The catalytic asymmetric Claisen rearrangement (CAC) in natural product synthesis: Synthetic studies toward curvicollides A-C

Article abstract of DOI:10.1055/s-2005-922789

A catalytic asymmetric Claisen rearrangement has been utilized as key C-C connecting transformation for the synthesis of a building block in the projected total synthesis of the fungicidal polyketides curvicollide A-C. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2005-922789

LETTER
121
The Catalytic Asymmetric Claisen Rearrangement (CAC) in Natural Product
Synthesis: Synthetic Studies toward Curvicollides A–C
C
A
C
in
T
otal Synthesi
a
s: Curvicol
r
lide
s
leen Körner, Martin Hiersemann*
Institut für Organische Chemie, Technische Universität Dresden, 01062 Dresden, Germany
Fax +49(351)46333162; E-mail: martin.hiersemann@chemie.tu-dresden.de
Received 21 October 2005
tioselective total synthesis. Furthermore, considering the
Abstract: A catalytic asymmetric Claisen rearrangement has been
utilized as key C–C connecting transformation for the synthesis of
a building block in the projected total synthesis of the fungicidal
polyketides curvicollide A–C.
unsolved supply issue, an efficient synthetic access to the
natural products and non-natural diastereomers thereof
could enable the comprehensive evaluation of the curvi-
collides 1a–c as potential fungicidal lead compounds. In
this paper, we report the enantioselective synthesis of a
C8–C12 building block utilizing a catalytic asymmetric
Claisen rearrangement (CAC) as the key C–C connecting
transformation.
Key words: total synthesis, natural products, asymmetric catalysis,
pericyclic reactions, Lewis acids
The curvicollides A–C (1a–c) have been isolated in small
quantities (1a: 6.5 mg, 1b: 2.1 mg, 1c: 3.0 mg) from an
organic extract of solid substrate fermentation cultures of
an isolate of Podospora curvicolla that was originally
obtained from the surface of a sclerotium of Aspergillus
flavus buried in an Illinois cornfield for three years.1 NMR
studies enabled the assignment of the gross structure of
these polyketides (Figure 1). The relative configuration of
the C9–C11 stereotriad of curvicollide A (1a) was de-
duced from NOESY data. The relative configuration of
curvicollide B (1b) and C (1c) was then assigned in anal-
ogy based on the similarity of the corresponding NMR
spectra. Neither the relative configuration of the remain-
ing stereogenic carbon atoms nor the absolute configura-
tion of the curvicollides could be established. Curvicollide
A (1a) exerted antifungal activity against Aspergillus
flavus and Fusarium verticillioides. Due to sample limita-
tions, the fungicidal properties of curvicollides B (1b) and
C (1c) could not be evaluated.
Our synthetic approach toward the curvicollides 1a–c
relies on a convergent retrosynthetic analysis that requires
the enantioselective access to the central C8–C12 segment
of known relative configuration (Figure 2). We envision
connecting the C1–C7 and the C13–C20 moieties to the
central segment by olefination chemistry. Based on this
analysis, we have identified the highly substituted ketone
2 as suitable building block. Building block 2 features
three contiguous stereogenic carbon atoms including the
secondary hydroxyl group at C11, which we envisioned to
establish by a diastereoselective reduction of the a-keto
ester anti-3. From the outset, we intended to generate the
crucial non-heteroatom-substituted stereogenic carbon at-
oms C9 and C10 in anti-3 by a Claisen rearrangement of
the allyl vinyl ether (E,Z)-4. Based on a chair-like transi-
tion state geometry for the Claisen rearrangement,2 the
relative configuration of the rearrangement product anti-3
was expected to be determined by the double bond config-
uration of the allyl vinyl ether (E,Z)-4. However, since the
allyl vinyl ether (E,Z)-4 is achiral, this approach requires
an external chiral inductor to control the absolute config-
uration of the rearrangement product anti-3.3 Relying on
our previously developed catalytic asymmetric Claisen
rearrangement (CAC),4 we set out to identify a suitable
chiral Lewis acid that would fulfill the role as a chiral in-
ductor in catalytic amounts, thereby establishing the CAC
as the key C–C connecting transformation in the synthesis
of the central building block 2. The projected CAC re-
quires an efficient and diastereoselective synthesis of the
E,Z-configured allyl vinyl ether 4. The diastereoselective
synthesis of a vinyl ether double bond that is not part of a
ring system is often troublesome and no general solution
to this problem exists. We envisioned utilizing an in-
herently E-selective olefination reaction between the
phosphonate 5 and acetaldehyde for this purpose.
O
OH
R16
20
11
9
8
12
10
1
O
R19
O
R16
R19
H
Compd Curvicollide
CH2OH
CH3
1a
1b
1c
A
B
C
OH
H
CH3
Figure 1 Curvicollides A–C 1a–c, fungicidal polyketides from
Podospora curvicolla.
In light of the interesting structure and biological activity
of the curvicollides 1a–c, we have initiated a research
program aimed at the unambiguous assignment of the
absolute configuration of the curvicollides 1a–c by enan-
The realization of our initial strategy for the synthesis of
the allyl vinyl ether 4 is depicted in Scheme 1. Trimethyl
diazophosphonoacetate (6)5 was converted into the
SYNLETT 2006, No. 1, pp 0121–0123
Advanced online publication: 16.12.2005
DOI: 10.1055/s-2005-922789; Art ID: G31105ST
© Georg Thieme Verlag Stuttgart · New York
0
5.
0
1.
2
0
0
6

122
M. Körner, M. Hiersemann
LETTER
1a–c
O
LDA, THF, –78 °C
then add CH3CHO
OMe
O
OMe
8
11
9
9
OH
O
10
12
10
O
O
OBn
8
OTBS
O
OMe
BnO
BnO
CAC
1. MsCl, Et3N, CH2Cl2, r.t.
2. DBU, THF, r.t.
O
2
anti-3
OBn
9 (63 %, dr = 69:31)
preparative HPLC
O
O
MeO
12
11
O
11
(MeO)2P
OMe
4
(E,Z)-4 + (Z,Z)-4
8
10
9
(73%, E:Z = 38:62)
O
O
BnO
9
Scheme 2 An aldol strategy toward the 2-alkoxycarbonyl-substitu-
ted allyl vinyl ether 4. DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene.
8
OBn
(E,Z)-4
5
Figure 2 Retrosynthetic analysis of the central segment of the
curvicollides.
Having successfully established a synthetic access to the
allyl vinyl ether (E,Z)-4 on a multigram scale, the pivotal
CAC of (E,Z)-4 was investigated (Scheme 3). Under care-
fully optimized conditions, (E,Z)-4 underwent the desired
Claisen rearrangement in the presence of 7.5 mol%
[Cu{(S,S)-tert-Bu-box}](H2O)2(SbF6)2 (10)12 to provide
the a-keto ester (3R,4R)-313 as a single diastereomer
2-(allyloxy)phosphonoacetate (5) by a rhodium(II)-cata-
lyzed OH insertion6 using the allylic alcohol 7.7 The reac-
tion afforded the phosphonate 5 in excellent yield on a
small scale (1.8 mmol of 7). However, the yield decreased
significantly when the reaction was performed on a gram
scale. The subsequent Horner–Wadsworth–Emmons re-
action of the phosphonate 5 with acetaldehyde provided
the desired allyl vinyl ether (E,Z)-48 as a 8:1 mixture with
(Z,Z)-4 that was removed by chromatography.9,10
1
(based on H NMR analysis) and enantiomer (based on
HPLC analysis, Chiracel OD 14025, n-hexane–i-PrOH
99.8:0.2, 1 mL/min) in good yield even on a gram scale.
The absolute configuration of 3 was assigned in analogy
to the previously established stereochemical course of the
Cu(box)-catalyzed Claisen rearrangement of 2-alkoxy-
carbonyl-substituted allyl vinyl ether.4
4 mol% Rh2(OAc)4
DCE, reflux, 1 h
O
O
(MeO)2P
OMe
OH
We have recently reported that K-Selectride is particular-
ly useful to reduce 3-substituted 2-oxo ester with a high
diastereoselectivity in favor of the corresponding 2,3-
anti-configured a-hydroxy ester.4c Accordingly, treatment
of the a-keto ester 3 with K-Selectride provided the a-hy-
droxy ester (2R,3R,4R)-11 as a single diastereomer (based
N2
7
BnO
6
O
O
(MeO)2P
OMe
LDA, THF, –78 °C
then CH3CHO
O
OBn
1
on H NMR analysis) in very good yield. As reported
5 (86%)
previously,4c the stereochemical course of the reduction
can be explained by the application of the Cram–Felkin–
Anh model.14
MeO
O
preparative HPLC
O
BnO
(E,Z)-4
7.5 mol% 10
(E,Z)-4
4
(75%, E:Z = 8:1)
4 Å mol sieves
CH2Cl2, r.t., 12 h
(0.92 g)
Scheme 1 Synthesis of the allyl vinyl ether (E,Z)-4.
O
K[(sec-Bu)3BH]
OMe
THF
–100 °C, 10 min
Though the discussed route provided a diastereoselective
access to the desired E,Z-configured allyl vinyl ether 4, it
was hampered by our inability to scale up the RhII-cata-
lyzed OH-insertion and by the expensiveness of the RhII
catalyst. Therefore, we utilized an alternative route to the
allyl vinyl ether 4 which can easily be performed on large
scale (Scheme 2).11 In the event, aldol addition of the
enolate of the allyloxy-substituted acetic acid ester 8 with
acetaldehyde afforded the b-hydroxy ester 9 in moderate
non-optimized yield as a mixture of diastereomers.
Mesylation of 9 followed by DBU-mediated elimination
provided the desired allyl vinyl ether 4 as mixture of vinyl
ether double bond isomers that were conveniently separat-
ed by preparative HPLC.9
O
BnO
(3R,4R)-3
(82%, >90% de, 99% ee)
O
OMe
OH
(2R,3R,4R)-11
(98%, >90% de)
BnO
2
O
O
N
2 SbF6
N
Cu
t-Bu
t-Bu
H2O OH2
10
Scheme 3 The sequence of the two stereodifferentiating reactions.
Synlett 2006, No. 1, 121–123 © Thieme Stuttgart · New York

LETTER
CAC in Total Synthesis: Curvicollides
123
(4) (a) Abraham, L.; Körner, M.; Hiersemann, M. Tetrahedron
Lett. 2004, 3647. (b) Abraham, L.; Körner, M.; Schwab, P.;
Hiersemann, M. Adv. Synth. Catal. 2004, 1281.
In order to verify the assumed relative configuration of
(2R,3R,4R)-11, the d-lactone (2R,3R,4R)-12 was prepared
in situ under the conditions of the oxidative removal15 of
the benzyl protecting group (Equation 1). NOESY on the
d-lactone 12 provided NOE that unambiguously support
our mechanism-based assignment of the relative configu-
ration of (2R,3R,4R)-11.
(c) Hiersemann, M.; Abraham, L. Eur. J. Org. Chem. 2002,
1461. (d) Abraham, L.; Czerwonka, R.; Hiersemann, M.
Angew. Chem. Int. Ed. 2001, 4700.
(5) Prepared according to: Gois, P. M. P.; Afonso, C. A. M. Eur.
J. Org. Chem. 2003, 3798.
(6) Miller, D. J.; Moody, C. J. Tetrahedron 1995, 51, 10811.
(7) Gao, Y.; Klunder, J. M.; Hanson, R. M.; Masamune, H.; Ko,
S. Y.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765.
(8) (E,Z)-4: 1H NMR (300 MHz, CDCl3): d = 1.96 (d, 3 H,
J = 7.4 Hz), 3.78 (s, 3 H), 4.09 (d, 2 H, J = 4.2 Hz), 4.31 (d,
2 H, J = 4.0 Hz), 4.50 (s, 2 H), 5.36 (q, 1 H, J = 7.4 Hz),
5.78–5.81 (m, 2 H), 7.31–7.34 (m, 5 H). 13C NMR (75.5
MHz, CDCl3): d = 12.6, 51.8, 65.0, 65.8, 72.4, 112.9, 127.7,
127.8, 128.0, 128.4, 129.7, 138.0, 144.8, 164.1. IR (neat):
3035–3030, 2950–2860, 1725 cm–1. Anal. Calcd for
C16H20O4: C, 69.54; H, 7.30. Found: C, 69.21; H, 7.39.
(9) Preparative HPLC: Nu 50-7, 32 × 250 mm, heptane–EtOAc
9:1, 30 ml/min, tR (Z) = 7 min, tR (E) = 9 min, baseline
separation with 400 mg/injection.
(10) The sequence of OH-insertion and olefination was originally
developed and utilized by Ganem and Berchtold for the
synthesis of chorismate und its derivatives, see: (a) Ganem,
B.; Ikota, N.; Muralidharan, V. B.; Wade, W. S.; Young, S.
D.; Yukimoto, Y. J. Am. Chem. Soc. 1982, 104, 6787.
(b) Pawlak, J. L.; Berchtold, G. A. J. Org. Chem. 1987, 52,
1765. (c) Lesuisse, D.; Berchtold, G. A. J. Org. Chem. 1988,
53, 4992. (d) Wood, H. B.; Buser, H. P.; Ganem, B. J. Org.
Chem. 1992, 57, 178. (e) Mattia, K. M.; Ganem, B. J. Org.
Chem. 1994, 59, 720. (f) For an intramolecular application,
see: Kim, S.; Sutton, S. C.; Guo, C.; LaCour, T. G.; Fuchs,
P. L. J. Am. Chem. Soc. 1999, 121, 2056.
DDQ, CH2Cl2
buffer (pH 7)
O
OH
O
OMe
r.t., 24 h
OH
O
BnO
(2R,3R,4R)-11
(2R,3R,4R)-12 (81%)
Equation 1 Validation of the relative configuration of 11 is based
on NOESY studies of the d-lactone (2R,3R,4R)-12.
The synthesis of the key building block 2 was concluded
by a five-step sequence from 11 as depicted in Scheme 4.
Thus, the protection of the secondary hydroxyl group as a
silyl ether16 was followed by the reduction of the ester
function to the primary alcohol and a subsequent Dess–
Martin oxidation17 to afford the aldehyde 13. Grignard re-
action between the aldehyde 13 and methylmagnesium
iodide followed by Dess–Martin oxidation17 provided the
desired ketone 2.18
1. TBSOTf, 2,6-lutidine
CH2Cl2, 0 °C
11
O
OTBS
2. DIBAL-H, CH2Cl2
–78 °C to r.t.
BnO
13 (76%)
3. Dess–Martin periodinane
CH2Cl2, 0 °C to r.t.
(11) Hiersemann, M. Synthesis 2000, 1279.
(12) Evans, D. A.; Miller, S. J.; Lectka, T.; von Matt, P. J. Am.
Chem. Soc. 1999, 121, 7559.
1. MeMgI, Et2O
–78 °C to r.t.
(13) (3R,4R)-3: 1H NMR (300 MHz, CDCl3): d = 1.08 (d, 3 H,
J = 6.9 Hz), 2.83 (m, 1 H), 3.31–3.39 (m, 2 H), 3.45 (ddAB, 1
H, J = 9.6, 4.9 Hz), 3.62 (s, 3 H), 4.33 (dAB, 1 H, J = 12.0
Hz), 4.37 (dAB, 1 H, J = 12.0 Hz), 5.12–5.17 (m, 2 H), 5.45–
5.58 (m, 1 H), 7.22–7.34 (m, 5 H). 13C NMR (75.5 MHz,
CDCl3): d = 14.4, 43.1, 48.4, 52.5, 72.4, 72.7, 118.7, 127.6,
128.3, 135.3, 137.5, 161.5, 195.6. IR (neat): n = 3300–3150,
2950–2870, 1728 cm–1. Anal. Calcd for C16H20O4: C, 69.54;
H, 7.30. Found: C, 69.28; H, 7.38. [a]25D +39.7 (c 0.89,
CHCl3).
O
2. Dess–Martin periodinane
CH2Cl2, 0 °C to r.t.
OTBS
BnO
2 (79%)
Scheme 4 Final steps toward the desired building block 2.
In conclusion, we have established a highly enantioselec-
tive access to the ketone 2, the central C8–C12 synthon in
the projected total synthesis of the curvicollides 1a–c. We
have demonstrated the efficiency of the catalytic asym-
metric Claisen rearrangement (CAC) in target-oriented
synthesis. Based on the availability of the building block
2, efforts aimed at the completion of the total synthesis of
the curvicollides 1a–c are currently underway in our
laboratory.
(14) Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191.
(15) Oikawa, Y.; Yoshioka, T.; Yonemitsu, O. Tetrahedron Lett.
1982, 23, 885.
(16) Corey, E. J.; Cho, H.; Rücker, C.; Hua, D. H. Tetrahedron
Lett. 1981, 22, 3455.
(17) (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. (c) Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58,
2899.
(18) (3R,4R,5R)-2: 1H NMR (300 MHz, CDCl3): d = 0.00 (s, 3
H), 0.05 (s, 3 H), 0.75 (d, 3 H, J = 7.1 Hz), 0.94 (s, 9 H),
2.03–2.12 (m, 1 H), 2.14 (s, 3 H), 2.71–2.79 (m, 1 H), 3.46
(ddAB, 1 H, J = 9.7, 7.2 Hz), 3.53 (ddAB, 1 H, J = 9.7, 6.1 Hz),
3.79 (d, 1 H, J = 7.7 Hz), 4.45 (dAB, 1 H, J = 12.1 Hz), 4.55
(dAB, 1 H, J = 12.1 Hz), 5.04–5.14 (m, 2 H), 5.71 (ddd, 1 H,
J = 17.2, 10.5, 8.8 Hz), 7.29–7.41 (m, 5 H). 13C NMR (75.5
MHz, CDCl3): d = –5.0, –4.8, 11.3, 18.1, 25.3, 25.8, 37.0,
42.8, 71.6, 72.6, 81.4, 117.5, 127.5, 127.6, 128.3, 136.2,
138.5, 211.7. IR (neat): n = 3100–3060, 2970–2860, 1716
cm–1. Anal. Calcd for C22H36O3Si: C, 70.16; H, 9.63. Found:
C, 70.30; H, 9.72. [a]25D +38.3 (c 0.90, CHCl3).
Acknowledgment
Financial support by the DFG and the FCI is gratefully acknowled-
ged. M.H. is a Heisenberg Fellow of the DFG.
References and Notes
(1) Che, Y.; Gloer, J. B.; Wicklow, D. T. Org. Lett. 2004, 1249.
(2) Ziegler, F. E. Chem. Rev. 1988, 1423.
(3) For reviews concerning the asymmetric Claisen rearrange-
ment, see: (a) Enders, D.; Knopp, M.; Schiffers, R.
Tetrahedron: Asymmetry 1996, 7, 1847. (b) Ito, H.;
Taguchi, T. Chem. Soc. Rev. 1999, 28, 43. (c)Nubbemeyer,
U. Synthesis 2003, 961.
Synlett 2006, No. 1, 121–123 © Thieme Stuttgart · New York