Rapid access to epothilone analogs via semisynthetic degradation and reconstruction of epothilone D

Article abstract of DOI:10.1016/j.tetlet.2003.12.123

A facile and efficient route to epothilone analogs has been developed from the natural product epothilone D (1). Degradation of 1 via an oxidative cleavage sequence provides acid intermediate 4 rapidly in six steps. From 4, a variety of epothilone analogs have been prepared utilizing ring-closing metathesis to reconstruct the trisubstituted-12,13-double bond. Using this approach, we report a number of epothilone analogs with varying C-15 aromatic side chains and C-14 allylic substitutions and their biological activities.

Full text of DOI:10.1016/j.tetlet.2003.12.123

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 1945–1947
Rapid access to epothilone analogs via semisynthetic degradation
and reconstruction of epothilone D
Steven D. Dong,* Kurt Sundermann, Karen M. J. Smith, Joseph Petryka,
Fenghua Liu and David C. Myles
Department of Chemistry, Kosan Biosciences, Hayward, CA 94545, USA
Received 24 November 2003; revised 19 December 2003; accepted 22 December 2003
Abstract—A facile and efficient route to epothilone analogs has been developed from the natural product epothilone D (1). De-
gradation of 1 via an oxidative cleavage sequence provides acid intermediate 4 rapidly in six steps. From 4, a variety of epothilone
analogs have been prepared utilizing ring-closing metathesis to reconstruct the trisubstituted-12,13-double bond. Using this ap-
proach, we report a number of epothilone analogs with varying C-15 aromatic side chains and C-14 allylic substitutions and their
biological activities.
Ó 2004 Elsevier Ltd. All rights reserved.
The epothilone class of microtubule stabilizers has
attracted considerable attention as novel chemothera-
peutic agents.1 Among the most promising is epothilone
D (1), which displays superior anti-tumor activity
relative to paclitaxel and is currently undergoing
human clinical trials.2 A variety of total syntheses of 1
have been reported.3;4 Though these approaches are
amenable to analog production, they require a sub-
stantial synthetic investment to generate key intermedi-
ates.
macrolactone (Scheme 1). We now report a rapid and
efficient degradation approach to intermediates such as
2 and the elaboration of 2 into analogs of 1.
Our degradation strategy focused on the cleavage of the
12,13-alkene of 1 followed by base promoted b-elimi-
nation of the resulting carboalkoxyaldehyde to provide
the C1-12 acid 2. To this end, epothilone D (1) was first
protected as the bis-TBS ether (Scheme 2). Stoichio-
metric osmium mediated dihydroxylation provided the
12,13-diol in 75%.6 Oxidative cleavage of the vicinal diol
with Pb(OAc)4 provided the keto aldehyde, which could
be isolated if desired. We found that when this crude
aldehyde was treated directly with K2CO3, the elimina-
tion of the aromatic side chain proceeded smoothly to
provide acid 2. Thus, over only three steps, epothilone D
1 was advanced to intermediate 2.
Julien and Shah have developed a heterologous bacterial
expression system containing the epothilone D poly-
ketide synthase genes to produce 1 by fermentation.5
With substantial quantities of 1 available to us by this
process, we sought a means by which to generate ana-
logs from 1 via degradation and reconstruction of the
Scheme 1.
Keywords: Epothilone; Epothilone analogs; Degradation; Semisynthesis; Ring closing metathesis.
* Corresponding author. Tel.: +1-510-731-5282; fax: +1-510-732-8401; e-mail: dong@kosan.com
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.12.123

1946
S. D. Dong et al. / Tetrahedron Letters 45 (2004) 1945–1947
Scheme 2. Conditions and reagents: (a) (1) TBSOTf, NEt2, CH2Cl2, )78 °C, 85%; (2) OsO4, TMEDA, )78 °C then NaHSO3, 65 °C, 75%; (3)
Pb(OAc)4, PhH then K2CO3, MeOH; (4) TMSCHN2, MeOH, PhMe, 81% over two steps; (b) (1) Cp2TiMe2, PhMe, 80 °C, 74%; (2) LiOH, i-PrOH,
H2O, 86%; (c) A, B, or C, EDC, DMAP, CH2Cl2, 54–82%; (d) (1) see Table 1; (2) TFA, CH2Cl2, 0 °C to rt; or HF–pyridine, THF, 0 °C to rt, 32–59%.
Olefination of the 12-keto group of 2 would provide acid
4, an ideal intermediate for the reconstitution of the
macrolactone with a variety of alkenyl alcohols (Scheme
2). To facilitate purification and the subsequent olefin-
ation, the acid was esterified with TMSCHN2 to provide
methyl ester 3. The ketone of 3 was converted to the
terminal alkene with dimethyltitanocene in 74% yield.7
Hydrolysis of the methyl ester provided acid 4 in 86%
yield. Acid 4 is available in six steps and 52% overall
yield from epothilone D (1), demonstrating the efficiency
of this route.
epothilone as in 14-methyl-epothilone D has been shown
to promote the population of the active conformation as
well as improving esterase stability.12 To this end,
alkenyl alcohols A,13 B,14 and C12 were prepared as
previously described. These alcohols were coupled to
acid 4 under the agency of EDC and catalytic DMAP in
54–82% yield to provide esters 5A–C (Scheme 2).
Metathesis conditions and reaction outcomes for sub-
strates 5A–C are shown in Table 1. Ring closing
metathesis of ester 5A with Grubbs second-generation
ruthenium catalyst (6) provided a 1:1 mixture of E:Z
isomers (X,Y) of 1 in 59% overall yield. Similar results
were observed with the benzoxazole ester 5B demon-
strating the permissiveness of 6 to differing side chains.
However, in the presence of the additional steric bulk of
the allylic methyl of 5C, no ring closing was observed in
refluxing CH2Cl2 with 20 mol % 6. Only under more
forcing conditions did we observe ring closing to yield 8
as a 1:1.3 mixture of E:Z isomers in 35% overall yield.
The resulting ring-closed analogs were deprotected with
either trifluoroacetic acid or HFÆpyridine to provide the
final products in 9–18% overall yield from intermediate
4.15
A variety of reported total syntheses has relied on ring
closing metathesis to establish the trisubstituted-12,13-
double bond of epothilone using the Schrock molybde-
num catalyst.8;9 We decided to investigate the versatility
of this approach to making epothilone analogs further
by employing degradation intermediate 4. In particular
we examined a number of metathesis partners that differ
structurally (A–C, Scheme 2) to probe the generality of
this ring closing metathesis using Grubbs second-gene-
ration ruthenium catalyst (6).10
Structure–activity relationships in the epothilone class
clearly indicate that changes in the aromatic side chain
can afford potent cytotoxins.4 Thus, side chains A–C
were chosen not only as probes of steric sensitivity of the
ring closing metathesis reactions but because of their
potential as representatives of active classes of epothi-
lone analogs. Bicyclic aryl side chains such as benzox-
azole B are known to be excellent surrogates for the
naturally occurring vinyl thiazole of epothilone.11 Fur-
thermore, appropriate substitution at the 14-position of
The reconstituted epothilone analogs were assayed
against a panel of human tumor cell lines (Table 2). The
biological activity of the benzoxazole side chain (7)
indicates that it is an excellent replacement for the vinyl
thiazole. Though some activity is lost upon substitution
at the 14-position of epothilone (8), the 14-position
remains an intriguing position for analogs, particularly
as a site to improve in vivo stability. As has been noted,

S. D. Dong et al. / Tetrahedron Letters 45 (2004) 1945–1947
Table 1. Metathesis conditions and outcomes
1947
Compound
(X,Y)
R
R0
Substrate
concentration
(mM)
Catalyst
Solvent
Temperature Time
Yield (%)
E=Z
1
–H
1
6 (20 mol %)
6 (20 mol %)
CH2Cl2
CH2Cl2
Reflux
6 h
59
1:1
7
8
–H
2.9
2
Reflux
Reflux
8 h
60
35
1:1
–Me
6 (129 mol %) Toluene
5 days
1:1.3
Table 2. Cytotoxicity of epothilone analogs (IC50 nM)
4. Nicolaou, K. C.; Ritzen, A.; Namoto, K. Chem. Commun.
2001, 1523–1535.
Compound
MCF-7
NCI/ADR
A549
5. Julien, B.; Shah, S. Antimicrob. Agents Chemother. 2002,
46, 2772–2778.
6. Vedejs, E.; Kruger, A. W. J. Org. Chem. 1999, 64, 4790–
4797.
7. Payack, J. F.; Hughes, D. L.; Cai, D.; Cottrell, I. F.;
Verhoeven, T. R. Dimethyltitanocene. In Org. Syntheses;
Hegedus, L. S., Ed.; Wiley: New York, 2002; Vol. 79, pp
19–26.
1Z
1E
7Z
7E
8Z
8E
17
200
36
58
440
27
290
49
140
350
47
ꢀ1300
280
320
400
79
62
160
8. Meng, D.; Bertinato, P.; Balog, A.; Su, D.-S.; Kamenecka,
T.; Sorensen, E. J.; Danishefsky, S. J. J. Am. Chem. Soc.
1997, 119, 10073–10092.
9. May, S. A.; Grieco, P. A. Chem. Commun. 1998, 1597–
1598.
the Z isomers are consistently more potent than the E
isomer.3
10. Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett.
1999, 1, 953–956.
11. Altmann, K. H.; Bold, G.; Caravatti, G.; Florsheimer, A.;
Guagnano, V.; Wartmann, M. Bioorg. Med. Chem. Lett.
2000, 10, 2765–2768.
In conclusion, we have developed a facile and efficient
route to epothilone analogs via a degradation approach
to epothilone D (1). Acid intermediate 4 can be pro-
duced rapidly in six steps from 1. Employing ring-clos-
ing metathesis with Grubbs second-generation
ruthenium catalyst to reconstruct the trisubstituted-
12,13-double bond, we have prepared a variety of ste-
rically differentiated epothilone analogs. The outcome of
the metathesis is dependent on the nature of the sub-
stitution at the allylic position. Using this approach, we
now have ready access to numerous epothilone analogs
with varying C-15 aromatic side chains and C-14 allylic
substitutions. Furthermore, this strategy highlights the
power of degradation and semisynthesis when natural
products are available as building blocks.16
12. Taylor, R. E.; Chen, Y.; Beatty, A.; Myles, D. C.; Zhou,
Y. J. Am. Chem. Soc. 2003, 125, 26–27.
13. Nicolaou, K. C.; Ninkovic, S.; Sarabia, F.; Vourloumis,
D.; He, Y.; Vallberg, H.; Finlay, M. R. V.; Yang, Z.
J. Am. Chem. Soc. 1997, 119, 7974–7991.
14. B was prepared from 2-methylbenzoxazole-5-carbalde-
hyde in an analogous fashion to A.
1
15. 7E: H NMR (CDCl3, 400 MHz) d 1.00 (m, 2H), 1.04 (s,
3H), 1.17 (d, 3H, J ¼ 6:2 Hz), 1.28 (m, 6H), 1.62 (s, 5H),
1.73 (br, 1H), 1.99 (m, 1H), 2.15 (m, 1H), 2.46 (m, 2H),
2.64 (s, 5H), 3.27 (m, 1H), 3.68 (br, 1H), 4.27 (d, 1H,
J ¼ 9:2 Hz), 5.13 (br s, 1H), 6.03 (br s, 1H), 7.27 (d, 1H,
J ¼ 8:4 Hz), 7.44 (d, 1H, J ¼ 8:0 Hz), 7.69 (s, 1H); 13C
NMR (400 MHz, CDCl3) d 14.40, 14.50, 15.65, 15.95,
20.00, 20.57, 29.64, 31.61, 35.00, 37.76, 38.95, 39.44, 42.52,
52.76, 72.10, 75.71, 75.91, 110.15, 116.87, 119.66, 122.77,
136.57, 138.74, 150.46, 164.70, 170.40, 220.29; HRMS
calcd for C28H39NO6: 485.27774; found: 486.28648
(M+H).
8E: 1H NMR (CDCl3, 400 MHz) d 1.00–0.90 (m, 8H), 1.07
(s, 3H), 1.16 (m, 4H), 1.27 (s, 3H), 1.61 (s, 3H), 1.76 (m,
2H), 2.04 (s, 2H), 2.09 (s, 3H), 2.55–2.40 (m, 2H), 2.70 (s,
3H), 2.83 (m, 1H), 3.15 (m, 1H), 3.72 (s, 1H), 4.11 (m, 1H),
4.95 (d, 1H, J ¼ 8:40 Hz),), 5.02 (d, 1H, J ¼ 8:0 Hz), 6.55
(s, 1H), 6.97 (s, 1H); 13C NMR (CDCl3, 400 MHz) d 13.92,
14.98, 16.14, 17.81, 19.06, 21.33, 24.91, 31.68, 34.94, 38.19,
38.53, 39.03, 42.45, 52.75, 72.24, 75.04, 84.43, 116.36,
122.56, 127.34, 136.12, 170.72, 220.18; HRMS calcd for
C28H43NO5S: 505.2862; found: 506.29418 (M+H).
Acknowledgements
We would like to thank John Carney, Nina Viswana-
than, and Chau Q. Tran for analytical support.
References and notes
1. He, L.; Orr, G. A.; Horwitz, S. B. Drug Discovery Today
2001, 6, 1153–1164.
2. For more information about clinical trials of epothilone
D, visit: www.kosan.com.
3. Harris, C. R.; Danishefsky, S. J. J. Org. Chem. 1999, 64,
8434–8760.
16. Niggemann, J.; Michaelis, K.; Frank, R.; Zander, N.;
€
Hofle, G. J. Chem. Soc., Perkin Trans. 1 2002, 2490–2503.