Concise total syntheses of epothilone A and C based on alkyne metathesis

Article abstract of DOI:10.1039/b101669p

A ring closing alkyne metathesis reaction catalyzed by the molybdenum complex 26 followed by a Lindlar reduction of the resulting cycloalkyne product opens an efficient and stereoselective entry into epothilone A and C.

Full text of DOI:10.1039/b101669p

Concise total syntheses of epothilone A and C based on alkyne
metathesis
Alois Fürstner,* Christian Mathes and Karol Grela
Max-Planck-Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, D-45470 Mülheim/Ruhr, Germany.
E-mail: fuerstner@mpi-muelheim.mpg.de; Fax: ++49 208 306 2994; Tel: ++49 208 306 2342
Received (in Cambridge, UK) 20th February 2001, Accepted 8th May 2001
First published as an Advance Article on the web 25th May 2001
A ring closing alkyne metathesis reaction catalyzed by the
molybdenum complex 26 followed by a Lindlar reduction of
the resulting cycloalkyne product opens an efficient and
stereoselective entry into epothilone A and C.
Table 1 RCM approaches towards epothilone A and C: formation of (E,Z)-
mixtures (Ar = 2-methyl-4-thiazolyl)
Catalysta
Ru]
R1
R2
Yield
Z+E
Ref.
[
TBS
TBS
TBS
H
TBS
TBS
H
H
TBS
86%
94%
85%
65%
86%
1.7+1
1+1
4b
6a
5b
4b
4b
1
The discovery that epothilone A (1) and congeners share a
®
common mechanism of action with paclitaxel (Taxol ) in
1.2+1
1+2
triggering programmed cell death (apoptosis) and exert high
activity even against paclitaxel-resistant human cancer cell lines
in vitro has spurred considerable drug development programs
[Mo]
TBS
1+2
a
2 3 2 2 3 2 2
[Mo] = Mo(NNAr)(NCHCMe Ph)[OCMe(CF ) ] ; [Ru] = (PCy ) (Cl) -
2
worldwide. As a consequence, these compounds became the
RuNCHPh.
focal point of many preparative studies aiming at their total
synthesis as well as at a synthesis-driven mapping of the
structure–activity relationship of these promising natural
products.2,3
retains all the advantages of metathetic conversions† but allows
for the first time the gearing of the stereochemical issue to the
cyclization event. If combined with a Lindlar-type reduction,
9
this method opens a stereoselective entry into (Z)-alkenes
(
Scheme 1). We felt that epothilone A constitutes an ideal
testing ground for the scope of this emerging new methodology
(
Scheme 2).9 Described below is the successful reduction of
–11
this plan to practice.
Earlier studies had revealed that the selectivity gained in the
formation of the three contiguous stereocenters at C-6, C-7 and
C-8 by an aldol reaction strongly depends on the remote
functionalization of the enolate partner.2 The best results were
reported by Schinzer et al. who employed ethyl ketone 11
bearing a conformationally rigid and chelating 1,3-dioxane unit
,3
6
as control element for this purpose. We took recourse to this
In this context it is remarkable that the first three successful
approaches towards 1 were all based on ring closing alkene
metathesis (RCM) for the formation of the 16-membered ring.
Product 4 thus formed can be selectively epoxidized at the
12,13
D
-bond and hence constitutes an excellent precursor for
4–6
epothilone A.
Scheme 1 Stereoselective synthesis of (Z)-alkenes by ring closing alkyne
metathesis/Lindlar reduction.
Although these studies were early highlights showing the
enormous potential of RCM for advanced organic synthesis,7
they invariably suffered from the fact that there was little—if
any—selectivity in favor of the required (Z)-alkene 4 (Table 1).
As this serious problem arose only towards the very end of
rather laborious sequences and since the isomeric alkenes could
not be readily separated at this stage, it is hardly surprising that
subsequent total syntheses of 1 were largely based on strategies
other than RCM that ensure better control over all structural
elements of this target.8
Recently, our group was able to show that the ring closing
metathesis of diynes constitutes a promising alternative that
Scheme 2 Retrosynthetic analysis of epothilone C (2).
DOI: 10.1039/b101669p
Chem. Commun., 2001, 1057–1059
This journal is © The Royal Society of Chemistry 2001
1057

View Article Online
elegant solution, seeking, however, an improved and shorter
entry into the required key building block 11.
Our synthesis starts from commercially available 3-hydroxy-
propionitrile 5 which reacts with the zinc enolate derived from
bromo ester 6 to afford ketoester 7 in 71% yield on a multigram
scale (Scheme 3). This Reformatsky-type reaction is best
carried out with the assistance of ultrasound.12 Silylation of 7
with tert-butyldiphenylsilyl chloride under standard conditions
followed by an asymmetric hydrogenation of 8 catalyzed by
[((S)-binap)RuCl
2 3
](NEt ) in the presence of Dowex (H-form) to
ensure acidic conditions delivers the unprotected diol 9 in high
13
enantiomeric purity (ee = 94%). ‡ All attempts to perform the
reduction directly with the unprotected substrate 7 resulted in
rather poor conversion. Acetalization of 9 followed by reaction
of the resulting product 10 with EtMgBr in toluene in the
Scheme 5 Reagents and conditions: i, LDA, THF, 278 °C, 70%; ii, PPTS,
MeOH, 85%; iii, TBSOTf, 2,6-lutidine, 92%; iv, camphorsulfonic acid cat.,
CH Cl –MeOH (1+1), 78%; v, PDC, DMF, 83%.
3
presence of NEt affords compound 11 in excellent overall
yield. The presence of the base during the addition of the
Grignard reagent to the ester is essential, as it enolizes the
ketone primarily formed and thereby avoids the formation of the
corresponding tertiary alcohol by addition of a second equiva-
lent of EtMgBr.14
2
2
Scheme 6 Reagents and conditions: i, (+)-Ipc
imidazole, DMF, 89% (over both steps); iii, (1) OsO
Pb(OAc) , 86%; iv, CBr , PPh , CH Cl , 68%; v, n-BuLi, MeI, THF, 65%;
vi, TBAF·3H O, THF, 74%.
2
B(allyl); ii, TBSCl,
4
cat., NMO; (2)
4
4
3
2
2
2
(
ee > 97%). Oxidative cleavage of its terminal double bond
Scheme 3 Reagents and conditions: i, Zn, ultrasound, THF; then aq. HCl,
7
1%; ii, TBDPSCl, imidazole, DMF, 90%; iii, [((S)-binap)RuCl
mol%), H (65 bar), Dowex, EtOH, 80 °C, 71%; iv, 2,2-dimethoxypropane,
acetone, camphorsulfonic acid cat., 92%; v, EtMgBr, NEt , toluene, 70 °C,
8%.
2 3
](NEt ) (6
affords the somewhat unstable aldehyde 22 which is im-
mediately used for a subsequent Corey-Fuchs reaction.
Specifically, treatment of 22 with CBr and PPh gives the
1
6
2
3
4
3
6
2c
expected 1,1-dibromo derivative 23, which is converted into
alkyne 24 by means of n-BuLi in THF and trapping of the
acetylide anion thus formed with MeI. Desilylation under
standard conditions followed by esterification of the resulting
alcohol 25 with compound 19 sets the stage for the crucial
macrocyclization step. It should also be noted that all attempts
to form product 25 from aldehyde 20 by direct asymmetric
propargylation were unrewarding in terms of yield and optical
purity.
Having secured an improved access to this key building
block, the subsequent aldol reaction was carried out in close
6
analogy to that described by Schinzer et al. The required
aldehyde component 14 is readily formed as shown in Scheme
4
, exploiting the excellent facial guidance exerted by Op-
polzer’s bornane sultam in the alkylation of substrate 12 (d.r. =
15
96+4). Reaction of the lithium enolate derived from 11 with
compound 14 affords aldol 15 in 70% yield (Scheme 5). The
selectivity for the desired anti-Cram product was 7+1 (HPLC),
which is easily separated from the minor isomer by flash
chromatography. Further elaboration of this compound involv-
ing deprotection of the acetal, per-silylation of the resulting triol
We were pleased to see that diyne 27 is in fact smoothly
converted into the 16-membered cycloalkyne 28 in 80%
isolated yield on exposure to catalytic amounts of the molybde-
num amido complex 2617 in toluene–CH Cl at 80 °C (Scheme
2
2
7). This outcome is particularly noteworthy as it compares well
to the results obtained in the conventional RCM approaches
(Table 1) in terms of yield and reaction rate. Furthermore, it
clearly attests to the mildness and preparative relevance of the
method since (i) neither the basic N-atom nor the sulfur group
of the thiazole ring interfere with the catalyst, (ii) the labile aldol
substructure, the rather electrophilic ketone, as well as the ester-
and silyl ether groups are fully preserved, (iii) no racemization
of the chiral center a to the carbonyl is encountered, and (iv) the
rigorous chemoselectivity of the catalyst is confirmed, which
reacts smoothly with alkynes but leaves pre-existing alkene
moieties unaffected.§ Therefore, this particular example in
1
6, and regioselective cleavage of the primary TBS-ether in 17
5,6
is performed in analogy to literature routes. Oxidation of the
resulting alcohol 18 with PDC in DMF smoothly affords the
desired carboxylic acid 19 ready for esterification with a
suitable thiazol fragment.
9
–11
concert with the previous applications from our laboratory
Scheme 4 Reagents and conditions: i, n-BuLi, THF–HMPA, MeI, 278 ?
substantiates the notion that alkyne metathesis in general holds
great promise for target oriented synthesis.
2
60 °C, 94%; ii, LiAlH
4 4 4 2 2
, THF, 85%; iii, Pr NRuO cat., NMO, CH Cl ,
MS 4 Å, 90%.
Lindlar reduction of cycloalkyne 28 followed by cleavage of
the silyl ether groups in the resulting (Z)-alkene 29 by means of
The preparation of the latter (Scheme 6) starts with an
allylation of aldehyde 20 with (+)-Ipc B(allyl) as described
aq. HF in Et
2
O–MeCN as the reaction medium delivers
2
epothilone C 2 in 79% yield. Because the selective epoxidation
5
2–6
earlier, followed by silylation of the crude material with TBSCl
and imidazole, thus delivering the homoallyl alcohol derivative
of 29 has already been described by various groups,
this
approach also constitutes a formal total synthesis of epothilone
21 in 89% yield over both steps in excellent enantiomeric excess
A 1.
1058 Chem. Commun., 2001, 1057–1059

View Article Online
1
2
G. Höfle, N. Bedorf, H. Steinmetz, D. Schomburg, K. Gerth and H.
Reichenbach, Angew. Chem., Int. Ed. Engl., 1996, 35, 1567.
Reviews: (a) K. C. Nicolaou, F. Roschangar and D. Vourloumis, Angew.
Chem., Int. Ed., 1998, 37, 2014; (b) C. R. Harris and S. J. Danishefsky,
J. Org. Chem., 1999, 64, 8434; (c) K.-H. Altmann, G. Bold, G.
Caravatti, N. End, A. Flörsheimer, V. Guagnano, T. O’Reilly and M.
Wartmann, Chimia, 2000, 54, 612; (d) J. Mulzer, Monatsh. Chem.,
2
000, 131, 205 and lit. cited therein.
K. C. Nicolaou, N. P. King and Y. He, Top. Organomet. Chem., 1998,
, 73.
3
4
1
(a) P. Bertinato, E. J. Sorensen, D. Meng and S. J. Danishefsky, J. Org.
Chem., 1996, 61, 8000; (b) D. Meng, P. Bertinato, A. Balog, D.-S. Su,
T. Kamenecka, E. J. Sorensen and S. J. Danishefsky, J. Am. Chem. Soc.,
1
997, 119, 10 073.
5
6
(a) Z. Yang, Y. He, D. Vourloumis, H. Vallberg and K. C. Nicolaou,
Angew. Chem., Int. Ed. Engl., 1997, 36, 166; (b) K. C. Nicolaou, Y. He,
D. Vourloumis, H. Vallberg, F. Roschangar, F. Sarabia, S. Ninkovic, Z.
Yang and J. I. Trujillo, J. Am. Chem. Soc., 1997, 119, 7960.
(a) D. Schinzer, A. Limberg, A. Bauer, O. M. Böhm and M. Cordes,
Angew. Chem., Int. Ed. Engl., 1997, 36, 523; (b) D. Schinzer, A. Bauer,
O. M. Böhm, A. Limberg and M. Cordes, Chem. Eur. J., 1999, 5,
2
483.
7
8
(a) A. Fürstner, Angew. Chem., Int. Ed., 2000, 39, 3012; (b) R. H.
Grubbs and S. Chang, Tetrahedron, 1998, 54, 4413.
For other total syntheses of 1 see: (a) D. Sawada, M. Kanai and M.
Shibasaki, J. Am. Chem. Soc., 2000, 122, 10 521; (b) K. C. Nicolaou,
S. Ninkovic, F. Sarabia, D. Vourloumis, Y. He, H. Vallberg, M. R. V.
Finlay and Z. Yang, J. Am. Chem. Soc., 1997, 119, 7974; (c) A. Balog,
D. Meng, T. Kamenecka, P. Bertinato, D.-S. Su, E. J. Sorensen and S. J.
Danishefsky, Angew. Chem., Int. Ed. Engl., 1996, 35, 2801; (d) B. Zhu
and J. S. Panek, Org. Lett., 2000, 2, 2575; (e) J. W. Bode and E. M.
Carreira, J. Am. Chem. Soc., 2001, 123, 3611.
Scheme 7 Reagents and conditions: i, DCC, DMAP, CH
10 mol%), toluene–CH Cl , 80 °C, 8 h, 80%; iii, Lindlar catalyst,
quinoline, H (1 atm), CH Cl , quant.; iv, aq. HF, Et O–MeCN, 79%; v,
dimethyldioxirane, 70% (ref. 4).
2 2
Cl , 81%; ii, 26
(
2
2
2
2
2
2
9
A. Fürstner and G. Seidel, Angew. Chem., Int. Ed., 1998, 37, 1734.
1
0 A. Fürstner, C. Mathes and C. W. Lehmann, J. Am. Chem. Soc., 1999,
21, 9453.
1
Generous financial support by the Deutsche Forschungs-
gemeinschaft (Leibniz award to A. F.) and the Fonds der
Chemischen Industrie is gratefully acknowledged. K. G. thanks
the Alexander-von-Humboldt Foundation for a fellowship.
1
1 Previous applications in total synthesis: (a) A. Fürstner, O. Guth, A.
Rumbo and G. Seidel, J. Am. Chem. Soc., 1999, 121, 11 108; (b) A.
Fürstner, K. Grela, C. Mathes and C. W. Lehmann, J. Am. Chem. Soc.,
2000, 122, 11 799; (c) A. Fürstner and K. Grela, Angew. Chem., Int. Ed.,
2000, 39, 1234; (d) A. Fürstner and A. Rumbo, J. Org. Chem., 2000, 65,
2608; (e) A. Fürstner, K. Radkowski, J. Grabowski, C. Wirtz and R.
Mynott, J. Org. Chem., 2000, 65, 8758; (f) A. Fürstner and T. Dierkes,
Org. Lett., 2000, 2, 2463.
Notes and references
†
For a discussion of the strategic advantages of metathesis in general over
more conventional transformations see ref. 18.
The need to perform this reduction under slightly acidic conditions
12 (a) K. Narkunan and B.-J. Uang, Synthesis, 1998, 1713; (b) review: A.
Fürstner, Synthesis, 1989, 571.
13 (a) D. F. Taber and L. J. Silverberg, Tetrahedron Lett., 1991, 32, 4227;
(b) review: R. Noyori, Asymmetric Catalysis in Organic Synthesis,
Wiley, New York, 1994.
14 I. Kikkawa and T. Yorifuji, Synthesis, 1980, 877.
15 W. Oppolzer, R. Moretti and S. Thomi, Tetrahedron Lett., 1989, 30,
5603.
16 E. J. Corey and P. L. Fuchs, Tetrahedron Lett., 1972, 3769.
17 For the preparation and inorganic chemistry of complex 26 see: C. C.
Cummins, Chem. Commun., 1998, 1777.
‡
determines the choice of the protecting group for the primary alcohol; the
TBDPS group turned out to be optimal, whereas the TBS ether was found
to be too unstable.
§
Other available catalysts for alkyne metathesis are (i) Mo(CO)
chlorophenol and (ii) alkylidyne complexes such as (t-BuO) W·CCMe
System (i), however, requires very harsh conditions (!130 °C), whereas the
tungsten alkylidyne is sensitive towards basic nitrogen atoms or sulfur(II
6
–p-
3
3
.
)
groups in the diyne substrate. Therefore they are not appropriate for the
cyclization of 27 to 28. For a more detailed discussion see ref. 11a.
18 A. Fürstner, Synlett, 1999, 1523.
Chem. Commun., 2001, 1057–1059
1059