Total synthesis and biological evaluation of the cytotoxic resin glycosides ipomoeassin A-F and analogues

Article abstract of DOI:10.1002/chem.200901449

A multitasking C-silylation strategy using the readily available compound 26 as a surrogate for cinnamic acid represents the key design element of a total synthesis of all known members of the ipomoeassin family of resin glyosides. This protecting group maneuver allows the unsaturated acids decorating the glucose subunit of the targets to be attached at an early phase of the synthesis, prevents their participation in the ruthenium-catalyzed ring-closing metathesis (RCM) used to form the macrocyclic ring, and protects them against reduc tion during the hydrogenation of the resulting cycloalkene over Wilkinson's catalyst. As the C-silyl group can be concomitantly removed with the O-TBS substituent using tris(dimethylamino)sulfonium difluorotrimethylsilicate (TASF) in acetonitrile, no separate protecting group manipulations were necessary in the final stages, thus contributing to a favorable overall "economy of steps". In addition to the naturally occurring ipomoeassins, a small set of synthetic analogues has also been prepared by "diverted total synthesis". The cytotoxicity of these compounds was assayed with two different cancer cell lines. The recorded data confirm previous findings that the acylation- and oxygenation pattern of these amphiphilic glycoconjugates is highly correlated with their biological activity profile. Ipomoeassin F turned out to be the most promising member of the series, showing IC₅₀ values in the low nanomolar range.

Full text of DOI:10.1002/chem.200901449

FULL PAPER
DOI: 10.1002/chem.200901449
Total Synthesis and Biological Evaluation of the Cytotoxic Resin Glycosides
Ipomoeassin A–F and Analogues
Takashi Nagano,[a] Jirˇi Pospꢀsˇil,[a] Guillaume Chollet,[a] Saskia Schulthoff,[a]
Volker Hickmann,[a] Emilie Moulin,[a] Jennifer Herrmann,[b] Rolf Mꢁller,[b] and
Alois Fꢁrstner*[a]
Abstract: A multitasking C-silylation
strategy using the readily available
compound 26 as a surrogate for cin-
namic acid represents the key design
element of a total synthesis of all
known members of the ipomoeassin
family of resin glyosides. This protect-
ing group maneuver allows the unsatu-
rated acids decorating the glucose sub-
unit of the targets to be attached at an
early phase of the synthesis, prevents
their participation in the ruthenium-
tion during the hydrogenation of the
resulting cycloalkene over Wilkinsonꢀs
catalyst. As the C-silyl group can be
concomitantly removed with the O-
TBS substituent using tris(dimethyl-
amino)sulfonium difluorotrimethylsili-
cate (TASF) in acetonitrile, no sepa-
rate protecting group manipulations
were necessary in the final stages, thus
“economy of steps”. In addition to the
naturally occurring ipomoeassins,
a
small set of synthetic analogues has
also been prepared by “diverted total
synthesis”. The cytotoxicity of these
compounds was assayed with two dif-
ferent cancer cell lines. The recorded
data confirm previous findings that the
acylation- and oxygenation pattern of
these amphiphilic glycoconjugates is
highly correlated with their biological
activity profile. Ipomoeassin F turned
out to be the most promising member
of the series, showing IC50 values in the
low nanomolar range.
contributing to
a favorable overall
Keywords:
glycolipids
cytotoxicity
macrocycles
·
·
catalyzed
ring-closing
metathesis
·
(RCM) used to form the macrocyclic
ring, and protects them against reduc-
metathesis · protecting groups
Introduction
They consist of differently acylated oligosaccharides glycosy-
lated with hydroxylated fatty acid derivatives which are usu-
ally tied back to form macrolactone rings of various sizes.[1–4]
The resulting amphiphilic architectures caught the imagina-
tion of preparative chemists, with the conquest of the tricol-
orin family[5–7] and the total synthesis of woodrosin being
representative.[8–10]
Resin glycosides exhibit a broad spectrum of biological
activities, ranging from laxative, purgative, haemolytic, anti-
bacterial, antifungal or plant growth inhibitory properties to
significant cytotoxicity.[1–4] However, the understanding for
their mode of action is largely missing,[11] as are systematic
studies into structure–activity relationships. The available in-
formation, however, suggests that these compounds offer
ample opportunity for optimization of their bioactivity pro-
files. An instructive case is the ipomoeassin family, which
was recently isolated from the morning glory Ipomoea squa-
mosa collected in the Suriname rainforest during a biodi-
versity program (Figure 1).[12,13] The recorded data for the
six naturally occurring members indicate that variations in
the peripheral oxygenation and acylation pattern alter the
cytotoxicity against the human A2780 OVCAR cell line to a
Plants of the morning glory family (Convolvulaceae) are
known as rich sources of alkaloids and resin glycosides and
have been extensively used in traditional medicine as herbal
remedies for various diseases.[1] Although chemical investi-
gations into their sugar components had already been under-
taken in the 19th century, it was not until the advent of the
powerful modern spectroscopic techniques that the intrigu-
ing structures of these glycoconjugates could be elucidated.
ˇ
[a] Dr. T. Nagano, Dr. J. Pospꢁsil, Dr. G. Chollet, S. Schulthoff,
Dipl.-Chem. V. Hickmann, Dr. E. Moulin, Prof. A. Fꢂrstner
Max-Planck-Institut fꢂr Kohlenforschung
45470 Mꢂlheim/Ruhr (Germany)
Fax : (+49)208-306-2994
E-mail: fuerstner@mpi-muelheim.mpg.de
[b] Dipl.-Chem. J. Herrmann, Prof. R. Mꢂller
Saarland University
Department of Pharmaceutical Biotechnology
66041 Saarbrꢂcken (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901449.
Chem. Eur. J. 2009, 15, 9697 – 9706
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9697

our previous excellent experiences,[25,26] it seemed likely that
RCM would allow us to forge the backbone of the ipo-
moeassins. Yet, the need to reduce the resulting disubstitut-
ed C8=C9 double bond in the tether during the further elab-
oration of the targets either mandates that the unsaturated
acids decorating the d-glucose ring are introduced at a late
stage, or requires a chemo- and regioselective hydrogenation
of the cycloalkene without affecting this unsaturated periph-
ery. In an attempt to minimize the necessary protecting
group manipulations, we opted for the second strategy, in
particular since we had noticed in the past that a macrocy-
clic backbone may impose serious constraints onto an oligo-
saccharide that render even seemingly trivial manipulations
of the sugar periphery highly capricious.[8,9,27]
The plan to selectively hydrogenate the macrocycle in the
presence of lateral unsaturation, however, is not without
risk either. Model studies had shown that a disubstituted
(cyclo)alkene can be reduced over Wilkinsonꢀs catalyst with-
out difficulty in the presence of a 2-methyl-2-butenoic acid
ester (tiglic acid ester), but that a cinnamate is hydrogenated
with similar rates.[22] A metal-free reduction using diimine
led to an even less favorable outcome. Therefore it was
planned to have the tiglate in place, whereas the cinnamate
would be installed only after RCM and saturation of the
macrocycle. The resulting synthesis blueprint, as spelled out
for ipomoeassin E in Scheme 1, might be reduced to practice
Figure 1. The family of ipomoeassin resin glycosides and survey of their
surprisingly different cytotoxicity data (IC50) against the human ovarian
cancer cell line A2780 published in the literature.[12,13,15]
significant extent.[12] “Hot spots” on the pharmacophoric
landscape seem to be the 4’-O site of the d-fucose moiety as
well as C-5 within the fatty acid tether, as evident from the
fact that ipomoeassin D (4) and F (6) are up to two orders
of magnitude more potent than their congeners. However,
peracetylation of 4 drastically reduces the activity.[12] Over-
all, the available information suggests that ipomoeassin F
(6)[13] incorporating a fatty acid that is two methylene units
longer than that of its sister compounds,[14] is the most prom-
ising member, which seems to indicate that the lipophilicity
of the backbone exerts an additional subtle influence on the
bioactivity of such amphiphilic glycoconjugates. The IC50
values of 6 against various human cancer cell lines lie in the
single-digit nanomolar range.[13,15] Moreover, the discoverers
of the ipomoeassins claimed that the cytotoxicity profile of
these compounds has no significant COMPARE correlation
with that of the recorded anticancer agents in the National
Cancer Institute (NCI) database, thus raising questions as to
their possible mode of action.[16]
As part of our investigations into the chemistry and biol-
ogy of complex glycolipids,[3,7–9,17–21] our group has reported
in 2007 the first total synthesis of two members of the ipo-
moeassin family based on a ring-closing olefin metathesis
(RCM) reaction for the formation of the macrocycle at the
[22]
ꢀ
C8 C9 bond. The exact same strategy was later used by a
team from the Eisai Research Institute, which disclosed a
synthesis of ipomoeassin F earlier this year.[23] Outlined
below is a full account of our work in this area, which pro-
vided access to all naturally occurring ipomoeassins as well
as to a few selected analogues. Additional data on the bioac-
tivity of these compounds are also presented.
Scheme 1. Original retrosynthetic analysis exemplified for ipomoeassin E.
Results and Discussion
using compound B as a key intermediate, in which the sec-
ondary alcohols of the glucose are already adequately differ-
entiated. B, in turn, can be accessed by reductive opening of
the corresponding 4,6-O-benzylidene acetal derivative D
Strategic considerations: Ring-closing olefin metathesis
(RCM) has emerged over the last years as a powerful
method for the formation of macrocyclic rings.[24] Based on
9698
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9697 – 9706

Natural Products
FULL PAPER
(n=1). Whereas ipomoeassin A–D should be available from
the very same building block, preparation of ipomoeassin F
would require the homologous compound D’ (n=3) featur-
ing the appropriately elongated appendix.
Assembly of the disaccharide building block: The optically
pure (99% ee) heptenol 10a was conveniently prepared on
treatment of commercial (S)-epichlorohydrine
7 with
EtMgBr in the presence of CuCN (10 mol%), treatment of
the resulting adduct 8a with powdered NaOH or KOH in
Et2O, and opening of the resulting epoxide 9a with vinyl-
magnesium bromide, which was again catalyzed by CuCN
(Scheme 2).[28] Subsequent silver-mediated[29] glycosylation
Scheme 3. a) Cl3CCN, Cs2CO3 cat., CH2Cl2, 86–89%; b) compound 13a,
BF3·Et2O cat., CH2Cl2/pentane 1:1, ꢀ208C, 16a (n=1, 75–77%); or:
compound 13b, TMSOTf (1.05 equiv), CH2Cl2, ꢀ308C, 16b (n=3, 82–
95%); c) KOMe cat., MeOH, 17a (n=1, 82–84%); 17b (n=3, 73–
77%); d) tiglic acid, DCC, DMAP, CH2Cl2, 18a (n=1, 55–78%); 18b
(n=3, 66–73%); e) TBSOTf, 2,6-lutidine, CH2Cl2, 19a (n=1, 93–96%);
19b (n=3, 91–97%); f) NaBH3CN, TMSCl, MS 4 ꢄ, MeCN, see Table 1.
TBS-protected to give compound 19a ready for reductive
opening of the substituted benzylidene acetal.
Scheme 2. a) CH
(CH2)nMgCl, CuCN (10 mol%), THF, ꢀ78 ! ꢀ208C;
b) NaOH, Et2O; c) CH2=CHMgBr, CuCN (10 mol%), THF, ꢀ78 ! 08C,
10a (n=1, 69–77% overall, 99% ee); 10b (n=3, 74–80% overall, 99%
ee); d) AgOTf, 2,6-di-tert-butylpyridine, MS 4 ꢄ, CH2Cl2, 08C ! RT, 12a
(n=1, 84%); 12b (n=3, 67%); e) KOMe cat., MeOH; f) 2,2-dimethoxy-
propane, pTsOH.H2O cat., acetone, 13a (n=1, 95–98% over both steps);
13b (n=3, 79–83% over both steps).
Despite a host of literature precedence and extensive ex-
perimentation, this seemingly routine step turned out to be
problematic and ultimately enforced a change in strategy.
Although many reagent combinations are known to convert
sugar 4,6-O-benzylidene acetals into the corresponding 4-O-
benzyl ether derivatives,[31] most of them simply decomposed
substrate 19a. These failures are ascribed to the incompati-
bility of the reducing agents with the ester group and/or
double bonds of this substrate. Only the use of NaBH3CN in
combination with TMSCl[38] furnished a mixture of the cor-
responding PMB-ether derivatives in up to 84% combined
yield with a 4:1 isomeric ratio (Table 1). In contrast to the
literature,[32,38] however, the 6’’-O-PMB ether 21 rather than
with fucosyl bromide 11 afforded compound 12a in good
yield, which was transformed into product 13a by saponifi-
cation of the acetates followed by selective ketalization of
the released triol. The homologous building block 13b re-
quired for the synthesis of ipomoeassin F was prepared anal-
ogously.
The necessary glucosyl donor 15 was obtained from the
hemiketal 14, which is accessi-
ble in multigram quantities by
following the literature route
Table 1. Reductive opening of the benzylidene acetal in 19.
Entry Reducing agent Conditions
(Scheme 3).[18]
A
BF3·Et2O-
n
21/20 Isolated yield [%] (21) Ref.
mediated reaction with alcohol
13a furnished disaccharide 16a
in excellent yield,[30] which was
deacetylated in preparation for
the esterification with tiglic
acid. As expected, this reaction
occurred regioselectively at the
more nucleophilic O-3’’ site of
the b-glucoside.[7] The remain-
[33]
1
2
3
4
5
6
7
8
9
BH3·THF
BH3·THF
Dibal-H
PMHS
Et3SiH
Cu
G
1
1
1
1
1
1
1
1
3
decomposition
decomposition
reduction of ester
decomposition
decomposition
48
[34]
Bu2BOTf (1 equiv), CH2Cl2, RT, 4 h
CH2Cl2, 08C, 5 min
AlCl3, CH2Cl2/Et2O, RT, 15 h
PhBCl2, MS 4 ꢄ, CH2Cl2, ꢀ788C, 1 h
TMSCl, 08C
[35]
[36]
[37]
[38]
NaBH3CN[a]
NaBH3CN[a]
NaBH3CN[a]
NaBH3CN[a]
2.0:1
3.5:1
4.0:1
4.0:1
TMSCl, 08C ! RT (1 h)
61–65
81–84
51
TMSCl, 08C/RT[b]
TMSCl, 08C/RT[b]
[a] NaBH3CN (10 equiv), TMSCl (10 equiv) in the presence of MS 4 ꢄ in MeCN; [b] The temperature was
ing -OH group in 18a was then quickly raised to room temperature immediately after the addition of TMSCl.
Chem. Eur. J. 2009, 15, 9697 – 9706
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9699

A. Fꢂrstner et al.
the 4’’-O-PMB derivative 20 was formed as the major prod-
uct. Steric hindrance is the most likely reason for this unex-
pected regiochemical outcome, as the bulky aglycone in 19
might disfavor ligation of the Lewis acid with the 6’’-O posi-
tion of the benzylidene group relative to coordination to the
4’’-O-site. Even though the conformation of 19 in solution is
unknown, the structure of the precursor compound 17a in
the solid state (Figure 2) seems to corroborate this interpre-
tation.
Scheme 4. Revised retrosynthetic plan exemplified for ipomoeassin E
based on a multitasking C-silylation strategy.
Figure 2. Structure of compound 17a in the solid state. Anisotropic dis-
placement parameters are drawn at the 50% probability level. The
oxygen atoms of the 4,6-O-benzylidene acetal are shown in red; they cor-
respond to the oxygen atoms in the derived compound 19 which compete
for coordination to the Lewis acid during the reductive ring opening re-
action.
Revised strategy and completion of the total synthesis of
ipomoeassin A and B: The inability to procure larger
amounts of the envisaged key intermediate 20 by reductive
opening of the benzylidene precursor 19 enforced a change
in strategy with regard to the planned RCM/hydrogenation
sequence. Rather than opting for further orthogonal manip-
ulation of the hydroxy groups, it was envisaged to protect
the cinnamate moiety of the targeted ipomoeassin by a C-
silyl substituent (Scheme 4). As a trisubstituted alkene, such
a cinnamic acid surrogate should neither interfere with the
metathetic ring closure nor should it be touched during the
hydrogenation of the alkene in the macrocyclic tether.
Moreover, it is expected that a C-silyl group can be concom-
itantly removed with the 2’’-O-TBS ether under mild condi-
tions, thus avoiding any extra manipulations during the end
game of the synthesis.[39]
The required acid 26 was conveniently obtained as shown
in Scheme 5. Hydroalumination of 22 followed by addition
of iodine[40] and subsequent O-silylation under standard con-
ditions furnished product 23. Lithium for halogen exchange
triggered a retro-Brook rearrangement that installed the C-
silyl substituent and released the primary alcohol in 24,[41]
which was then oxidized to the corresponding acid 26 in ex-
cellent overall yield.
Scheme 5. a) NaAlH2(OCH2CH2OMe)2, Et2O, then I2, 86–88%; b)
PhMe2SiCl, Et3N, DMAP cat, CH2Cl2, 95–97%; c) tBuLi, THF, ꢀ788C
! RT, 50–57%; d) (COCl)2, DMSO, Et3N, CH2Cl2, ꢀ788C; e) NaClO2,
2-methyl-2-butene, NaH2PO4, tBuOH/H2O, 88–93% (over both steps).
With this C-silylated cinnamic acid in hand, it was the
major isomer derived from the benzylidene cleavage that
could be used en route to the target. Specifically, alcohol
21a[42] was esterified with 26 under Yamaguchi conditions[43]
to give product 27, which was then subjected to oxidative
PMB cleavage[44] followed by attachment of the 4-oxo-8-
nonenoic acid ester segment (Scheme 6).[45] In line with our
expectations, treatment of the resulting diene 29 with cata-
lytic amounts of the commercial “second-generation” ruthe-
nium alkylidene complex 30[46,47] in refluxing CH2Cl2 afford-
ed macrocycle 31 as an E/Z mixture in high yield, which was
hydrogenated with the aid of [RhClACTHNUTRGNEUNG(PPh3)3] without affect-
ing the lateral sites of unsaturation.[48] The C-silyl group and
the O-TBS ether in 32 were then concurrently removed with
tris(dimethylamino)sulfonium
difluorotrimethylsilicate
(TASF) in MeCN[49–51] before the remaining isopropylidene
acetal in 33 was cleaved off with dilute trifluoroacetic acid.
9700
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9697 – 9706

Natural Products
FULL PAPER
Scheme 6. Total synthesis of ipomoeassin A and B: a) 26, 2,4,6-trichlorobenzoyl chloride, Et3N, DMAP, toluene, 79–83% (over both steps from 19a);[42]
b) DDQ, CH2Cl2/H2O; c) 4-oxo-8-nonenoic acid, 2,4,6-trichlorobenzoyl chloride, Et3N, DMAP, toluene, 78% (over both steps); d) complex 30 (10%),
CH2Cl2, reflux, 71%; e) H2 (1 atm), [RhClACHTUGNTRNE(NUGN PPh3)3] (20%), EtOH, 81%; f) TASF, MeCN; g) trifluoroacetic acid, CH2Cl2, 45% (over both steps); h) MeC-
ACHTUNGTRENNUNG(OEt)3, camphersulfonic acid cat; i) aq. HOAc, 95%.
Hence, the novel multitasking C-silylation strategy has fully
met our expectations. The analytical and spectral properties
of synthetic ipomoeassin B (2) thus obtained were in excel-
lent agreement with those reported in the literature.[12]
Moreover, 2 could be further elaborated into ipomoeassin A
(1) by treatment with MeCACHTNUTRGENNG(U OEt)3 in the presence of cam-
phersulfonic acid, followed by an HOAc-induced rearrange-
ment of the resulting orthoester 34 to the conspicuous axial
acetate group present in this particular target.[52] The data of
the synthetic samples of 1 were again in full accord with
those of the natural product.[12]
As an initial foray into the preparation of synthetic ana-
logues by “diverted total synthesis”,[53–55] product 38 was tar-
geted in which the ketone residing in the lipophilic tether
was formally edited out (Scheme 7). This incremental but
deep seated structural change was easily secured by esterifi-
cation of 21a with commercial 8-nonenoic acid followed by
ring-closing metathesis of the resulting diene 35 and elabo-
ration of the isomeric cycloalkenes by the selective hydroge-
nation/deprotection sequence outlined above. Although
these steps were not fully optimized, a good overall yield of
4-deoxy-ipomoeassin B (38) was obtained, thus illustrating
the robustness of the chosen route.
Scheme 7. Preparation of 4-deoxy-ipomoeassin B. a) 8-nonenoic acid,
2,4,6-trichlorobenzoyl chloride, Et3N, DMAP, toluene, 72%; b) complex
30 (10%), CH2Cl2, reflux, 93%; c) H2 (1 atm), [RhClACTHNUGTRENUNG(PPh3)3] (20%),
EtOH, 80%; d) TASF, MeCN; e) trifluoroacetic acid, CH2Cl2, 56%
(over both steps).
Chem. Eur. J. 2009, 15, 9697 – 9706
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9701

A. Fꢂrstner et al.
Total synthesis of ipomoeassin D and E: The success of the
C-silylation strategy in combination with the flexibility in-
herent to RCM brings all other members of the ipomoeassin
series into reach. The necessary acid fragment 44 required
for the synthesis of ipomoeassin D (4) and E (5) was best
prepared by oxidative rearrangement of furyl alcohol 40,[56]
which was obtained in optically pure form (>99% ee) by
Treatment of 43 with TASF and esterification of the re-
sulting acid 44 with disaccharide 28 gave diene 45 ready for
ring closure.[60] As expected, this transformation was again
highly productive, as was the selective hydrogenation of the
resulting E/Z-alkene mixture over Wilkinsonꢀs catalyst. Fol-
lowing the protocol outlined above, product 46 was then de-
silylated prior to acid-catalyzed removal of the isopropyli-
dene acetal in 47. Finally, aliquots of synthetic ipomoeassin
E (5) thus obtained were effectively converted to ipomoeas-
sin D (4) by the orthoester rearrangement process[52] de-
scribed above.
the
Sharpless-type
kinetic
resolution
shown
in
Scheme 8.[57,58] Treatment of (ꢀ)-40 with catalytic amounts
A striking and previously unrecognized spectral property
of ipomoeassin E is the strong dependence of its proton
NMR spectrum in C6D6 on the concentration and quality of
the solvent (freshly distilled versus non-distilled). Figure 3
shows the remarkable shift and broadening of the signal at-
tributed to H-3’’ which is ascribed to a hydrogen-bonding
event locking the alcoholic protons in its neighborhood.
Ipomoeassin C: When the alcohol formed upon opening of
lactone 42 with trimethylsilyl ethanol is protected by a PMB
group rather than an acetate, the very same sequence opens
access to ipomoeassin C (3) (Scheme 9). Subsequent treat-
Scheme 8. Total syntheses of ipomoeassin D and E. a) 1-bromo-3-pen-
tene, Mg, THF, 82–89%; b) Ti
tBuOOH, CH2Cl2, ꢀ208C, 47 % (=94% theoretical yield), >99 ee; c)
tBuOOH, VO(acac)2 (2 mol%), CH2Cl2, 71–73%; d) CrO3, H2SO4, ace-
tone, 08C; e) Zn, HOAc, CHCl3, 75–78% (over both steps); f) (i) HO-
(CH2)2SiMe3, pTsOH·H2O cat., CH2Cl2; ii) Ac2O, DMAP cat., CH2Cl2,
(OiPr)4, d-(ꢀ)-diisopropyltartrate (DIPT),
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
86–93%, 97% ee (over both steps); g) TASF, DMF, 68%; h) 28, 2,4,6-tri-
chlorobenzoyl chloride, Et3N, DMAP, toluene, 87%; i) complex 30 (10–
Scheme 9. Total synthesis of ipomoeassin C: a) HO
ACHTNUGERTN(NUGN CH2)2SiMe3,
15 mol%), CH2Cl2, reflux, 85–96%; j) H2 (1 atm), [RhCl
ACHTUNGTRNE(NUNG PPh3)3]
pTsOH.H2O cat., CH2Cl2; b) PMBO(C=NH)CCl3, TfOH cat., Et2O, 0 8C,
AHCTUNGTRENNUNG
(20 mol%), EtOH, 83%; k) TASF, MeCN; l) TFA, CH2Cl2, 63% (over
63% (over both steps); c) TASF, DMF, 76%; d) 28, 2,4,6-trichlorobenzo-
yl chloride, Et3N, DMAP, toluene, 73%; e) complex 30 (10 mol%),
CH2Cl2, reflux, 88%; f) H2 (1 atm), [RhClACHTNUTRGNE(UNG PPh3)3] (20%), EtOH, 86%;
both steps); m) i) MeC
90%.
ACHTUNGTRENNUNG(OEt)3, camphersulfonic acid cat; ii) aq. HOAc,
g) DDQ, CH2Cl2/H2O; h) TASF, MeCN; i) TFA, CH2Cl2, j) MeC
camphersulfonic acid cat, then aq. HOAc, 30% (over steps g–j).
ACHTUNGTRENNUNG(OEt)3,
of VOACHTUNGTRENNUNG(acac)2 and tBuOOH as the stoichiometric oxidant
gave 41 which was converted to lactone 42 using Jones con-
ditions followed by selective reduction of the conjugated
double bond with zinc dust in acetic acid.[59] Subsequent
transesterification with trimethylsilyl ethanol and acetylation
of the released secondary hydroxyl group provided the nec-
essary building block 43.
ment of 48 with TASF and attachment of the released acid
49 to the common disaccharide platform 28 gave diene 50.
Its further elaboration by RCM and selective hydrogenation
of the resulting cycloalkene mixture was uneventful. At this
9702
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9697 – 9706

Natural Products
FULL PAPER
stage, the PMB ether in the
tether of 51 was cleaved with
DDQ[44] and the resulting prod-
uct 52 transformed into the nat-
ural product by following the
established route.
Ipomoeassin F and regioisomer:
Ipomoeassin F (6) differs from
its sister compound ipomoeas-
sin A (1) only by the presence
of two additional methylene
units in the tail hanging off the
fucoside anomeric center.[13] Al-
though this formal homologa-
tion required the preparation of
a
different fucoside building
block (13b, Scheme 2), no stra-
tegic changes were necessary en
route to this promising target.
Even though not all individu-
al steps shown in Scheme 10
were fully optimized, synthetic
6 was obtained in respectable
overall yield. From the bioactiv-
ity standpoint, its progenitor
compound 59 formed by cleav-
age of the isopropylidene group
in 58 is also noteworthy.
Though not a natural product
itself, it constitutes a valuable
probe for testing the subtle cor-
relation between the presence
of an acetyl group at 4’-OH and
the cytotoxicity of the ipo-
moeassins, which the published
data seem to imply (cf. Intro-
duction).[12,15,23] Additional in-
formation should be obtained
from the regioisomeric com-
pound 60, which was accidental-
ly isolated during our initial at-
tempts to drive the late-stage
acetal cleavage by using higher
concentrations of trifluoroacetic
acid. Although these conditions
ultimately lead to degradation,
it was noticed that the process
can be stopped at a stage where
compound 60 has accumulated.
Scheme 10. Total synthesis of ipomoeassin F and isomer: a) 26, 2,4,6-trichlorobenzoyl chloride, Et3N, DMAP,
toluene, 71–83%; b) DDQ, CH2Cl2/H2O, 75–84%; c) 4-oxo-8-nonenoic acid, 2,4,6-trichlorobenzoyl chloride,
Et3N, DMAP, toluene, 92–96%; d) complex 30 (20 mol%), CH2Cl2, reflux, 78–81%; e) H2 (1 atm), [RhCl-
ACHTUNGTRENNUNG(PPh3)3] (40 mol%), EtOH, 87–89%; f) TASF, MeCN, 43%; g) trifluoroacetic acid, CH2Cl2; h) MeCACHTUGNTREN(NUGN OEt)3,
camphersulfonic acid cat., then aq. HOAc, 96% (overall); i) trifluoroacetic acid, CH2Cl2, 54% (cf. text).
A
careful NMR analysis
showed unambiguously that
both ester groups on the glu-
cose have migrated along the
periphery.
isomer may provide further in-
sights into the relationship be-
Therefore
this
1
Figure 3. Concentration dependent H NMR spectra of ipomoeassin E recorded in C6D6 (0.6 mL each). Top to
bottom: 14 mg (commercial C6D6); 4 mg (commercial C6D6); 6 mg (distilled C6D6); 3 mg (distilled C6D6);
“commercial” refers to the solvent provided in ampules by the supplier.
Chem. Eur. J. 2009, 15, 9697 – 9706
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9703

A. Fꢂrstner et al.
tween the lateral acylation pattern and the cytotoxicity of
such resin glycosides.
ther statistical analysis, approximately 90 and 70% of all re-
corded events were gated after 24 and 48 h treatment, re-
spectively, in order to generally exclude smaller cell debris
(cf. Figure 4).
Biological assessment: All natural products and the small
assortment of non-natural analogues prepared during this
campaign were assayed for cytotoxicity against HeLa cells
and/or the mouse L-929 tumor cell line. Relevant results are
compiled in Table 2 and compared with the published data
for the A2780 human ovarian cancer cell line[12,13] and the
human leukemic monocyte lymphoma U937 cell line, which
had turned out to be the most sensitive amongst the tested
cell lines.[23]
Table 2. Antitumor activity of the ipomoeassins and analogues, and com-
parison with published data for the A2780 and U937 cell lines.[12,13,23]
Compound
HeLa
L-929[a]
A2780
U937
1
2
3
4
5
6
38
59
60
64 nm
2.5 mm
1.5 mm
32 nm
4.3 mm
77.8 nm
0.5 mm
0.4 mm
2.9 mm
35 nm
3.3 mm
36 nm
20.2 nm
134 nm
> 1 mm
135 nm
> 1 mm
7.4 nm
>1 mm
290 nm
580 nm
7.9 nm
163 nm
2.6 nm
7.0 mm
[a] Values represent the average of two measurements, incubation time:
5 d, in-screen validation against known cytotoxic compounds.
Our assays confirmed previous observations that the anti-
tumor activity of the individual ipomoeassin resin glycosides
is very strongly correlated with their acylation and oxygena-
tion pattern. In particular, acetylation of the axial 4’-OH
group of the fucopyranose is highly beneficial, as evident
from the comparison of the results obtained for the sister
compounds 1/2, 4/5, and 6/59. Thus, acetylation of this sec-
ondary alcohol upgrades the cytotoxicity of the compound
by up to two orders of magnitude. Moreover, comparison of
the pair 1 and 3 shows that hydroxylation of the C-5 posi-
tion in the lipophilic tether is detrimental, even though ace-
tylation of this site, as borne out in compound 4, largely re-
stores or even enhances the cytotoxicity. On the other hand,
our synthetic analogue 38, in which the carbonyl group of
the tether has been removed, is largely inactive, showing
that the ketone is essential. Likewise, our synthetic ana-
logues 59 and 60 reveal that the exact location of the two
unsaturated esters on the glucose ring has a strong influence
on the bioactivity.
Ipomoeassin F (6) consistently turned out to be the most
active member of this series, showing IC50 values in the low
nanomolar range against all cell lines tested so far. As 6 in-
corporates only two more methylene groups in the alkyl
chain than its shorter homologue 1, it is evident that the lipo-
philicity/hydrophilicity balance is another subtle but critical
parameter that deserves further careful inspection.
A cell cycle analysis showed that incubation of L-929 cells
with those ipomoeassins that have IC50 values <1 mm invari-
ably led to an accumulation of cells in the G1 phase. For fur-
Figure 4. Influence of different ipomoeassins and ipomoeassin analogues
on the cell cycle of mouse L-929 cells upon 24 h (top) and 48 h (bottom)
treatment. Untreated cultures showed a high peak in the G1 phase. In the
presence of ipomoeassins (varying concentrations according to formerly
determined IC50 values, see Table 2) cells further accumulated in the G1
phase, also showing a sub-G1 population after 48 h incubation. Color
code: light blue: MeOH control (final concentration 0.5% v/v); red: 6
(50 ngmLꢀ1); green: 1 (0.5 mgmLꢀ1); blue: 59 (1 mgmLꢀ1); brown: 4
(1 mgmLꢀ1); purple: 60 (5 mgmLꢀ1).
In untreated cultures, however, most of the cells (ca.
60%) were in the G1 phase, whereas approximately 25%
were in the G2/M phase. After 24 h treatment with ipo-
moeassins, the number of cells in the G1 phase increased to
about 73%, accompanied with a decreased number of L-929
mouse fibroblasts in the G2/M phase (ca. 18%). Besides the
observed G1 arrest, it could be shown that after 48 h treat-
ment a significant number of sub-G1 cells (DNA content
smaller than in G1, cf. Figure 4) and of apoptotic cells was
noticed. Only ca. 8% of the L-929 cells were in G2/M,
whereas the amount of cells in the S phase was still nearly
unaffected (ca. 10%).
9704
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9697 – 9706

Natural Products
FULL PAPER
[19] a) A. Fꢂrstner, M. Albert, J. Mlynarski, M. Matheu, E. DeClercq, J.
Am. Chem. Soc. 2003, 125, 13132–13142; b) A. Fꢂrstner, M. Albert,
J. Mlynarski, M. Matheu, J. Am. Chem. Soc. 2002, 124, 1168–1169;
c) A. Fꢂrstner, J. Mlynarski, M. Albert, J. Am. Chem. Soc. 2002,
124, 10274–10275.
Experimental Section
All experimental details can be found in the Supporting Information.
The details include compound characterization, bioassay and X-ray de-
tails.
[20] A. Fꢂrstner, J. Ruiz-Caro, H. Prinz, H. Waldmann, J. Org. Chem.
2004, 69, 459–467.
[21] a) A. Fꢂrstner, I. Konetzki, J. Org. Chem. 1998, 63, 3072–3080;
b) A. Fꢂrstner, I. Konetzki, Tetrahedron 1996, 52, 15071–15078.
[22] A. Fꢂrstner, T. Nagano, J. Am. Chem. Soc. 2007, 129, 1906–1907.
[23] M. H. D. Postema, K. TenDyke, J. Cutter, G. Kuznetsov, Q. Xu, Org.
Lett. 2009, 11, 1417–1420.
Acknowledgements
[24] a) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18–29;
b) A. Fꢂrstner, Angew. Chem. 2000, 112, 3140–3172; Angew. Chem.
Int. Ed. 2000, 39, 3013–3043; c) R. R. Schrock, A. H. Hoveyda,
Angew. Chem. 2003, 115, 4740–4782; Angew. Chem. Int. Ed. 2003,
42, 4592–4633; d) M. Schuster, S. Blechert, Angew. Chem. 1997,
109, 2124–2144; Angew. Chem. Int. Ed. Engl. 1997, 36, 2036–2056;
e) A. Fꢂrstner, Top. Catal. 1997, 4, 285–299; f) K. C. Nicolaou, P. G.
Bulger, D. Sarlah, Angew. Chem. 2005, 117, 4564–4601; Angew.
Chem. Int. Ed. 2005, 44, 4490–4527; g) A. Gradillas, J. Perez-Cas-
tells, Angew. Chem. 2006, 118, 6232–6247; Angew. Chem. Int. Ed.
2006, 45, 6086–6101; h) A. Deiters, S. F. Martin, Chem. Rev. 2004,
104, 2199–2238.
Generous financial support by the MPG, the Chemical Genomics Center
(CGC) of the MPG, and the Fonds der Chemischen Industrie is gratefully
acknowledged. We thank Dr. R. Mynott, B. Gabor and C. Wirtz for ex-
cellent NMR support, A. Deege, H. Hinrichs and R. Leichtweiß for their
help with many HPLC analyses, and Dr. C. W. Lehmann for solving the
crystal structure of compound 17a. We also thank Dr. B. Ellinger, MPI
fꢂr Molekulare Physiologie, Dortmund, for his help with some of the cy-
totoxicity assays.
[1] R. Pereda-Miranda, M. Bah, Curr. Top. Med. Chem. 2003, 3, 111–
131.
[2] V. M. Dembitsky, Chem. Biodiversity 2004, 1, 673–781.
[3] A. Fꢂrstner, Eur. J. Org. Chem. 2004, 943–958.
[4] J. Furukawa, N. Sahairi, Trends Glycosci. Glycotechnol. 2001, 13, 1–
10.
[5] a) D. P. Larson, C. H. Heathcock, J. Org. Chem. 1996, 61, 5208–
5209; b) D. P. Larson, C. H. Heathcock, J. Org. Chem. 1997, 62,
8406–8418; c) M. Brito-Arias, R. Pereda-Miranda, C. H. Heathcock,
J. Org. Chem. 2004, 69, 4567–4570.
[6] a) S.-F. Lu, Q. Oꢀyang, Z.-W. Guo, B. Yu, Y.-Z. Hui, Angew. Chem.
1997, 109, 2442–2444; Angew. Chem. Int. Ed. Engl. 1997, 36, 2344–
2346; b) S.-F. Lu, Q. Oꢀyang, Z.-W. Guo, B. Yu, Y.-Z. Hui, J. Org.
Chem. 1997, 62, 8400–8405.
[25] Early sudies: a) A. Fꢂrstner, K. Langemann, J. Org. Chem. 1996, 61,
3942–3943; b) A. Fꢂrstner, N. Kindler, Tetrahedron Lett. 1996, 37,
7005–7008; c) A. Fꢂrstner, K. Langemann, Synthesis 1997, 792–803;
d) A. Fꢂrstner, K. Langemann, J. Am. Chem. Soc. 1997, 119, 9130–
9136; e) A. Fꢂrstner, T. Mꢂller, Synlett 1997, 1010–1012; f) A. Fꢂrst-
ner, T. Gastner, H. Weintritt, J. Org. Chem. 1999, 64, 2361–2366;
g) A. Fꢂrstner, G. Seidel, N. Kindler, Tetrahedron 1999, 55, 8215–
8230; h) A. Fꢂrstner, O. R. Thiel, N. Kindler, B. Bartkowska, J. Org.
Chem. 2000, 65, 7990–7995; i) A. Fꢂrstner, O. R. Thiel, L. Acker-
mann, Org. Lett. 2001, 3, 449–451; j) A. Fꢂrstner, J. Grabowski,
C. W. Lehmann, J. Org. Chem. 1999, 64, 8275–8280; k) A. Fꢂrstner,
O. R. Thiel, J. Org. Chem. 2000, 65, 1738–1742.
[7] a) A. Fꢂrstner, T. Mꢂller, J. Am. Chem. Soc. 1999, 121, 7814–7821;
b) A. Fꢂrstner, T. Mꢂller, J. Org. Chem. 1998, 63, 424–425.
[8] A. Fꢂrstner, F. Jeanjean, P. Razon, Angew. Chem. 2002, 114, 2203–
2206; Angew. Chem. Int. Ed. 2002, 41, 2097–2101.
[9] a) A. Fꢂrstner, F. Jeanjean, P. Razon, C. Wirtz, R. Mynott, Chem.
Eur. J. 2003, 9, 307–319; b) A. Fꢂrstner, F. Jeanjean, P. Razon, C.
Wirtz, R. Mynott, Chem. Eur. J. 2003, 9, 320–326.
[10] For the total synthesis of other resin glycosides see: a) Z.-H. Jiang,
A. Geyer, R. R. Schmidt, Angew. Chem. 1995, 107, 2730–2734;
Angew. Chem. Int. Ed. Engl. 1995, 34, 2520–2524; b) J. Furukawa, S.
Kobayashi, M. Nomizu, N. Nishi, N. Sakairi, Tetrahedron Lett. 2000,
41, 3453–3457; c) X.-M. Zhu, L.-L. He, G.-L. Yang, M. Lei, S.-S.
Chen, J.-S. Yang, Synlett 2006, 3510–3512.
[11] Based on the structure of tricolorin A in the solid state, it was pro-
posed that this cytotoxic resin glycoside may exert its function by
forming pores in the cell membrane, cf. A. Rencurosi, E. P. Mitchell,
G. Cioci, S. Pꢅrez, R. Pereda-Miranda, A. Imberty, Angew. Chem.
2004, 116, 6044–6048; Angew. Chem. Int. Ed. 2004, 43, 5918–5922;
however, this assumption does not readily explain why closely simi-
lar compounds show greatly different activities.
[26] For recent total syntheses based on RCM-macrocyclizations report-
ed by our group, see: a) A. Fꢂrstner, L. C. Bouchez, J.-A. Funel, V.
Liepins, F.-H. Porꢅe, R. Gilmour, F. Beaufils, D. Laurich, M.
Tamiya, Angew. Chem. 2007, 119, 9425–9430; Angew. Chem. Int.
Ed. 2007, 46, 9265–9270; b) A. Fꢂrstner, C. Nevado, M. Waser, M.
´
Tremblay, C. Chevrier, F. Teply, C. Aꢆssa, E. Moulin, O. Mꢂller, J.
Am. Chem. Soc. 2007, 129, 9150–9161; c) A. Fꢂrstner, C. Nevado,
´
M. Tremblay, C. Chevrier, F. Teply, C. Aꢆssa, M. Waser, Angew.
Chem. 2006, 118, 5969–5974; Angew. Chem. Int. Ed. 2006, 45, 5837–
5842; d) A. Fꢂrstner, L. C. Bouchez, L. Morency, J.-A. Funel, V. Lie-
pins, F.-H. Porꢅe, R. Gilmour, D. Laurich, F. Beaufils, M. Tamiya,
Chem. Eur. J. 2009, 15, 3983–4010; e) A. Fꢂrstner, C. Mꢂller, Chem.
Commun. 2005, 5583–5585; f) B. Scheiper, F. Glorius, A. Leitner,
A. Fꢂrstner, Proc. Natl. Acad. Sci. USA 2004, 101, 11960–11965;
g) C. Aꢆssa, R. Riveiros, J. Ragot, A. Fꢂrstner, J. Am. Chem. Soc.
2003, 125, 15512–15520; h) A. Fꢂrstner, K. Radkowski, C. Wirtz, R.
Goddard, C. W. Lehmann, R. Mynott, J. Am. Chem. Soc. 2002, 124,
ˇ
7061–7069; i) J. Pospꢁsil, C. Mꢂller, A. Fꢂrstner, Chem. Eur. J. 2009,
15, 5956–5968.
[27] The total synthesis of 6 reported by the Eisai group later confirmed
this forecast. Specifically, the introduction of the cinnamate at O-4’’
required 6–7 evaporative cycles to give 62% combined yield, and
the attachment of the tiglic acid residue to O-3’’ gave only 45% of
the desired product.
[12] S. Cao, R. C. Guza, J. H. Wisse, J. S. Miller, R. Evans, D. G. I. King-
ston, J. Nat. Prod. 2005, 68, 487–492.
[13] S. Cao, A. Norris, J. H. Wisse, J. S. Miller, R. Evans, D. G. I. King-
ston, Nat. Prod. Res. 2007, 21, 872–876.
[14] 11-Hydroxy-hexadecoanoic acid (jalapinolic acid) present in ipo-
moeassin F is by far the most common fatty acid component of the
resin glycosides known to date, cf. ref. [1–4].
[28] For a related sequence, see: N. Holub, J. Neidhçfer, S. Blechert,
Org. Lett. 2005, 7, 1227–1229.
[29] H. Paulsen, Angew. Chem. 1982, 94, 184–201; Angew. Chem. Int.
Ed. Engl. 1982, 21, 155–173.
[30] a) X. Zhu, R. R. Schmidt, Angew. Chem. 2009, 121, 1932–1967;
Angew. Chem. Int. Ed. 2009, 48, 1900–1934; b) R. R. Schmidt,
Angew. Chem. 1986, 98, 213–236; Angew. Chem. Int. Ed. Engl.
1986, 25, 212–235; c) R. R. Schmidt, K.-H. Jung, in Preparative Car-
[15] D. G. I. Kingston, S. Cao, US 2006/0264383A1; Nov. 23, 2006.
[16] D. G. I. Kingston, J. Org. Chem. 2008, 73, 3975–3984.
[17] C. W. Lehmann, A. Fꢂrstner, T. Mꢂller, Z. Kristallogr. 2000, 215,
114–117.
[18] A. Fꢂrstner, K. Radkowski, J. Grabowski, C. Wirtz, R. Mynott, J.
Org. Chem. 2000, 65, 8758–8762.
Chem. Eur. J. 2009, 15, 9697 – 9706
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9705

A. Fꢂrstner et al.
bohydrate Chemistry (Ed.: S. Hanessian), Dekker, New York, 1997,
pp. 283–312.
[51] TBAF was found unsuitable in this case as it also cleaves the tiglate
ester.
[52] a) R. U. Lemieux, H. Driguez, J. Am. Chem. Soc. 1975, 97, 4069–
4075; b) see also: R. Chevalier, B. Colsch, C. Afonso, N. Baumann,
J.-C. Tabet, J.-M. Mallet, Tetrahedron 2006, 62, 563–577.
[53] R. M. Wilson, S. J. Danishefsky, J. Org. Chem. 2006, 71, 8329–8351.
[54] Diverted total synthesis” allows for the preparation of analogues
that cannot (or hardly) be obtained by chemical manipulation of the
lead compound itself. Therefore it should be distinguished from the
traditional methods of derivatization of natural products, as com-
monly practiced in medicinal chemistry.
[31] a) P. J. Garegg, in Preparative Carbohydrate Chemistry (Ed.: S. Ha-
nessian), Dekker, New York, 1997, pp. 53–67; b) for a compilation
of more recent methodology, see: K. Daragics, P. Fꢂgedi, Tetrahe-
dron Lett. 2009, 50, 2914–2916; c) mechanistic studies: R. Johnsson,
D. Olsson, U. Ellervik, J. Org. Chem. 2008, 73, 5226–5232.
[32] P. Y. Chong, P. A. Petillo, Org. Lett. 2000, 2, 1093–1096.
[33] C.-R. Shie, Z.-H. Tzeng, S. S. Kulkarni, B.-J. Uang, C.-Y. Hsu, S.-C.
Hung, Angew. Chem. 2005, 117, 1693–1696; Angew. Chem. Int. Ed.
2005, 44, 1665–1668.
[55] For previous exercises in “diverted total synthesis” from this group,
see inter alia: a) A. Fꢂrstner, D. Kirk, M. D. B. Fenster, C. Aꢆssa, D.
De Souza, O. Mꢂller, Proc. Natl. Acad. Sci. USA 2005, 102, 8103–
8108; b) A. Fꢂrstner, D. De Souza, L. Turet, M. D. B. Fenster, L.
Parra-Rapado, C. Wirtz, R. Mynott, C. W. Lehmann, Chem. Eur. J.
2007, 13, 115–134; c) A. Fꢂrstner, D. Kirk, M. D. B. Fenster, C.
Aꢆssa, D. De Souza, C. Nevado, T. Tuttle, W. Thiel, O. Mꢂller,
Chem. Eur. J. 2007, 13, 135–149; d) A. Fꢂrstner, A. Korte, Chem.
Asian J. 2008, 3, 310–318; e) A. Fꢂrstner, T. Nagano, C. Mꢂller, G.
Seidel, O. Mꢂller, Chem. Eur. J. 2007, 13, 1452–1462; f) A. Fꢂrstner,
K. Grela, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 2000, 122,
11799–11805; g) A. Fꢂrstner, S. Flꢂgge, O. Larionov, Y. Takahashi,
T. Kubota, J. Kobayashi, Chem. Eur. J. 2009, 15, 4011–4029; h) A.
Fꢂrstner, E. Kattnig, G. Kelter, H.-H. Fiebig, Chem. Eur. J. 2009, 15,
4030–4043; i) A. Fꢂrstner, K. Radkowski, H. Peters, G. Seidel, C.
Wirtz, R. Mynott, C. W. Lehmann, Chem. Eur. J. 2007, 13, 1929–
1945; j) A. Fꢂrstner, F. Feyen, H. Prinz, H. Waldmann, Tetrahedron
2004, 60, 9543–9558; k) P. Buchgraber, M. M. Domostoj, B. Scheip-
er, C. Wirtz, R. Mynott, J. Rust, A. Fꢂrstner, Tetrahedron 2009, 65,
6519–6534.
[34] X. Wu, R. R. Schmidt, Eur. J. Org. Chem. 2004, 2826–2832.
[35] D. Crich, A. Banerjee, Org. Lett. 2005, 7, 1395–1398.
[36] S. Chandrasekhar, Y. R. Reddy, C. R. Reddy, Chem. Lett. 1998,
1273–1274.
[37] M. Sakagami, H. Hamana, Tetrahedron Lett. 2000, 41, 5547–5551.
[38] R. Johansson, B. Samuelsson, J. Chem. Soc. Perkin Trans. 1 1984,
2371–2374.
[39] For a discussion of the “economy of steps” as a major design princi-
ple, see: A. Fꢂrstner, Synlett 1999, 1523–1533.
[40] Attempts to trap the organoaluminum intermediate directly with
PhMe2SiCl were low yielding.
[41] Y. Ukaji, K. Sada, K. Inomata, Chem. Lett. 1993, 1227–1230.
[42] Although the regiomers 20 and 21 can be separated by preparative
HPLC, the mixture can be used in this step. The resulting acylated
products are conveniently separated by routine flash chromatogra-
phy.
[43] J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull.
Chem. Soc. Jpn. 1979, 52, 1989–1993.
[44] Y. Oikawa, T. Yoshioka, O. Yonemitsu, Tetrahedron Lett. 1982, 23,
885–888.
[56] T.-L. Ho, S. G. Sapp, Synth. Commun. 1983, 13, 207–211.
[57] a) M. Kusakabe, Y. Kitano, Y. Kobayashi, F. Sato, J. Org. Chem.
1989, 54, 2085–2091; b) T. Kametani, M. Tsubuki, Y. Tatsuzaki, T.
Honda, J. Chem. Soc. Perkin Trans. 1 1990, 639–646.
[58] Alternative but less satisfactory syntheses of (ꢀ)-40 relied on a
Noyori-transfer hydrogenation of the corresponding ketone which
afforded the desired product in 77% with 93% ee, whereas an asym-
metric addition of bis(3-butenyl)zinc to furan-2-carbaldehyde medi-
[45] Prepared by reaction of 4-pentenylmagnesium bromide with succinic
anhydride, cf. G. Lhommet, S. Frꢅville, V. Thuy, H. Petit, J. P. Cꢅlꢅri-
er, Synth. Commun. 1996, 26, 3897–3901. It was found that the reac-
tion is cleaner in the presence of FeACTHNUTRGNENUG(acac)3 (10 mol%) in THF at
ꢀ788C, even though the unoptimized yield was only 41%. For iron
catalyzed formations of functionalized ketones, see: B. Scheiper, M.
Bonnekessel, H. Krause, A. Fꢂrstner, J. Org. Chem. 2004, 69, 3943–
3949.
ated by TiACHTUNRGTNE(NUG OiPr)4 (1.2 equiv) and trans-N,N’-1,2-cyclohexandiyl bis-
[46] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953–
956.
trifluoromethanesulfonamide (5 mol%) in toluene at ꢀ208C gave
only 70% ee.
[47] For closely related second generation catalysts that were almost si-
multaneously reported by our group, see: a) L. Ackermann, A.
Fꢂrstner, T. Weskamp, F. J. Kohl, W. A. Herrmann, Tetrahedron
Lett. 1999, 40, 4787–4790; b) L. Ackermann, D. El Tom, A. Fꢂrst-
ner, Tetrahedron 2000, 56, 2195–2202; c) A. Fꢂrstner, O. R. Thiel, L.
Ackermann, H.-J. Schanz, S. P. Nolan, J. Org. Chem. 2000, 65, 2204–
2207; d) A. Fꢂrstner, L. Ackermann, B. Gabor, R. Goddard, C. W.
Lehmann, R. Mynott, F. Stelzer, O. R. Thiel, Chem. Eur. J. 2001, 7,
3236–3253.
[59] M. P. Georgiadis, A. Tsekouras, S. I. Kotretsou, S. A. Haroutounian,
M. G. Polissiou, Synthesis 1991, 929–932.
[60] Attempts to form diene 45 directly by reaction of 28 with lactone 42
were low yielding.
[61] S.-K. Kang, D.-C. Park, H.-S. Rho, C.-M. Yu, J.-H. Hong, Synth.
Commun. 1995, 25, 203–214.
[62] K. C. Nicolaou, C. W. Hummel, M. Nakada, K. Shibayama, E. N.
Pitsinos, H. Saimoto, Y. Mizuno, K.-U. Baldenius, A. L. Smith, J.
Am. Chem. Soc. 1993, 115, 7625–7635.
[48] A. J. Birch, D. H. Williamson, Org. React. 1976, 24, 1–186.
[49] a) R. Noyori, I. Nishida, J. Sakata, J. Am. Chem. Soc. 1983, 105,
1598–1608; b) K. A. Scheidt, H. Chen, B. C. Follows, S. R. Chemler,
D. S. Coffey, W. R. Roush, J. Org. Chem. 1998, 63, 6436–6437.
[50] A. Fꢂrstner, M. D. B. Fenster, B. Fasching, C. Godbout, K. Radkow-
ski, Angew. Chem. 2006, 118, 5636–5641; Angew. Chem. Int. Ed.
2006, 45, 5510–5515.
[63] R. Hong, Y. Chen, L. Deng, Angew. Chem. 2005, 117, 3544–3547;
Angew. Chem. Int. Ed. 2005, 44, 3478–3481.
[64] D. M. Hodgson, P. A. Stupple, F. Y. T. M. Pierard, A. H. Labande, C.
Johnstone, Chem. Eur. J. 2001, 7, 4465–4476.
Received: May 29, 2009
Published online: August 20, 2009
9706
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9697 – 9706