A practical synthesis of (+)-discodermolide and analogues: Fragment union by complex aldol reactions

Article abstract of DOI:10.1021/ja011211m

A practical stereocontrolled synthesis of (+)-discodermolide (1) has been completed in 10.3% overall yield (23 steps longest linear sequence). The absolute stereochemistry of the C₁-C₆ (7), C₉-C₁₆ (8), and C

Full text of DOI:10.1021/ja011211m

J. Am. Chem. Soc. 2001, 123, 9535-9544
9535
A Practical Synthesis of (+)-Discodermolide and Analogues:
Fragment Union by Complex Aldol Reactions
Ian Paterson,* Gordon J. Florence, Kai Gerlach, Jeremy P. Scott, and Natascha Sereinig
Contribution from the UniVersity Chemical Laboratory, Cambridge UniVersity, Lensfield Road,
Cambridge CB2 1EW, UK
ReceiVed May 16, 2001
Abstract: A practical stereocontrolled synthesis of (+)-discodermolide (1) has been completed in 10.3% overall
yield (23 steps longest linear sequence). The absolute stereochemistry of the C1-C6 (7), C9-C16 (8), and
C17-C24 (9) subunits was established via substrate-controlled, boron-mediated, aldol reactions of the chiral
ethyl ketones 10, 11, and 12. Key fragment coupling reactions were a lithium-mediated, anti-selective, aldol
reaction of aryl ester 8 (under Felkin-Anh induction from the aldehyde component 9), followed by in situ
reduction to produce the 1,3-diol 40, and a (+)-diisopinocampheylboron chloride-mediated aldol reaction of
methyl ketone 7 (overturning the inherent substrate induction from the aldehyde component 52) to give the
(7S)-adduct 58. The flexibility of our overall strategy is illustrated by the synthesis of a number of diastereomers
and structural analogues of discodermolide, which should serve as valuable probes for structure-activity studies.
Introduction
Discodermolide (1) is a unique polyketide isolated by
Gunasekera and co-workers at the Harbor Branch Oceanographic
Institute in 1990 from the Caribbean deep sea sponge Disco-
dermia dissoluta.1,2 Its gross structure was determined by
extensive spectroscopic studies and the relative stereochemistry
was assigned by single-crystal X-ray crystallography. Structur-
ally, it bears 13 stereogenic centers, a tetrasubstituted δ-lactone
(C1-C5), one di- and one trisubstituted (Z)-double bond, a
pendant carbamate moiety (C19), and a terminal (Z)-diene (C21-
C24). In the solid state, discodermolide adopts the U-shaped
conformation shown in Figure 1, where the two internal (Z)-
olefins in the side chain act as conformational locks by
minimizing A(1,3) strain between their respective substituents
in concert with the avoidance of syn-pentane interactions, while
the tetrasubstituted δ-lactone prefers a boatlike conformation.
The 24-membered polyketide carbon skeleton of discodermolide
appears to be made up of eight propionate and four acetate units.
It is likely that this dodecaketide is biosynthesized by a
symbiotic microorganism associated with Discodermia dissoluta
species, involving a modular polyketide synthase.
Discodermolide was initially found to be a potent immuno-
suppressive agent, both in vivo and in vitro, as well as displaying
antifungal activity.3 It inhibited T-cell proliferation with an IC50
of 9 nM and graft versus host disease in transplanted mice.
Further biological screening of this compound revealed startling
cytotoxicity, causing cell cycle arrest in the G2/M phase in a
variety of human and murine cell lines.4 Discodermolide joins
the group of antimitotic agents shown in Figure 2 now known
Figure 1.
to act by microtubule stabilization whose number include Taxol
(2),5a epothilones A and B (3, 4),5b eleutherobin (5),5c and most
recently, laulimalide (6).5d Discodermolide has been recognized
(4) ter Haar, E.; Kowalski, R. J.; Hamel, E.; Lin, C. M.; Longley, R. E.;
Gunasekera, S. P.; Rosenkranz, H. S.; Day, B. W. Biochemistry 1996, 35,
243.
(1) Gunasekera, S. P.; Gunasekera, M.; Longley, R. E.; Schulte, G. K.
J. Org. Chem. 1990, 55, 4912. Additions and corrections: J. Org. Chem.
1991, 56, 1346.
(2) Gunasekera, S. P.; Pomponi, S. A.; Longley, R. E. U.S. Patent No.
US5840750, Nov 24, 1998.
(3) (a) Longley, R. E.; Caddigan, D.; Harmody, D.; Gunasekera, M.;
Gunasekera, S. P. Transplantation 1991, 52, 650. (b) Longley, R. E.;
Caddigan, D.; Harmody, D.; Gunasekera, M.; Gunasekera, S. P. Trans-
plantation 1991, 52, 656. (c) Gunasekera, S. P.; Cranick, S.; Longley, R.
E. J. Nat. Prod. 1989, 52, 757.
(5) (a) Nicolaou, K. C.; Dai, W.-M.; Guy, R. K. Angew. Chem., Int. Ed.
Engl. 1994, 33, 15. (b) Bollag, D. M.; McQueney, P. A.; Zhu, J.; Hensens,
O.; Koupal, L.; Liesch, J.; Goetz, M.; Lazarides, E.; Woods, C. M. Cancer
Res. 1995, 55, 2325. (c) Long, B. H.; Carboni, J. M.; Wasserman, A. J.;
Cornell, L. A.; Casazza, A. M.; Jensen, P. R.; Lindel, T.; Fenical, W.;
Fairchild, C. R. Cancer Res. 1998, 58, 1111. (d) Mooberry, S. L.; Tien,
G.; Hernandez, A. H.; Plubrukarn, A.; Davidson, B. S. Cancer Res. 1999,
59, 653.
10.1021/ja011211m CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/05/2001

9536 J. Am. Chem. Soc., Vol. 123, No. 39, 2001
Paterson et al.
development, though, is severely hampered by the extremely
scarce supply of discodermolide (0.002% w/w frozen sponge)
from the natural source (a rare, deep-sea sponge only found in
the Caribbean that requires the use of manned submersibles for
collection). Thus, total synthesis presently provides the only
viable route to useful quantities of this novel cytotoxic
polyketide. Consequently, there has been considerable synthetic
effort toward discodermolide, culminating in several total
syntheses10 and numerous fragment syntheses.11 Indeed, the
absolute configuration of discodermolide was established by
Schreiber and co-workers by their initial syntheses of both (+)-
and (-)-discodermolide.10a,b Herein, we report full details of
the development of a novel, aldol-based, total synthesis12a of
(+)-discodermolide to provide useful quantities of this important
marine natural product, with improved fragment syntheses and
couplings, and some novel structural analogues.12b Notably, our
optimized synthesis involves a highly effective fragment coupl-
ing strategy, distinct from all previously reported approaches,
which leads to a 10.3% overall yield (over 23 linear steps) and
offers the potential for producing substantial quantities of (+)-
discodermolide, thus helping to relieve the supply problem and
enabling its further development in cancer chemotherapy.
Results and Discussion
Synthesis Plan. At the outset, a revised synthetic strategy
was designed to overcome the problems associated with our
earlier route,11a,h which were encountered in fragment coupling
and installation of the synthetically challenging trisubstituted
(Z)-alkene. Specific objectives were to design a practical
synthesis employing a high level of convergency, involving
efficient and scaleable chemistry, that both has the potential to
deliver multigram quantities of discodermolide and is amenable
to analogue synthesis. Our resulting retrosynthesis of discoder-
molide (Scheme 1) is based on two key aldol-type disconnec-
tions, across C6-C7 and C16-C17, leading back to the C1-C6
methyl ketone 7, the C9-C16 aryl ester 8,13 and the C17-C24
diene aldehyde 9. These three fragments are of similar stereo-
chemical and functional group complexity. We viewed these
subunits as being readily accessible by boron-mediated anti-
Figure 2.
as one of the most potent tubulin polymerizing agents presently
known. Despite having no apparent structural similarities,
discodermolide has been found to stabilize microtubules more
potently than Taxol (2, paclitaxel) and competitively inhibit its
binding to tubulin polymers.4,6 The growth of Taxol-resistant
ovarian and colon cancer cells is inhibited by discodermolide
with an IC50 of <2.5 nM,7 while the timing and type of DNA
fragmentation induced is consistent with the induction of
apoptosis.8
In recent comparative studies of discodermolide, the epothilones
and eleutherobin against a Taxol-dependent human lung carci-
noma cell line (A549-T12),9 it was found that discodermolide
was unable to act as a substitute for Taxol, whereas the
epothilones and eleutherobin were able to maintain the viability
of the cell line. Significantly, the presence of low concentrations
of Taxol amplified the cytotoxicity of discodermolide 20-fold
against this cell line. However, this synergistic effect in vitro
was not observed with combinations of the epothilones or
eleutherobin with Taxol.
(10) (a) Hung, D. T.; Nerenberg, J. B.; Schreiber, S. L. J. Am. Chem.
Soc. 1996, 118, 11054. (b) Nerenberg, J. B.; Hung, D. T.; Somers, P. K.;
Schreiber, S. L. J. Am. Chem. Soc. 1993, 115, 12621. (c) Smith, A. B., III;
Beauchamp, T. J.; LaMarche, M. J.; Kaufman, M. D.; Qiu, Y. P.; Arimoto,
H.; Jones, D. R.; Kobayashi, K. J. Am. Chem. Soc. 2000, 122, 8654. (d)
Smith, A. B., III; Kaufman, M. D.; Beauchamp, T. J.; LaMarche, M. J.;
Arimoto, H. Org. Lett. 1999, 1, 1823. Additions and corrections, Org. Lett.
2000, 2, 1983. (e) Smith, A. B., III; Qiu, Y. P.; Jones, D. R.; Kobayashi,
K. J. Am. Chem. Soc. 1995, 117, 12011. (f) Harried, S. S.; Yang, G.; Strawn,
M. A.; Myles, D. C. J. Org. Chem. 1997, 62, 6098. (g) Marshall, J. A.;
Johns, B. A. J. Org. Chem. 1998, 63, 7885. (h) Halstead, D. P. Ph.D. Thesis,
Harvard University, Cambridge, 1998.
(11) (a) Paterson, I.; Wren, S. P. J. Chem. Soc., Chem. Commun. 1993,
1790. (b) Clark, D. L.; Heathcock, C. H. J. Org. Chem. 1993, 58, 5878. (c)
Golec, J. M. C.; Jones, S. D. Tetrahedron Lett. 1993, 34, 8159. (d) Evans,
P. L.; Golec, J. M. C.; Gillespie, R. J. Tetrahedron Lett. 1993, 34, 8163.
(e) Golec, J. M. C.; Gillespie, R. J. Tetrahedron Lett. 1993, 34, 8168. (f)
Yang, G.; Myles, D. C. Tetrahedron Lett. 1994, 35, 2503. (g) Yang, G.;
Myles, D. C. Tetrahedron Lett. 1994, 35, 1313. (h) Paterson, I.; Schlapbach,
A. Synlett 1995, 498. (i) Miyazawa, M.; Oonuma, S.; Maruyama, K.;
Miyashita, M. Chem. Lett. 1997, 1191. (j) Miyazawa, M.; Oonuma, S.;
Maruyama, K.; Miyashita, M. Chem. Lett. 1997, 1193. (k) Marshall, J. A.;
Lu, Z. H.; Johns, B. A. J. Org. Chem. 1998, 63, 817. (l) Misske, A. M.;
Hoffmann, H. M. R. Tetrahedron 1999, 55, 4315. (m) Evans, D. A.;
Halstead, D. P.; Allison, B. D. Tetrahedron Lett. 1999, 40, 4461. (n) Filla,
S. A.; Song, J. J.; Chen, L. R.; Masamune, S. Tetrahedron Lett. 1999, 40,
5449.
The highly encouraging biological profile of discodermolide
(1) makes it a promising candidate for clinical development as
a chemotherapeutic agent for Taxol-resistant breast, ovarian,
and colon cancer and other multi-drug-resistant cancers. Clinical
(6) Hung, D. T.; Chen, J.; Schreiber, S. L. Chem. Biol. 1996, 3, 287.
(7) Kowalski, R. J.; Giannakakou, P.; Gunasekera, S. P.; Longley, R.
E.; Day, B. W.; Hamel, E. Mol. Pharm. 1997, 52, 613.
(8) Balachandran, R.; ter Haar, E.; Welsh, M. J.; Grant, S. G.; Day, B.
W. Anti-Cancer Drugs 1998, 9, 67.
(9) Martello, L. A.; McDaid, H. M.; Regl, D. L.; Yang, C. H.; Meng,
D.; Pettus, T. R. R.; Kaufman, M. D.; Arimoto, H.; Danishefsky, S. J.;
Smith, A. B., III; Horwitz, S. B. Clin. Cancer Res. 2000, 6, 1978.
(12) (a) Paterson, I.; Florence, G. J.; Gerlach, K.; Scott, J. P. Angew.
Chem., Int. Ed. 2000, 39, 377. (b) Paterson, I.; Florence, G. J. Tetrahedron
Lett. 2000, 41, 6935.

Synthesis of (+)-Discodermolide and Analogues
Scheme 1. Retrosynthetic Analysis
J. Am. Chem. Soc., Vol. 123, No. 39, 2001 9537
Scheme 2a
aldol reactions of the chiral ethyl ketones (S)-10,14 (S)-11,15 and
(S)-12,16 which should serve to construct the requisite stereo-
chemical motifs in a rapid and efficient manner.17
Synthesis of the C1-C6 Subunit, 7. The R-chiral ethyl
ketone (S)-10 required for the sequence was readily synthesized
in three steps (71% overall yield) from commercially available
methyl (S)-2-methyl-3-hydroxypropionate, (S)-13.14a Enolization
of (S)-10 under standard conditions, (c-Hex)2BCl/Et3N, led to
the selective generation of the (E)-boron enolate 14 (Scheme
2), which underwent addition to acetaldehyde through a highly
ordered chairlike transition state (TS-1).14 This was followed
by in situ reduction of the intermediate boron aldolate 15 with
LiBH4, via axial hydride delivery (TS-2), to provide, after
oxidative workup, the 1,3-syn diol 16 (86%, >97% ds).18,19
Thus, the initial substrate-controlled aldol addition is extended
to allow the one-pot assembly of the specific stereotriad found
in the C1-C6 unit. Treatment of the 1,3-syn diol 16 with
TBSOTf and 2,6-lutidine in CH2Cl2 provided the bis-TBS ether
17 in 99% yield. Selective removal of the less sterically
encumbered TBS ether at C5 was then achieved by using
a Key: (a) BnOC(dNH)CCl3, TfOHcat., Et2O, 20 °C; (b) iPrMgCl,
MeNH(OMe)‚HCl, THF, -30 f -10 °C; (c) EtMgBr, THF, 0 °C; (d)
(i) c-Hex2BCl, Et3N, Et2O, 0 °C; MeCHO, -78 f -20 °C; (ii) LiBH4,
-78 °C; (iii) H2O2(30% aq)/MeOH, NaOH(10% aq), 0 °C; (e) TBSOTf,
2,6-lutidine, CH2Cl2, -78 °C; (f) CSA, MeOH/CH2Cl2, 0 °C; (g) 20%
Pd(OH)2/C, H2, EtOH, 20 °C; (h) (i) (COCl)2, DMSO, CH2Cl2, -78
f -50 °C; (ii) Et3N, -50 f -20 °C; (i) (i) NaClO2, NaH2PO4,
2-methyl-2-butene, t-BuOH, H2O, 20 °C; (ii) CH2N2, Et2O, 20 °C.
catalytic CSA in 1:2 MeOH/CH2Cl2 at 0 °C, to provide alcohol
18 in 82% yield, along with diol 16 and recovered bis-TBS
ether 17, which could be recycled accordingly. Subsequent
hydrogenolysis of the benzyl ether 18 provided diol 19 in
quantitative yield. The diol was cleanly converted into the
required C1-C6 fragment 7 in a further three steps. In general,
this sequence was carried out without isolation of the intermedi-
ates, affording the C1-C6 fragment 7 in excellent yield (93%).
Double oxidation of 19 to the corresponding keto aldehyde was
possible by using a modified Swern protocol.20 Further oxida-
tion21 with sodium chlorite provided the intermediate carboxylic
acid, which was esterified directly with diazomethane to give
7. This sequence was readily performed on a multigram scale
to provide the C1-C6 fragment 7 in 46% overall yield from the
starting ester (S)-13.
(13) (a) Montgomery, S. H.; Pirrung, M. C.; Heathcock, C. H. Organic
Syntheses; Wiley: New York, 1990; Collect. Vol. VII, p 190. (b) Pirrung,
M. C.; Heathcock, C. H. J. Org. Chem. 1980, 45, 1727. (c) Heathcock, C.
H. Aldrichim. Acta 1990, 23, 99. (d) Paterson, I. Tetrahedron Lett. 1983,
24, 1311.
(14) (a) Paterson, I.; Norcross, R. D.; Ward, R. A.; Romea, P.; Lister,
M. A. J. Am. Chem. Soc. 1994, 116, 11287. (b) Paterson, I.; Lister, M. A.
Tetrahedron Lett. 1988, 29, 585. (c) Paterson, I.; Channon, J. A. Tetrahedron
Lett. 1992, 33, 797. (d) Paterson, I.; Tillyer, R. D. J. Org. Chem. 1993, 58,
4182. (e) Paterson, I.; Goodman, J. M.; Isaka, M. Tetrahedron Lett. 1989,
30, 7121. (f) Paterson, I. Pure Appl. Chem. 1992, 64, 1821.
(15) Paterson, I.; Arnott, E. A. Tetrahedron Lett. 1998, 39, 7185.
(16) (a) Paterson, I.; Wallace, D. J.; Cowden, C. J. Synthesis 1998, 639.
(b) Paterson, I.; Wallace, D. J.; Vela´zquez, S. M. Tetrahedron Lett. 1994,
35, 9083. (c) Paterson, I.; Wallace, D. J. Tetrahedron Lett. 1994, 35, 9087.
(17) For a review of asymmetric boron-mediated aldol reactions, see:
Cowden, C. J.; Paterson, I. Org. React. 1997, 51, 1.
Synthesis of the C9-C16 Subunit, 8. One of the most
synthetically challenging aspects of discodermolide is the
efficient introduction of the C13-C14 trisubstituted (Z)-olefin.
Previous syntheses have adopted conventional methods for its
(18) (a) Paterson, I.; Perkins, M. V. Tetrahedron 1996, 52, 1811. (b)
Paterson, I.; Perkins, M. V. Tetrahedron Lett. 1992, 33, 801.
(19) The relative stereochemistry of 1,3-syn diol 16 and 1,3-anti diol 22
was established by 13C NMR analysis of the 1,3-diol acetonides which were
consistent with the analyses reported: (a) Rychnovsky, S. D.; Skalitzky,
D. J. Tetrahedron Lett. 1990, 31, 945. (b) Rychnovsky, S. D.; Rogers, B.;
Yang, G. J. Org. Chem. 1993, 58, 3511. (c) Rychnovsky, S. D.; Rogers,
B.; Richardson, T. I. Acc. Chem. Res. 1998, 31, 9. (d) Evans, D. A.; Reiger,
D. L.; Gage, J. R. Tetrahedron Lett. 1990, 31, 7099.
(20) (a) Omura, K.; Swern, D. Tetrahedron Lett. 1978, 34, 1651. (b)
Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem. 1978, 43, 2231. (c)
Oppolzer, W.; De Brabander, J.; Walther, E.; Bernardinelli, G. Tetrahedron
Lett. 1995, 36, 4413.
(21) (a) Mann, J.; Thomas, A. Tetrahedron Lett. 1986, 27, 3533. (b)
Bal, B. S.; Childers, W. E. J.; Pinnick, H. W. Tetrahedron 1981, 37, 2091.

9538 J. Am. Chem. Soc., Vol. 123, No. 39, 2001
Paterson et al.
Scheme 3a
Scheme 4a
a Key: (a) NaOMe, MeOH, 0 °C f 20 °C; (b) TBSOTf, 2,6-lutidine,
CH2Cl2, -78 °C; (c) KOH (1 M aq), MeOH, reflux; (d) 2,6-
dimethylphenol, DCC, 4-DMAP, CH2Cl2, 20 °C. Ar ) 2,6-dimeth-
ylphenyl.
(K2CO3, MeOH) then provided the 1,3-anti diol 22 (97%).
Alternatively, reduction of â-hydroxy ketone 20 was achieved
by using Me4NBH(OAc)3 in MeCN and AcOH to provide
directly the 1,3-anti diol 22 (94%, >97% ds). While the latter
method shortens the synthetic route, it was somewhat less
amenable to material throughput than the two-step procedure.
Following the Holmes protocol,22e the phenylselenoacetal 23
was accessed by acetal exchange of 2-phenylselenoacetaldehyde
diethylacetal with diol 22 under acidic conditions. This gave
the expected Claisen precursor 23 as an inconsequential 2:1
diastereomeric mixture at the acetal carbon (94%). Oxidation
of selenide 23 to the selenoxide was readily achieved with
NaIO4. The crude product was then directly subjected to the
conditions of Claisen rearrangement providing the desired eight-
membered lactone 24 in 82% yield, along with recovered
selenide 23 (18%), which could be recycled. The exclusive
formation of the C13-C14 (Z)-trisubstituted olefin can be
attributed to the preferred bicyclic-chair conformation adopted
by the ketene acetal 25, as shown in Scheme 3. These conditions
were amenable to the production of multigram quantities of the
eight-membered lactone 24.
The conversion of lactone 24 to the aryl ester 8 was now
required (Scheme 4). Initially, a four-step sequence was
employed, whereby methanolysis of 24 and TBS protection of
the resulting hydroxy ester provided 26 (84%, 2 steps), and
saponification and esterification with 2,6-dimethylphenol under
Steglich conditions25 afforded the key fragment 8 (78%, 2 steps).
Subsequently, a shorter three-step sequence was developed to
provide the key fragment 8 in 81% overall yield. This involved
direct opening of the lactone 24 to the intermediate hydroxy
acid and esterification25 with 2,6-dimethylphenol to provide the
hydroxy ester 27, followed by TBS protection to give 8. In
summary, the optimized multigram synthesis of the C9-C16
fragment 8 was completed in 10 steps with 43% overall yield
from the starting ester (S)-13.
a Key: (a) PMBOC(dNH)CCl3, TfOHcat., Et2O, 20 °C; (b) iPrMgCl,
MeNH(OMe)‚HCl, THF, -15 °C; (c) EtMgBr, THF, 0 °C; (d) (i)
c-Hex2BCl, Et3N, Et2O; H2CdC(Me)CHO, -78 f -20 °C; (ii)
H2O2(30% aq)/MeOH, pH 7 buffer, 0 °C; (e) SmI2, EtCHO, THF, -10
°C; (f) K2CO3, MeOH, 20 °C; (g) Me4NBH(OAc)3, MeCN/AcOH, -40
f -23 °C; (h) PhSeCH2CH(OEt)2, PhMe, PPTScat., reflux; (i) (i) NaIO4,
NaHCO3, MeOH/H2O, 20 °C; (ii) DBU, H2CdC(OMe)OTBS, xylenes,
reflux.
construction, e.g. Wittig olefinations, but these have often proved
unreliable, exhibiting variable yields and selectivities.10c,g Our
C9-C16 fragment 8 required a carbonyl functionality attached
to C16 to allow aldol-based fragment coupling. The elegant
Claisen-type ring expansion methodology, developed extensively
by Holmes and co-workers for the synthesis of medium ring
lactones, allows the installation of this functionality and the C13-
C14 trisubstituted olefin simultaneously.22
The synthesis of the C9-C16 aryl ester fragment 8 started
from the PMB protected ketone (S)-11 (Scheme 3).15 This was
accessed in three steps (77% overall yield, 0.2 mol scale) from
methyl (S)-3-hydroxy-2-methylpropionate, (S)-13, under similar
conditions to that described above for the benzyl protected
ketone (S)-10. Under our standard conditions (c-Hex)2BCl/
Et3N,14 (E)-enolization of the ketone (S)-11, and reaction with
methacrolein gave, after oxidative workup, the expected anti-
aldol product 20 (95%, >97% ds).23 Substrate-controlled 1,3-
anti reduction of the â-hydroxy ketone was achieved under the
Evans-Tischenko conditions,24 with SmI2 and propionaldehyde,
to provide 1,3-diol monoester 21 (96%, >97% ds). Methanolysis
Synthesis of the C17-C24 Subunit, 9. The synthesis of the
C17-C24 fragment 9 required the union of two chiral coupling
partners to configure the stereotriad (Scheme 5). The ethyl
ketone (S)-12 was prepared in three steps from commercial ethyl
(S)-lactate in 65% yield, as previously described.16a Enolization
of the ketone (S)-12, using (c-Hex)2BCl/Me2NEt, generates the
(E)-boron enolate 28 exclusively. Addition of the R-chiral
aldehyde 29 (prepared from (S)-13) provides, after oxidative
(22) (a) Carling, R. W.; Holmes, A. B. J. Chem. Soc., Chem. Commun.
1986, 325. (b) Curtis, N. R.; Holmes, A. B.; Looney, M. G. Tetrahedron
1991, 47, 7171. (c) Congreve, M. S.; Holmes, A. B.; Looney, M. G. J. Am.
Chem. Soc. 1993, 115, 5815. (d) Fuhry, M. A. M.; Holmes, A. B.; Marshall,
D. R. J. Chem. Soc., Perkin Trans. 1 1993, 2743. (e) Burton, J. W.; Clark,
J. S.; Derrer, S.; Stork, T. C.; Bendall, J. G.; Holmes, A. B. J. Am. Chem.
Soc. 1997, 119, 7483. (f) Harrison, J.; Holmes, A. B. Synlett 1999, 972.
(23) The 1,2-anti relative stereochemistry of the aldol bond construction
in 20 was supported by the observed vicinal coupling constant (3J ) 8.5
Hz): Heathcock, C. H.; Pirrung, M. C.; Sohn, J. E. J. Org. Chem. 1979, 44,
4294.
(25) Nieses, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17,
(24) Evans, D. A.; Hoveyda, A. H. J. Am. Chem. Soc. 1990, 112, 6447.
522.

Synthesis of (+)-Discodermolide and Analogues
J. Am. Chem. Soc., Vol. 123, No. 39, 2001 9539
Scheme 5a
Scheme 6a
a Key: (a) (i) c-Hex2BCl, Me2NEt, Et2O, 0 °C; -78 f -20 °C; (ii)
H2O2(30% aq)/MeOH, 0 °C; (b) PMBOC(dNH)CCl3, TfOHcat., Et2O,
20 °C; (c) LiAlH4, THF -78 f -20 °C; (d) (i) NaBH4, MeOH, H2O,
0 f 20 °C; (ii) K2CO3, MeOH, 20 °C.
workup, the expected anti-aldol product 30 in excellent yield
and selectivity (99%, >97% ds).16,26 Optimum conditions
required a 1.5- to 2-fold excess of aldehyde 29 with respect to
the enolate. After considerable experimentation, protection of
the â-hydroxy group was achieved by using high-purity p-
methoxybenzyl-trichloroacetimidate and triflic acid (0.3 mol %)
to provide the ketone 31 in 93% yield.27 With the protected
ketone in hand, conversion to the 1,2-diol 32 was now required.
This was achieved by one of two methods: (i) reduction with
LiAlH4 (85%) or (ii) ketone reduction (NaBH4) followed by
benzoate hydrolysis (K2CO3, MeOH) (86%), where the latter
procedure proved somewhat more amenable to multigram
synthesis.16
Introduction of the terminal C21-C24 (Z)-diene unit of
discodermolide was achieved in a highly efficient manner
following our previously developed protocol.11h This required
first, the oxidative cleavage of the 1,2-diol 32 with NaIO4 to
afford the aldehyde 33 (Scheme 6). Nozaki-Hiyama reaction
between crude aldehyde 33 and allyl chromium reagent 34,
generated in situ from 1-bromo-1-trimethylsilyl-2-propene 35
and chromium(II) chloride in THF,28 provided (via TS-3) the
intermediate anti â-hydroxy silanes 36. These crude products
were then directly subjected to Peterson-type syn elimination,29
with KH in THF, to provide the desired (Z)-diene 37 exclusively
in excellent yield (98% from 32). Deprotection of the TBS ether
(CSA, MeOH/CH2Cl2) and Dess-Martin oxidation30 of the
intermediate alcohol completed the synthesis of the C17-C24
diene aldehyde 9 (84%, 2 steps). In summary, the optimized
synthesis of the key fragment 9 from ethyl (S)-lactate was
achieved in 10 steps with 42% overall yield, on a multigram
scale.
a Key: (a) NaIO4, MeOH, H2O, 20 °C; (b) (i) CrCl2, THF, 20 °C;
(ii) KH, THF, 0 °C; (c) CSAcat., MeOH/CH2Cl2, 20 °C; (d) Dess-
Martin periodinane, CH2Cl2, 20 °C.
strategy was totally different from that adopted by other
researchers, efficient aldol-based protocols needed to be devel-
oped in this demanding case. We first concentrated on realizing
the C16-C17 bond connection. The stereoselective, lithium-
mediated, aldol reaction of C9-C16 fragment 8 with the C17-
C24 aldehyde 9 was expected to afford the desired anti-aldol
adduct 38a with high levels of Felkin-Anh selectivity (Scheme
7).13d,31 Selective enolization of the Heathcock-type aryl ester
8 was best achieved by employing LiTMP/LiBr32 at -100 °C
(external bath temperature) to provide the (E)-lithium enolate
39. Addition of aldehyde 9 then proceeded via TS-4 preferen-
tially, which provided on workup the desired aldol product 38a
in an optimized 81% yield with >97% ds.33 In practice, a 2-fold
excess of enolate was employed relative to aldehyde 9, as partial
R-epimerisation of the aldehyde component was periodically
observed. The low reaction temperature (-100 °C) proved a
(31) Initial studies into a stereoselective C16-C17 boron-mediated aldol
coupling of a tert-butyl thioester derivative 67 and a truncated C17-C21
aldehyde 68 led to the formation of the desired aldol product 69 (20-
30%) and the â-OPMB elimination product 70.
Fragment Union and Completion of the Total Synthesis
of (+)-Discodermolide. With the three key subunits in place,
attention was now focused on the stereocontrolled union of these
components to advance our synthesis. As our fragment coupling
(26) The 1,2-anti relative stereochemistry of the aldol bond construction
in 30 was supported by the observed vicinal coupling constant (3J ) 9.5
Hz) in accord with ref 23.
(27) (a) Iverson, T.; Bundle, D. R. J. Chem. Soc., Chem. Commun. 1981,
1240. (b) Nakajima, N.; Horita, K.; Abe, R.; Yonemitsu, O. Tetrahedron
Lett. 1988, 29, 4139.
(28) (a) Cintas, P. Synthesis 1992, 248. (b) Hodgson, D. M.; Wells, C.
Tetrahedron Lett. 1992, 33, 4761. (c) Andringa, H.; Heus Kloos, Y. A.;
Brandsma, L. J. Organomet. Chem 1987, 336, C41.
(29) For a review of Peterson-type eliminations, see: Ager, D. Org.
React. 1990, 38, 1.
Conditions: (a) (i) 67, (c-Hex)2BBr, Et3N, Et2O, 0 °C.; 68, -78 f -20
°C; (ii) H2O2 (30%)/MeOH.
(30) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155. (b)
Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(32) Hall, P. L.; Gilchrist, J. H.; Collum, D. B. J. Am. Chem. Soc. 1991,
113, 9571.

9540 J. Am. Chem. Soc., Vol. 123, No. 39, 2001
Paterson et al.
Scheme 7a
a Key: (a) 8, LiTMP, LiBr, THF, -100 °C; 9, -100 °C; (b) LiAlH4, THF, -30 °C; (c) 2,4,6-Me3(C6H2)SO2Cl, Et3N, CH2Cl2, 20 °C; (d) LiAlH4,
THF, -10 °C; (e) TBSOTf, Et3N, CH2Cl2, 20 °C; (f) DDQ, CH2Cl2/pH 7 buffer, 20 °C; Ar ) 2,4,6-trimethylphenyl.
key determinant, as significant levels of R-elimination of the
enolate, generating the 2,6-dimethylphenolate, were observed
at higher temperatures. However, no other aldol products were
observed, which suggested that elimination only occurred from
the enolate and equilibration was avoided. With the aldol adduct
38a in hand, LiAlH4 reduction of the aryl ester provided the
1,3-diol 40 in 88% yield. To further streamline the synthesis,
we subsequently developed an alternative in situ aldol/reduction
sequence. By now employing an excess of aldehyde 9 (2 equiv)
over ester 8, the lithium aldol coupling was performed in a
similar fashion but instead of working up as above to give 38a,
the intermediate lithium aldolate 38b was treated directly with
LiAlH4. This convenient one-pot procedure gave the 1,3-diol
40 in 62% isolated yield (from 8) with complete diastereose-
lectivity (cf. 71% overall yield for the two-pot process), along
with the alcohol corresponding to the precursor of fragment 9
which was easily recyclable.
With the 1,3-diol 40 in hand, controlled deoxygenation was
now required to introduce the C16 methyl group. The regiose-
lective derivatization of the primary hydroxyl in diol 40 was
achieved by using 2,6-mesitylenesulfonyl chloride and Et3N to
afford the mono-sulfonate 41 in excellent yield with complete
selectivity. The deoxygenation sequence was then completed
by hydride displacement of 41 with LiAlH4 to give alcohol 42
in 97% yield.34 The C17-OH was readily protected as its TBS
ether to provide 43 (99%). Subsequent deprotection of both
(33) The 1,2-anti relative stereochemistry of the aldol bond construction
in 38 was supported by the observed vicinal coupling constant (3J ) 8.5
Hz) in accord with ref 23. The absolute stereochemistry was not proven
directly but by correlation with the related aldol product 71, which was
readily transformed into the corresponding PMP-acetal 72 (Oikawa, Y.;
Yoskioka, T.; Yonemitsu, O. Tetrahedron Lett. 1982, 23, 889). Strong nOe
contacts were observed between H17, H19 and the acetal proton, consistent
with their axial orientations in the preferred chairlike conformation. The
small vicinal coupling constants between H17 and H18 (3J17-18 ) 1.1 Hz),
and H18 and H19 (3J18-19 ) 0.7 Hz) established the 1,3-syn relative
stereochemistry between the C17 and C19 oxygen functionalities and the
absolute stereochemistry at C17.
(34) Interestingly, in preliminary investigations into the deoxygenation
of 41 treatment with Super-Hydride led to the exclusive formation of an
oxetane 73.
Conditions: (a) Super-Hydride, Et3N, THF, -10 f 20 °C.

Synthesis of (+)-Discodermolide and Analogues
J. Am. Chem. Soc., Vol. 123, No. 39, 2001 9541
Scheme 8a
a Key: (a) 24, LiTMP, LiBr, THF, -100 °C; RCHO, -100 °C; (b) LiAlH4, THF, -30 °C; (c) 2,4,6-Me3(C6H2)SO2Cl, Et3N, CH2Cl2, 20 °C; (d)
LiAlH4, THF -10 °C; (e) TBSOTf, Et3N, CH2Cl2, 20 °C. Ar ) 2,4,6-trimethylphenyl.
Scheme 9a
PMB ethers at C9 and C19 with DDQ then proceeded unevent-
fully to give diol 44 (91%).
In efforts to further shorten our synthetic route, the viability
of a C16-C17 aldol coupling with the eight-membered lactone
24 was investigated (Scheme 8). First, the π-facial bias of the
lithium enolate 45 derived from lactone 24 was determined by
reaction with an achiral aldehyde. Enolization of lactone 24
using LiTMP/LiBr32 at -100 °C generated the lithium enolate
45 and addition of isobutyraldehyde gave the aldol product 46
in 60% yield with 80% ds. Under these same conditions, the
lithium-mediated aldol reaction of lactone 24 and the C17-C24
aldehyde 9 gave the two diastereomeric adducts 47 (55%) and
48 (39%). This result was not unexpected, as the enolate facial
bias was expected to oppose the Felkin-Anh influence of the
aldehyde 9 in this situation.35 To demonstrate the applicability
of this result in our synthesis, the aldol product 48, bearing the
correct C16-C17 stereochemistry, was converted to the known
intermediate 43. Following LiAlH4 reduction, the triol 49 was
converted to the sulfonate 50 in 73% yield. Deoxygenation with
LiAlH4 then provided the diol 51 (96%). Finally, bis-TBS
protection of 51 was achieved using TBSOTf and Et3N to
provide 43 in 82% yield. While representing a saving of two
steps over the one-pot aldol/reduction of aryl ester 8, this
sequence gives an unacceptable overall yield of 43 (17% in 5
steps from the lactone 24). Although this leads overall to the
realization of a 20-step synthesis of (+)-discodermolide, the
longer route was obviously preferable, providing 43 in 46% yield
over 7 steps from 24, and proved suitable for multigram
synthesis.
a Key: (a) TEMPO, PhI(OAc)2, CH2Cl2, 20 °C; (b) (CF3CH2O)2P(O)-
CH2CO2Me, 18-crown-6, K2CO3, PhMe, -20 f 0 °C; (c) (i)
Cl3CC(O)NCO, CH2Cl2, 20 °C; (ii) K2CO3, MeOH, 20 °C; (d) DIBAL,
CH2Cl2, -78 °C; (e) Dess-Martin periodinane, CH2Cl2, 20 °C.
At this stage, the two distinct stereochemical arrays found in
the C7-C24 section of discodermolide were complete and
elaboration of the C9-C24 fragment 44 was now required to
furnish the requisite C7-C24 aldehyde 52 (Scheme 9). The
introduction of the C8-C9 (Z)-olefin required the selective
oxidation of the C9 terminus and (Z)-selective olefination.
Employing catalytic TEMPO (0.2 equiv, iodobenzene diacetate,
CH2Cl2, 20 °C) as a sterically demanding oxidant,36 the diol 44
(35) Anderson, E. A.; Holmes, A. B.; Collins, I. Tetrahedron Lett. 2000,
41, 117.

9542 J. Am. Chem. Soc., Vol. 123, No. 39, 2001
Paterson et al.
Scheme 10a
a Key: (a) (i) 7, c-Hex2BCl, Et3N, Et2O, 0 °C; 52, -78 f -20 °C; (ii) H2O2(30% aq)/MeOH, 0 °C; (b) (i) 7, (+)-Ipc2BCl, Et3N, Et2O, 0 °C; 52,
-78 f -20 °C; (ii) H2O2(30% aq)/MeOH, 0 °C; (c) Me4NBH(OAc)3, MeCN/AcOH, -20 f 0 °C; (d) HF‚pyr, THF, 20 °C; (e) 3 N HCl, MeOH,
20 °C.
was cleanly converted to the intermediate aldehyde without any
detectable oxidation of the hindered C19 hydroxyl group. The
(Z)-olefin was then introduced selectively under the milder HWE
conditions (K2CO3, 18-crown-6) reported by Still and Gennari37
to provide the desired (Z)-enoate 53, exclusively, in 87% yield
over two steps. The C7-C24 aldehyde 52 was completed in a
further three steps. The C19 carbamate moiety was installed
following a modification of the Kocovsky protocol.10g,38 Reac-
tion of the hydroxy ester 53 with trichloroacetylisocyanate,
followed by methanolysis (K2CO3/MeOH), provided the car-
bamate ester 54 in 98% yield. Chemoselective reduction of 54
with DIBAL at -78 °C then furnished allylic alcohol 55 in
92% yield on a gram scale. Finally, the alcohol 55 was oxidized
cleanly to the C7-C24 aldehyde 52 in 96% yield, using Dess-
Martin periodinane,30 in preparation for the final, and most
challenging, C6-C7 aldol coupling.
In planning our synthesis of discodermolide, we envisaged
the late stage C6-C7 aldol coupling between the (Z)-enal 52,
containing the entire C7-C24 section of discodermolide, with a
suitable enolate derivative of 7 (Scheme 10). To avoid compet-
ing intramolecular Claisen condensation of the methyl ketone
with the ester in 7, it was essential to identify a metal enolate
with the appropriate reactivity/selectivity characteristics, such
as were anticipated to be provided by use of the corresponding
boron enolate. However, the stereochemical requirement for
setting up the (7S) carbinol center by si-face attack in this
complex aldol coupling situation proved particularly challenging.
While extensive studies of structurally related boron enolates
are available,17,39 the π-facial bias of such an unusual chiral
aldehyde substrate as 52 was an unknown quantity. Following
a detailed stereochemical investigation,12b,40 it was apparent that
the boron-mediated C6-C7 aldol coupling of ketone 7 and the
C7-C24 subunit 52 would be a mismatched situation. Enolization
of 7 with (c-Hex)2BCl/Et3N provided the boron enolate 56 (L
) c-Hex). Reaction of this preformed enolate 56 with (Z)-enal
52 gave, after oxidative work up, the (7R) adduct 57 (major) in
71% yield with high levels of remote 1,4-stereoinduction (88%
(39) (a) Paterson, I.; Goodman, J. M.; Lister, M. A.; Schumann, R. C.;
McClure, C. K.; Norcross, R. D. Tetrahedron 1990, 46, 4663. (b) Paterson,
I.; Cumming, J. G.; Smith, J. D.; Ward, R. A. Tetrahedron Lett. 1994, 35,
441. (c) Paterson, I.; Oballa, R. M.; Norcross, R. D. Tetrahedron Lett. 1996,
37, 8581. (d) Paterson, I.; Gibson, K. R.; Oballa, R. M. Tetrahedron Lett.
1996, 37, 8585. (e) Paterson, I.; Oballa, R. M. Tetrahedron Lett. 1997, 38,
8241. (f) Mulzer, J.; Berger, M. J. Am. Chem. Soc. 1999, 121, 8393.
(40) Initial boron-mediated aldol reactions of 7 with the model truncated
(Z)-enal 74 demonstrated the viability of controlling this C6-C7 coupling
step (ref 12b).
Conditions: (a) (i) 7, c-Hex2BCl, Et3N, Et2O, 0 °C; 74, -78 f -20 °C;
(ii) H2O2(30% aq)/MeOH, 0 °C; (b) (i) 7, (+)-Ipc2BCl, Et3N, Et2O, 0 °C;
74, -78 f -20 °C; (ii) H2O2(30% aq)/MeOH, 0 °C; (c) (i) 7, (-)-Ipc2-
BCl, Et3N, Et2O, 0 °C; 74, -78 f -20 °C, 19 h; (ii) H2O2(30% aq)/MeOH,
0 °C.
(36) De Mico, A.; Margarita, R.; Parlanti, L.; Vescovi, A.; Piancatelli,
G. J. Org. Chem. 1997, 62, 6974.
(37) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.
(38) Kocovsky, P. Tetrahedron Lett. 1986, 27, 5521.

Synthesis of (+)-Discodermolide and Analogues
J. Am. Chem. Soc., Vol. 123, No. 39, 2001 9543
ds) in the wrong sense for discodermolide. This undesired
outcome can be attributed to preferential re-face addition of
the boron enolate 56 (L ) c-Hex) to the γ-chiral aldehyde. The
π-facial selectivity can be rationalized by steric control,
considering the preferred s-trans conformation that the aldehyde
52 adopts, where A(1,3) strain is minimized (TS-5).41 Addition
of the boron enolate 56 (L ) c-Hex) will then occur favorably
from the less sterically congested re-face of the aldehyde.
Gratifyingly, when (+)-Ipc2BCl/Et3N39,42 was employed for
enolization, the dominating selectivity of the aldehyde 52 could
be overturned in favor of si-facial attack, leading to isolation
of the desired (7S) adduct 58 in 74% yield (84% ds). Notably,
this represents the first successful example of this chiral boron
reagent overturning the intrinsic substrate selectivity of a
complex aldol coupling between two chiral carbonyl compo-
nents.17 The desired (7S) aldol product 58 was readily separable
by chromatography from the minor (7R) epimer 57. To achieve
good overall conversion of the valuable aldehyde component,
it was best to employ a 10-fold excess of the ketone. The excess
ketone 7 could be recovered in 90-95% yield without any
degradation. In this way, optimum diastereoselection and
conversion could be realized in this highly challenging, mis-
matched aldol reaction.
Scheme 11. Total Synthesis of (+)-Discodermolide (1).
exploration of analogues, inter alia with permutations in the
δ-lactone section.
Following the successful union of 7 and 52, the C1-C24
carbon skeleton of (+)-discodermolide was now in place. The
completion of the synthesis required the substrate-directed
reduction of the C5 ketone and subsequent global deprotection
with concomitant δ-lactonization (Scheme 10). The 1,3-anti
reduction of the â-hydroxy ketone 58 was conveniently achieved
under the Evans-Saksena conditions,43 using Me4NBH(OAc)3
in MeCN and AcOH, to provide the 1,3-anti diol 59 in
quantitative yield, introducing the final stereocenter at C5 with
>97% diastereoselectivity. Final deprotection and δ-lactonisa-
tion was achieved by treatment of the protected C1-C24
precursor 59 with HF‚py in THF over 16 h, or 3 N HCl/MeOH
for 4 days, to provide (+)-discodermolide (1) in 85% and 81%
yield, respectively. Our synthetic (+)-discodermolide was
identical in all respects by 1H NMR (CDCl3), 13C NMR (CDCl3),
IR, and TLC to a natural sample. The specific rotation of our
material at the sodium D-line was measured as +13.0 (c 1.1,
MeOH), in fair agreement with that reported for authentic
material +7.2 (c 0.7, MeOH),1 while matching better that
reported by Schreiber: +14.0 (c 0.6, MeOH),10a,b along with
that of other research groups.10c,g Our synthetic discodermolide
was also shown to be equipotent to natural material in tubulin
binding assays conducted by Novartis Pharma AG.44a In
summary, a highly efficient and practical stereocontrolled
synthesis of (+)-discodermolide (1) has been completed in
10.3% overall yield over 23 steps (longest linear sequence via
fragment 8) from methyl (S)-3-hydroxy-2-methylpropionate, (S)-
13. As summarized in Scheme 11, this improved route involves
the formation of the three subunits 7, 8, and 9 and their efficient
union, exploiting stereocontrolled aldol reactions. To date, this
synthesis has been used to produce gram quantities of the
advanced C7-C24 intermediate 55, thus enabling the SAR
Application to the Synthesis of Discodermolide Analogues.
The foregoing approach to discodermolide presents a variety
of options for analogue chemistry. We have the ability to
selectively access both C7 epimers in the final C6-C7 aldol
coupling and this should then allow the synthesis of three further
epimeric discodermolides 60-62 by stereocontrolled reduction
of the C5 ketone and subsequent deprotection (Figure 3). These
three epimers may in turn provide valuable information about
the effect that altering the configuration at C5 and the internal
C7-hydroxyl, and hence the conformation of the C1-C6 segment,
has on the biological profile.44b
(41) Hoffmann, R. W. Chem. ReV. 1989, 89, 1841.
(42) Depending on the precise experimental conditions, the (+)-Ipc2-
BCl mediated aldol coupling between 7 and 52 leads to selectivities in the
range 3.5-6:1 in favor of 58. The Ipc2BCl reagents are available from
Aldrich under the name DIPCl.
(43) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc.
1988, 110, 3560.
(44) (a) We are grateful to Dr. Walter Fuhrer of Novartis Pharma AG
(Basel) for arranging these studies. (b) The biological results for these
analogues will be reported elsewhere.
Figure 3.
The synthesis of the 5,7-bis-epi-discodermolide (60) first
required the stereoselective 1,3-anti reduction of the (7R) aldol
product, 57 (Scheme 12).12b In an analogous fashion to the
approach used to complete discodermolide, the â-hydroxy
ketone 57 was reduced by using Me4NBH(OAc)3 to provide

9544 J. Am. Chem. Soc., Vol. 123, No. 39, 2001
Paterson et al.
Scheme 12a
a Key: (a) Me4NBH(OAc)3, MeCN/AcOH, -20 f 0 °C; (b) (i) c-Hex2BCl, Et3N, THF, -78 °C; LiBH3(OMe), -78 °C; (ii) H2O2(30% aq)/
MeOH, 0 °C; (c) 3 N HCl, MeOH, 20 °C.
Scheme 13a
Conclusions
The stereocontrolled synthesis of compounds 60-62 and 66
demonstrates the ready applicability of our synthetic route to
analogues. The use of the intrinsic 1,4-stereoinduction from the
aldehyde 52 should also allow the synthesis of structurally less
complex discodermolide analogues,12b where the full C7-C24
skeleton is retained while the C1-C6 unit can be varied, probing
the role of the C1-C5 δ-lactone moiety. Moreover, such
stereoinduction in other nucleophilic additions to γ-chiral (Z)-
enals may be found to be a useful process for achieving remote
stereocontrol in acyclic systems. Work is now underway to
explore this effect, along with further studies into the tubulin
binding and polymerization properties of (+)-discodermolide
(1) and related structural analogues. Finally, the practical route
to (+)-discodermolide described herein should be amenable to
the preparation of multigram quantities, enabling further bio-
logical and clinical studies in cancer chemotherapy.
a Key: (a) 3 N HCl/MeOH, 20 °C.
the 1,3-anti diol 63 with 75% ds.43 Chromatographic separation,
deprotection, and concomitant lactonisation of 63, employing
3 N HCl in MeOH, provided 5,7-bis-epi-discodermolide (60)
in 91% yield (63% yield, over 2 steps). To furnish 7-epi-
discodermolide (61) and 5-epi-discodermolide (62), stereocon-
trolled 1,3-syn reductions of the â-hydroxy-ketones 57 and 58
were performed utilizing a modified Narasaka-Prasad proto-
col.18,45,46 Treatment of the â-hydroxy ketones 57 and 58 with
(c-Hex)2BCl/Et3N reformed the corresponding boron aldolates
and subsequent reduction with LiBH3(OMe) gave, after oxida-
tive workup, the expected 1,3-syn diols 64 and 65 in 72% and
65% yield, respectively, both with >97% diastereoselectivity.
The conversion of diols 64 and 65 to the corresponding epimeric
discodermolides 61 and 62 was achieved by exposure to 3 N
HCl in MeOH, in 74% and 63% yield, respectively. To further
address the role of the C1-C5 δ-lactone moiety in the binding
of discodermolide to tubulin, the truncated C7-C24 analogue
66 was prepared from alcohol 55 in 74% yield, by deprotection
under acidic conditions (Scheme 13).
Acknowledgment. We thank the EPSRC (Grants GR/
M46686 and L41646 and studentships for G.J.F. and J.P.S.),
the DFG (Postdoctoral Fellowship for K.G.), the EC (Marie
Curie Postdoctoral Fellowship for N.S. and HPRN-CT-2000-
00018), Novartis Pharma AG, and Merck for support, as well
as Dr. S. P. Gunasekera (Harbor Branch Oceanographic
Institute) for kindly providing a sample of natural (+)-
discodermolide. We also thank Prof. A. B. Holmes (Cambridge)
for helpful discussions regarding the Claisen rearrangement
chemistry and Dr. P. A. Wallace for contributions to early
studies.
Supporting Information Available: Experimental details
and analytical data for all new compounds and comparison data
for natural and synthetic (+)-discodermolide (1) (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
(45) (a) Narasaka, K.; Pai, F.-C. Tetrahedron 1984, 40, 2233. (b) Chen,
K.-M.; Hardtmann, G. E.; Prasad, K.; Repic, O. Tetrahedron Lett. 1987,
28, 155.
(46) Paterson, I.; Donghi, M.; Gerlach, K. Angew. Chem., Int. Ed. 2000,
39, 3315.
JA011211M