Towards the total synthesis of saponaceolides: Synthesis of cis-2,4- disubstituted 3,3-dimethylmethylenecyclohexanes

Article abstract of DOI:10.1002/(SICI)1521-3773(19991216)38:24<3662::AID-ANIE3662>3.0.CO;2-1

Controlling diastereoselectivity by metal coordination: In the case of 3,3-dimethylmethylenecyclohexanes 2 - a structural unit present in many interesting natural products - the thermodynamically less stable cis-2,4- disubstituted compounds become availab

Full text of DOI:10.1002/(SICI)1521-3773(19991216)38:24<3662::AID-ANIE3662>3.0.CO;2-1

COMMUNICATIONS
molecular carbametalation via conformer
6
[Eq. (1);
Towards the Total Synthesis of Saponaceolides:
Synthesis of cis-2,4-Disubstituted
1
TBDMS ˆ tert-butyldimethylsilyl] is 2.8 kcalmol lower in
energy than the alternative reaction leading to the trans
product. In principal, this intermediate can be accessed by a
cycloisomerization of enyne 4 or a Heck-type process with
vinyl bromide 5.
3,3-Dimethylmethylenecyclohexanes**
Barry M. Trost,* James R. Corte, and
Mark S. Gudiksen
A variety of natural products are characterized by the
cyclohexane structural unit 1. For example, the cycloiridals are
prized for their fragrance, and they show interesting activities
such as potent piscicidal activity.[1] Carotenoids such as
sarcinaxanthin also have this core unit.[2] The high cytotoxicity
of saponaceolides A ± D (2a ± d) towards a large number of
RO
R1O
OTBDMS
OTBDMS
RO
R1O
PdX
OTBDMS
(1)
4
RO
R1O
6
Br
5
RO
R1O
7
OTBDMS
Scheme 1 outlines the synthesis of the cyclization precur-
sors 4 (i.e., 12) and 5 (i.e., 13) in racemic form (the latter was
OTBDMS
d
a-c
AcO
OTBDMS
tumors focused our interest on the core unit of this family of
compounds.[3, 4] While natural products have both cis and trans
2,4-substitution in 1, all of the above-mentioned compounds
show cis substitution. The fact that the cis isomer is the
thermodynamically less stable isomer has hampered its
access.[5] A recent report of the synthesis of 2-epi-saponaceo-
lide B highlights this issue[6] and stimulated us to report our
studies directed at this unit in the context of a total synthesis
of the saponaceolides.
83% x 63% x 93%
78%
8
R1O
CO2CH3
j
e
RO
CH3O2C
OTBDMS
OTBDMS
H
86%
89%
9
10
a: R = R1 = H
b: R = H, R1 = Ac
g, 94%
f
h, 90%
i, 90%
c: R = TBDPS, R1 = Ac
d: R = TBDPS, R1 = H
83%
For a synthesis of the saponaceolides, the cyclohexyl system
3 serves as our target. Traditional strategies to structures like 1
based upon carbonyl chemistry that set the relative stereo-
chemistry by base-catalyzed epimerization invariably led to
the trans isomer.[5] Our interest in palladium-catalyzed cyclo-
isomerizations of enynes[7, 8] led us to consider an alternative
strategy to methylenecyclohexanes and raised the question of
the geometrical constraints imposed by an intramolecular
carbametalation. Simple modeling suggested that the intra-
H
RO
12
k
TBDPSO
OR1
67%
O
H
a: R = H, R1 = TBDMS
l
95%
b: R = R1 = TBDMS
TBDPSO
OTBDMS
m
89%
H
c: R = TBDMS, R1 = H
11
H
R1O
RO
13
OH
n
Br
80%
H
[*] Prof. B. M. Trost, Dr. J. R. Corte, M. S. Gudiksen
Department of Chemistry
a: R = TBDPS, R1 = H
b: R = TBDPS, R1 = TBDMS
c: R R1 = C(CH3)2
o
p
83%
Stanford University
87%
Stanford, CA 94305-5080 (USA)
Fax: (‡1)650-725-0002
Scheme 1. a) n-C4H9Li, THF, 788C then CH3COCH3; b) Red-Al, Et2O,
08C; c) Ac2O, 5% DMAP, (C2H5)3N, RT; d) CH2(CO2CH3)2, BSA, 5%
[Mo(CO)6], PhCH3, reflux; e) LAH, THF, 08C; f) NaH, THF, 558C then
TBDPSCl, RT; g) PCL (Amano), CH2 CHOAc, THF, RT; h) TBDPSCl,
imidazole, DMF, 08C; i) Ba(OH)2, CH3OH, RT; j) 5% TPAP, NMO, 3-Š
E-mail: bmtrost@leland.stanford.edu
[**] We thank the National Science Foundation and the National Institutes
of Health, General Medical Sciences, for their generous support of
our programs. J.R.C. held a NSF predoctoral fellowship and a Veatch
Fellowship. The X-ray crystal structure was determined by Dr. Athar
Masood in the crystallography laboratory at Stanford University.
Mass spectra were provided by the Mass Spectrometry Facility of the
University of California-San Francisco, supported by the NIH
Division of Research Resources. We are grateful to Prof. Dr. G.
Vidari for copies of spectra of the natural product.
ˆ
ˆ ˆ
MS, CH2Cl2, 08C; k) CH2 C CHMgBr, Et2O,
788C; l) TBDMSO-
SO2CF3, 2,6-lutidine, CH2Cl2, 08C; m) C2H5OH, PPTS, 508C;
n) CH2 C(Br)CH2Br, Sn, HBr, Et2O, H2O, RT; o) same as (l) then
C2H5OH, PPTS, 558C; p) TBAF, THF, RT, then CH3COCH3, amberlyst
ˆ
‡
15-H , RT. DMAP ˆ 4-dimethylaminopyridine, BSA ˆ N,O-bis(trimethyl-
silyl)acetamide, TBDPS ˆ tert-butyldiphenylsilyl, TPAP ˆ tetrapropylam-
monium perruthenate, NMO ˆ N-methylmorpholin-N-oxide, PPTS ˆ pyr-
idinium p-toluenesulfonate, TBAF ˆ tetrabutylammonium fluoride.
Supporting information for this article is available on the WWW
under http://www.wiley-vch.de/home/angewandte/ or from the author.
3662
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3824-3662 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 24

COMMUNICATIONS
obtained in enantiomerically enriched form as well). Intro-
duction of the malonate moiety regioselectively by allylic
alkylation with acetate 8 takes advantage of the bias of
molybdenum catalysts to direct nucleophilic attack towards
the more highly substituted allyl terminus; regioisomer 9 is
the exclusive product.[9]
illustrated in Equation (3) in which a cis:trans ratio of 2:1
was observed (dppb ˆ 1,4-bis(diphenylphosphanyl)butane,
DMA ˆ N,N-dimethylacetamide). Methanolysis of 16 result-
ed in conversion into 15a.
2.5% [Pd2(dba)3]•CHCl3
7% dppb
The racemic silyl ether 10d is available directly from the
diol 10a by the method of McDougal et al.[10] Enantiomeri-
cally enriched monoacetate 10b is obtained by enzymatic
transesterification with 86% ee (45% yield, 91% yield based
on recovered starting material) with Pseudomonas fluorescens
lipase (PFL).[11] With Pseudomonas cepacia lipase (PCL)[12]
higher conversions (68% yield, 94% yield based on recovered
starting material) are obtained, but with a slightly lower ee
value (81% ee). The absolute configuration was established
by the known stereochemical preferences of the enzymes
employed and verified by correlation with a known com-
pound.[13] Interestingly, organometallic additions to aldehyde
11 proceeded completely diastereoselectively with allenyl-
magnesium bromide[14] to give 12 and with tin and 2,3-
dibromopropene[15] under Barbier conditions to give 13. The
relative configuration in both cases was established by
converting the 1,3-diol units into cyclic acetonides (13c in
the latter case), and NMR spectroscopy verifies the trans
relationship of the two methine hydrogen atoms (for 13c, J ˆ
9.2 Hz). This stereochemistry corresponds to a Felkin ± Anh
mode of addition.[16]
Previous work suggests that the cycloisomerization of
substrates like 12 should produce 1,3- rather than 1,4-
dienes.[7, 17] In contrast to this expectation, palladium-cata-
lyzed cycloisomerization of 12b proceeded to form exclu-
sively the 1,4-diene product, an enol silyl ether, which is
directly hydrolyzed to the aldehyde 14 [Eq. (2); dba ˆ di-
benzylideneacetone]. Alternatively, cycloisomerization of
12c provides the aldehyde 14 directly as the sole product. In
both cases, the product is isolated as a 2:1 mixture in which the
major isomer indeed has the desired cis configuration (see
below).
Ag3PO4, CaCO3
CH3OH
PPTS
H
H
O
DMA
15a
(3)
13c
O
CHO
79%
69%
H
16
To establish the relative configuration, the aldehyde 16 was
converted into 17 [Eq. (4)], which corresponds to the right-
hand portion of deoxysaponaceolide D. X-ray crystallography
established the full stereochemistry as depicted. The diol 15a
may serve as a general intermediate for the synthesis of
O
Ph3P
O
O
O
THF
92%
O
16
(4)
H
H
O
17
saponaceolides. While saponaceolide D requires retention of
both hydroxyl groups, the syntheses of saponaceolides A ± C
require the removal of the free hydroxyl group from the
monosilyl ether 15b. For this purpose, we turned to free
radical chemistry [Eq. (5)]. The reaction proceeded unevent-
S
S
N
N
N
N
N
O
OCH3
OCH3
N
15b
PhCH3
97%
H
H
TBDMSO
(5)
18
1) (C4H9)3SnH
(C2H5)3B/O2
PhCH3, 81%
To compare the two processes, the corresponding Heck
reactions of vinyl bromides 13b and 13c were examined.
Cyclization of 13b to 14 proceeded smoothly to generate a
poorer 1:1 ratio of the cis:trans products 14 [Eq. (2)].[18] On
the other hand, the rigid acetonide 13c leads to the cyclization
product 16 in a cis:trans ratio as high as 2.4:1. Efforts to
further increase this ratio failed. Preparatively, the highest
yield (53%) of the cis isomer stemmed from the conditions[19]
OCH3
OCH3
2) TBAF/THF
94%
H
H
HO
19
fully via thionocarbamate 18 to give the saponaceolide
intermediate 19.[20] The pure cis isomer can be isolated at
the stage of the diol 15a, its monosilyl ether 15b, or the
monoalcohol 19. Practically, the
separation was best done with
alcohol 19.
The palladium-catalyzed cy-
cloisomerization and Heck reac-
tions described here provide the
best route to date for the forma-
tion of the requisite cis-3,3-di-
methyl-2,4-disubstituted methyl-
enecyclohexyl unit 1 common to
many natural products. The un-
usual nature of this core unit is
further illustrated by the com-
Angew. Chem. Int. Ed. 1999, 38, No. 24
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3824-3663 $ 17.50+.50/0
3663

COMMUNICATIONS
[18] Conditions analogous to those of Grigg et al.: R. Grigg, P. Stevenson,
T. Worakun, J. Chem. Soc. Chem. Commun. 1984, 1073.
[19] In our case the effect of silver salts on the regioselectivity of the b-
hydrogen elimination is opposite that normally observed; see T.
Jeffrey, Tetrahedron Lett. 1993, 34, 1133.
[20] See: D. H. R. Barton, S. W. McCombie, J. Chem. Soc. Perkin Trans. 1
1975, 1574; K. Pankiewicz, A. Matsuda, K. A. Watanabe, J. Org.
Chem. 1982, 47, 485.
pletely different regioselectivity, wherein a 1,4- rather than a
1,3-diene was obtained in the cycloisomerization and in the
silver-promoted Heck cyclization. The route constitutes a
reasonable strategy for the synthesis of cyclohexyl cores that
can provide access to all the currently known saponaceolides.
Its practicality is highlighted by its use in a successful synthesis
of saponaceolide B.[21]
[21] B. M. Trost, J. R. Corte, Angew. Chem. 1999, 111, 3947 ± 3949; Angew.
Chem. Int. Ed. 1999, 38, 3664 ± 3666.
Received: June 23, 1999 [Z13611IE]
German version: Angew. Chem. 1999, 111, 3945 ± 3947
Keywords: C C coupling ´ cyclohexanes ´ cycloisomeriza-
tions ´ enynes ´ natural products
Total Synthesis of (‡)-Saponaceolide B**
Barry M. Trost* and James R. Corte
[1] H. Miyake, H. Ito, T. Yoshida, Can. J. Chem. 1997, 75, 734; H. Ito, Y.
Miyake, T. Yoshida, Chem. Pharm. Bull. 1995, 43, 1260; F.-J. Marner,
W. Krick, B. Gellrich, L. Jaenicke, W. Winter, J. Org. Chem. 1982, 47,
2531; W. Krick, F.-J. Marner, L. Jaenicke, Helv. Chim. Acta 1984, 67,
318.
[2] O. B. Weeks, A. R. Montes, A. G. Andrewes, J. Bacteriol. 1980, 141,
1272; S. Hertzberg, S. Liaaen-Lensen, Acta Chem. Scand. 1977, 31, 215.
[3] Z. Pang, K. E. Berquist, O. Sterner, Acta Chem. Scand. 1994, 48, 453;
C. Geraci, M. Piattelli, C. Tringali, Magn. Reson. Chem. 1991, 29, 603;
M. De Bernardi, L. Garlaschelli, L. Toma, G. Vidari, P. Vita Finzi,
Tetrahedron 1991, 47, 7109; M. De Bernardi, L. Garlaschelli, G. Gatti,
G. Vidari, P. Vita Finzi, Tetrahedron 1988, 44, 235.
The saponaceolides A ± D (1a ± d), discovered by Vidari
and co-workers from the northern Italian mushroom Tricho-
loma saponaceum,[1] possess antitumor activity in 60 human
cancer cell lines.[2] They possess several challenging structural
OH
2'
a: R1 = OH, R2 = R3 = H
b: R1 = R2 = R3 = H
R3
3
O
O
6'
O
2
O
5'
6
1
O
c: R1 = R2 = OH, R3 = H
d: R1 = R3 = OH, R2 = H
9'
H
H
R2
10
[4] G. Vidari, G. Lanfranchi, P. Sartori, S. Serra, Tetrahedron: Asymmetry
1995, 6, 2977.
R1
1
[5] M. Lanz, B. Bartels, H. Pfander, Helv. Chim. Acta 1997, 80, 804; H.
Monti, G. Audran, J.-P. Monti, G. Leandri, J. Org. Chem. 1996, 61,
6021; C. Chapius, R. Brauchli, Helv. Chim. Acta. 1993, 76, 2070; J. P.
Ferezou, M. Julia, Tetrahedron 1990, 46, 475; C. Nussbaumer, G.
Frater, Helv. Chim. Acta 1988, 71, 619; F. Leyendecker, M. T. Comte,
Tetrahedron 1987, 43, 85; T. Kawanobe, M. Iwamoto, K. Kozami, M.
Matsui, Agric. Biol. Chem. 1987, 51, 791; T. Kitahara, K. Tanida, K.
Mori, Agric. Biol. Chem. 1983, 47, 581.
elements including the unique tricyclic trioxaspiroketal and a
cis 2,6-disubstituted 1-methylene-3,3-dimethylcyclohexane
ring. While a number of synthetic efforts have targeted
portions of these structures,[2, 3] until recently only a synthesis
of 2-epi-saponaceolide B has appeared, where the difficulties
of controlling the ring configuration were highlighted.[4] We
report here our efforts that have culminated in a synthesis of
(‡)-saponaceolide B (1b), a member that exhibits significant
activity in four human cancer cell linesÐleukemia K-562,
nonsmall cell lung NCI-H23, melanoma LOX-IMVI, and SK-
MEL-5.[2]
[6] G. Vidari, N. Pazzi, G. Lanfranchi, S. Serra, Tetrahedron Lett. 1999, 40,
3067.
[7] M. J. Krische, B. M. Trost, Tetrahedron 1998, 54, 3693; B. M. Trost, Y.
Li, J. Am. Chem. Soc. 1996, 118, 6625; B. M. Trost, D. L. Romero, F.
Rise, J. Am. Chem. Soc. 1994, 116, 4268; B. M. Trost, G. J. Tanoury, M.
Lautens, C. Chan, D. T. MacPherson, J. Am. Chem. Soc. 1994, 116,
4255.
[8] Review: B. M. Trost, M. J. Krische, Synlett 1998, 1.
[9] B. M. Trost, M. Lautens, J. Am. Chem. Soc. 1987, 109, 1469; B. M.
Trost, M. Lautens, Tetrahedron 1987, 43, 4817.
[10] P. G. McDougal, J. G. Rico, Y.-I. Oh, B. D. Condon, J. Org. Chem.
1986, 51, 3388.
[11] Y. Terao, M. Akamatsu, K. Achiwa, Chem. Pharm. Bull. 1991, 39, 823.
Also known as lipase PS or lipase P, it is claimed by Amano to be
identical to PPS; see R. J. Kazlauskas, A. N. E. Weissfloch, A. T.
Rappaport, L. A. Cuccia, J. Org. Chem. 1991, 56, 2656.
[12] Reviews: Enzyme Catalysis in Organic Synthesis (Eds.: K. Drauz, H.
Waldmann), VCH, New York, 1995; T. Itoh, Y. Takagi, H. Tsukube, J.
Mol. Catal. B 1997, 3, 261.
[13] Monoacetate 10b was converted into (R)-2,2-dimethyl-3-hydroxy-
methyl-g-butyrolactone, which was also prepared from known (R)-2-
acetoxymethyl-4-penten-1-ol; see K. Tsuji, Y. Tecao, K. Achiwa,
Tetrahedron Lett. 1989, 30, 6189.
Scheme 1 illustrates the retrosynthetic analysis into the
three subunits 2 ± 4. The central core unit 3 represents a
significant challenge since the cis configuration at C2 and C6,
which is thermodynamically less stable than the trans config-
uration, has proven difficult to accessÐa fact emphasized by
the reported synthesis of only the 2-epi isomer.[4] The
accompanying paper describes a synthesis of this unit.[13]
[*] Prof. B. M. Trost, Dr. J. R. Corte
Department of Chemistry
Stanford University
Stanford, CA 94305-5080 (USA)
Fax: (‡1)650-725-0002
E-mail: bmtrost@leland.stanford.edu
[**] We thank the National Institutes of Health, General Medical
Sciences, for their generous support of our programs. J.R.C. held a
NSF predoctoral Fellowship and a Veatch Fellowship. Mass spectra
were provided by the Mass Spectrometry Facility of the University of
California-San Francisco, supported by the NIH Division of Research
Resources. We are grateful to Prof. Dr. G. Vidari for copies of spectra
of the natural product.
[14] H. Hopf, I. Bohm, J. Kleinschroth, Org. Synth. 1981, 60, 41.
[15] T. Mandai, J. Nokami, T. Yano, Y. Yoshinaga, J. Otera, J. Org. Chem.
1984, 49, 172.
[16] O. Arjona, R. Perez-Ossorio, A. Perez-Rubalcaba, M. L. Quiroga, J.
Chem. Soc. Perkin Trans. 2 1981, 597.
[17] B. M. Trost, P. A. Hipskind, J. Y. L. Chung, C. Chan, Angew. Chem.
1989, 101, 1559; Angew. Chem. Int. Ed. Engl. 1989, 28, 1502; B. M.
Trost, J. Y. L. Cung, J. Am. Chem. Soc. 1985, 107, 4586.
Supporting information for this article is available on the WWW
under http://www.wiley-vch.de/home/angewandte/ or from the author.
3664
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3824-3664 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 24