. Sci.
USA, 2001, 98, 7510; (d) Y. Wan, A. Saghatelian, L. W. Chong, C.-
L. Zhang, B. F. Cravatt and R. M. Evans, Genes & Dev., 2007, 21,
1895; (e) M. Toborek, Y. W. Lee, R. Garrido, S. Kaiser and B. Hennig,
Am. J. Clin. Nutr., 2002, 75, 119.
General procedure for the retro-Diels–Alder reactions: synthesis of
adducts 15a to 15q
7 (a) M. H. Beale and J. L. Ward, Nat. Prod. Rep., 1998, 533; (b) E. J.
Corey and S. P. T. Matsuda, Tetrahedron Lett., 1987, 28, 4247.
8 (a) F. Bohlmann, R. K. Gupta, R. M. King and H. Robinson,
Phytochem., 1981, 20, 1417; (b) F. Bohlmann, N. Borhtakur, R. M.
King and Robinson, Phytochem., 1982, 21, 125; (c) F. Bohlmann, P.
Singh, J. Jakupovic, R. M. King and H. Robinson, Phytochem., 1982,
21, 371.
9 (a) M. Krischke, C. Loeffler and M. J. Mueller, Phytochem., 2003, 62,
351; (b) I. Thoma, M. Krischke, C. Loeffler and M. J. Mueller, Chem.
Phys. Lipids, 2004, 128, 135.
10 (a) H. Kikuchi, Y. Tsukitani, K. Iguchi and Y. Yamada, Tetrahedron
Lett., 1983, 24, 1549; (b) H. Nagaoka, K. Iguchi, T. Miyakoshi, N.
Yamada and Y. Yamada, Tetrahedron Lett., 1986, 27, 223.
11 (a) M. E. Jung, J. A. Berliner, D. Angst, D. Yue, L. Koroniak, A. D.
Watson and R. Li, Org. Lett., 2005, 7, 3933; (b) H. P. Acharya and Y.
Kobayashi, Tetrahedron Lett., 2005, 46, 8435; (c) G. Subbanagounder,
J. W. Wong, H. Lee, K. F. Faull, E. Miller, J. L. Witztum and J. A.
Berliner, J. Biol. Chem., 2002, 277, 7271.
4-Methyl-5-[1-phenylmeth-(E)-ylidene]cyclopent-2-enone, 15a.
Under nitrogen, a solution of E-14a (250 mg, 1.0 mmol, 1.0 eq.)
and maleic anhydride (490 mg, 5.0 mmol, 5.0 eq.) in DCM (10 cm3)
was treated with a 1.0 M solution of MeAlCl2 in hexane (1.1 cm3,
1.1 mmol, 1.1 eq.). This mixture was heated to reflux for 6 h.
On cooling, silica (ca. 2.5 g) was added and the solvent was
removed under reduced pressure. Flash column chromatography
(Hex–EtOAc; 3 : 1) gave the title◦compound E-15a (138 mg, 75%)
as a colourless solid. Mp 64–66 C; Rf 0.25 (Hex–EtOAc; 3 : 1);
(Found C, 84.66; H, 6.60%, C13H12O requires C, 84.78; H, 6.57%);
n
max/cm-1 3077, 2983, 2944, 2886, 2340, 1680, 1623, 1580, 1446,
1380; dH (400 MHz, CDCl3) 1.22 (3H, d, J 7.0 Hz, CH3), 3.84–
4.00 (1H, m, CH), 6.40 (1H, dd, J 1.75, 5.75 Hz, CH), 7.39–7.44
(4H, ArH), 7.54 (2H, d, J 7.0 Hz, ArH), 7.60 (1H, ddd, J 1.0, 2.5,
5.75 Hz, CH); dC (100 MHz, CDCl3) 16.3, 38.8, 128.7, 129.3, 130.6,
12 H. Singh, R. Sen, D. Baltimore and P. A. Sharp, Nature, 1986, 319,
154.
+
131.7, 133.6, 134.8, 138.3, 163.9, 197.4; m/z (CI) 202 (NH4 , 20%),
13 A. Hoffmann and D. Baltimore, Immunological Rev., 2006, 210, 171.
14 (a) A. Rossi, P. Kapahi, G. Natoli, T. Takahashi, Y. Chen, M. Karin
and M. G. Santoro, Nature, 2000, 403, 103; (b) 15-deoxy-D12,14-PGJ2
has also been shown to affect NF-kB/DNA binding through covalent
modification of the p50 subunit: E. Cernuda-Morollon, E. Pineda-
Molina, F. J. Canada and D. Pe´rez-Sala, J. Biol. Chem., 2001, 276,
35530; (c) D. S. Straus, G. Pascual, M. Li, J. S. Welch, M. Ricote, C. H.
Sengchanthalangsy, L. L. Ghosh and C. K. Glass, Proc. Natl. Acad.
Sci. USA, 2000, 97, 4844.
185 (MH+, 100%); Found 185.09654, C13H13O requires 185.09665
(-0.6 ppm).
Acknowledgements
We thank the Ministry of Science and Technology and NIBGE,
Pakistan for a Ph.D. scholarship awarded to M. I., the Irish
research council for Science Engineering and Technology (IRC-
SET) for a postgraduate scholarship awarded to P. D. and
the Science Foundation Ireland (SFI) for financial support.
Financial support from Spanish MEC (CTQ2005-623) is also
gratefully acknowledged. The pFA plasmids, containing the ligand
binding domains of the PPAR proteins, were kindly provided by
F. Gregoire (Metabolex Inc., Hayward CA, USA).
15 (a) H. R. Saibil, Curr. Opin. Struc. Bio., 2008, 18, 35; (b) M. G. Santoro,
Trends Microbiol., 1997, 5, 276.
16 R. M. Evans, Science, 1988, 240, 889.
17 (a) T. M. Willson, P. J. Brown, D. D. Sternbach and B. R. Henke, J. Med.
Chem., 2000, 43, 527; (b) L. A. Sorbera, L. Leeson, L. Martin and J.
Castaner, Drugs Future, 2001, 26, 354; (c) A. B. Jones, Med. Res. Rev.,
2001, 21, 540; (d) S. A. Kliewer and T. M. Willson, Curr. Opin. Genet.
Dev., 1998, 8, 576; (e) T. M. Willson and W. Wahli, Curr. Opin. Chem.
Biol., 1997, 1, 235; (f) B. Staels and J. Auwerx, Curr. Pharm. Des., 1997,
3, 1; (g) B. M. Spiegelman, Diabetes, 1998, 47, 507; (h) K. Schoonjans,
G. Martin, B. Staels and J. Auwerx, Curr. Opin. Lipidol., 1997, 8, 159;
(i) B. R. Henke, J. Med. Chem., 2004, 47, 4118.
18 (a) B. M. Forman, P. Tontonoz, J. Chen, R. P. Brun, B. M. Spiegelman
and R. M. Evans, Cell, 1995, 83, 803; (b) S. A. Kliewer, J. M. Lenhard,
T. M. Willson, I. Patel, D. C. Morris and J. M. Lehman, Cell, 1995, 83,
813.
References
1 (a) For recent review articles on the chemistry and biology of cyclopen-
tenone prostanoids see: M. Negishi and H. Katoh, Prostaglandins &
other Lipid Mediators, 2002, 68–69, 611; (b) D. S. Straus and C. K.
Glass, Med. Res. Rev., 2001, 21, 185; (c) S. M. Roberts, M. G. Santoro
and E. S. Sickle, J. Chem. Soc., Perkin Trans. 1, 2002, 1735.
2 (a) M. Suzuki, M. Mori, T. Niwa, R. Hirata, K. Furuta, T. Ishikawa
and R. Noyori, J. Am. Chem. Soc., 1997, 119, 2376; (b) M. Renedo, J.
Gayarre, C. A. Garc´ıa-Dom´ınguez, A. Pe´rez-Rodr´ıguez, A. Prieto, F. J.
Can˜ada, J. M. Rojas and D. Pe´rez-Sala, Biochemistry, 2007, 46, 6607;
(c) J. L. Oliva, D. Pe´rez-Sala, A. Castrillo, N. Mart´ınez, F. J. Can˜ada,
L. Bosca´ and J. M. Rojas, Proc. Natl. Acad. Sci. USA, 2003, 100,
4772; (d) A.-L. Levonen, A. Landar, A. Ramachandran, E. K. Ceaser,
D. A. Dickinson, G. Zanoni, J. D. Morrow and V. M. Darley-Usmar,
Biochem. J., 2004, 378, 373.
19 (a) W. S. Powell, J. Clin. Invest., 2003, 112, 828; (b) L. C. Bell-Parikh,
T. Ide, J. A. Lawson, P. McNamara, M. Reilly and G. A. FitzGerald,
J. Clin. Invest., 2003, 112, 945.
20 (a) M. Suzuki, T. Kiho, K. Tomokiyo, K. Furuta, S. Fukushima, Y.
Takeuchi, M. Nakanishi and R. Noyori, J. Med. Chem., 1998, 41,
3084; (b) B. Cox, L. J. Murphey, W. E. Zackert, R. Chinery, R. Graves-
Deal, O. Boutaud, J. A. Oates, R. J. Coffey and J. D. Morrow, Biochim.
Biophys. Acta, 2002, 1584, 37; (c) I. Hubatsch, M. Ridderstrom and B.
Mannervik, Biochem. J., 1998, 330, 175; (d) I. Hubatsch, B. Mannervik,
L. Gao, L. J. Roberts, Y. Chen and J. D. Morrow, Chem. Res. Toxicol.,
2002, 15, 1114.
21 A. M. O’Hara and F. Shanahan, EMBO Reports, 2006, 7, 688.
22 (a) C. Loeffler, S. Berger, A. Guy, T. Durand, G. Bringmann, M. Dreyer,
U. V. Rad, J. Durner and M. J. Mueller, Plant Physiol., 2005, 137, 328;
(b) E. E. Farmer and C. Davoine, Curr. Opin. Plant Biol., 2007, 10, 380;
(c) S. Mueller, B. Hilbert, K. Dueckershoff, T. Roitsch, M. Krischke,
M. J. Mueller and S. Berger, Plant Cell, 2008, 20, 768; (d) M. Iqbal, P.
Evans, A. Lledo´, X. Verdaguer, M. A. Perica`s, A. Riera, C. Loeffler,
A. K. Sinha and M. J. Mueller, ChemBioChem, 2005, 6, 276.
3 (a) For general references see: C. D. Funk, Science, 2001, 294, 1871;
(b) R. Noyori and M. Suzuki, Science, 1993, 259, 44; (c) P. W. Collins
and S. W. Djuric, Chem. Rev., 1993, 93, 1533; (d) R. Noyori and M.
Suzuki, Angew. Chem., Int. Ed. Engl., 1984, 23, 847; (e) S. Das, S.
Chandrasekhar, J. S. Yadav and R. Gre´e, Chem. Rev., 2007, 107, 3286.
4 F. A. Fitzpatrick and M. A. Wynalda, J. Biol. Chem., 1983, 258, 11713.
5 (a) J. D. Morrow, J. A. Awad, H. J. Boss, I. A. Blair and L. J. Roberts,
Proc. Natl. Acad. Sci. USA, 1992, 89, 10721; (b) Y. Chen, J. D. Morrow
4660 | Org. Biomol. Chem., 2008, 6, 4649–4661
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
23 (a) For representative recent syntheses of this class of compound see:
H. Tanaka, T. Hasegawa, N. Kita, H. Nakahara, T. Shibata, S. Oe, M.
Ojika, K. Uchida and T. Takahashi, Chem. Asian J., 2006, 1, 669; (b) C.
Kuhn, E. Roulland, J.-C. Madelmont, C. Monneret and J.-C. Florent,
Org. Biomol. Chem., 2004, 2028; (c) M. Schelwies, P. Du¨bon and G.
Helmchen, Angew. Chem. Int. Ed., 2006, 45, 2466; (d) H. P. Acharya
and Y. Kobayashi, Tetrahedron, 2006, 62, 3329; (e) G. Zanoni, A. Porta
and G. Vidari, J. Org. Chem., 2002, 67, 4346; (f) J. F. Bickley, V. Jadhav,
S. M. Roberts, M. G. Santoro, A. Steiner and P. W. Sutton, Synlett,
2003, 1170; (g) K. M. Brummond, P. C. Sill and H. F. Chen, Org. Lett.,
2004, 6, 149.
24 (a) G. Stork, G. L. Nelson, F. Rouessac and O. Gringgore, J. Am. Chem.
Soc., 1971, 93, 3091; (b) J.-B. Wiel and F. Rouessac, J. Chem. Soc. Chem.
Commun., 1976, 446.
25 A. J. H. Klunder, J. Zhu and B. Zwanenburg, Chem. Rev., 1999, 99,
1163.
26 (a) O. Tsuge, S. Kanemasa and Y. Ninomiya, Chem. Lett., 1984, 1993;
(b) J. Taneka, S. Kanemasa, Y. Ninomya and O. Tsuge, Bull. Chem.
Soc. Jpn., 1990, 63, 466; (c) J. Tanaka, H. Kobayashi, S. Kanemasa and
O. Tsuge, Bull. Chem. Soc. Jpn., 1989, 62, 1193.
27 M. Iqbal and P. Evans, Tetrahedron Lett., 2003, 44, 5741.
28 M. Iqbal, N. Vyse, J. Dauvergne and P. Evans, Tetrahedron Lett., 2002,
43, 7859.
Lett., 1985, 26, 5005; (c) M. B. Gillies, J. E. Tonder, D. Tanner and
R.-O. Norrby, J. Org. Chem., 2002, 67, 7378.
36 P. Knochel and R. D. Singer, Chem Rev., 1993, 93, 2117.
37 M. Kitamura, T. Miki, K. Nakano and R. Noyori, Bull. Chem. Soc.
Jpn., 2000, 73, 999.
38 W. H. Sikarski and G. J. Reich, J. Am. Chem. Soc., 2001, 123, 6527.
39 S. Reddy, H. Krishna, C. Kazuhiro, S. Yongmao and K. D. Moeller,
Tetrahedron, 2001, 57, 5183.
40 C. Shih, E. L. Fritzen and J. S. Swenton, J. Org. Chem., 1980, 45, 4462.
41 M. Ohno and M. Otsuka, Org. React., 1989, 37, 1.
42 (a) J. Sola`, M. Reve´s, A. Riera and X. Verdaguer, Angew. Chem. Int. Ed.,
2007, 46, 5020; (b) X. Verdaguer, A. Moyano, M. A. Perica`s, A. Riera,
M. Angel Maestro and J. Mah´ıa, J. Am. Chem. Soc., 2000, 122, 10242;
(c) X. Verdaguer, A. Lledo´, C. Lo´pez-Mosquera, M. Angel Maestro,
M. A. Perica`s and A. Riera, J. Org. Chem., 2004, 69, 8053.
43 For the synthesis of 49 see: D. Ranganathan, S. Ranganathan and
M. M. Mehrotra, Tetrahedron, 1980, 36, 1869.
44 M. Suzuki, A. Yanagisawa and R. Noyori, J. Am. Chem. Soc., 1988,
110, 4718.
45 M. Iqbal, Y. Li and P. Evans, Tetrahedron, 2004, 60, 2531.
46 (a) G. Kru¨ge, C. Hard and F. Bohlmann, Tetrahedron Lett., 1985,
26, 6027; (b) Z.-Y. Liu and X.-J. Chu, Tetrahedron Lett., 1993, 34,
3885; (c) Z.-Y. Liu, H. Dong and X.-Y. Chu, Tetrahedron, 1994, 50,
12337.
29 For a recent review see: J. B. Urgoiti, L. An˜orbe, L. P. Serrano, G.
Dom´ınguez and J. P. Castells, Chem. Soc. Rev., 2004, 33, 32.
30 Crystal structural data for compound 14a: C18H18O M = 250.32;
47 G. Williams and J. C. Pickup, Handbook of Diabetes, 3rd edn, Blackwell
Publishing Ltd., Oxford, 2004.
˚
˚
orthorhombic, Pbcn; a = 8.8975(5) A, b = 12.1847(7) A, c = 24.8677(4)
48 T. Allen, F. Zhang, S. A. Moodie, L. E. Clemens, A. Smith, F. Gregoire,
A. Bell, G. E. O. Muscat and T. A. Gustafson, Diabetes, 2006, 55,
2523.
3
˚
˚
A; U = 2696.6(3) A ; T = 150 K; Z = 8; 15692 reflections measured,
3200 unique (Rint = 0.0192). The final wR2 was 0.1339 (all data). CCDC
reference number 219462.
49 For a report suggesting that Z12-D12,14-15 deoxy-PGJ2 4 more effectively
activates PPAR-g see: K. M. Maxey, E. Hessler, J. MacDonald and L.
Hitchingham, Prostaglandins & other Lipid Med., 2000, 62, 15.
50 H. Zhang, A. Zhang, D. E. Kohan, R. D. Nelson, F. J. Gonzalez and
T. Yang, Proc. Natl. Acad. Sci. USA, 2005, 102, 9406.
51 Y. Li, Z. Wang, N. Furukawa, P. Escaron, J. Weiszmann, G. Lee,
M. Lindstrom, J. Liu, X. Liu, H. Xu, O. Plotnikova, V. Prasad, N.
Walker, R. M. Learned and L.-J. Chen, J. Biol. Chem., 2008, 283,
9168.
31 (a) H. Sasaki, S. Niimi, M. Akiyama, T. Tanaka, A. Hazato, S.
Kurozumi, S. Fukushima and M. Fukushima, Cancer Res., 1999, 59,
3919; (b) F. Shoji, T. Yoshikazu, K. Shuichi, Y. Shoichi, U. Kenji, S.
Satoshi, S. Masaaki, F. Kyoji, N. Ryoji, S. Hiroshi, K. Yoshihiro, K.
Tsunekazu, Y. Takao, S. Junichi, K. Michio, H. Atsuo, K. Seiji and F.
Masanori, Anti Cancer Drugs, 2001, 12, 221.
32 M. Suzuki, T. Kawagishi, A. Yanagisawa and T. Suzuki, Bull. Chem.
Soc. Jpn., 1988, 61, 1299.
33 P. A. Grieco and N. Abood, J. Chem. Soc. Chem. Commun., 1990, 410.
34 (a) B. H. Lipshutz, D. H. Aue and B. James, Tetrahedron Lett., 1996,
37, 8471; (b) S. Woodward, Chem. Soc. Rev., 2000, 29, 393.
35 (a) L. F. van Staden, D. Gravestock and D. J. Ager, Chem. Soc. Rev.,
2002, 31, 195; (b) R. K. Boeckman, Jr. and R. L. Chinn, Tetrahedron
52 R. T. Nolte, G. B. Wisely, S. Westin, J. E. Cobb, M. H. Lambert, R.
Kurokawa, M. G. Rosenfeld, T. M. Willson, C. K. Glass and M. V.
Milburn, Nature, 1998, 395, 137.
53 For a review on protein prenylation see: F. L. Zhang and P. J. Casey,
Ann. Rev. Biochem., 1996, 65, 241.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4649–4661 | 4661
©