Quassinoid support studies: Fused carbocycle synthesis from benzoic acid derivatives via 5-hexynyl and 6-heptynyl radical cyclizations

Article abstract of DOI:10.1021/ol061802n

Eight bromoalkynes were prepared from substituted benzoic acids and treated with n-Bu₃SnH to provide trans-fused perhydroindans or cis-and trans-fused perhydronaphthalenes. Atom-transfer reactions that accompany the free radical reactions resul

Full text of DOI:10.1021/ol061802n

ORGANIC
LETTERS
2006
Vol. 8, No. 20
4577-4580
Quassinoid Support Studies: Fused
Carbocycle Synthesis from Benzoic
Acid Derivatives via 5-Hexynyl and
6-Heptynyl Radical Cyclizations
Valerie Cwynar, Matthew G. Donahue, David J. Hart,* and Dexi Yang
Department of Chemistry, The Ohio State UniVersity, Columbus, Ohio 43210
hart@chemistry.ohio-state.edu
Received July 21, 2006
ABSTRACT
Eight bromoalkynes were prepared from substituted benzoic acids and treated with n-Bu3SnH to provide trans-fused perhydroindans or cis-
and trans-fused perhydronaphthalenes. Atom-transfer reactions that accompany the free radical reactions resulted in several tandem radical
cyclizations with formation of up to three carbon−carbon bonds in a single reaction. The relationship between these reactions and an approach
to the quassinoid family of natural products is also described.
We have shown that benzoic acids can be converted to highly
functionalized fused carbocycles via a sequence that involves
sequential reductive alkylation, halolactonization, and free
radical cyclization reactions.1 An alkene has served as the
radical acceptor in most systems studied to date.2 This paper
describes a variation of this sequence in which alkynes are
used as the radical acceptor. The motivation for conducting
this research was to extend the methodology toward a unified
approach to quassinoids containing either perhydroindan or
perhydronaphthalene substructures. The quassinoids, exem-
plified in Figure 1 by chaparrinone (1) and (5R)-polyandrane
(2), remain of interest as targets for synthesis due to their
complex structures and broad biological activity.3-5 Extrapo-
lation of this methodology and some surprising observations
made along the way are described herein.
Figure 1. Quassinoids and cyclization substrates.
Preparation of Cyclization Substrates. Substrates select-
ed for the aforementioned free radical cyclization studies
were of types 3 and 4 (Figure 1), which were expected to
provide perhydroindans and perhydronaphthalenes, respec-
tively. Substrates where the C11 substituent (X) was a methyl
(1) Chuang, C.-P.; Hart, D. J. J. Org. Chem. 1983, 48, 1782. Chuang,
C.-P.; Gallucci, J. C.; Hart, D. J.; Hoffman, C. J. Org. Chem. 1988, 53,
3218. Hart, D. J.; Huang, H. C.; Krishnamurthy, R.; Schwartz, T. J. Am.
Chem. Soc. 1989, 111, 7507.
(2) For oxime ethers, hydrazones, and nitriles, see: Hart, D. J.; Seely,
F. L. J. Am. Chem. Soc. 1988, 110, 1631. Ellis, D. A.; Hart, D. J.; Zhao, L.
Tetrahedron Lett. 2000, 41, 9357. Chenera, B.; Chuang, C.-P.; Hart, D. J.;
Hsu, L.-Y. J. Org. Chem. 1985, 50, 5409.
(3) For lead reviews, see: Spino, C. Synlett 2006, 23. Kawada, K.; Kim,
M.; Watt, D. S. Org. Prep. Proced. Int. 1989, 21, 521. Polonsky, J. J.
Fortschr. Chem. Org. Naturst. 1985, 47, 221.
10.1021/ol061802n CCC: $33.50
© 2006 American Chemical Society
Published on Web 08/31/2006

group were studied first because we suspected their synthesis
would be simpler than substrates with a C11 acetal (X )
OCH3) due to the known sensitivity of the latter type of
compound.6
oxetane with acetylide 7 gave a separable mixture of alkynol
15 (45%) and its C7 diastereomer (38%).12
The synthesis of substrates of type 3 and 4 with a C11
methoxy group required a different point of departure
(Scheme 2). Ester 16 was prepared in 100-g quantities using
We began with the known aldehyde 5, prepared in four
steps from methyl 3,5-dimethylbenzoate (Scheme 1).7 Treat-
Scheme 2. More Cyclization Substrates
Scheme 1. Synthesis of Cyclization Substrates
ment of 5 with dimethylsulfonium methylide gave a separable
2:1 mixture of epoxide 6 and its C7 diastereomer, respec-
tively, in 75% combined yield.8 The epoxide was opened
with a series of acetylides (7-9), in the presence of boron
trifluoride etherate, to provide homopropargylic alcohols
10-12 in variable yields (43-95%).9 Treatment of 12 with
tetra-n-butylammonium fluoride also provided cyclization
substrate 13 (81%).10 Treatment of 5 with boron dimethyl-
sulfoxonium methylide under more pressing conditions
provided an inseparable 1:1 mixture of oxetane 14 and its
C7 diastereomer in 85% combined yield.11 Opening of the
a known procedure.13 Birch reduction of 16 was followed
by alkylation of the intermediate enolate with iodomethyl
pivalate.14 Reduction of the resulting 17 with lithium alum-
inum hydride gave crystalline diol 18 in 50% overall yield
from 16. Bromoetherification of 18 using NBS in dichlo-
romethane in the presence of powdered potassium carbonate
gave 19. The base was critical to the success of this reaction.
Bromo ether 19 was very sensitive and thus, was oxidized
directly to aldehyde 20 using the Swern conditions.15
Whereas 20 proved more stable than 19, it was also
sensitive and was treated with dimethylsulfonium methylide
to provide a 2:1 mixture of epoxide 21 and its C7 diastere-
omer in combined 85% yield. The two epoxides could be
separated by column chromatography on a several gram
scale. Treatment of pure 21 with acetylides 7 and 9 gave
cyclization substrates 22 (66%) and 23 (50%). Conversion
of epoxide 21 to oxetane 24 was accomplished using
(4) For some total syntheses of quassinoids, see: Shing, T. K. M.; Yeung,
Y. Y. Angew. Chem., Int. Ed. 2005, 44, 7981 (samaderine Y). Stojanac,
N.; Valenta, Z. Can. J. Chem. 1991, 69, 853 (quassin). Kim, M.; Kawada,
K.; Gross, R. S.; Watt, D. S. J. Org. Chem. 1990, 55, 504 (quassin). Grieco,
P. A.; Ferrino, S.; Vidari, G. J. Am. Chem. Soc. 1980, 102, 7586 (quassin).
Sasaki, M.; Murae, T.; Takahashi, T. J. Org. Chem. 1990, 55, 528
(bruceantin). Grieco, P. S.; Collins, J. L.; Moher, E. D.; Fleck, T. J.; Gross,
R. S. J. Am. Chem. Soc. 1993, 115, 6078 (chaparrinone). Walker, D. P.;
Grieco, P. A. J. Am. Chem. Soc. 1999, 121, 9891 (polyandranes).
(5) For a lead reference on biological activity, see: Fukamiya, N.; Lee,
K.-H.; Ilias, M.; Murakami, C.; Okano, M.; Harvey, I.; Pelletier, M. Cancer
Lett. 2005, 220, 37.
(11) Okuma, K.; Tanaka, Y.; Kaji, S.; Ohta, H. J. Org. Chem. 1983, 48,
8, 5133.
(12) Yamaguchi, M.; Nobayashi, Y.; Hirao, I.; Tetrahedron 1984, 40,
4261. Kurek-Tyrlik, A.; Wicha, J.; Zarecki, A.; Snatzke, G. J. Org. Chem.
1990, 55, 3484.
(6) Chenera, B.; Chuang, C.-P.; Hart, D. J.; Lai, C.-S. J. Org. Chem.
1992, 57, 2018.
(7) Donahue, M. G.; Hart, D. J. Can. J. Chem. 2004, 82, 314.
(8) Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1352.
Kutsuma, T.; Nagayama, I.; Okazaki, T.; Sakamoto, T.; Akaboshi, S.
Heterocycles 1977, 8, 397.
(9) Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1983, 24, 391.
(10) Nakamura, E.; Kuwajima, I. I. Angew. Chem., Int. Ed. Engl. 1976,
15, 498. Mohr, P. Tetrahedron Lett. 1991, 32, 2223.
(13) Turner, F. A.; Gearien, J. E. J. Org. Chem. 1959, 24, 1952. Frutos,
O.; Atienza, C.; Echavarren, A. M. Eur. J. Chem. 2001, 163.
(14) Van Bekkum, H.; Van Den Bosch, C.; Van Minnenpathuis, G.; De
Mos, J. C.; van Wijk, A. M. Recuil 1971, 90, 137. Schultz, A. G.; Taylor,
R. E. J. Am. Chem. Soc. 1992, 114, 3937.
(15) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43,
3, 2480.
4578
Org. Lett., Vol. 8, No. 20, 2006

dimethyloxosulfonium methylide in 78% yield. Opening of
the oxetane with acetylide 7 gave alkynol 25 in 92% yield.
Free Radical Cyclizations. Cyclization results with
compounds of type 3 are summarized in Scheme 3. Treat-
Scheme 4. Preparation of Perhydroindans
Scheme 3. Cyclization of 10-13
in benzene gave a mixture of products from which 32 and
33 were isolated in 42% and 31% yields, respectively.
Whereas 32 was clearly a cyclization-reduction product, 33
appeared to result from cyclization of the initially formed
radical, 1,6-hydrogen atom transfer from the methoxy group
to the intermediate vinyl radical, and 6-endo cyclization of
the resulting alkoxymethyl radical (34) onto the intermediate
R,â-unsaturated ester.21,22 A labeling experiment using n-Bu3-
SnD indicated that 32 was almost entirely derived from 34.23
This suggested that at high concentrations of tri-n-butyltin
hydride 32 would be the major product and that at low
concentrations of tri-n-butyltin hydride 33 would become the
major product. This was the case. Thus, treatment of 22 (10
mM) with excess n-Bu3SnH (100 mM) gave 32 and 33 in
75% and 10% isolated yields, respectively. On the other
hand, slow addition of n-Bu3SnH to a solution of 22 (10
mM in PhH) gave 32 and 33 in 6% and 64% isolated yields,
respectively. The behavior of alkynylsilane 23 was qualita-
tively similar to that of 22. The standard cyclization
conditions provided 35 (11%, E/Z ) 3:1) and 36 (18%), but
substantial amounts of reduction product 37 (60%) were also
obtained.24 It was possible to obtain 36 in 63% yield (along
with 6% of 35 and 3% of 37) under high dilution conditions.
The cyclization results with substrates of type 4 (X ) CH3
and X ) OCH3) are shown in Scheme 5. Propiolate 15 reacts
under standard conditions to afford 39 (23%), 40 (46%), and
41 (12%). Thus, the initially formed radical partitions
approximately 2:1 between trans-fused (40 and 41) and cis-
fused (39) perhydronaphthalenes. The structure of 39 was
established by X-ray crystallography.16 Whereas the origin
of 40 is clear, 41 appears to result from a 1,5-hydrogen atom
transfer of an intermediate vinyl radical, followed by a rare
5-endo cyclization of the resulting 5-pentenyl radical.
ment of 10 with n-Bu3SnH and AIBN in benzene under
reflux provided crystalline trans-perhydroindan 26 in 75%
yield. The structure of this material was clear on the basis
of spectroscopy and was confirmed by X-ray crystal-
lography.16 The cyclization of substrate 11 also proceeded
in high yield (95%) but gave a separable 1:1 mixture of
geometrical isomers 27 and 28.17 Likewise, the free radical
cyclization of TMS-alkyne 12 gave a separable 4:1 mixture
of geometrical isomers 29 and 30 in 85% combined yield.
Terminal alkyne 13 gave a low yield of the expected
cyclization product 31 (24%).18 We note that 31 was best
prepared (70% yield) by protodesilylation of the 4:1 mixture
of vinylsilanes 29 and 30 using CuCl2 in ethanol.19
Several aspects of these cyclizations are notable. First, they
all produced trans-fused perhydroindans. This is in accord with
results obtained with the aforementioned alkene substrates
and can be rationalized in much the same manner (presence
of cis-fused 2-oxabicyclo[3.3.0]octane substructure in the pro-
duct).1 Second, the clean olefin geometry obtained in the cy-
clization of 10 was unexpected. Thus, we conducted a label-
ing experiment in which the cyclization of 10 was initiated
using n-Bu3SnD. The resulting 26 was equally labeled with
deuterium at the vinylic C1 position and C11 methyl group, in-
dicating that the presumed intermediate vinyl radical is, in part,
reduced intramolecularly by a 1,5-hydrogen atom transfer.20
The cyclization results with substrates of type 3 (X )
OCH3) are also shown in Scheme 4. Treatment of bromide
22 under the standard cyclization conditions (initial [n-Bu3-
SnH] ) 18 mM, initial [22] ) 9 mM, 0.06 equiv of AIBN)
(16) We thank Dr. Judith Gallucci (OSU) for performing this X-ray
cyrstallographic analysis.
(17) The ratio of products was determined by NMR spectroscopy of the
crude product mixture. The isomers were not completely separated by
chromatography, but pure samples of each isomer were obtained (see the
Supporting Information).
(18) Impure materials that appear to be derived from initial addition of
tri-n-butylstannyl radical to the terminal acetylene were also detected.
(19) To our knowledge, this is a new method for protodesilylation of
vinylsilanes. Protodesilylation using p-toluenesulfinic acid (Buchi, G.;
Wuest, H. Tetrahedron Lett. 1977, 18, 4305) proceeded in 43% yield.
(20) The ratio of deuterated products was determined by integration of
signals appearing at δ 6.08 (dCD) and δ 1.59 (CH2D) in the 2H NMR
spectrum deuterated 26.
(21) For 5-hexynyl radical cyclizations of ynoates, see: Lowinger, T.
B.; Weiler, L. J. Org. Chem. 1992, 57, 6099.
(22) For 5-hexynyl radical cyclization-radical translocation reactions,
see: Martinex-Grau, A.; Curran, D. P. Tetrahedron 1997, 53, 5698.
Rochigneux, I.; Fontaneux, M.-L.; Malanda, J.-C.; Doutheau, A. Tetrahedron
Lett. 1991, 32, 2017. Choi, J.-K.; Hart, D. J. Tetrahedron 1985, 41, 3159.
(23) The ratio of deuterated products (93% deuteration on the methoxy
group and 7% deuteration at C1) was determined by integration of signals
appearing at δ 7.0 (dCD) and δ 3.3 (OCH2D) in the 2H NMR spectrum of
deuterated 33.
(24) Alkyne 37 was actually characterized after desilylation to provide
38 (see the Supporting Information).
Org. Lett., Vol. 8, No. 20, 2006
4579

addition, due to ring-size differences the chemistry that
follows the 5-hexynyl radical cyclizations (from 3) is less
complex than that following 6-heptynyl radical cyclizations
(from 4). This is due to differences in the proximity of the
intermediate C11 methyl and alkoxymethyl radicals (from 3
and 4, respectively) to the intermediate C10-alkylidene groups.
In terms of the targets that stimulated this research, perhy-
droindans 33 and 36 both have features that render them
potential intermediates in an approach to C19-quassinoids
such as 2.25 On the other hand, stereochemical issues at C9
will have to be addressed before the potential of the
6-heptynyl radical cyclizations for the synthesis of C20-
quassinoids such as 1 can be fully realized.25,26
Scheme 5. Preparation of Perhydronaphthalenes
Acknowledgment. This paper is dedicated to the memory
of Nabi Magomedov, his wife Natalya Scherbinina, and their
son Amir Magomedov who died in a tragic accident on
February 7, 2006.
Substrate 25 provided cis-fused perhydronaphthalene 42
(19%) and trans-fused perhydronaphthalene 43 (41%). The
origin of 43 is most likely alkoxymethyl radical 44.
Supporting Information Available: Spectroscopic and
analytical data for new compounds, crystallographic data,
and experimental procedures. This material is available free
of charge via the Internet at http://pubs.acs.org.
The behavior of radicals derived from 3 and 4 are
qualitatively similar, with two important differences. Radicals
derived from 3 give clean stereochemistry at C9 while
radicals from 4 lead to mixtures of stereoisomers at C9. This
is a consequence of the absence of an oxabicyclo[3.3.0]-
octane substructure in products derived from 4, and empha-
sizes that it is this feature that renders the cyclizations of
radicals derived from 3 highly diastereoselective at C9. In
OL061802N
(25) The conversions of 16 to 33 or 36 each require only seven steps
and proceed in 9% and 7% overall yields, respectively.
(26) We note that the C7 diastereomers of cyclization substrates 12, 15,
and 25 behave similarly to the substrates described herein.
4580
Org. Lett., Vol. 8, No. 20, 2006