On Pd-catalyzed cycloisomerization versus cycloreduction. A general strategy for drimane synthesis and a short total synthesis of siccanin

Article abstract of DOI:10.1021/ja9543098

The Pd-catalyzed reductive cyclization of diynes represents an attractive alternative to the Pd-catalyzed cycloisomerization to 1,3-dienes. By minimization of steric constraints, cyclizations that fail in the latter case may now succeed. For the cycloredu

Full text of DOI:10.1021/ja9543098

5146
J. Am. Chem. Soc. 1996, 118, 5146-5147
Scheme 1. Synthesis of Monocyclic Diola
On Pd-Catalyzed Cycloisomerization versus
Cycloreduction. A General Strategy for Drimane
Synthesis and a Short Total Synthesis of Siccanin
Barry M. Trost,* Fred J. Fleitz, and William J. Watkins
Department of Chemistry, Stanford UniVersity
Stanford, California 94305-5080
a
(a) LDA, THF, I(CH2)3Cl, -78° to rt, workup with HCl, H2O. (b)
ReceiVed December 28, 1995
CH2(CO2C2H5)2, TiCl4, C5H5N, CCl4, THF, 0°. (c) TMSCtCCH2CH2I,
n-C4H9Li, CuCN, ether, -78° to 0°. (d) 10% NaI, Cs2CO3, 3 Å MS,
acetone, reflux. (e) LAH, THF, 0°.
Our recently developed Pd-catalyzed cycloisomerization of
enynes has offered an opportunity to develop novel synthetic
strategies for complex targets.1 The drimanes are an important
terpenoid class whose members exert broad biological activities
including antibacterial, antifungal, cytotoxic, insecticidal, etc.2
To illustrate the types of new strategic insight proffered and to
explore the scope of the cycloisomerizations illustrated in eq
1, path a, an effort directed at the evolution of a general approach
to this family was undertaken. Under a wide variety of
conditions, the cis isomer 1 (R ) H or TMS) gave none of the
cyclized product and the trans isomer gave only a 30% yield
for R ) H [2.5% (dba)3Pd2, 5% HOAc, 10% (o-C7H7)3P, PhH,
80°]. In contradistinction to this result, an enyne metathesis to
form 3, which presumably involves similar intermediates,
succeeds [5% TCPCTFE, (i-C3H7O)3P, PhH, 68% yield].3
Clearly, the stereoelectronics of seemingly related processes are
different. A possible source of the problem lies in the
geometrical restrictions of a vinyl group in allowing intramo-
lecular carbametalations. We, therefore, turned to consideration
of a different new reaction being developed in our laboratories,
the catalytic diyne reductive cyclization,4 wherein its ability to
form six-membered rings and its stereoelectronic requirements
vis-a-vis the cycloisomerization would be compared.
Scheme 2. Synthesis of Diynea
a (a) PCC, celite, CH2Cl2, from 6a Q, from 9, 92%. (b) (CH3O)2P-
(O)CHN2, NaHMDS, THF, -78° to rt, 77% to 11, 91% to 12a. (c) (i)
Morpholine, I2, PhH, rt, (ii) Dimethyl orcinol, n-C4H9Li, THF, -78°
to 0 °C, CuCN, 96%; (d) K2CO3, CH3OH, rt, 97%.
The synthesis of 4, both trans and cis, requires differentiation
of the hydroxy groups of the diol 5 prepared as shown in Scheme
1 in 48% overall yield. Surprisingly, acylation with either
pivaloyl chloride or benzoyl chloride (C5H5N, CH2Cl2, rt) led
to preferential reaction with the “more hindered” hydroxyl group
with the latter giving better results. The monoester 6b, isolated
in 77% yield, easily and cleanly separates from the other
products which are the diester 7b (10%) and the alternate
monoester (10%). Since the minor products can be converted
back to 5, a single recycle can boost the yield to 92%. On the
other hand, silylation gives the monosilyl derivatives 8 and 9
in 40% and 58% isolated yields, respectively. Recycling the
easily separated minor product 8 by acidic desilylation (TsOH,
CH3OH, rt, 96%) to 5 and silylation as before raises the yield
of the desired silyl ether 9 to 80%. NOE experiments establish
the chemoselectivity which is confirmed by the completed
synthesis. Thus, 6 and 9 provide potential entry to either ring
fusion series.
Scheme 2 outlines the synthesis of the various diynes.
Formation of the alkyne from the aldehyde failed with the
Corey-Fuchs protocol.5 Olefination via phosphonates6 proved
superior with excellent results obtained with dimethyl phos-
phonodiazomethane6a,b thereby providing the cyclization sub-
strates 11 and 12a. Subjection of 11 or 12a to our reductive
cyclization conditions [(o-C7H7)3P, (dba)3Pd2‚CHCl3, HOAc,
(C2H5)3SiH, PhH]4 failed. Among the various permutations
explored, two variables proved important, ligand and acid. Use
of poorer donor ligands like trifurylphosphine, Ph3As, and Ph3-
Sb led to some product 2a but in low yields (eq 1). Switching
from acetic acid to the slightly stronger acid, formic acid,
presumably to help shift the equilibrium between the Pd(0)
catalyst and the active hydridopalladium catalyst toward the
latter, proved to be the major key. For example, trans diyne
11 cycloreduced to give 2a in 87% yield [2.5% (dba)3Pd2‚CHCl3,
10% Ph3As, (C2H5)3SiH, HCO2H, PhCH3, 80 °C, 20 min], but
(1) Trost, B. M.; Krische, M. J. J. Am. Chem. Soc. 1996, 118, 233. Trost,
B. M.; Phan, L. T. Tetrahedron Lett. 1993, 34, 4735. Trost, B. M.; Tasker,
A. S.; Ru¨hter, G.; Brandes, A. J. Am. Chem. Soc. 1991, 113, 670.
(2) Jansen, B. M. M.; de Groot, A. Nat. Prod. Rep. 1991, 8, 309. Ibid.
1991, 8, 319.
(3) Trost, B. M.; Tanoury, G. J. J. Am. Chem. Soc. 1988, 110, 1636.
Trost, B. M.; Trost, M. K. J. Am. Chem. Soc. 1991, 113, 1850.
(4) Trost, B. M.; Lee, D. C. J. Am. Chem. Soc. 1988, 110, 7255. Also
see Trost, B. M.; Shi, Y. J. Am. Chem. Soc. 1993, 115, 12491 in which
six-membered rings form in a cycloisomerization of enediynes.
(5) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769.
(6) Seyferth, D.; Marmor, R. S.; Hilbert, P. J. Org. Chem. 1971, 36,
1379. Gilbert, J. C.; Weerasooriya, U. J. Org. Chem. 1979, 44, 4997. Reich,
H. J.; Willis, W. W. J. Am. Chem. Soc. 1980, 102, 5967. Comassetto, J.;
Petragnani, N. J. Organomet. Chem. 1978, 152, 295. Miwa, K.; Aoyama,
T.; Shioiri, T. Synlett 1994, 107.
S0002-7863(95)04309-5 CCC: $12.00 © 1996 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 118, No. 21, 1996 5147
Scheme 3. A Synthesis of Siccanina
Gratifyingly, the reductive cyclization saw no ill effects of the
aryl substitution and provided 13 as a geometrically homoge-
neous crystalline diene in 79% isolated yield (see Scheme 3).
Unambiguous creation of the alkene geometry sets the stage
for the three remaining stereogenic centers, a critical benefit of
the diyne cycloreduction. Attempts to directly cyclize 14 to
siccanin methyl ether (16) with protonic acids normally lead to
siccanochromene E methyl ether9 and with sodium bisulfate to
the tetrahydrofuran 15. On the other hand, BF3 etherate effects
direct cyclization to siccanin methyl ether 16.13 Demethylation
forms siccanin, identical spectroscopically in every respect
(except optical rotation) with an authentic sample. This
sequence requires 14 steps and proceeds in 3% overall yield.
Alternatively, cyclization of tetrahydrofuran 15 proceeds more
satisfactorily to siccanin methyl ether (16) under the same
conditions. The tetrahydrofuran 15 is preferably available from
13 by inverting the sequence, i.e., first protonic acid desilylation
and initial cyclization followed by O-demethylation. The
sequence via 15 requires 15 steps and proceeds in 5% overall
yield.
The Pd-catalyzed reductive cyclization of diynes represents
an attractive alternative to the Pd-catalyzed cycloisomerization
to 1,3-dienes. By minimization of steric constraints, cyclizations
that fail in the latter case may now succeed. For the cyclore-
duction, a more general protocol in which formic acid is the
key has now emerged. In the first test of this reductive
cyclization to solve a problem of complex synthesis, the reaction
performed exceptionally well. The dialkylidenecycloalkanes
that result have great versatility, especially with respect to the
type of functionality commonly found in the drimanes. Both
the cis and trans ring fused drimanes might be available.
Furthermore, asymmetric syntheses also should be readily
available using the protocol of Mukaiyama, who has developed
an oxazepandione as a chiral version of the alkylidenemalonate
for cuprate additions.14 Further work to explore the generality
of this strategy for the syntheses of other biologically interesting
drimanes will be the focus of future efforts.
a
(a) 2.5% (dba)3Pd2‚CHCl3, 10% trifurylphosphine, 2 equiv
(C2H5)3SiH, 2 equiv HCO2H, PhCH3, 80°. (b) NaH, C2H5SH, DMF,
100-120°. (c) NaHSO4, acetone, 60°. (d) BF3‚ether, CH2Cl2, ether, rt.
(e) TsOH, CH3OH, rt.
cis diyne gave 2b in <15% yield. In the latter case, use of
formic acid with no ligand proved much better, raising the yield
to 98% under otherwise identical conditions.
With this new protocol, we targeted one of the most complex
drimanes, siccanin (17, Scheme 3), a clinically important
antifungal agent.7 In spite of a number of efforts, only one
synthesis was successful to date.8,9 The effectiveness of a
strategy emanating from this methodology may be compared
to this successful route. Installation of the required aryl group
by various permutations of the widely used Pd-catalyzed cross-
coupling reaction failed.10 On the other hand, the more classical
Cu-catalyzed coupling,11,12 as shown in Scheme 2, proceeded
nearly quantitatively.
Cycloreduction of 12c really tests the mettle of this new
protocol because of the highly hindered nature of the product.
(7) (a) Ishibashi, K. J. Antibiot., Ser. A. 1962, 15, 161. (b) Ishibashi, K.;
Hirai, K.; Arai, M.; Sugasawa, S.; Endo, A.; Yasuma, A.; Matsuda, H.;
Muramutsu, T. Ann. Sankyo Res. Lab. 1970, 22, 1. (c) Bellotti, M. G.;
Riviera, L. Chemioterapia 1985, 4, 431. (d) Hirai, K.; Nozoe, S.; Tsuda,
K.; Iitake, Y.; Shirasawa, M. Tetrahedron Lett. 1967, 2177. (e) Hirai, K.;
Suzuki, K. T.; Nozoe, S. Tetrahedron 1971, 27, 6057. (f) Sugawara, S.
Antimicrob. Agents Chemother. 1969, 253. (g) Arai, M. Ishibashi, K.;
Okazaki, H. Antimicrob. Agents Chemther. 1969, 247. (h) Crippa, D.;
Albanese, C. G.; Sala, G. P. G. Ital. Dermatol. Venereol. 1985, 120, LVII
and references therein. Also, see: Kitano, N.; Kondo, F.; Kusano, K.;
Ishibashi, K. German Patent 2458886; Chem. Abstr. 1975, 83, 197813.
Kondo, K.; Nakada, S. Japanese Patent 02204413; Chem. Abstr. 1991, 114,
30165. Wood, L.; Calton, G. J. U.S. Patent 5260066; Chem. Abstr. 1994,
120, 200438.
(8) For some synthetic efforts, see: Nozoe, S.; Hirai, K. Tetrahedron
1971, 27, 6073. Oida, S.; Ohashi, Y.; Yoshida, A.; Ohki, E. Chem. Pharm.
Bull. 1972, 20, 2634. Yoshida, A.; Oida, S.; Ohasi, Y.; Tamura, C.; Ohki,
E. Chem. Pharm. Bull. 1972, 20, 2642. Oida, S.; Ohashi, Y.; Ohki, E. Chem.
Pharm. Bull. 1973, 21, 528. Liu, H.; Ramani, B. Tetrahedron Lett. 1988,
29, 6721. Kato, M.; Matsumura, Y.; Heima, K.; Yoshikoshi, A. Bull. Chem.
Soc. Jpn. 1988, 61, 1991.
(9) For a successful synthesis, see: Kato, M.; Heima, K.; Matsuma, Y.;
Yoshikoshi, A. J. Am. Chem. Soc. 1981, 103, 2434. Kato, M.; Matsuma,
Y.; Heima, K.; Fukamiya, N.; Kabuto, C.; Yoshikoshi, A. Tetrahedron 1987,
43, 711.
(10) Cf: Hirthhammer, M.; Volhardt, K. P. C. J. Am. Chem. Soc. 1986,
108, 2481. Cassar, L. J. Organomet. Chem. 1975, 93, 253. Dieck, H. A.;
Heck, R. F. Ibid. 1975, 93, 259. Sanogashire, K.; Tohda, Y.; Hagihara, N.
Tetrahedron Lett. 1975, 4467.
(11) Oliver, R.; Walton, D. R. M. Tetrahedron Lett. 1972, 5209.
Verboom, W.; Westmijze, H.; Bos, H. J. T. Tetrahedron Lett. 1978, 1441.
Letyn, J. M.; Spronck, H. J. W.; Salemink, C. A. Recl. TraV. Chim. Pays-
Bas 1978, 97, 187.
(12) For iodination of the alkyne, see: Kayama, M. J. Med. Chem. 1987,
30, 552. Southwick, P. L.; Kirchner, J. R. J. Org. Chem. 1962, 27, 3305.
Acknowledgment. We thank the National Institutes of Health,
General Medical Sciences Institute, and the National Science Foundation
for their generous support of our programs. F.J.F. held a NSF
predoctoral fellowship for part of his stay, and W.J.W. held a NATO
postdoctoral fellowship administered by the SERC of the United
Kingdom. Mass spectra were provided by the Mass Spectrometry
Facility at the University of CaliforniasSan Francisco. We are indebted
to Johnson Matthey Alfa Aesar for their generous loan of palladium
salts and to Professor Michiharu Kato of Tohoku University for a
generous gift of natural siccanin.
Supporting Information Available: Characterization data for ethyl
2-ethoxycarbonyl-3-(4-trimethylsilyl-3-butynyl)-4,4-dimethyl-7-chloro-
heptanoate, 1,1-bis(ethoxycarbonyl)-2-(4-trimethylsilyl-3-butynyl)-3,3-
dimethylcyclohexane, 5, 9, and 12-17 and experimental procedure for
cyclization of 12c to 13 (5 pages). Ordering information is given on
any current masthead page.
JA9543098
(13) Partial 1H NMR data have been recorded for siccanin methyl ether;
see ref 7e. Our data do not agree. Our spectral data are in full accord with
the assignment. Moreover, the successful completion of our synthesis and
a direct comparison of our synthetic siccanin with an authentic sample
confirm the correctness of our assignment.
(14) Mukaiyama, T.; Takida, T.; Fumimoto, K. Bull. Chem. Soc. Jpn.
1978, 51, 3368.