Synthesis of (-)-dactylolide and 13-desmethylene-(-)-dactylolide and their effects on tubulin

Article abstract of DOI:10.1021/ol100665m

An efficient new synthesis has been elaborated for non-natural (-)-dactylolide ((-)-2) and its 13-desmethylene analogue 4, employing a HWE-based macrocyclization approach with β-keto-phosphonate/aldehyde 19 and the respective 13-desmethylene derivative as the key intermediates. Both (-)-2 and 4 as well as the corresponding C20 alcohols inhibit human cancer cell proliferation with IC₅₀ values in the sub-micromolar range and induce the polymerization of tubulin in vitro.

Full text of DOI:10.1021/ol100665m

ORGANIC
LETTERS
2010
Vol. 12, No. 10
2302-2305
Synthesis of (-)-Dactylolide and
13-Desmethylene-(-)-dactylolide and
Their Effects on Tubulin
Didier Zurwerra,† Ju¨rg Gertsch,‡ and Karl-Heinz Altmann*,†
Swiss Federal Institute of Technology (ETH) Zu¨rich, HCI H405, Wolfgang-Pauli-Str. 10,
CH-8093 Zu¨rich, Switzerland, and UniVersity of Bern, Bu¨hlstrasse 28,
CH-3012 Bern, Switzerland
karl-heinz.altmann@pharma.ethz.ch
Received March 20, 2010
ABSTRACT
An efficient new synthesis has been elaborated for non-natural (-)-dactylolide ((-)-2) and its 13-desmethylene analogue 4, employing a
HWE-based macrocyclization approach with ꢀ-keto-phosphonate/aldehyde 19 and the respective 13-desmethylene derivative as the key
intermediates. Both (-)-2 and 4 as well as the corresponding C20 alcohols inhibit human cancer cell proliferation with IC50 values in the
sub-micromolar range and induce the polymerization of tubulin in vitro.
(-)-Zampanolide ((-)-1) and (+)-dactylolide ((+)-2) are
structurally related polyketide-based macrolides, which are
characterized by a highly unsaturated 20-membered macro-
lactone core structure containing a syn-2,6-disubstituted
tetrahydropyran ring with an exocyclic methylene group
(Figure 1). Compound (-)-1 was first isolated in 1996 by
Tanaka and co-workers from the marine sponge Fasciospon-
gia rimosa at Cape Zampa in Okinawa;1 the compound was
shown to exhibit high antiproliferative activity against
different human cancer cell lines in vitro, with IC50’s in the
low nanomolar range (2-10 nM). Compound (-)-1 features
an unusual N-acylhemiaminal side chain that is not found
in (+)-2. The latter was isolated in 2001 by Cutignano and
co-workers from the sponge Dactylospongia sp. at the
Vanuatu Islands.2 In contrast to (-)-1, (+)-2 is only a
moderately potent inhibitor of human cancer cell growth with
Figure 1. Structures of natural (-)-zampanolide ((-)-1), non-
natural (-)-dactylolide ((-)-2), and analogue structures 3-5.
IC50’s in the low micromolar range.2 Rather intriguingly,
however, it was subsequently shown by Smith and co-
workers3 that the absolute configuration of the macrolactone
core structure in (-)-1 is opposite to that found in natural
† Swiss Federal Institute of Technology (ETH) Zu¨rich.
‡ University of Bern.
(2) Cutignano, A.; Bruno, I.; Bifulco, G.; Caspullo, A.; Debitus, C.;
(1) Tanaka, J.; Higa, T. Tetrahedron Lett. 1996, 37, 5535.
Gomez-Paloma, L.; Riccio, R. Eur. J. Org. Chem. 2001, 775.
10.1021/ol100665m  2010 American Chemical Society
Published on Web 04/23/2010

(+)-2. Interestingly, based on work by Ding and Jennings,
synthetic (-)-2 appears to be slightly more active than natural
(+)-2 (although a direct comparison is available only for a
single cell line).4 More recently, (-)-1 was also isolated from
the Togan sponge Cacospongia mycofijiensis by Northcote
and co-workers, who demonstrated the compound to be an
efficient promotor of tubulin assembly.5 This suggests that
(-)-1 inhibits cancer cell growth through the same mecha-
nism of action as Taxol or epothilones.6
lective epoxide opening with lithiated vinyl iodide 9. The
latter was to be obtained by Prins-type reaction of alkyne
10 to deliver a 4-iodotetrahydropyran derivative; the iodo
substituent would then be elaborated into the desired meth-
ylene group. Finally, 10 would be derived from epoxide 11
which, in turn, is accessible from (R)-aspartic acid.12 Acid
6 was envisaged to be accessed from a protected Z-vinyl
iodide via reaction with epichlorohydrin followed by conver-
sion of the resulting chlorohydrin to a new oxirane, epoxide
opening with lithiated diethylphosphite, and finally, HWE
reaction and ester hydrolysis.
While a number of stereoselective syntheses of (-)-17 and
(+)-28 have appeared in the literature, little work has been
reported on analogue structures and their biological activity.9
It also remains to be shown whether (-)-2 or (+)-2, like
(-)-1, may promote tubulin polymerization and stabilize
microtubules. Intrigued by the divergent stereochemistry of
natural (-)-1 and (+)-2 and in light of the distinct lack of
SAR data for these structures, we had initiated a program
on the synthesis of natural (-)-1 and (+)-2, their non-natural
counterparts (+)-1 and (-)-2, and analogue structures for
SAR studies, even before the recent discovery of the tubulin-
polymerizing activity of (-)-1. In this paper, we now report
on a new synthesis of (-)-210 and of dactylolide analogues
3-5 (Figure 1), all of which were found to have similar
antiproliferative activity and to induce tubulin polymerization
in vitro.
Scheme 2. Synthesis of Vinyl Iodide 9
Scheme 1. Retrosynthetic Analysis for (-)-2
The synthesis of vinyl iodide 9 started with the Cu-
mediated regioselective epoxide opening of 11 (available
from (R)-aspartic acid in three steps in 78% overall yield)12
with vinyl-MgBr in excellent yield (98%; Scheme 2).
Compound 12 was then elaborated into tetrahydropyran 13
(3) Smith, A. B., III; Safonov, I. G.; Corbett, R. M. J. Am. Chem. Soc.
2002, 124, 11102.
(4) Ding, F.; Jennings, M. P. J. Org. Chem. 2008, 73, 5965.
(5) Field, J. J.; Singh, A. J.; Kanakkanthara, A.; Halafihi, T.; Northcote,
P. T.; Miller, J. H. J. Med. Chem. 2009, 52, 7328.
(6) For a review on microtubule-stabilizing natural products, see:
Altmann, K.-H.; Gertsch, J. Nat. Prod. Rep. 2007, 24, 327.
(7) (a) Smith, A. B., III; Safonov, I. G.; Corbett, R. M. J. Am. Chem.
Soc. 2001, 123, 12426. (b) Smith, A. B., III; Safonov, I. G.; Corbett, R. M.
J. Am. Chem. Soc. 2002, 124, 11102. (c) Hoye, T. R.; Hu, M. J. Am. Chem.
Soc. 2003, 125, 9576. (d) Uenishi, J.; Iwamoto, T.; Tanaka, J. Org. Lett.
2009, 11, 3262. Studies toward (-)-1: (e) Loh, T.-P.; Yang, J.-Y.; Feng,
L.-C.; Zhou, Y. Tetrahedron Lett. 2002, 43, 7193. (f) Troast, D. M.; Porco,
J. A., Jr. Org. Lett. 2002, 4, 991. (g) Troast, D. M.; Yuan, J.; Porco, J. A.,
Jr. AdV. Synth. Catal. 2008, 350, 1701.
Our retrosynthesis for (-)-2 is outlined in Scheme 1 and
is centered around an intramolecular HWE reaction for the
closure of the 20-membered macrolide ring. The requisite
ꢀ-keto phosphonate/aldehyde percursor would be obtained
via esterification of acid 6 and alcohol 7, followed by silyl
ether cleavage and oxidation. While HWE-based macrocy-
clizations involving the formation of the CdC double bond
in R,ꢀ-unsaturated ketone units are well precedented in
natural product synthesis, they have not been used exten-
sively;11 in particular, and quite surprisingly, this strategy
has not been employed in the context of zampanolide or
dactylolide syntheses. Alcohol 7 was envisioned to be
accessible from protected (R)-glycidol 8 through regiose-
(8) (a) Smith, A. B., III; Safonov, I. G. Org. Lett. 2002, 4, 635. (b)
Aubele, D. L.; Wan, S. Y.; Floreancig, P. E. Angew. Chem., Int. Ed. 2005,
44, 3485. (c) Sanchez, C. C.; Keck, G. E. Org. Lett. 2005, 7, 3053.
(9) A notable exception is the work by Uenishi et al. (ref 7d), who have
shown the C20 epimer of (-)-1 to be ca. 10-fold less active than (-)-1.
(10) For previous syntheses of (-)-2, see: (a) Louis, I.; Hungerford,
N. L.; Humphries, E. J.; McLeod, M. D. Org. Lett. 2006, 8, 1117. (b)
Reference 4.
(11) For examples, see: (a) Nicolaou, K. C.; Seitz, S. P.; Pavia, M. R.
J. Am. Chem. Soc. 1982, 104, 2030. (b) Paterson, I.; Yeung, K.-S.
Tetrahedron Lett. 1993, 34, 5347. (c) Paterson, I.; Yeung, K.-S.; Watson,
C.; Ward, R. A.; Wallace, P. A. Tetrahedron 1998, 54, 11935. (d) Kadota,
I.; Hu, Y.; Packard, G. K.; Rychnovsky, S. D. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 11192. (e) Berger, G. O.; Tius, M. A. J. Org. Chem. 2007, 72,
6473.
(12) Frick, J. A.; Klassen, J. B.; Bathe, A.; Abramson, J. M.; Rapoport,
H. Synthesis 1992, 621.
Org. Lett., Vol. 12, No. 10, 2010
2303

in a highly stereoselective Prins-type reaction employing a
segment coupling approach as developed by Rychnovsky.13
In a first step, this involved esterification of 12 with
2-butynoic acid,14 which was followed by DIBALH reduc-
tion of the ester and trapping of the aluminated intermediate
at -78 °C with Ac2O to furnish 10 in excellent yield (78%
from 12, dr ca. 1.6:1). Treatment of 10 with TMSI gave
substituted tetrahydropyran 13 with the desired 2,6-syn
relationship and the iodine located anti to the substituents
in the 2- and 6-positions.15 This is in line with previous
observations by Rychnovsky for related transformations.16
The use of other Lewis acids gave either lower conversion
(TMSBr, BF3·Et2O, BF3·(HOAc)2, TFA) or led to partial
epimerization (SnBr4) at C6 (see numbering of 13 in Scheme
2). Attempts to conduct the Prins reaction directly with an
appropriate aldehyde and a homoallylic alcohol failed
completely. Iodide displacement from 13 with CsOAc/18-
c-6 was accompanied by elimination, with both possible
olefin products formed in significant amounts.17 Elimination
could be minimized, however, by conducting the reaction at
lower temperature (55 °C instead of 90 °C), although
extended reaction times (up to 3-4 days) were required
under these conditions for full conversion. Base-induced
cleavage of the acetate formed in the displacement step,
Swern oxidation of the alcohol to the ketone, and Wittig
reaction then served to install the exocyclic CdC double
bond in 14 in good overall yield (58%, four steps from 13).
Conversion of 13 to the desired E-vinyl iodide 9 was
Scheme 3. Synthesis of Unsaturated Acid 6
such as THF or Et2O or mixtures thereof. This rather
unexpected finding turned out to be crucial at a later stage
of the synthesis.
Treatment of 16 with base followed by BF3·OEt2-mediated
opening of the resulting epoxide with lithiated diethylphos-
phite gave ꢀ-hydroxy phosphonate 17. TBS protection of
17 followed by PMB removal under oxidative conditions
(DDQ) resulted in a mixture of aldehyde 18 and the
corresponding allylic alcohol, which was oxidized under
Swern conditions to deliver 18 as the sole product. HWE
reaction followed by basic hydrolysis then afforded the
desired acid 6.
18
achieved with in situ generated Bu3Sn(Bu)CuCNLi2 fol-
The final assembly of (-)-2 commenced with the reaction
of vinyl iodide 9 with PMB-protected (R)-glycidol 8 (Scheme
4). As for the reaction of 15 with epichlorohydrin, and in
lowed by Sn-I exchange with NIS. Alternative methods
investigated for the formation of the metalated vinyl inter-
mediate required for this transformation, such as Pd-mediated
hydrostannylation,18,19 silylcupration with Fleming’s re-
agent,20 or the use of the Schwartz reagent21 afforded the
desired product only in low yields and as mixtures of
regioisomers (E/Z ratios between 5:1 and 1:1).
Scheme 4. Final Steps in the Total Synthesis of (-)-2
The preparation of unsaturated acid 6 started with BF3·OEt2-
mediated coupling between lithiated Z-vinyl iodide 15
(obtained in two steps from 2-butynol in 79% yield)22 and
epichlorohydrin in toluene (Scheme 3).23 No reaction was
observed between epichlorohydrin and 15 in etheral solvents
(13) (a) Rychnovsky, S. D.; Hu, Y.; Ellsworth, B. Tetrahedron Lett.
1998, 39, 7271. (b) Rychnovsky, S. D.; Kopecky, D. J. J. Org. Chem. 2000,
65, 191. (c) Jaber, J. J.; Mitsui, K.; Rychnovsky, S. D. J. Org. Chem. 2001,
65, 4679. (d) Rychnovsky, S. D.; Marumoto, S.; Jaber, J. J. Org. Lett. 2001,
3, 3815. (e) Marumoto, S.; Jaber, J. J.; Vitale, J. P.; Rychnovsky, S. D.
Org. Lett. 2002, 4, 3919. (f) Jasti, R.; Vitale, J.; Rychnovsky, S. D. J. Am.
Chem. Soc. 2004, 126, 9904. (g) Vitale, J. P.; Wolckenhauer, S. A.; Nga,
M. D.; Rychnovsky, S. D. Org. Lett. 2005, 7, 3255. (h) Jasti, R.;
Rychnovsky, S. D. J. Am. Chem. Soc. 2006, 128, 13640.
(14) Trost, B. M.; Rudd, M. T. J. Am. Chem. Soc. 2005, 127, 4763.
(15) No NOEs were observed between C(4)H (see numbering of 13 in
Scheme 2) with the protons R to the tetrahydropyran oxygen.
(16) For a previous example of an axial-selective Prins reaction, see ref
13f.
(17) See also ref 13g. In our case, the use of AgOCOCF3 only led to
elimination of HI.
(18) Betzer, J.-F.; Delaloge, F.; Muller, B.; Pancrazi, A.; Prunet, J. J.
Org. Chem. 1997, 62, 7768.
spite of literature precedence for related transformation (see,
e.g., refs 7b and 7d), 7 was only formed when toluene was
used as the solvent. No traces of product were observed if
THF, Et2O, or mixtures thereof were used, independent of
the precursor for the lithiated vinyl nucleophile (e.g., vinyl
(19) Liron, F.; Knochel, P. Chem. Commun. 2004, 304
.
(20) Fleming, I.; Newton, T. W.; Roessler, F. J. Chem. Soc., Perkin
Trans. 1 1981, 2527.
(21) Schwartz, J.; Labinger, J. A. Angew. Chem., Int. Ed. 1976, 15, 333.
(22) Chau, A.; Paquin, J.-F.; Lautens, M. J. Org. Chem. 2006, 71, 1924.
2304
Org. Lett., Vol. 12, No. 10, 2010

iodide, vinyl bromide, vinyl tributylstannane); likewise, the
use of cuprates, variations in Lewis acid (SnCl4, MgBr2·OEt2,
TMSOTf), or the addition of HMPA or NMP did not result
in any improvement. Gratifyingly, the reaction between 8
and 9 under optimized conditions in toluene delivered alcohol
7 as a single isomer in yields of up to 61%. Esterification of
7 with acid 6 under Yamaguchi conditions (DCC or EDCI
werelesseffective)followedbyglobaldesilylation(HF·pyridine)
and a one-step oxidation of the resulting monoprotected triol
afforded the desired HWE substrate 19 in good yields. Initial
attempts at base-induced macrocyclization of 19 involved
the use of NaHMDS in THF as the solvent, and while 20
was indeed obtained under these conditions, extended
reaction times were required (2-4 d) and yields were highly
variable (20-80%). After significant experimentation it was
finally discovered that the use of activated Ba(OH)211b,c not
only substantially reduced the reaction time (to 0.5-1 h at
0 °C) but, importantly, also led to increased and reproducible
yields in the range of 80-85%. Under all conditions
investigated, the macrolactone with the desired C8-C9 E
configuration was the only detectable isomer.
Compound (-)-2, its 13-desmethylene derivative 4, as well
as alcohols 3 and 5 were assessed for their antiproliferative
and potential tubulin-polymerizing effects. All four com-
pounds inhibit human cancer cell growth with comparable
potency and IC50 values in the sub-micromolar range (Table
1). In addition, (-)-2, 3, 4, and 5 all show significant tubulin-
Table 1. Antiproliferative Activity of (-)-2, 3, 4, and 5 (IC50
values (nM)a after 72 h Exposure Time)25
cell line
(-)-2
4
3
5
A549
301.5 ( 4.3 149.0 ( 12.8 127.5 ( 2.9 189.0 ( 19.3
247.6 ( 2.6 68.0 ( 5.6 106.0 ( 3.6 114.4 ( 10.2
74.1 ( 1.5
MCF-7
HCT116 210.4 ( 4.7 249.5 ( 28.2 155.8 ( 2.1
a GI50 values of 198 and 346 nM have been reported for (-)-2 against
the MCF-7 and HCT116 cell lines, respectively.4 No data are available in
ref 4 for A549 cells.
polymerizing activity (41%-74% induction of tubulin po-
lymerization relative to the effect of 25 µM of epothilone B
(10 µM tubulin, 2 µM test compound) vs 82% for epothilone
A). Thus, (-)-2, like natural (-)-zampanolide, is a new
microtubule stabilizer, with neither the methylene group at
C13 nor the aldehyde functionality at C20 being essential
for antiproliferative and tubulin-polymerizing activity.
In conclusion, we have established an efficient new
synthesis of (-)-dactylolide ((-)-2). Utilizing an advanced
intermediate from this synthesis, we have then followed the
same HWE-based macrocyclization strategy to prepare 13-
desmethylene-(-)-dactylolide (4). Both compounds as well
as the corresponding alcohols 3 and 5 are potent inhibitors
of human cancer cell growth and promote tubulin polym-
erization in vitro. Efforts to exploit these findings in the
design and synthesis of new potent analogues are currently
ongoing in our laboratory.
After the successful implementation of the crucial ring-
closure step, the synthesis of (-)-2 was completed by PMB
removal, to furnish 3, and DMP oxidation of the primary
alcohol in 64% yield for the two steps from 20.
Based on the chemistry developed for the synthesis of (-)-
2, we were then able to efficiently address the construction
of 13-desmethylene-(-)-dactylolide (4), again building on
Prins product 13 as a key intermediate (Scheme 5). In spite
Scheme 5. Synthesis of 13-Desmethylene-(-)-dactylolide (4)
Acknowledgment. We thank the Roche Research Foun-
dation for a Ph.D. fellowship to D.Z. We are indebted to
Dr. Bernhard Pfeiffer and Fabienne Gaugaz (ETHZ) for
NMR support, to Kurt Hauenstein (ETHZ) for technical
advice, and to Louis Bertschi (ETHZ, LOC MS-Service) for
HRMS spectra acquisition.
Supporting Information Available: Synthetic procedures,
1
complete spectroscopic data, H and 13C NMR data for all
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
of the potential susceptibility of the alkyne moiety toward
radical hydrostannylation, 13 could be selectively dehalo-
genated with Bu3SnH (1.1 equiv) and AIBN. In analogy to
the synthesis of (-)-2, carbostannylation/iodination and
reaction with epoxide 8 delivered homoallylic alcohol 21.
Yamaguchi esterification of 21 with acid 6, global desily-
lation with HF·pyridine, oxidation with DMP, and HWE-
based macrocyclization then afforded macrolactone 22.24
PMB removal and oxidation completed the synthesis of target
structure 4, which is the first backbone-modified zampano-
lide/dactylolide analogue reported so far.
OL100665M
(23) For literature precedence for the coupling of a lithiated aromatic
compound with epichlorohydrin in toluene, see: Okano, K.; Tokuyama, H.;
Fukuyama, T. J. Am. Chem. Soc. 2006, 128, 7136.
(24) For 22, macrocyclization so far has only been performed with
NaHMDS as a base; the reaction was hampered by the same reproducibility
problems as the cyclization of 19. Due to a lack of material, the use of
Ba(OH)2 has not yet been investigated in this case.
(25) For experimental details, see: Dietrich, S. A.; Lindauer, R.; Stierlin,
C.; Gertsch, J.; Matesanz, R.; Notararigo, S.; D´ıaz, J. F.; Altmann, K.-H.
Chem.sEur. J. 2009, 15, 10144.
Org. Lett., Vol. 12, No. 10, 2010
2305