Studies on iejimalide B: Preparation of the seco acid and identification of the molecule's "Achilles Heel"

Article abstract of DOI:10.1002/anie.200601859

(Chemical Equation Presented) Limitations of models in total synthesis are illustrated by a study towards the potent cytotoxic macrolide iejimalide B. Although the Yamaguchi protocol allowed for the esterification of elaborate segments, attempted macrolactonization of the seco acid met with failure (see scheme, Boc = tert-butyloxycarbonyl). The assembly of the seco acid involves some of the most advanced applications of the Julia olefination known to date.

Full text of DOI:10.1002/anie.200601859

Communications
Total Synthesis
DOI: 10.1002/anie.200601859
Studies on Iejimalide B: Preparation of the Seco
Acid and Identification of the Moleculeꢀs
“Achilles Heel”**
Alois Fürstner,* Christophe Aïssa, Carine Chevrier,
´
Filip Teply, Cristina Nevado, and Martin Tremblay
Extraction of the tunicate Eudistoma cf. rigida collected off Ie
island, Okinawa province, Japan, led to the isolation of a
family of novel 24-membered polyene macrolides designated
iejimalides A–D (1–4).[1] Although the structure of these
extremely scarce cytotoxic metabolites (0.0003–0.0006% of
the tunicates wet weight) could be established, it was only
after a reextraction from a Cystodytes sp. that enough
´
[*] Prof. A. Fürstner, Dr. C. Aïssa, Dr. C. Chevrier, Dr. F. Teply,
Dr. C. Nevado, Dr. M. Tremblay
Max-Planck-Institut für Kohlenforschung
45470 Mülheim/Ruhr (Germany)
Fax: (+49)208-306-2994
E-mail: fuerstner@mpi-muelheim.mpg.de
[**] Generous financial support from the MPG, the Fonds der
Chemischen Industrie, the Alexander-von-Humboldt Foundation
(fellowship to C.N.), and the NSERC, Canada (fellowship to M.T.) is
gratefully acknowledged. We thank Dr. R. Riveiros for assistance in
the initial phase of this project, and Dr. R. Mynott, B. Gabor, and C.
Wirtz for their invaluable help with the structure assignment of
several key compounds.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5832
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837

Angewandte
Chemie
material was accumulated to allow for the establishment of
their relative and absolute stereochemistry; at the same time,
ꢀ
the configuration of the C13 C14 double bond was corrected
to Z (rather than E as originally assigned).[2]
The scarcity of the iejimalides also delayed in-depth
studies of their physiological properties; details reported
recently, however, are very encouraging. Specifically, the data
for iejimalide A, disclosed in 2005 by the National Cancer
Institute (NCI), illustrates the truly remarkable potency of 1
against the panel of 60 standard human cancer cell lines, with
the concentrations required to inhibit growth by 50% (GI50)
and tumor gene index (TGI) values in the low nanomolar
range.[3,4] Equally remarkable is the report by Kobayashi and
co-workers that the activity profile of 1–4 does not correlate
with those of other anticancer drugs, which might indicate an
unprecedented mode of action.[2] The same authors also
demonstrated the potent in vivo activity of 3 and 4 against
P388 leukemia.[2a]
Scheme 1. “First generation” retrosynthetic analysis and Kobayashi’s
Collectively, these data suggest that the iejimalides may
be candidates for further development in a (pre)clinical
setting. Spurred on by this prospect, we ventured into a
synthesis-driven investigation of these enticing targets.[5]
Therefore, it was necessary to prepare sufficient amounts
for further testing. Presented in this and the following
Communication is a preliminary report on our endeavors in
the field, which not only led to the first total synthesis of
iejimalide B (2) as the most active member of the series, but
also unraveled many of the peculiar chemical characteristics
of these delicate lead compounds.[6,7]
numbering scheme of iejimalide B (2).
A cursory inspection of 1–4 shows that five of the six chiral
centers reside at allylic or even doubly allylic sites. Appre-
hensive that this feature likely potentiates the lability of their
polyunsaturated backbones, our first approach to 2 gravitated
toward methodology that is believed to cope with such
potentially fragile structural elements.
Specifically, it was planned to incorporate the peptide
residue at the very end of the synthesis, because of the
proclivity of N-formylserine derivatives toward racemization,
formyl cleavage, and oxazoline formation. The success of the
venerable Yamaguchi macrolactonization in complex set-
tings[8,9] and our excellent experiences with the Julia olefina-
tion for the elaboration of sensitive, and even enolizable
compounds, suggested the use of these transformations in the
Scheme 2. a) NaHMDS, THF, ꢀ788C!RT (“Barbier conditions”),
89% (E/Z=4:1); b) HCl (5% in EtOH), EtOH, 08C, 60%; c) 1-phenyl-
1H-tetrazol-5-thiol, DEAD, PPh3, THF, 82%; d) cat. [Mo7O24(NH4)6]
·4H2O, aq H2O2, EtOH, 75%. HMDS=1,1,1,3,3,3-hexamethyldisila-
zane, DEAD=diethylazodicarboxylate, PTS=1-phenyl-1H-tetrazol-5-
thiyl.
present context.[10,11] Complemented by Pd-catalyzed C C
ꢀ
bond formations,[12] this ensemble of reliable and mild
chemical reactions promised a flexible access route to 2
startingfrom five synthons of similar size and moderate
complexity (Scheme 1).
The required buildingblock 10 (Scheme 2) was prepared
by a Julia olefination of sulfone 5 (derived from the
commercial Roche ester)[13] and aldehyde 6 (derived from
commercial (S)-2-hydroxybutyrolactone)[13] to give alkene 7
in almost quantitative yield in a diastereomeric ratio of E/Z =
4:1. Since attempted isomerization with PhSSPh/azobisisobu-
tyronitrile (AIBN) led to a complex mixture, (E)-7 and (Z)-7
were separated by preparative HPLC prior to removal of the
terminal tert-butyldimethylsilyl (TBS) ether and installation
of a phenyltetrazolyl sulfone terminus in 10 by the standard
two-step protocol. It is notable that the cleavage of the silyl
ether in 7 worked only with dilute HCl in EtOH; the use of
tetra-n-butylammonium fluoride (TBAF)/HOAc affected the
integrity of the conjugated diene. Although this finding
forecasted there would be subtle stability issues, it was not
until the late stages of the synthesis that this aspect was fully
appreciated.
The complementary hemisphere of 2 was accessible by
adaptation of a literature route,[7a] involvingan effective Heck
reaction of 11 and 12 as the key step,[14] deliveringmultigram
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5833

Communications
amounts of aldehyde 15 (Scheme 3). This compound was then
cible results on fairly large scale reactions.[17] Oxidative
cleavage of the double bond followed by a modified Still–
Gennari olefination of the resultingaldehyde 21 afforded the
required Z-configured enoate 22 as a single isomer.[18,19]
Cleavage of the silyl group with K2CO3 in MeOH,[20]
exhaustive reduction/carbometalation of the resultingprod-
uct 23 with an excess of [Cp2Zr(H)Cl],[21] and an iodolytic
workup completed the synthesis of alkenyl iodide 24, which
was obtained in 34% yield over seven readily scalable
operations.
subjected to Brown crotylation[15] to install the anti-config-
ured homoallylic motif of 16 in excellent chemical yield and
remarkable optical purity (anti/syn 98:2, 95 % ee).
Since the homoallylic terminus renders 16 directly amen-
able to cross-couplingunder Heck conditions, this particular
ꢀ
C C bond formation was preferred over other conceivable
reactions. Despite its impressive track record,[14] attempted
fusion of iodide 24 with alkene 16 under various experimental
conditions led to a morass of isomers. A successful couplingof
these components was only possible after switchingto the
“cationic manifold”, which is operative upon addition of
AgOAc (1.2 equiv)[22] (Scheme 5). The desired E,E-config-
ured diene 25 could be separated from the E,Z isomer by
careful chromatography.
Scheme 3. a) Pd(OAc)2 (3 mol%), P(o-tol)3 (6 mol%), Et3N, 1008C,
84%; b) trifluoroacetic acid, CH2Cl2, 87%; c) 1. DIBAL-H, CH2Cl2,
ꢀ788C, 97%; 2. DMSO, (COCl)2, Et3N, CH2Cl2, ꢀ788C!RT, 79%;
d) (+)-((E)-crotyl)-B(Ipc)2, THF, ꢀ788C, 82% (95% ee). Boc=tert-
butyloxycarbonyl, tol=tolyl, DIBAL-H=diisobutylaluminum hydride,
Ipc=isopinocampheyl.
The couplingpartner 24 was prepared by asymmetric
transfer hydrogenation of ketone 17[13] followingconditions
developed by Noyori and co-workers and furnished propargyl
alcohol 19 in 98% yield and 98.8% ee, without affectingthe
olefinic and the acetylenic sites in the vicinity (Scheme 4).[16]
In contrast, however, the seemingly trivial conversion of 19
into methyl ether 20 was surprisingly troublesome. After
considerable experimentation, it was found that the use of
DMSO as a dipolar cosolvent gave satisfactory and reprodu-
Scheme 5. a) Pd(OAc)2 (10 mol%), AgOAc, DMF, RT, 46% (25) +
13% (D[20,21] isomer); b) cat. tBuOK, cat. CuBr2, cat. 2,2’-bipyridine,
cat. TEMPO, O2 (1 atm), MeCN/H2O (2:1), 94%; c) 10, NaHMDS,
THF, ꢀ788C!RT, 57% (E/Z>10:1); d) Me3SnOH (40 equiv), 1,2-
dichloroethane, 808C, 94%. TEMPO=2,2,6,6-tetramethyl-1-piperidi-
noxyl (free radical), TES=triethylsilyl, PMB=para-methoxybenzyl.
Although both alcohol groups in 25 are allylic, the
selective oxidation of the primary one was accomplished
with the aid of a procedure co-catalyzed by copper and
Scheme 4. a) 18 (0.6 mol%), iPrOH, 98% (98.8% ee); b) 1. nBuLi,
MeI, THF, ꢀ788C; 2. DMSO, ꢀ258C!RT; c) cat. OsO4, NaIO4, 2,6-
lutidine, aq 1,4-dioxane, 74% (over two steps);
d) (CF3CH2O)2P(O)CH(Me)COOMe, KHMDS, [18]crown-6 (0.7 equiv),
toluene, ꢀ208C, 87%; e) K2CO3, MeOH, 80%; f) [Cp2Zr(H)Cl]
(3.1 equiv), THF, then I2, 80%. Cp=cyclopentadienyl, Ts=toluene-4-
sulfonyl.
TEMPO, usingO as the ultimate oxidant.[23] The resulting
2
aldehyde 26 was amenable to Julia–Kocienski olefination[10,24]
with tetrazolylsulfone 10, provided that this reagent was
deprotonated with an excess of NaHMDS in the presence of
the aldehyde (“Barbier conditions”).
5834
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837

Angewandte
Chemie
The success of this transformation attests to the maturity
of the Kocienski methodology, especially if one considers that
an unprotected OH group,[25] a carbamate, and potentially
Despite this encouraging precedent, attempted lactoniza-
tion of seco acid 28 failed to afford the desired macrocycle
(Scheme 8). Rather, the C1–C6 unit of the substrate was
ꢀ
acidic C H sites are present in the reaction partners. The
desired alkene 27, which represents the C1–C28 carbon
backbone of 2 in a fully functional form, was obtained in a
reproducible yield of 57% with excellent selectivity for the
required E isomer (E/Z > 10:1, NMR).
At this stage, iejimalide B seemed to be just a few
standard manipulations away; the most important lessons,
however, had yet to be learned. First, attempted conversion of
ester 27 into the required seco acid 28 turned out to be very
difficult. While saponification of the north segment 9, naively
considered as a valid model, worked as expected (Scheme 6),
Scheme 8. Attempted macrocyclization under Yamaguchi conditions:
a) 1. 2,4,6-trichlorobenzoyl chloride, Et3N, THF; 2. cat. DMAP, toluene,
see text.
Scheme 6. a) aq LiOH (1m), THF/MeOH (1:1), RT, 91%.
aromatized to a phenol ring, as shown by extensive 2D NMR
spectroscopic investigations. The contrast between this out-
come and the intermolecular esterification depicted in
Scheme 7 is striking. Although phenol 31 was isolated in
rather low yields (25–40%), it was the only distinct product
that was reproducibly formed under the standard Yamaguchi,
as well as the “one-pot” Yonemitsu conditions. Unaware of
any precedent, we propose the mechanism depicted in
Scheme 9 to explain this unusual result. Attack of the base
(most likely DMAP)[29] on the a,b-unsaturated mixed anhy-
dride 32 initially formed may outperform the attack by the
hindered alcohol at C23. Collapse of the resultingintermedi-
ate 33 is thought to release ketene 34, which eliminates the
dimethylpyridinium moiety to give the polyunsaturated
attempted saponification of 27 compromised the seemingly
remote C18–C23 region of the molecule without touching the
ester function at all; other standard conditions were equally
unsuitable. Only the procedure recently outlined by Nicolaou
et al. afforded the desired product 28,[26,27] although a large
excess of Me3SnOH and prolonged (ꢁ 5 days) heatingwere
necessary to obtain satisfactory results.
Prior to subjectingthis valuable material to macrocycli-
zation, the arsenal of feasible methods[8b] was validated in
various model reactions. By far the best results over the entire
series were obtained with the Yonemitsu variant of the
Yamaguchi method.[8,28] The most advanced and instructive
case is the successful esterification of the northern domain 29
with the elaborate south-eastern hemisphere 25b (R = TBS),
which afforded product 30 in a respectable 72% yield
(Scheme 7).
Scheme 7. An intermolecular esterification serving as a model for the
projected macrocyclization: a) 2,4,6-trichlorobenzoyl chloride, Et3N,
cat. DMAP, toluene, 72%. DMAP=4-dimethylaminopyridine.
Scheme 9. Proposed mechanism for the observed phenol formation
under Yamaguchi conditions.
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5835

Communications
Lett. 1997, 38, 6375 – 6378; c) T. M. Pedersen, E. L. Hansen, J.
Kane, T. Rein, P. Helquist, P.-O. Norrby, D. Tanner, J. Am. Chem.
Soc. 2001, 123, 9738 – 9742.
ketene 35, which is set-up for a subsequent 6p-electrocycliza-
tion with formation of phenol 36.[30] This compound is finally
acylated to give the observed product 31.
[8] a) J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi,
Bull. Chem. Soc. Jpn. 1979, 52, 1989 – 1993; b) review: A.
Parenty, X. Moreau, J.-M. Campagne, Chem. Rev. 2006, 106,
911 – 939.
[9] For advanced applications from our research group see: a) O.
Lepage, E. Kattnig, A. Fürstner, J. Am. Chem. Soc. 2004, 126,
15970 – 15971; b) J. Mlynarski, J. Ruiz-Caro, A. Fürstner, Chem.
Eur. J. 2004, 10, 2214 – 2222.
Despite considerable experimentation, this unfavorable
reaction channel could not be avoided.[29] Gentle warmingof
the mixture with the hope of renderingthe attack by the
tethered alcohol competitive, only led to the rapid destruction
of the startingmaterial. Similarly, applications of other
standard macrolactonization strategies[8b] failed.
At this point, we had to conclude that the lactone entity of
2 constitutes a highly vulnerable “Achilles heel” of this
unusually sensitive marine natural product in a synthetic
context. Even though the required seco acid 28 could be
secured by a convergent approach involving one of the most
advanced applications of the Julia–Kocienski olefination
known to date,[10] our inability to lactonize the seco acid
enforced a relaunch of the entire project. Since the crucial
ester clearly must be installed intermolecularly, only a
[10] Review: P. R. Blakemore, J. Chem. Soc. Perkin Trans. 1 2002,
2563 – 2585.
[11] a) A. Fürstner, M. Albert, J. Mlynarski, M. Matheu, E.
DeClercq, J. Am. Chem. Soc. 2003, 125, 13132 – 13142; b) A.
Fürstner, J. Mlynarski, M. Albert, J. Am. Chem. Soc. 2002, 124,
10274 – 10275; c) A. Fürstner, F. Feyen, H. Prinz, H. Waldmann,
Tetrahedron 2004, 60, 9543 – 9558.
[12] K. C. Nicolaou, P. G. Bulger, D. Sarlah, Angew. Chem. 2005, 117,
4516 – 4563; Angew. Chem. Int. Ed. 2005, 44, 4442 – 4489.
[13] For details see the SupportingInformation.
[14] S. Bräse, A. de Meijere in Metal-Catalyzed Cross-Coupling
Reactions, Vol. 1 (Eds.: A. de Meijere, F. Diederich), Wiley-
VCH, Weinheim, 2004, pp. 217 – 315.
[15] H. C. Brown, K. S. Bhat, J. Am. Chem. Soc. 1986, 108, 293 – 294.
[16] a) K. Matsumura, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am.
Chem. Soc. 1997, 119, 8738 – 8739; b) R. Noyori, S. Hashiguchi,
Acc. Chem. Res. 1997, 30, 97 – 102; c) for a recent application
from this research group see: A. Fürstner, P. Hannen, Chem.
Eur. J. 2006, 12, 3006 – 3019.
ꢀ
macrocyclization by C C-bond formation can forge the 24-
membered ringof 2. Amongthe various conceivable options,
ring-closing metathesis (RCM) ultimately proved viable,[31]
even though application of this method may seem counter-
intuitive in view of the necessity to activate a polyunsaturated
and likely very fragile cyclization precursor by the catalyst in
a strictly regio- and chemoselective fashion. The following
Communication reports how this ambitious plan was achieved
in practice, although success came only after experiencing yet
another “Achilles heel” of the iejimalides that went unde-
tected in the studies summarized above.[6]
[17] R. Suzuki, H. Tsukuda, N. Watanabe, Y. Kuwatani, I. Ueda,
Tetrahedron 1998, 54, 2477 – 2496.
[18] W. C. Still, C. Gennari, Tetrahedron Lett. 1983, 24, 4405 – 4408.
[19] S. J. Mickel, G. H. Sedelmeier, D. Niederer, F. Schuerch, M.
Seger, K. Schreiner, R. Daeffler, A. Osmani, D. Bixel, O.
Loiseleur, J. Cercus, H. Stettler, K. Schaer, R. Gamboni, A.
Bach, G.-P. Chen, W. Chen, P. Geng, G. T. Lee, E. Loeser, J.
McKenna, F. R. Kinder, Jr., K. Konigsberger, K. Prasad, T. M.
Ramsey, N. Reel, O. Repi›, L. Rogers, W.-C. Shieh, R.-M. Wang,
L. Waykole, S. Xue, G. Florence, I. Paterson, Org. Process Res.
Dev. 2004, 8, 113 – 121.
Received: May 11, 2006
Published online: August 4, 2006
Keywords: Heck reaction · lactones · macrolides · olefination ·
.
total synthesis
[20] The use of TBAF and other fluoride sources scrambled the
trisubstituted olefin of the enoate.
[21] D. W. Hart, J. Schwartz, J. Am. Chem. Soc. 1974, 96, 8115 – 8116.
[22] M. Kalesse, K. P. Chary, M. Quitschalle, A. Burzlaff, C. Kasper,
T. Scheper, Chem. Eur. J. 2003, 9, 1129 – 1136.
[23] P. Gamez, I. W. C. E. Arends, R. A. Sheldon, J. Reedijk, Adv.
Synth. Catal. 2004, 346, 805 – 811.
[24] P. R. Blakemore, W. J. Cole, P. J. Kocienski, A. Morley, Synlett
1998, 26 – 28.
[25] Note that an alcohol might interfere with the Smiles rearrange-
ment, which constitutes the crucial second step of the Kocienski
variant of the Julia olefination, see. Ref. [24].
[26] K. C. Nicolaou, A. A. Estrada, M. Zak, S. H. Lee, B. S. Safina,
Angew. Chem. 2005, 117, 1402 – 1406; Angew. Chem. Int. Ed.
2005, 44, 1378 – 1382.
[27] Attempts to form the macrocycle by intramolecular transesteri-
fication followinga closely related method (cat. Bu 2SnO,
toluene, reflux) were unsuccessful; compare: P. Baumhof, R.
Mazitschek, A. Giannis, Angew. Chem. 2001, 113, 3784 – 3786;
Angew. Chem. Int. Ed. 2001, 40, 3672 – 3674.
[1] a) J. Kobayashi, J. Cheng, T. Ohta, H. Nakamura, S. Nozoe, Y.
Hirata, Y. Ohizumi, T. Sasaki, J. Org. Chem. 1988, 53, 6147 –
6150; b) Y. Kikuchi, M. Ishibashi, T. Sasaki, J. Kobayashi,
Tetrahedron Lett. 1991, 32, 797 – 798.
[2] a) K. Nozawa, M. Tsuda, H. Ishiyama, T. Sasaki, T. Tsuruo, J.
Kobayashi, Bioorg. Med. Chem. 2006, 14, 1063 – 1067; b) M.
Tsuda, K. Nozawa, K. Shimbo, H. Ishiyama, E. Fukushi, J.
Kawabata, J. Kobayashi, Tetrahedron Lett. 2003, 44, 1395 – 1399.
[3] The activity data are available from the NCI homepage (http://
www.dtp.nci.nih.gov/docs/dtp_search.html).
[4] Although the NCI data show no particular selectivity,
a
preliminary report from the Walther Cancer Research Center
(http://www.nd.edu/ ꢂ science/documents/waltherbrochure.pdf)
claims spectacular effects against colon cancer and stunning
morphological changes upon injection of iejimalides into solid
tumors.
[5] For a related endeavor see: A. Fürstner, D. Kirk, M. D. B.
Fenster, C. Aïssa, D. De Souza, O. Müller, Proc. Natl. Acad. Sci.
USA 2005, 102, 8103 – 8108.
´
[6] A. Fürstner, C. Nevado, M. Tremblay, C. Chevrier, F. Teply, C.
[28] M. Hikota, Y. Sakurai, K. Horita, O. Yonemitsu, Tetrahedron
Lett. 1990, 31, 6367 – 6370.
Aïssa, M. Waser, Angew. Chem. 2006, 118, 5969 – 5974; Angew.
Chem. Int. Ed. 2006, 45, 5837 – 5842.
[29] Attempts to suppress this pathway by replacement of DMAP
[7] For fragment syntheses see: a) M. Cottard, N. Kann, T. Rein, B.
Škermark, P. Helquist, Tetrahedron Lett. 1995, 36, 3115 – 3118;
b) M. T. Mendlik, M. Cottard, T. Rein, P. Helquist, Tetrahedron
with 2,6-di-tert-butylpyridine as
a more bulky base were
unsuccessful.
5836
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837

Angewandte
Chemie
[30] Formation of phenols by electrocyclizations of polyunsaturated
ketenes: a) M. J. Heileman, H. W. Moore, Tetrahedron Lett.
1998, 39, 3643 – 3646; b) W. F. Austin, Y. Zhang, R. L. Danheiser,
Org. Lett. 2005, 7, 3905 – 3908.
[31] a) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18 – 29;
b) A. Fürstner, Angew. Chem. 2000, 112, 3140 – 3172; Angew.
Chem. Int. Ed. 2000, 39, 3012 – 3043.
Angew. Chem. Int. Ed. 2006, 45, 5832 –5837ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5837